EPOS/POSNA Abstract Book (2024)

Radiological, clinical, and functional outcome of children with traumatic hip dislocation: review of 66 cases

Sara De Salvo, Shunyou Chen, Fabio Sammartino, Jeanne-Agathe Mujadiki Luesa, Yunan Lu, Wentao Wang, Liwei Shi, Lianyong Li, Vito Pavone, Federico Canavese, France-China-Italy Traumatic Hip Dislocation Study Group

CHU Lille, Lille, France

LOE-Prognostic-Level III

Purpose: Current understanding of the outcomes of traumatic hip dislocation (THD) in pediatric patients is still limited. This study aims to assess the radiological, clinical, and functional results of patients with THD, whether in isolation or in conjunction with acetabular (ACF) or proximal femur (PFF) fracture, and to identify potential risk factors for unfavorable outcomes.

Methods: This is a retrospective study of pediatric patients with THD consecutively enrolled at three different institutions between January 2016 and July 2023. Patients were categorized into three groups: isolated THD (Group A), THD with ACF (Group B), and THD with PFF (Group C). Standard radiographs were utilized to classify each THD and identify the presence of any other associated bone lesions. Clinical and functional outcomes were assessed via the Harris Hip score (HHS). Avascular necrosis (AVN) was determined according to the Ratliff criteria. The association between outcome and associated injuries, age at trauma (≤10 versus > 11 years), traumatic mechanism (low energy versus high energy), reduction type (open versus closed), and dislocation’s direction (posterior versus anterior) was evaluated.

Results: Sixty-six cases of unilateral THD (63 posterior) were analyzed, consisting of 42 males and 24 females, with an average age of 10.5 years (3–18 years). Out of these, 25 patients were aged ≤10 years (37.88%), of which 18 had low-energy trauma. Meanwhile, 41 patients were >11 years old (62.12%), of which 25 had high-energy trauma (p < 0.0001). It was observed that Group A patients were significantly younger than those in groups B and C (p < 0.0001). Group A had 5/31 patients with labral injury (16.1%), Group B had 2/22 patients with AC fractures misdiagnosed >3 weeks after injury (18%), and Group C had 3/13 patients with AVN (23%). Most patients exhibited a favorable mean HHS (92.17; SD: ±7.16), except for individuals with AVN and those with misdiagnosed AC fractures (p < 0.05).

Conclusions: The outcome of THD is worse in patients with AVN secondary to PF fracture, misdiagnosed AC fractures, high-energy trauma, and older age at the time of injury. A computed tomography (CT) scan is necessary to rule out AC fractures in isolated dislocations. If pain persists after reduction, a magnetic resonance imaging (MRI) is needed to uncover labral injury. Timely diagnosis and treatment of these lesions typically lead to a favorable outcome.

Significance: In more than 50% of THD cases, there are associated injuries that must be eliminated being significant risk factors for negative outcomes.

Diaphyseal femur fractures in children under the age of 3—risk factors for non-accidental trauma: a CORTICES multi-center study

Manya Bali, Patricia E. Miller, Benjamin J. Shore, Scott B. Rosenfeld, CORTICES

Boston Children’s Hospital, Boston, MA, USA

LOE-Prognostic-Level III

Purpose: The American Academy of Orthopaedic Surgeons (AAOS) recommends that children (<36 months) who present with a diaphyseal femur fracture (DFF) undergo Non-Accidental Trauma (NAT) workup. Compliance with this recommendation has been reported to be poor and identification of risk factors associated with positive NAT diagnosis, elusive. The purpose of this study was to analyze a multicenter national database to calculate the frequency of NAT screening, NAT diagnosis, and risk factors associated with a positive NAT diagnosis in children (<36 months) presenting with a DFF.

Methods: A total of 1263 children (<36 months) with DFF presenting to one of 15 CORTICES hospitals were identified. Positive NAT was recorded when a child was removed from their caregivers by CPS. Percent compliance of NAT screening and proportion of patients who received a positive NAT diagnosis were calculated with 95% confidence interval (CI). Multivariable logistic regression analysis was used to determine factors associated with positive NAT.

Results: The median age was 23 months (0.2–36 months). Seventy-one percent were male. Fifty-six percent (704/1263) underwent NAT evaluation (95% CI, 52.9–58.5). Fifteen percent of all patients had a positive NAT diagnosis (185/1263; 95% CI = 12.8–16.7). Of the 704 patients who underwent NAT evaluation, 26% had a positive NAT diagnosis (95% CI = 23.1–29.7). Multivariable logistic regression analysis found children <15.4 months to be 5.5 times (odds ratio (OR) = 5.52; 95% CI = 4.14–7.37; p < 0.001) more likely of being diagnosed with NAT compared to older children. Patients who identified as non-White had two times the odds of a positive NAT diagnosis compared to patients identifying as White (OR = 2.04; 95% CI = 1.53–2.71; p = 0.01). For each 10-unit increase in Area Deprivation Index (ADI), the odds of a positive NAT diagnosis increased by 18% (OR = 1.18; 95% CI = 1.17–1.19; p = 0.007). Patients with conflicting mechanism of injury had 3.5 times the odds (OR = 3.49; 95% CI = 2.45–4.98; p < 0.001) of a positive NAT diagnosis.

Conclusions: Only 56% of patients were evaluated for NAT; however, 26% of those evaluated were positive for NAT. Younger age, race, socioeconomic factors (defined here by ADI), and unknown mechanism of injury are risk factors to consider when assessing for NAT in children (<36 months) with DFF.

Significance: This is the largest report on the risk of NAT associated with children <36 months presenting with DFF. It is the first use multicenter/multi-region data to confirm previously reported risk factors. It is also the first report to link socioeconomic factors (ADI) to NAT in this population. Improved national compliance with NAT screening is necessary to improve the care we provide to our children.

Increased odds of non-accidental traumatic fractures in pediatric patients with intellectual disability disorder: a stratified analysis

Rishi Gonuguntla, David Momtaz, Mehul Mittal, Beltran Torres-Izquierdo, Pooya Hosseinzadeh

Washington University in St. Louis, St. Louis, MO, USA

LOE-Prognostic-Level III

Purpose: Pediatric patients with an intellectual disability disorder (IDD) are known to be at a higher risk of experiencing child abuse. In this study, we aim to determine the odds of a pediatric patient experiencing a non-accidental traumatic fracture (NATF) and stratify this by degree of intellectual disability when compared to accidental traumatic fractures (ATFs).

Methods: A large, nationally representative sample of 15,802,711 children was identified, with 16,441 of those patients sustaining NATF and 15,786,270 sustaining an ATF between 1 January 2003 and 31 December 2022. Of these patients, 260 of the NATF patients had some degree of IDD, and 3450 of the ATF patients had some degree of IDD. Patients were then stratified by degree of IDD. Retrospective cohort analysis was performed to determine odds ratios of patients in each category to sustain an NATF compared to an ATF.

Results: When comparing patients with profound intellectual disability to those without intellectual disability, patients had a 184.28 times greater odds of sustaining an NATF (p < 0.0001), patients with severe intellectual disability had a 145.523 times greater odds of sustaining an NATF (p < 0.0001), patients with moderate intellectual disability had a 36.34 times greater odds of sustaining an NATF (p < 0.0001), and patients with a mild intellectual disability had a 35.32 times greater odds of sustaining an NATF (p < 0.0001) (Figure 1).

Conclusions: Patients with increasingly severe intellectual disability had increasing odds to sustain an NATF relative to patients without intellectual disability. It is critical that physicians are aware of this relationship to identify patients who may be at increased risk of experiencing abuse.

Significance: The stark escalation in odds of NATF with increasing severity of intellectual disability underscores a dire need for enhanced physician awareness, improved preventive measures, and stringent child protection protocols.

Ischial tuberosity avulsion fractures: treatment and return to sport in athletes with displaced fragments

Jayson Saleet, Eduardo Novais, Yi-Meng Yen, Mininder S. Kocher, Lyle J. Micheli, Benton E. Heyworth

Boston Children’s Hospital, Boston, MA, USA

LOE-Therapeutic-Level III

Purpose: Ischial tuberosity avulsion fractures are relatively rare injuries that typically occur in young athletes during activity. Some patients may experience significant morbidity, often through chronic pain and changes in sports participation. The influence of fragment displacement on optimal treatment and proper return to sport protocol has not been well investigated. This study sought to evaluate the relationship between displacement and both choice of treatment and time to return to sport.

Methods: Retrospective analysis of ischial tuberosity avulsion fractures in pediatric patients between 2010 and 2021 at Boston Children’s Hospital was completed. Information regarding demographics, treatment, and recovery were extracted from patient charts. Plain radiographs and magnetic resonance imaging (MRI) were analyzed to measure fragment displacement at the ischial apophysis. Treatment protocol, return to sport (RTS) time, and complications were analyzed.

Results: Eighty-eight patients with an ischial tuberosity avulsion fracture were identified. Sixty-four (73%) were treated non-operatively, 8 (9%) were treated acute operatively (within 6 weeks), and 16 (18%) were treated chronic operatively (>6 weeks). Imaging for measurement of fragment displacement was available for 86 (98%) patients. The median displacement was 0.2 cm for the non-operative group, 2.2 cm for the acute operative group, and 1.5 cm for the chronic operative group. Of the 86 patients with available imaging, 67 (78%) had an RTS time. The median RTS was 3.6 months in the non-operative group, 5.5 months in the acute operative group, and 13.3 months in the chronic operative group.

Conclusions: In this retrospective cohort study, nearly all fractures with displacement ≥1.0 cm were surgically treated, either acutely or at a delayed time point because of failure of non-operative treatment. All patients who underwent acute operation had displacement >1.0 cm, while 87% of those undergoing delayed operation had displacements >1.0 cm. The vast majority of those treated non-operatively had displacement <1 cm, with only three patients in the series having successful return to sport despite displacement >1 cm. Overall, non-operatively treated patients returned to sport approximately 1 month earlier than those undergoing acute operation and several months prior to those with delayed surgical intervention. These data suggest that patients with displacement <1.0 cm may be successfully treated non-operatively while greater displacement most commonly underwent surgical to ultimately optimize outcome and return to sport.

Significance: This study suggests that there may be an approximate threshold measure of displacement at which ischial tuberosity avulsion fractures may be considered for operative treatment to result in the most successful outcomes.

Pre-existing femoro-acetabular impingement is associated with pelvic avulsion fractures in adolescents: a matched cohort study

Miles Batty, Samantha L. Ferraro, Munif Hatem, Patricia E. Miller, Benton E. Heyworth, Sarah D. Bixby, Eduardo Novais

Boston Children’s Hospital, Boston, MA, USA

LOE-Not Applicable-Level III

Purpose: Apophyseal pelvic avulsion fractures occur almost exclusively in the adolescent athlete population. A possible relationship between femoro-acetabular impingement (FAI) and pelvic avulsion fractures has been discussed in the literature, but no prior studies have investigated the hip morphology of patients with pelvic avulsion fractures. The purpose of this study is to use computed tomography (CT) imaging to compare morphologic femoral and acetabular measurements of adolescents who sustained a pelvic avulsion fracture to those of gender- and age-matched controls. We hypothesize that a higher incidence of radiographic FAI will be found in patients with pelvic avulsion fractures than in the matched control patients.

Methods: Cases of pelvic avulsion fractures were retrospectively reviewed from the electronic medical records at a tertiary care pediatric center. Patients with CT imaging of the hip near the time of injury were included for analysis. The avulsion fracture cohort was matched in a 1:2 ratio to non-fractured hips with CT imaging based on age and sex. Cranial acetabular version, central acetabular version, lateral center edge angle, and alpha angles—anterior, anterior-superior, superior, posterior-superior, posterior—were measured for all subjects. Radiographic measurements were compared between the avulsion cohort and matched controls using independent sample Wilcoxon rank-sum tests. A logistic regression model estimated the effect of the presence of avulsion fracture on the likelihood of CAM impingement (any alpha angle >55°).

Results: Thirty-four hips with avulsion fractures were matched to 68 hips without avulsion fractures for analysis. Avulsion fracture subjects had 2.5 times the odds of impingement, compared to controls (odds ratio (OR) = 2.5; p = 0.04). The anterior-superior alpha angle was 5° higher for the avulsion cohort compared to controls (median, 58° versus 53°; p = 0.03). Additional findings include a 2° lower superior alpha angle (median, 40° versus 42°; p = 0.03) and 3° higher central acetabular version (median, 17° versus 14°; p = 0.03) for the avulsion group compared to controls.

Conclusions: Our findings demonstrate that patients with an apophyseal pelvic avulsion fracture are significantly more likely to show radiographic evidence of FAI when compared to matched controls. These data may represent a morphologic predisposition to avulsion fractures among adolescents with pre-existing FAI.

Significance: This is the first study to identify pre-existing morphologic features of the hip that are associated with adolescent pelvic avulsion fractures. In addition, this adds to the list of injuries that have been shown in the literature to be associated with FAI.

An emerging healthcare crisis: trends in pediatric firearm injuries over time—analysis of over 1100 cases

Claire Sentilles, Elizabeth Lane Whitman, Keith Jayson Orland, Abu Mohd. Naser, Jonathan Rowland, Jeffrey R. Sawyer, Benjamin West Sheffer, David D. Spence, William C. Warner, Derek M. Kelly

Campbell Clinic, Germantown, TN, USA

LOE-Not Applicable-Level III

Purpose: Behind motor vehicle accidents, gunshot injuries are the leading cause of injury-related death in the pediatric population in the United States. The goal of this study was to analyze trends among children with firearm injuries to better understand if certain age groups, mechanisms of injury, and type of firearms were driving the increasing rates of pediatric gun-related trauma.

Methods: This study includes pediatric gunshot wound patients who presented to a single free-standing level 1 pediatric hospital or died at the scene from 2010 to 2021. All the hospital data were collected prospectively as part of a level 1 trauma system database. All the death data were collected from the local coroner’s office. Patients were grouped according to age (0–1 years, 1–4 years, 5–9 years, 10–14 years, 15+ years), mechanism of injury (suicide, assault, accident, unknown), and weapon type (handgun, air gun, BB gun, shotgun, hunting rifle, unknown, other).

Results: From 2010 to 2021, there were 1126 pediatric gunshot injuries in the geographic region studied. The incidence of firearm trauma in pediatric patients has increased rapidly over the time studied. Rise in injury rates occurred most strikingly in the older adolescent age groups 3 and 4 (Figure 1a), in the category of assault (Figure 1b), and among handguns (Figure 1c). Younger age groups, suicides, accidents, and injury by hunting rifle remained relatively stable over time.

Conclusions: Firearm injuries are the second leading cause of injury-related death in the pediatric population in the United States, and they are increasing in frequency year-to-year at an alarming rate. These findings demonstrate that overall rates of firearm injuries in the pediatric and adolescent populations have increased over the period of interest with most of the increase occurring in older children from assault with handguns. These data can hopefully help inform civic leaders, law enforcement officials, and legislators to design strategies to combat these areas of concern.

Significance: Analysis of specific groups, types of firearms, and mechanism of injuries will hopefully allow for risk stratification and future interventions. The findings of this study will hopefully provide insight on how best to initiate change and where to focus prevention strategies.

Loss of reduction in pediatric distal radius fractures: risk factors from a prospective multicenter registry

Apurva S. Shah, Zoe Elizabeth Belardo, Mark Leland Miller, Michael Willey, Susan T. Mahan, Divya Talwar, Rebecca Aguiar, Sana Bouajaj, Aspen Miller, Joshua Marino, Donald S. Bae, Pediatric Distal Radius Fracture Registry

Children’s Hospital of Philadelphia, Philadelphia, PA, USA

LOE-Therapeutic-Level II

Purpose: The purpose of this study was to report risk factors for loss of reduction of pediatric distal radius fractures from a large prospective multicenter cohort. We hypothesized that fractures with increased translation and angulation at the time of injury would be more likely to lose reduction.

Methods: Children aged 4–18 years old undergoing closed reduction for a displaced distal radius fracture and enrolled in the Pediatric Distal Radius Fracture (PDRF) Registry at four academic children’s hospitals were included. Children with open fractures, torus fractures, incomplete radiographs, or less than 4 weeks of follow-up were excluded. Radiographic loss of reduction (LOR) was defined as a change in angulation ≥10 degrees or an increase in translation greater than 50% of the radial width in any plane. Demographic, clinical, and radiographic variables were analyzed; bivariate analysis and multivariate logistic regressions were performed.

Results: In total, 616 children (69% male) were eligible with a mean age of 10.4 ± 3.4 years. Most subjects sustained bicortical distal radius fractures (68%), followed by Salter-Harris II physeal fractures (26%). The rate of radiographic LOR was 44% (271/616). LOR was more likely in patients <11 years old (54% versus 34%), in metaphyseal bicortical fractures compared to physeal fractures (53% versus 24%), when initial fracture translation exceeded 100%, and in patients with concomitant ulnar fracture (48% versus 27%) (all p < 0.001; Table 1). On multivariate analysis, metaphyseal bicortical radial fractures (odds ratio (OR) = 2.57), ≥51% translation on injury films (OR = 3.68), and non-anatomic closed reductions (OR = 3.00) significantly increased the odds of LOR. Ultimately, 47/616 children (8%) underwent secondary procedures, including repeat closed reduction or operative treatment. Children with self-identified White race (p < 0.001), increased pre-reduction translation (p = 0.002), and pre-reduction angulation (p = 0.013) were more likely to undergo a secondary procedure (Table 1). On multivariate analysis, non-anatomic closed reductions increased odds of secondary procedures (OR = 3.13), while self-identified Black race decreased these odds (OR = 0.16).

Conclusions: Bicortical metaphyseal fractures, ≥51% radial translation on pre-reduction films, and concomitant metaphyseal ulna fracture independently increase the odds of LOR for pediatric distal radius fractures. Patient body mass index (BMI) does not appear to influence LOR. Future investigations are warranted to study the relationships between LOR, secondary procedures, and patient race.

Significance: Orthopedic providers are advised that radiographic loss of reduction of ≥10 degrees is not uncommon for pediatric distal radius fractures. Bicortical fractures with ≥51% shaft width translation and concomitant ulna fracture should be closely monitored in clinic for LOR.

Modified elastic stable intramedullary nailing, a new approach for distal metaphyso-diaphyseal junction of forearm in children

Elie Georges Saliba, Pauline Savidan, Clement Munoz, Yan Lefèvre

Hopital Des Enfants CHU de Bordeaux, Bordeaux, France

LOE-Therapeutic-Level II

Purpose: Elastic stable intramedullary nailing (ESIN) is the treatment of choice for diaphyseal forearm fractures in children. However, when the fracture is located at the distal metaphyso-diaphyseal junction (DMDJ), the radial nail has only limited support on the distal fragment. This study aims to evaluate a modified ESIN method to treat these fractures.

Methods: The technique is initially carried out identically to a classic ESIN. Once its ascent is complete, the radial nail is moved back 4 cm and angled at 90° in its extra-osseous portion. The nail is then pushed back so that the entire angulated part penetrates the distal radius, increasing the pressure on the distal medial cortex. This has the effect of stabilizing the fracture in a perfectly aligned position. Between 2010 and 2015, 27 patients were included. Surgical treatment was accepted for fractures with an angulation greater than 20° and instability after reduction.

Results: Modified ESIN achieved a good and stable reduction of the fracture, with no displacement until late follow-up at 6 months.

Conclusions: This study confirms the feasibility and effectiveness of the modified ESIN technique as a treatment for these fractures.

Significance: Modified ESIN offers a new and reliable approach for DMDJ.

Take it or leave it: prevalence and complications from hardware removal following pediatric fracture fixation

Vineet Desai, Scott J. Mahon, Lucas Hauth, Amanda Pang, Apurva S. Shah, Jason Anari

The Children’s Hospital of Philadelphia, Philadelphia, PA, USA

LOE-Therapeutic-Level IV

Purpose: Hardware removal (HR) is one of the most common surgical procedures in pediatric orthopedics. Surgeons advocate for HR for a variety of reasons, including to limit peri-implant fracture risk, native anatomy for adult reconstruction surgery, and to mitigate implant-related pain/irritation. To our knowledge, no recent study has investigated the characteristics and complications of HR in pediatric orthopedics. The goal of this study is to report the prevalence and complications of HRs across all pediatric fractures.

Methods: A retrospective case series was conducted of all HRs from 2014 to 2022 for patients who suffered a fracture requiring surgical intervention at a single urban tertiary care children’s hospital. Cases were identified using CPT codes/billing records. Spinal hardware and elective deformity correction hardware were excluded. Patient demographic and clinical data were recorded. For patients with multiple HRs, each case was recorded independently.

Results: Five hundred and fourteen HR cases for 478 patients met the study criteria. The median post-operative follow-up time was 1.58 weeks (interquartile range (IQR): 0.56–5.35). In total, 346 (67.3%) of the cohort were male. The mean age at HR was 12.05 (SD: 3.91), and the median BMI was 18.33 (IQR: 15.4–21.8). Pain (36.5%), surgeon preference/planned removal (36.1%), and parent/patient preference (14.6%) were the most common indications for HR. 85.4% of patients who underwent HR for pain experienced pain relief post-operatively. 50.4% cases were upper extremity while 49.6% were lower extremity. Flexible intramedullary nails (39.8%), single screws (39.5%), and compression plates with screws (15.2%) were the most common types of hardware removed. Fifty-seven (11.0%) of HRs had a complication. Seven (1.4%) cases had a surgical site infection. Seven (1.4%) cases had new, persistent pain after HR. Six (1.2%) cases had re-fracture (five forearm/wrist, one femur/knee). Of the forearm/wrist re-fracture cases, two occurred at prior fracture site, and three occurred adjacent to flexible intramedullary nail location. Six (1.2%) cases had incomplete HR (two syndesmosis screws, one distal tibia screw, one T-condylar screw, one supracondylar screw, one ulna flexible intramedullary nail). Four (0.8%) patients had hardware breakage while four (0.8%) had nerve palsy/injury (one each for radial, median, ulnar, and plantar nerves). Two (0.4%) patients suffered from a seroma/hematoma and one (0.2%) required re-operation.

Conclusions: HR following fracture fixation in pediatric orthopedics is associated with an overall complication rate of 11.0%.

Significance: Removal of hardware following semi-elective pediatric fracture surgery is not benign, and the 11% complication rate should be a part of the informed consent process.

Use of the bioabsorbable Activa IM-Nail™ in pediatric diaphyseal forearm fractures: a prospective cohort study with at least 1-year follow-up*

David Goettsche, Morten Jon Andersen

Copenhagen University Hospital—Herlev and Gentofte, Copenhagen Denmark

LOE-Therapeutic-Level IV

Purpose: Pediatric diaphyseal forearm fractures (PDFF) are very common injuries. Fractures needing surgery are most often treated with metal elastic stable intramedullary nails (ESIN). Metal nail removal is widely advocated; however, it is a burden on the child, the family, and healthcare economy. Bioabsorbable intramedullary nails (BIN) made from poly lactic-co-glycolic acid (PLGA) have been developed for some of the same indications as metal ESIN. The aim of this study was to evaluate the feasibility and safety of BIN (Activa IM-Nail™, Bioretec, Tampere, Finland) for PDFF.

Methods: From 1 May 2021 to 30 September 2022, children with unstable PDFF requiring surgery were sought to be prospectively and consecutively recruited. Surgical indications were complete fracture and displacement >50% of bone width or angulation >10° in any plane or irreducible or unstable after closed reduction. If consented, patients were operated with the Activa IM-Nail™ followed by 4 weeks of cast. Primary outcome was radiographic healing at 3 months assessed by the Radiographic Union Score (RUS). Secondary outcomes were pain, wound appearance, neurovascular status, bilateral elbow, forearm range of motion (ROM), fracture angulation and displacement, and any adverse device effects (ADE).

Results: In total, 27 children were eligible for inclusion and 9 children were excluded (2 declined participation, 6 fractures were reduced and casted, 1 had metal ESIN due to narrow intramedullary canal). Eighteen children were operated with BIN, 8 girls and 10 boys, mean age 10 years (4–14 years). Operating time was a mean of 66 (24–115) minutes. RUS was at least 8 at 3 months after surgery, 10 at 6 months, and 12 at 1 year. Three months after surgery, all patients had pain free and normal ROM. There were no serious adverse device effects (SADEs). Two children developed minor postoperative fracture angulation that did not require manipulation. One child sustained a non-displaced radius fracture 5 months after surgery and was treated in a cast. Another child sustained a displaced radius fracture more than 12 months after surgery and was treated with metal ESIN. One child with an open fracture had a pain free non-union and is awaiting further treatment.

Conclusions: The use of BIN showed results similar to metal ESIN. Three months after surgery, all, but one child, had solid healing and all children had pain free and normal ROM.

Significance: The use of BIN for PDFF is feasible, seems very safe, and eliminates the need for further surgery to remove implants.

*Indicates a presentation in which the Food and Drug Administration (FDA) has not cleared the drug and/or medical device for the use described (i.e. the drug or medical device is being discussed for an “off label” use.)

Biodegradable magnesium implants—a game changer in pediatric trauma care

Romy Marek, Tavishi Singh, Nicole Gabriele Sommer, Annelie-Martina Weinberg

Medical University of Graz, Graz, Austria

LOE-Not Applicable-Not Applicable

Purpose: What is needed to adequately treat pediatric fractures? A healthy and degradable implant material without harmful elements to overcome implant-related issues such as hypersensitivities or accumulation of metals in tissues. Another major advantage of degradable materials is the obviation of hardware removal after fracture healing. Hence, implantation through the active physis without affecting longitudinal bone growth, by using resorbable magnesium (Mg)-based implants, could broaden the spectrum of pediatric fracture treatment. Hence, the aim of the study was to investigate the local and systemic response of trans-epiphyseally implanted Mg-based ZX00 screws (Mg synthesized with <0.5 wt% Zn <0.5% Ca; l = 40 mm; d = 3.5 mm) on the active physis in a sheep model.

Methods: Three-month-old female, juvenile sheep (n = 8) underwent monocortical implantation with ZX00 through the epiphysis of the right tibia, whereas the left tibia was either implanted with a titanium (Ti) screw or was left untreated to serve as a control. In order to monitor ZX00 degradation and physis defects, in vivo clinical computed tomography (cCT) was performed at 3, 6, 12, 24, 52, 104, and 156 weeks post-surgery. All animals were sacrificed 180 weeks post-surgery. Tibiae and soft tissues were excised for further analysis. To quantitatively calculate the limb length difference (LLD) and physeal defect area, ex vivo high-resolution micro-computed tomography (µCT; 20 µm per voxel) was performed. Undecalcified tibiae with implants were embedded in Technovit for hard qualitative histology. To examine biocompatibility, organs were collected for soft tissue histology.

Results: First, we observed ZX00 screw breakage between 12 and 24 weeks post-surgery, resulting in ZX00 movement away from the physis, which additionally showed remarkable healing capacity, resulting in the absence of defect and axial deviation at 2 years. However, at 104 and 156 weeks post-surgery, Ti legs were significantly shorter (p < 0.05) when compared to ZX00 and Ctrl legs. In regard to biocompatibility, bone tissue did not show any signs of foreign body reactions or encapsulation. Moreover, qualitative assessment of histologically stained soft tissues did not reveal any harmful effect due to implantation and degradation.

Conclusions: These findings indicate the advantages of Mg-based implants for pediatric internal fracture fixation.

Significance: ZX00 implants are promising due to their biocompatibility and biodegradability, thereby minimizing complications, such as growth discrepancies, foreign body reactions, or encapsulations, making them superior for pediatric interventions.

Biodegradable intramedullary nailing of severely displaced distal pediatric metaphyseal radius fractures

Marcell Benjamin Varga, Gergo Józsa, Zsófia Krupa, Tamás Kassai

Manninger Jenő Baleseti Központ, Budapest, Hungary

LOE-Therapeutic-Level II

Purpose: Severely displaced and shortened distal pediatric forearm fractures may require closed reduction. Many authors recommend osteosynthesis if the fracture remains unstable. The gold standard method is percutaneous pinning with Kirschner wires. K-wire-related complications like migration of the pins, superficial infections, and skin irritation are relatively frequent. We present the results of a prospective multicenter study comparing the K-wire technique and a new bioresorbable intramedullary nailing procedure.

Methods: We compared the results of K-wire osteosynthesis (KW groups) and biodegradable nailing (BR groups) of three Level I pediatric trauma centers. We compared the incidence of minor and major complications (superficial skin infection, tendon injury, nerve injury, secondary displacement, deep infection, skin irritation), differences in operative and radiation time, and unplanned medical check-up visits. Inclusion criteria were the clinical diagnosis of an unstable distal radial/forearm metaphyseal fracture with complete displacement, the presence of open growth plates, the child’s age under 14 years, and a minimal follow-up of 1 year.

Results: A total of 160 patients met the criteria indicated above (82 in the BR group, 78 in the KW group). No significant differences in age or right- or left-hand involvement were observed in either group. There was no significant difference between operative time. Radiation time was slightly reduced in the BR group. The number of minor complications was significantly reduced in the BR group. Secondary intervention (implant removal) was not necessary at all in the BR group, while in the KW group the K-wire had to be removed from all children. The number of unplanned medical examinations in the first 6 weeks was significantly lower in the BR group. After the 1-year follow-up, there was no difference in function between the two groups. Growth disturbance was not observed in either group after 1 year.

Conclusions: The biodegradable nailing technique eliminates the need for new implant removal interventions and reduces minor complications and the number of repeated, unplanned medical control examinations.

Significance: The paper is about a new alternative to a very common intervention. In addition to the health benefits, the absence of further intervention can also increase cost-effectiveness.

Can scoliosis-specific exercises be performed with wearing brace in treating adolescent idiopathic scoliosis: an alternative use of scoliosis-specific exercises without sacrificing bracing hours

Charlene Fan, Michael To, Jason Cheung, Kenneth M.C. Cheung

The University of Hong Kong—Shenzhen Hospital, Shenzhen, People’s Republic of China

LOE-Therapeutic-Level I

Purpose: Bracing is the standardized care in treating progressive adolescent idiopathic scoliosis (AIS). Scoliosis-specific exercises (SSE), as an add-on to bracing treatment, show the promising effects in preventing curve progression. However, the general SSE is performed while patients take off the brace, which sacrifices the bracing hour that is a vital influencing factor of bracing success. Objective is to compare the difference between the SSE with wearing brace (SSE with brace-on), SSE without wearing brace (SSE with brace-off), and brace alone in preventing curve progression of AIS.

Methods: This was an assessor- and patient-blinded, randomized control trial with 4-year follow-up. One hundred and twenty patients undertaking Cheneau bracing treatment participated and randomly allocated into three groups: SSE with brace-on (group 1), SSE with brace-off (group 2), and brace-alone (group 3). SSE protocol was standardized with seven exercises with and without wearing the brace, 1.5 h/day for 5 days a week for patients in groups 1 and 2. Initial curve magnitude (Cobb angle at baseline before bracing treatment), in-brace correction, and the follow-up outcomes (Cobb angle without brace at every 6 months till skeletal maturity) were assessed by two surgeons blinded to this study. Exercise compliance and bracing compliance were monitored by tele questionnaire bi-weekly. Scoliosis Research Society 22-item questionnaire (SRS-22) was studied in addition to the changes of Cobb angle. Intention-to-treat (ITT) and per-protocol (PP) analyses were performed with mixed model analysis of variance (ANOVA).

Results: Patients showed no differences of initial curve magnitude, curve pattern, in-brace correction, and demographic characteristics between groups at baseline. Eighteen patients dropped out of this study; thus, outcomes of 102 patients were analyzed. The average bracing hour was highest in group 3 (group 1: 18 ± 2.2 h/d; group 2: 17 ± 3.8 h/d; group 3: 19 ± 0.6 h/d, p = 0.02). ITT analysis showed that the Cobb angle of the major curve, after brace weaned, was comparable between group 1 and group 2 (29° ± 5.4° versus 30° ± 3.6°, p > 0.05), yet was higher in group 3 (36° ± 7.8°, p < 0.01). PP analysis revealed that the highest reduction of Cobb angle was observed in group 2 (group 1: 27° ± 4.2°; group 2: 25° ± 5.3°; group 3: 33° ± 7.2°, p < 0.01), but the differences between group 1 and group 2 were within measurement error.

Conclusions: Bracing with SSE was better than bracing alone in treating progressive AIS. SSE performed with wearing brace was comparable to SSE performed without wearing brace in preventing curve progression.

Significance: Bracing with SSE shows promising effects in treating progressive AIS.

Surgeon contoured versus pre-contoured patient-specific rods in adolescent idiopathic scoliosis: assessing global sagittal alignment

Sahir Jabbouri, Peter Joo, Wyatt David, Seongho Jeong, Jay Moran, Anshu Jonnalagadda, Dominick A. Tuason

Yale School of Medicine, New Haven, CT, USA

LOE-Therapeutic-Level III

Purpose: Surgery for adolescent idiopathic scoliosis (AIS) typically involves posterior spinal fusion (PSF) using rods contoured by the surgeon, which may be time-consuming and may not reliably restore optimal coronal and sagittal alignment. With the use of artificial intelligence (AI) and preoperative planning software to develop pre-contoured patient-specific rods, restoration of more optimal spinal alignment may be achieved. However, the literature lacks studies with a comparative group investigating outcomes of pre-contoured rods in AIS surgery.

Methods: This is a retrospective review of prospectively collected data of two consecutive groups of AIS patients who underwent PSF with surgeon contoured or patient-specific rods. Demographics, Lenke classification, fused levels, osteotomies, estimated blood loss (EBL), surgical time, and last follow-up Scoliosis Research Society (SRS)-22 scores were obtained via chart review. Radiographic measurements including coronal Cobb angle, T5-T12 thoracic kyphosis (TK), lumbar lordosis (LL), pelvic incidence (PI), PI-LL mismatch, and T1 pelvic angle (TPA) were obtained pre-operatively, postoperatively, and at last follow-up. Patients were excluded if follow-up was less than 6 months. Adaptive spine intelligence software was utilized to predict post-operative sagittal alignment which was compared with observed measurements. Student’s and paired t-tests were performed to determine significant mean differences for continuous variables and chi-square for categorical variables.

Results: Thirty-six patients in the surgeon contoured cohort (average last follow-up = 11.3 months) and 22 in the pre-contoured cohort (average last follow-up = 9.7 months) were included. No significant differences were noted with regard to demographics, Lenke classification, preoperative radiographic measurements, number of fused levels and osteotomies, EBL, surgical time, and overall total SRS-22 scores at last follow-up. TK was between 20 and 40 degrees in 95.5% of patients with pre-contoured rods versus 61.1% of patients with surgeon contoured rods at last follow-up (p = 0.004). PI-LL mismatch was within 10 degrees postoperatively in 72.7% of patients in the pre-contoured cohort versus 33.3% of patients in the surgeon contoured cohort (p = 0.004). Other radiographic measurements were similar. Average differences between AI predicted and observed values for the PC group were 3.7 for TK (p = 0.005), 8.7 for LL (p < 0.001), −7.6 for PI-LL mismatch (p = 0.002), and −2.6 for TPA (p = 0.112).

Conclusions: AI and pre-contoured rods help achieve global sagittal balance with excellent accuracy and notably improved kyphosis restoration and PI-LL mismatch than surgeon contoured rods in AIS patients.

Significance: Patient-specific pre-contoured rods help achieve more optimal global sagittal balance than surgeon contoured rods in AIS.

Pulmonary function at minimum 10 years after segmental pedicle screw instrumentation for thoracic adolescent idiopathic scoliosis

Linda Helenius, Matti Mikael Ahonen, Johanna Syvänen, Ilkka J. Helenius

Helsinki University Hospital, Helsinki, Finland

LOE-Therapeutic-Level II

Purpose: Adolescent idiopathic scoliosis (AIS) with thoracic curves is associated with reduced pulmonary function preoperatively. It remains unclear how much pulmonary improvement can be obtained using pedicle screw instrumentation at long-term follow-up.

Methods: Out of 64 consecutively surgically treated patients with thoracic AIS (Lenke 1-4, 6) using pedicle screw instrumentation, 50 (mean age at surgery 14.7 ± 1.9 years, 44 females) participated in a prospective 10-year follow-up study (mean FU, 12.2 years). Preoperative major curve averaged 57° (SD, 8.5°) with a remaining curve of 15° (SD, 6.3°) at 10-year follow-up. They were evaluated using clinical examination, spinal radiographs, and spirometry preoperatively and at 10-year follow-up. The preoperative percentage predicted values were adjusted for the height loss caused by the scoliosis according to curve size.

Results: Preoperatively, 49% (20/41) had forced vital capacity (FVC) or forced expiratory volume in one second (FEV1) below 80% of the predicted normal values representing pulmonary function impairment. FVC improved from preoperative 3.29 L (SD, 0.78 L) to 3.87 L (0.79 L) at 10-year follow-up (p < 0.001). This improvement averaged 510 mL (SD, 560 mL) in patients having both preoperative and 10-year follow-up measurements available. The percentage predicted values for FVC showed an improvement from 83% preoperatively to 86% at 10-year follow-up (p = 0.048). At 10-year follow-up, 38% (19/50) of the patients had FVC or FEV1 below 80% of the predicted values.

Conclusions: Lung volumes improved by a mean of 510 mL from preoperative to 10-year follow-up in patients undergoing pedicle screw instrumentation for thoracic AIS. Despite improvement of absolute values, one-third of the patients fulfilled the criteria for pulmonary function impairment at 10-year follow-up.

Significance: Preoperatively, 49% of the AIS patients with thoracic curves showed impairment in pulmonary function. After surgical correction of the scoliosis and a 10-year follow-up, 38% still fulfill the criteria for restrictive lung disease.

Osteotomies at the time of graduation surgery: how much do we get from them?

Tyler A. Tetreault, Tiffany Phan, Tishya A.L. Wren, Michael J. Heffernan, John B. Emans, Lawrence I. Karlin, Amer F. Samdani, Michael G. Vitale, Ilkka J. Helenius, Lindsay Andras, Pediatric Spine Study Group

Children’s Hospital Los Angeles, Los Angeles, CA, USA

LOE-Therapeutic-Level III

Purpose: Following growing instrumentation for early-onset scoliosis (EOS), patients often undergo PCO at the time of conversion to definitive fusion. Their efficacy may be limited following prior instrumentation. Our purpose was to determine whether PCO at the time of conversion to definitive fusion in EOS graduates impacts outcomes.

Methods: Patients from a multicenter database undergoing growing rod instrumentation conversion to definitive fusion were grouped by those that did or did not have PCO. Patients with inadequate radiographs, <2 years of follow-up, or three-column osteotomies at the time of fusion were excluded.

Results: A total of 839 EOS patients who underwent conversion to definitive fusion met the inclusion criteria. In total, 176 (21%) patients had PCOs and 663 (79%) did not. The age at index surgery was younger (6.6 versus 7.4 years, p = 0.0009) and the mean duration of growing instrumentation was greater (6.1 versus 5.5 years, p = 0.009) in the PCO group. Prior to fusion, curve magnitude was similar between the groups (PCO = 61.9 degrees, no PCO = 59.3 degrees, p = 0.18). For the PCO group, on average, 4.4 osteotomies (range: 1–12) were performed. EBL (PCO = 821 cc versus no PCO = 752 cc, p < 0.01) and surgical time (PCO = 403 min versus no PCO = 348 min, p < 0.01) were greater in the PCO group. Postoperatively, mean major curve magnitude and curve correction were similar between the two groups (mean correction PCO = 16.5, no PCO = 14.4 degrees, p = 0.19). However, accounting for preoperative curve magnitude, there was a relationship between number of PCOs and curve correction (p = 0.04). Mean correction was 14.2 degrees with 1–4 osteotomies, 19.0 degrees with 5–8 osteotomies, and 23.9 degrees with 9–12 osteotomies. Overall, mean correction per osteotomy was 4.5 degrees. There was no relationship between degrees of correction per osteotomy and duration of instrumentation prior to conversion (p = 0.12). Postoperative complications at 2 years were similar between groups (PCO = 25% versus no PCO = 27%, p = 0.63).

Conclusions: EOS graduates achieved minimal correction at the time of conversion regardless of whether PCOs were performed. PCOs were associated with increased EBL and operative time, but similar complication rate. More PCOs resulted in more correction, albeit <10 degrees/osteotomy anticipated in a previously uninstrumented spine.

Significance: PCOs at the time of conversion to fusion in growing rod graduates resulted in less than half of the correction reported in previously uninstrumented spines.

Is bracing after completion of Mehta casting worthwhile?

Tiffany Thompson, Carlos Monroig-Rivera, Mike (M) O’Sullivan, Charles E. Johnston

Scottish Rite for Children, Dallas, TX, USA

LOE-Therapeutic-Not Applicable

Purpose: Following Mehta cast treatment for idiopathic early-onset scoliosis (IEOS), there is no consensus on the value of continued orthotic management, specifically regarding preventing progression. We wished to compare outcomes for patients treated with a brace following cast discontinuation compared to those who were simply observed.

Methods: From a single institution database of IEOS patients treated by casting, we identified 73 patients treated initially by Mehta casts. At completion of the casting program, patients were either observed or prescribed a thoracolumbar sacral orthosis (TLSO) to continue non-operative treatment, at the discretion of the treating physician. We recorded patient demographics, cast details and length of treatment, and curve magnitudes at time points pre-cast, at cast discontinuance, and at last follow-up or until another form of treatment was initiated. We hypothesized that patients doing braces post-casting would achieve/maintain better curve control long term. Bracing was defined “successful” if no other treatment was required.

Results: Seventy-three patients with mean age of 22.2 months (7–62 months) at cast initiation were studied. Five patients underwent surgery after cast discontinuance. Twelve patients with initial mean curve of 36° (17°–70°) were corrected to 16.1° (1°–41°) after final cast and then observed. Fifty-six patients with mean curve of 46° (21°–89°) at initiation were corrected to 28° (1°–79°) at cast discontinuance and then prescribed a TLSO which was worn for a mean of 5.8 years (0–15 years). At last follow-up, 44 braced cases were considered “success” (79%) with mean curve of 19.4°, while 12 braced cases failed with curves measuring 48°, and further treatment (11 surgery, 1 re-cast) needed. Braces were discontinued in 25 patients with 16 remaining stable (no progression, mean curve 13.5°) and 9 progressing 15°, while 14/19 continuing bracing remained stable. One of 12 observed patients compared to 11/56 braced patients required surgery.

Conclusions: Bracing following cast treatment for IEOS was nearly 80% successful in avoiding further treatment even though curves at cast discontinuance were larger (28 versus 16) compared to observed cases.

Significance: When cast treatment does not achieve correction <20°, continuing treatment with a brace appears to have benefit in avoiding surgery.

Limited fusion for congenital scoliosis: is it truly one and done?

Brandon Yoshida, Tyler A. Tetreault, Luke Christian Drake, Tiffany Phan, Jacquelyn Nicole Valenzuela-Moss, Tishya A. L. Wren, Lindsay Andras, Michael J. Heffernan

Children’s Hospital Los Angeles, Los Angeles, CA, USA

LOE-Therapeutic-Level III

Purpose: Limited spinal fusion is a common surgical strategy in the treatment of congenital scoliosis. Due to the heterogeneity of deformity and treatment strategies, long-term outcomes, including the need for additional surgery, are poorly understood. The purpose of the study was to understand the rate and risk factors for reoperation following limited fusion for congenital scoliosis.

Methods: A single-center retrospective review of congenital scoliosis patients who were initially treated with a limited instrumented fusion was conducted. Curve progression, short and long-term complications, and need for additional spinal surgery were assessed. Risk factors for additional surgery were compared between patients who underwent additional surgery and patients who did not require additional surgery.

Results: Thirty-six patients underwent limited instrumented fusion at an average age of 5 years (SD: 2.5 years). The mean number of levels fused was 3.5 (SD: 1.3). Index surgery location included cervical (6%), thoracic (64%), thoracolumbar (50%), lumbar (42%), and lumbosacral (11%). Seven (19.4%) of the patients had two areas of fusion (separated by unfused segments) treated at the index procedure. Average follow-up was 7.6 years (range: 2.2–14.1, SD: 3.4). Average curve magnitude improved from 50° to 26°, with a mean improvement of 18° (range: 11.6°–23.9°, SD: 18°, p < 0.0001). Nineteen patients (53%) required reoperation at a mean of 5.2 years. The mean number of additional surgeries was 3.6 (SD: 3.4). Preoperative curve magnitude was greater in patients who had reoperation compared to patients who did not require reoperation (62 ± 4.9 versus 38 ± 4.7, p = 0.002). Age at index surgery, medical comorbidities, if all congenital anomalies were addressed during index procedure, number of levels fused, location of the index procedure, and postoperative curve magnitude did not differ between groups (Table 1).

Conclusions: Limited fusion for congenital scoliosis resulted in a 53% reoperation rate at a mean of 5.2 years. Preoperative curve magnitude was associated with reoperation, while age, location, number of levels fused, and addressing all congenital anomalies at the index procedure did not differ between groups. These results suggest that surgeons should consider intervention at smaller curve magnitudes in patients with congenital scoliosis and counsel families that a one and done procedure is less likely with larger curves.

Significance: The results of this study suggest that limited fusion should be considered prior to significant curve progression in patients with congenital scoliosis to minimize the need for additional surgery.

Greater implant density does not improve pelvic obliquity and major curve correction in neuromuscular scoliosis

Patrick Thornley, Arlene R. Maheu, Kenneth Rogers, Paul D. Sponseller, Peter O. Newton, A. Noelle Larson, Joshua Pahys, Peter G. Gabos, M. Wade Shrader, Tracey P. Bastrom, Suken A. Shah, Harms Study Group

Nemours Children’s Health, Wilmington, DE, USA

LOE-Therapeutic-Level II

Purpose: Neuromuscular scoliosis (NMS) is the second most prevalent pediatric scoliosis etiology. The NMS population present unique challenges with larger curves and associated pelvic obliquity (PO), osteopenia, and greater comorbidities. The influence of implant density (ID) has undergone much investigation in the adolescent idiopathic scoliosis literature. These investigations have shown no significant difference in correction between low and high ID cohorts. It is however unknown how ID affects initial and long-term correction in NMS. We hypothesized that greater ID in posterior instrumented fusions (PIF) to the pelvis in NMS patients will achieve better coronal and sagittal correction with long-term durability.

Methods: A retrospective review of a prospectively collected multicenter NMS registry database was performed. All NMS patients undergoing upper thoracic to pelvis PIF with a minimum follow-up of 2 years were included. ID was defined as the number of screws per level fused. Patients were divided into three groups: low density (LD; ≤1.3 screws/level), medium density (MD; ≥1.3 and ≤1.6 screws/level), and high density (HD; ≥1.6 screws/level). Demographic and radiographic data were collected to perform independent sample t-tests and cluster analysis for correlation.

Results: A total of 235 patients (56 LD, 84 MD, 95 HD) were included. Baseline demographic characteristics were similar between cohorts including CPChild scores. There were mild preoperative major Cobb angle differences with no difference in PO (Table). Perioperative data was comparable between groups, with significantly increased operative times with higher ID (p < 0.001) (Table). Two-year postoperative coronal plane and PO corrections demonstrated no between-group difference in correction regardless of ID (Table). Sagittal plane measurements demonstrated no statistically significant difference between groups at all time points. Similarly, no functional outcome differences as measured by CPChild were present at any time point in either group (Table).

Conclusions: Lower ID in NMS demonstrates comparable major curve correction and PO correction to higher ID constructs at 2 years postoperatively with reduced operative time and equivalent functional outcomes.

Significance: Efforts to maximize balanced corrections achieve improved sitting balance, while minimizing operative time, complications, and cost are imperative in NMS management. Lower ID should be considered in neuromuscular scoliosis correction as both an effective surgical strategy and a potential source of significant cost reduction.

The effect of traction and spinal cord morphology on intraoperative neuromonitoring alerts

Evan Fene, Lydia Klinkerman, Charles E. Johnston, Jaysson T. Brooks, Megan Johnson

Scottish Rite for Children, Dallas, TX, USA

LOE-Prognostic-Level II

Purpose: Patients with type 3 spinal cords are at greater risk for intraoperative neuromonitoring (IONM) alerts when undergoing thoracic scoliosis correction. The use of intraoperative traction during deformity correction is also associated with an increased risk of IONM alerts. The purpose of this study is to examine the interplay between spinal cord morphology and intraoperative traction.

Methods: An institutional review board (IRB)-approved retrospective review of patients with major thoracic curves ≥70° who underwent spinal fusion from 2016 to 2022 at a single institution were reviewed. Patients without a preoperative magnetic resonance imaging (MRI) were excluded, yielding 102 patients for review. Spinal cord morphology was determined by consensus of four observers using the criteria by Sielatycki et al.

Results: Seventy-five (73.5%) patients were female, with an average age of 13.7 ± 2 years at surgery. Mean thoracic Cobb was 85° ± 13°. The average number of levels fused was 12 ± 1.5, implant density 1.5 ± 0.3, estimated blood loss (EBL) 895 ± 645 mL, and total surgical time 314 ± 112 min. Eighty-seven patients had posterior spinal fusion (PSF), while 15 had an anterior spinal fusion (ASF)/PSF. Intraoperatively, traction was used in 15 patients. Sixteen patients had type 1 cords, 71 type 2, and 15 type 3. Seventeen patients had an IONM alert and surgery was aborted in three patients due to persistent loss of IONM. Patients with type 3 cords were more likely to have an IONM alert than those with type 2 cords (46.7% versus 14.1%, odds ratio (OR) = 5.3, p = 0.004). One hundred percent of patients with type 3 cords placed in traction intraoperatively had an IONM alerts versus 20% without traction (p = 0.007). Multivariate regression found cord type (OR = 6.3, p = 0.02) and intraoperative traction (OR = 8.9, p = 0.02) to be significant risk factors for an IONM alert.

Conclusions: Type 2 and 3 spinal cords are found in 84% of patients with curves ≥70°. An IONM alert is six times more likely in patients with type 3 cord. Patients with a type 2 or type 3 cord are nine times more likely to have an IONM alert when traction is used intraoperatively and 100% of patients with a type 3 cord and intraoperative traction in this cohort experienced an IONM alert.

Significance: The use of intraoperative traction should be carefully considered in patients with a thoracic curve ≥70°, especially if a preoperative MRI demonstrates type 3 cord morphology.

What kind of kyphosis? Stratifying thoracolumbar kyphosis in achondroplasia

Luiz Carlos Almeida Da Silva, Yusuke Hori, Colleen P. Ditro, Kenneth Rogers, J. Richard Bowen, William G. Mackenzie, Stuart Mackenzie

Nemours Children’s Hospital—Delaware, Wilmington, DE, USA

LOE-Therapeutic-Level IV

Purpose: When treating thoracolumbar kyphosis (TLK) in children with achondroplasia, predicting outcomes, correction, and surgical complications is essential. There is limited literature clearly describing the treatment and risk profiles among different types of TLK. This study aimed to classify deformities of the thoracolumbar junction, describe effective operative treatment, and compare surgical complications.

Methods: A single-center retrospective review at a quaternary referral center was performed from 2004 to 2021. Inclusion criteria consisted of achondroplasia diagnosis, TLK, symptomatic spinal stenosis, surgical treatment, and at least 1 year of follow-up. Patient demographics, clinical notes, and radiographs were analyzed. Post-operative complications were classified by Clavien-Dindo System (CDS). Intra-observer reliability was measured with Cohen’s kappa coefficient, and interobserver reliability with Fleiss’ kappa coefficient.

Results: Forty patients with achondroplasia and TLK requiring surgical spinal decompression and fusion were treated at an average age of 16 ± 5 years, with a mean follow-up of 5 ± 4 years. We identified four classes of patients: Class 1 involves gradual multi-level TLK with diffuse stenosis; Class 2 involves severe focal TLK associated with apical vertebral body wedging and localized stenosis, with a flattened or lordotic thoracic spine above; Class 3 involves severe post-laminectomy kyphosis; and Class 4 is a heterogeneous group with additional diagnoses related to TLK. The intra-observer kappa index of this classification was 0.90 (p < 0.001), and the inter-observer was 0.83 (p < 0.001). The Class 1 patients were treated with posterior spinal decompression and fusion (PSDF), while all others required a combination of PSDF, anterior spinal fusion, and posterior spinal osteotomies. Kyphosis correction rate averaged 81%, with 52.5% (21/40) rate of intraoperative complication. Class 1 patients had shorter surgical times and lowered estimated blood losses (Table 1). Classes 3 and 4 presented more major complications when compared to classes 1 and 2 (p = 0.45).

Conclusions: Treatment of TLK and symptomatic spinal stenosis in patients with achondroplasia can involve complex correction techniques and has a high rate of complication. Surgical options and risk can be stratified based on this reliable TLK classification system. With improved communication and preoperative planning, patients with achondroplasia and TLK can be successfully managed with excellent correction and managed risk.

Significance: This novel classification of TLK in achondroplasia has high inter-observer reliability and can help surgeons better educate and treat these challenging patients.

Distribution of curve flexibility in idiopathic scoliosis—a descriptive study

Simon Blanchard, Matan Malka, Ritt Givens, Michael G. Vitale, Benjamin D. Roye

New York-Presbyterian Morgan Stanley Children’s Hospital, New York, NY, USA

LOE-Prognostic-Level II

Purpose: Curve flexibility is an important variable for adolescent idiopathic scoliosis (AIS) outcomes. This study sought to determine the distribution of flexibility in a multicenter AIS cohort and investigate associated characteristics. It was hypothesized that flexibility would be normally distributed and would correlate with patient characteristics such as body mass index (BMI), curve pattern, and skeletal maturity.

Methods: Surgical AIS patients enrolled in a Harms Study Group (HSG) center were included. Flexibility was measured using lateral bending radiographs. The distribution was graphed using SPSS, and the relationship between flexibility and other parameters was assessed using t and chi-square tests. Pearson’s R was used for regression analysis.

Results: A total of 4574 patients (mean age: 14.5 ± 2.2 years, 80.5% female) met the inclusion criteria. Mean flexibility was 72.0% ± 11.6%. Both primary thoracic (Lenke 1 + 2, N = 2895) and primary lumbar curves (Lenke 5 + 6, N = 1030) followed a normal distribution. Mean thoracic curve flexibility was significantly higher than primary lumbar curve flexibility (71.4% versus 68.98%, p < 0.001). Flexibility was inversely correlated with both higher age (p = 0.002) and increased skeletal maturity (p = 0.003). Patients who have a healthy weight had more flexible curves than those who were overweight (p = 0.019). Interestingly, and unanticipated, better Scoliosis Research Society (SRS)-22 and SRS-24 scores were associated with lower flexibility (p < 0.03, p < 0.01). Not associated with curve flexibility were sex (p = 0.124), weight (p = 0.536), and height (p = 0.121). Surprisingly, thoracic curves were more flexible than lumbar curves (p < 0.001). Increased BMI was associated with decreased flexibility in thoracic curves (p < 0.001), but not with lumbar curves (p = 0.963). In addition, for thoracic curves, those with a lumbar A modifier were significantly more flexible than B (p < 0.001) and C (p < 0.001) types. For sagittal parameters, normo-kyphotic and hypo-kyphotic curves were more flexible than hyperkyphotic curves (p = 0.001).

Conclusions: Flexibility follows a normal distribution in AIS patients, as shown in the attached figure. Older, more skeletally mature patients had less flexibility. Interestingly, thoracic curves were more flexible than lumbar curves, and only thoracic flexibility varied with BMI. In addition, higher SRS scores were associated with less flexible curves.

Significance: This study describes a normal distribution of flexibility in a large population of surgical AIS patients. It also confirms previously described relationships between flexibility and various patient characteristics (age, maturity, BMI) with a significantly larger sample size than previous studies, as well as describing several new factors that correlate with curve flexibility.

When is growth greatest? Spine and total body growth in idiopathic scoliosis through Sanders maturation stages 2, 3A, 3B, and 4

Yusuke Hori, Bryan Menapace, Burak Kaymaz, Luiz Carlos Almeida Da Silva, Norihiro Isogai, Sadettin Ciftci, Kenneth Rogers, Petya Yorgova, Andrea Mary Elsby, Peter G. Gabos, Suken A. Shah

Nemours Children’s Hospital, Wilmington, DE, USA

LOE-Prognostic-Level III

Purpose: The ability to accurately predict growth is crucial in treating the growing spine. The Sanders Maturation Stage (SMS) 2–4 represents periods of rapid growth and are key stages for growth modulation surgery. A detailed assessment of spine growth during these stages is imperative and is lacking. This study aimed to evaluate the spine and total body height growth through SMS 2, 3A, 3B, and 4 and to assess scoliosis progression during this period.

Methods: In this single-center retrospective case-control study, consecutive patients with idiopathic scoliosis staged SMS 2–4 were analyzed from January 2013 to December 2022. T1-S1 spine height, total body height, and curve magnitude were measured at each visit. Spine and total body height velocity and curve progression rate were assessed between the initial and first follow-up visit (6–12 months). For those observed until skeletal maturity (Risser 4+ or 5), height gain and curve progression were evaluated. Considering height loss due to scoliosis, spine and total body height were corrected for curve magnitude using validated formulas. For statistical comparisons between the groups, the Kruskal–Wallis test was used, followed by Bonferroni post hoc analysis.

Results: The study included 525 patients (68% girls, mean age: 12.4 ± 1.4) and 2082 radiographs. Spine height velocity peaked during SMS 3A (2.6 ± 0.8 mm/month), being approximately 1.3 times that of SMS 2 (2.0 ± 0.9 mm/month), 1.5 times that of SMS 3B (1.7 ± 0.6 mm/month), and 1.7 times that of SMS4 (1.5 ± 0.7 mm/month) (Figure 1a). SMS 2 and SMS 3A had comparable total body height velocities (6.3 ± 2.2 mm/month; 6.4 ± 1.7 mm/month, respectively), both surpassing SMS 3B (4.8 ± 2.0 mm/month) and SMS 4 (3.5 ± 1.6 mm/month) (Figure 1b). Curve progression rates were consistent across the subtypes. Among those observed until skeletal maturity, SMS 2 had the highest spine and total body height gains, with declines noted as SMS advanced. Curve progression was greatest in SMS 2 (16° ± 14°), followed by SMS 3A (11° ± 12°), with SMS 3B and SMS 4 being similar (7° ± 9°; 6° ± 9°, respectively) (Figure 2).

Conclusions: This investigation revealed the spine grows fastest during SMS 3A and suggests that leg growth peaks at SMS 2. SMS 2 demonstrated the highest potential for both spine and total body height growth, as well as curve progression. These insights are crucial for determining intervention timing in scoliosis treatment, including decisions on bracing, surgery, and selecting between fusion or growth modulation.

Significance: This study identifies the pivotal SMS where spine and leg growth peak, facilitating more precise timing of treatment in idiopathic scoliosis and potentially guiding more successful outcomes.

The true cost of late referral in adolescent idiopathic scoliosis: a 5-year follow-up study

Emma Nadler, Jennifer Dermott, Dorothy Kim, David E. Lebel

Hospital for Sick Children, Toronto, ON, Canada

LOE-Economic-Level III

Purpose: Bracing moderate curvatures in skeletally immature idiopathic scoliosis patients is known to minimize the risk of curve progression to surgical range. Unfortunately, late referrals are common with more patients presenting as surgical versus ideal brace candidates. This study analyzes the total treatment cost differential between adolescent idiopathic scoliosis (AIS) patients who at initial consultation are ideal brace candidates versus late referrals.

Methods: This is a retrospective review and cost analysis of AIS patients seen for initial consultation in 2014 who were either a late referral or an ideal brace candidate. Late referrals were defined as 50+° curvatures or over 40° and ≤Risser 2. Ideal brace candidates were defined as 25°–40° curves, ≤Risser 2. Patient medical records were reviewed to determine progression in brace candidates to surgical range, up to 5 years after initial visit or time. Total costs assumed all curves 50+° would at some point have surgery. In addition to the cost of a brace or a surgical procedure, total treatment costs considered typical number of clinic visits (including associated travel, parking, meals, and loss of income costs for family) and spine X-rays obtained.

Results: Within this cohort, 63 patients met ideal brace indications (17%) compared to 103 patients considered late referral (28%) (Scenario 1). The average total cost per patient for brace treatment was calculated at $13,459 versus $68,009 for surgical treatment. Within the 5-year study period, 10 ideal brace candidates progressed to 50+° curvatures and 5 were lost to follow-up. In 2014, the theoretical total cost for ideal brace candidates was $1,868,052 versus $700,492 for late referrals (Scenario 1). If the late referrals seen that year were instead seen as ideal brace candidates, the total cost of treatment, including surgical costs for patients expected to progress despite brace treatment, would be $4,954,554, a cost savings of $3,918,373.

Conclusions: The total treatment cost of a late AIS referral is 56% more than the treatment cost of an ideal brace candidate. Early diagnosis and treatment are significantly more cost-effective.

Significance: These findings underscore the financial consequences of late AIS diagnosis, emphasizing the need for early detection to reduce the cost burden. It is expected that the cost differential is underestimated as longer than recommended wait times, if applicable, and the inherent re-operation rate are not considered.

Battle of the braces: a comparison of brace efficacy in patients with adolescent idiopathic scoliosis treated with Providence, Boston-style, and Rigo-Cheneau braces

Leigh Davis, Amy Bridges, Julie Hantak, Hilary Harris, Sofie-Ellen Stroeva, Nikolay Braykov, Afrin Jahan, Nicholas D. Fletcher

Children’s Healthcare of Atlanta, Atlanta, GA, USA

LOE-Therapeutic-Level IV

Purpose: A variety of brace options exist for the management of adolescent idiopathic scoliosis (AIS). While institutional or surgeon’s preferences abound, there is sparse literature comparing brace efficacy.

Methods: A retrospective review of patients aged 10–18 with AIS of 20°–45° who were Risser 0–2 at initiation of treatment treated at a single center using a single orthotics team was performed. Patients were included if they had either completed brace treatment AND had a minimum of 6-month out of brace follow-up OR had undergone surgery. Brace success was defined as curve progression <6 degrees without the need for surgery. Patients were treated with a Providence nighttime brace (PNB), Boston-style TLSO (BB), or Rigo-Cheneau (RC) custom brace. Brace wear was determined by the treating physician.

Results: Patients treated with 32 PNB, 37 BB, and 30 RC braces met all inclusion criteria. Overall, 44 patients (44%) progressed >6° and 18 patients (18 %) progressed to surgery. Eighty-eight percent were female, with an average age at brace initiation of 12.0 years. Sixty-two percent of females were premenarchal at the onset of bracing, 66% were Risser 0, and 34% Risser 1 or 2. Thoracic, thoracolumbar, and lumbar curves averaged 27° ± 7°, 28° ± 6°, and 26° ± 6° at initiation. Thoracic curve correction was 46% ± 34% compared to thoracolumbar curves 73% ± 36%. PB were worn at nighttime only (8–10 h) while BB and RC were worn for part of the day and night. Twenty-seven percent of patients treated with a BB progressed to surgery compared to 13% of RC and 13% of PNB patients (p = 0.21). Among Risser 0 patients, 37% of BB patients progressed to surgery compared to 11% of RC (p = 0.07) and 10% of PNB patients (p = 0.03). Similar trends toward less efficacy in patients with BB were seen using curve magnitude increase >6°; however, these did not meet statistical significance. Multivariate regression analysis found that skeletally immature (R0) patients treated with a BB had a 5.6-fold higher risk of surgery than RC or PNB (p = 0.04) and a 2.8-fold higher risk of curve progression (p = 0.09). This difference was not seen in Risser 1 or 2 patients.

Conclusions: Skeletally immature patients (Risser 0) treated with a Boston brace were nearly six times more likely to require surgery compared to those treated with a Rigo-Chenault or Providence brace. Patients who were Risser 1 or 2 were successfully treated in any of the three braces.

Significance: Rigo-Chenault and Providence bracing were more successful than Boston braces for skeletally immature patients with AIS.

Improvement in axial rotation with bracing reduces risk of curve progression in patients with adolescent idiopathic scoliosis

Michael Fields, Christina Carin Rymond, Matan Malka, Ritt Givens, Matthew E. Simhon, Hiroko Matsumoto, Gerard F. Marciano, Afrain Z. Boby, Benjamin D. Roye, Michael G. Vitale

Children’s Hospital of New York, New York, NY, USA

LOE-Therapeutic-Level II

Purpose: While in-brace coronal plane correction is commonly used as a proxy for brace efficacy, emerging evidence supports the importance of three-dimensional (3D) in-brace correction for adolescent idiopathic scoliosis (AIS) patients. This study investigated the relationship between axial plane parameters and treatment failure in patients at a single center undergoing brace treatment for AIS. We hypothesize that AIS patients with large in-brace axial vertebral rotation (AVR) and/or poor improvement in AVR with bracing would have increased risk of treatment failure.

Methods: AIS patients (Sanders 1–5) undergoing Rigo-Chêneau bracing at a single institution were included. AVR was determined by utilizing pre-brace and in-brace (3D) spinal reconstructions based on biplanar low-dose EOS® radiographs. The primary outcome was treatment failure defined as coronal curve progression >5°. Minimum follow-up was 2 years.

Results: Seventy-five patients (61/75, 81% female) were included in the final cohort. Mean age at bracing initiation was 12.8 ± 1.3 years and patients had a pre-brace major curve of 31.0°°± 6.5°. Twenty-five (33%; six males, 19 females) patients experienced curve progression >5°, and 18/25 required surgical intervention. The treatment failure group had larger in-brace absolute AVR than the success group (5.8°± 4.1° versus 9.9°± 7.6°, p = 0.003), but also larger initial coronal curve measures. The magnitude of in-brace AVR did not appear to be associated with treatment failure after adjusting for pre-brace major curve (hazard ratio (HR): 0.99, 95% confidence interval (CI): 0.94–1.05, p = 0.833). After adjusting for pre-brace major coronal curve, patients with improvement of AVR with bracing had an 85% risk reduction in treatment failure versus those without improvement (HR: 0.15, 95% CI: 0.02–1.13, p = 0.066). At final follow-up, 42/50 (84%) patients who did not progress had a Sanders ≥7.

Conclusions: While absolute in-brace rotation was not an independent predictor of curve progression (due to its correlation with curve magnitude), improved AVR with bracing was a significant predictor of curve progression.

Significance: This study is the first step toward investigating the interplay between 3D parameters, skeletal maturity, compliance, and brace efficacy, setting the stage for a future prospective multi-center study with adequate design and power.

PROMIS-based assessment of brace compliance

Carlos Monroig-Rivera, David C. Thornberg, Chan-Hee Jo, Megan Johnson

Scottish Rite for Children, Dallas, TX, USA

LOE-Not Applicable-Level III

Purpose: Bracing is the primary non-surgical treatment for patients with idiopathic scoliosis (IS). While the TLSO brace remains the gold standard of treatment, Rigo-Cheneau (RC) and Providence (PROV) nighttime braces are being used more often due to a perceived increased level of comfort/ease of wear. The goal of this study was to examine compliance rates and patient-reported outcomes (PROs) for each type of brace.

Methods: Retrospective review of patients treated in a brace for IS at a single institution from 2021 to 2023. Patients with compliance data, pre-treatment PROMIS scores, and post-treatment PROMIS scores were included. Those with prior treatment (casting, bracing) or non-idiopathic etiologies were excluded.

Results: One hundred and twenty-seven patients were identified, of whom 95 (78%) were female. Eighty patients were treated with a TLSO and 47 with a PROV. The average compliance rate was 83% and was higher for PROV (84%) than for TLSO (70%, p = 0.03). Patients who were noncompliant with bracing (<80%) had worse PROMIS anxiety scores than those who were compliant (41 versus 37, p = 0.005) at brace prescription, 6 months after brace initiation (41 versus 36, p = 0.006), 1 year after brace initiation (40 versus 35, p = 0.04), and at their latest follow-up (39 versus 36, p = 0.03). Twenty-two of the 80 TLSO patients were treated with a RC. There was no difference in PROMIS scores between the TLSO and RC groups. There was a difference between compliance rates for PROV, TLSO, and RC patients (83.5% versus 75% versus 55% respectively, p = 0.006). Four patients in the cohort went on to surgery. All four were in the TLSO group and two were noncompliant with bracing. Patients who failed bracing had worse PROMIS anxiety scores than those who were still in their brace at latest follow-up (49.5 versus 38.2, p = 0.02).

Conclusions: Patients treated with nighttime only bracing have better compliance rates, but similar PROMIS scores compared to those treated with TLSO braces. Patients who are noncompliant with brace wear have worse PROMIS anxiety scores at all time points during bracing, including brace prescription. RC braces do not seem to offer any benefit over traditional TLSO braces in terms of PROMIS scores or brace compliance.

Significance: IS patients with abnormal PROMIS anxiety scores are more likely to be non-compliant with bracing. Identification of patients with anxiety prior to the start of bracing may allow for the opportunity to intervene to increase compliance and may also serve as a prognostic indicator for the success of bracing.

Can surgery be proposed to adolescent idiopathic scoliosis patients with structural lumbar curves associated with non-reducible iliolumbar angle?

Laurentiu-Cosmin Focsa, Louise Ponchelet, Mikael Finoco, Anne-Laure Simon, Brice Ilharreborde

Pediatric Orthopedic Department, CHU Robert Debré, Paris, France

LOE-Therapeutic-Level III

Purpose: The iliolumbar angle (ILA, measured between L4 superior endplate and bi-iliac crest line) is an important parameter in Lenke 5 and 6 curves. Assessing its flexibility preoperatively is key to define the surgical strategy and determine the distal level of fusion (LIV). Ending on L4 is always preferable for motion preservation, but an imperfect correction can also lead to distal disk degeneration or coronal imbalance. Reduction rates have recently improved in adolescent idiopathic scoliosis (AIS), due to a greater flexibility than adults and modern instrumentation. The objective of this study was therefore to assess the postoperative outcomes of AIS patients with structural lumbar curves, associated with non-reducible ILA.

Methods: All consecutive AIS patients with Lenke 5 or 6 curves, treated by a posterior fusion ending on L4 between January 2017 and January 2022, were included. A minimum follow-up of 18 months was required. Non-reducible ILA (NR group) was defined as >10° on standing radiographs AND a reducibility <50% on bending films. A control group of reducible ILA (R group) was also analyzed. After independent radiological analysis, postoperative results were classified as follows: excellent if postoperative ILA was <5°, acceptable between 5° and 10° with maintained coronal balance (T1-CSVL < 2 cm), and poor if >10° or if coronal imbalance >2 cm. Scoliosis Research Society (SRS) scores at follow-up were also compared.

Results: A total of 122 patients were enrolled, with 36 patients (30%) included in the NR group. Preoperative ILA was significantly greater in the NR group (20° ± 4° versus 15° ± 6°, p < 0.001). Mean lumbar Cobb angles were significantly reduced postoperatively in both groups, with no significant difference between NR and R groups (p = 0.2). ILA was significantly reduced in both groups after surgery (from 20° ± 4° to 5.7° ± 3° in NR group, p < 0.05, and from 15° ± 6° to 5.1° ± 3° in R group, p < 0.05). No significant difference was found between groups regarding postoperative ILA (p = 0.2). In the NR group, excellent, acceptable, and poor outcomes were achieved in, respectively, 52%, 32%, and 16% of the cases, without significant impact on SRS scores. R group had fewer poor outcomes at follow-up (5%), but the difference did not reach significance. No correlation was found between ILA correction and the demographic or preoperative radiological measures.

Conclusions: Surgery is efficient to restore L4 frontal tilt in AIS patients with main lumbar curves, and a fusion to L4 can still be proposed even if ILA appears non-reducible on preoperative radiographs.

Significance: Level of evidence—III—retrospective cohort study.

Utility of routine postoperative laboratory testing after posterior spinal fusion for adolescent idiopathic scoliosis

David Liu, Alexander Farid, Gabriel S. Linden, Danielle Cook, Craig Munro Birch, Michael T. Hresko, Daniel Hedequist, Grant Douglas Hogue

Boston Children’s Hospital, Boston, MA, USA

LOE-Diagnostic-Level II

Purpose: Perioperative blood loss in surgical treatment of adolescent idiopathic scoliosis (AIS) is a major risk. However, with advancements to blood management strategies (intraoperative cell salvage; tranexamic acid), risk of perioperative transfusion has diminished. Despite low rates of symptomatic acute anemia, routine laboratory testing on post-operative day 1 (POD1) and beyond are still commonplace.

Methods: We performed a retrospective cohort study of consecutive patients aged 11–19 with AIS who underwent posterior spinal fusion at a single institution. Univariable logistic regression was utilized to determine whether factors were associated with hematocrit ≤22% on POD1 or a postoperative transfusion. Firth’s penalized logistic regression was used for any separation in the data. Youden’s index was utilized to determine the optimal point on the receiver-operating characteristic (ROC) curve that maximizes both sensitivity and specificity.

Results: Among 527 patients included in this study, only eight had POD1 hematocrit ≤22%; none underwent transfusion. These patients had lower last intraoperative hematocrit levels compared to patients with POD1 hematocrit >22% (24.1% versus 31.5%, p < 0.001), and there was no difference in preoperative hematocrit levels in this group (38.2% versus 39.8%, p = 0.11). Four patients underwent postoperative transfusion. Both preoperative hematocrit levels (34.0% versus 39.9%, p = 0.001) and last intraoperative hematocrit levels (25.1% versus 31.4%, p = 0.002) were lower compared to patients who did not undergo transfusion (Table 1). Intraoperative hematocrit <26.2%, operative time of more than 35.8 minutes per level fused, or cell salvage <241 cc were significant risk factors for postoperative transfusion (relative risk of 105.6; 28.6; and 9.8, respectively) (Table 2). No other risk factors were associated with postoperative transfusion.

Conclusions: Postoperative transfusion after posterior spinal fusion for pediatric AIS is exceedingly rare. POD1 labs should be considered when last intraoperative hematocrit <26%, operative time per level fused >35 minutes, or cell salvage amount >241 cc. Otherwise, unless symptomatic, patients do not benefit from additional laboratory screening on POD1 and beyond.

Significance: Our results suggest that routine postoperative laboratory testing is unnecessary in patients undergoing surgery for AIS, particularly in the setting of advancements like TXA and cell salvage that have effectively limited the risk of blood loss during these procedures. Given the potential negative psychological effects of repeated blood draws on children, in addition to the risks typically associated with venipuncture, reducing the number of postoperative tests can improve the safety and the overall healthcare experience for both the child and their family.

A comparison of opioid-sparing versus opioid-containing postoperative pain management for idiopathic scoliosis

Michael Schallmo, Kayla Hietpas, Michael Paloski

Carolinas Medical Center/OrthoCarolina, Charlotte, NC, USA

LOE-Therapeutic-Level III

Purpose: Several standardized pain management protocols have been reported following instrumented posterior spinal fusion (PSF) for idiopathic scoliosis, aimed at hastening recovery while lowering opioid consumption. However, entirely opioid-free (OF) protocols have yet to be thoroughly evaluated in the pediatric population. The purpose of this study was to compare an OF pain management pathway with a traditional opioid-containing (OC) pathway in pediatric idiopathic scoliosis patients.

Methods: A database of patients who underwent instrumented PSF by a fellowship-trained pediatric orthopedic surgeon was reviewed retrospectively. Patients 10–20 years of age at the time of surgery who had a diagnosis of idiopathic scoliosis and underwent primary instrumented PSF were included. Patients were assigned to one of two groups, based on date of surgery: OF pathway (surgery between June 2019 and July 2020) or OC pathway (surgery between June 2018 and June 2019). Total morphine milligram equivalents (MME) were recorded, beginning once patients reached the inpatient unit postoperatively.

Results: A total of 93 patients (OF = 37, OC = 56) were included. Groups were similar with respect to age and body mass index (Table 1). Patients in the OF group had significantly more levels fused compared with patients in the OC group (11 versus 9.5, respectively; p = 0.0363), though both groups had a similar median length of stay (LOS; 2 days, p = 0.9613). For patients in the OF group, 19/37 (51.4%) required “rescue” doses of opioid medications during admission; 11/37 (29.7%) were prescribed opioids at discharge, compared with 55/56 (98.2%) in the OC group (p < 0.0001).

Conclusions: An OF pathway following instrumented PSF for idiopathic scoliosis results in equivalent LOS and fewer opioids prescribed at discharge compared with an OC pathway.

Significance: To our knowledge, this is the first study to show that OF pain management is possible in this population.

An accelerated postoperative protocol for discharging posterior spinal fusions home in less than 2 days: comparison of two matched cohorts

Chase Bauer, Jeffrey Kessler

Kaiser Permanente, Los Angeles, CA, USA

LOE-Therapeutic-Level III

Purpose: Decadron has recently been shown to help significantly with pain control and decreased length of stay (LOS) in children with scoliosis undergoing posterior spinal instrumentation and fusion (PSIF). We have recently introduced a protocol utilizing postop decadron in addition to accelerated early physical therapy for both our idiopathic (IS) and neuromuscular (NM) scoliosis patients undergoing PSIF. The goal of the present study was to compare LOS in our most recent consecutive cohort of patients (WD) versus a previous patient cohort which did not have accelerated PT or decadron (ND). Secondary outcomes compared included morphine milligram equivalents (MME) and MME/kg in the first 24 h postop, pain score, and infection rate.

Methods: A retrospective chart review of consecutive IS and NM patients with PSIF was performed. Patient demographics along with all intraoperative data and postoperative pain scores and opioid/PCA usage was recorded. Univariate (UV) and multivariable logistic regression analysis (MVLRA) was done to compare the cohorts with 95% confidence intervals (CIs).

Results: Forty consecutive patients in the WD cohort were compared to a prior ND cohort composed of 68 consecutive patients. Cohorts were matched in all patient demographics, body mass index (BMI), levels fused, blood loss, and curve magnitude. Both UV and MVLRA showed no significant difference in maximum or median pain score or MME/kg. The WD cohort averaged 1.06 less days LOS (1.73 versus 2.79, p = 0.0007), and MVLRA also demonstrated that the WD protocol led to a 1 day decreased odds ratio (OR) of LOS (CI = −1.57 to −0.44, p = 0.006). MVLRA also showed 96% decreased OR of using a PCA (OR = 0.04, CI = 0.01–0.13, p < 0.0001), with only 37.5% of patients in the WD group requiring a PCA versus 89.7% in the ND group. The infection rate was 0% and 5%, respectively, in the ND and WD group; both UV and MVLRA showed no significant difference in infection rate between these groups.

Conclusions: This accelerated postoperative spine protocol led to a 38% decrease in LOS, with a 96% decreased OR of needing a PCA despite the MME/kg use not being significantly different. Pain scores were not significantly decreased in the WD group, likely due to these patients being more rapidly mobilized and thus more active during the first 24 h postop.

Significance: This study demonstrates that, for the first time, hospital discharge on postoperative day 1 may be possible, with 70% of all NM and IS patients being discharged on postoperative day 1 over the past 10 months.

Can surgery improve painful adolescent idiopathic scoliosis patients?

Arthur Poiri, Louise Ponchelet, Anne-Laure Simon, Florence Julien-Marsollier, Mikael Finoco, Brice Ilharreborde

Pediatric Orthopedic Department, Robert Debré University Hospital, Paris, France

LOE-Therapeutic-Level III

Purpose: Posterior fusion is the gold standard surgical procedure to correct progressive adolescent idiopathic scoliosis (AIS). Optimizing perioperative management is key to improve postoperative outcomes and to avoid the non-negligible risk of chronic pain reported after surgery. Preoperative back pain has been associated with poorer functional scores in adults, but the influence of this parameter remains unclear in the AIS population. The aim of this study was therefore to evaluate the incidence of pain in AIS surgical candidates and report the postoperative outcomes of patients with significant preoperative pain.

Methods: All consecutive AIS patients who underwent posterior fusion between 2015 and 2022 were included, with a minimum 1-year follow-up. Demographic data, VAS, and Scoliosis Research Society (SRS)-22 scores were analyzed. Back pain group (BP) was defined as a preoperative VAS >3, while the back pain free group (BPF) had a VAS ≤3. Radiological measurements were performed before and after surgery by an independent observer. Scores and radiological measures were compared between groups before and after surgery.

Results: A total of 319 patients were enrolled, with 120 patients included in the BP group (pain incidence 37.6%). Mean follow-up was 31 ± 18 months. No difference was found between groups regarding mean preoperative main Cobb and correction rates (p = 0.39 and p = 0.25, respectively). Preoperative VAS averaged 5.7 ± 1.4 in the BP group and 0.38 ± 0.9 in BPF (p ≤ 0.01). No correlation was found between pain and any demographic or radiological parameter. After surgery, VAS was significantly reduced in the BP group (p ≤ 0.01), and 68% of the patients reported non-significant pain at latest follow-up. Mean SRS total scores significantly improved in BP patients (from 3.48 ± 0.41 to 3.89 ± 0.65, p ≤ 0.01). SRS satisfaction domains were significantly greater in patients initially painful (from 3.11 ± 0.66 to 4.33 ± 0.78, p ≤ 0.01). BP and BPF patients had no significant difference in postoperative VAS, SRS total, and all domains’ scores after surgery.

Conclusions: Back pain is a frequent and underestimated symptom in AIS surgical candidates, not correlated with deformity magnitude. It can be significantly improved after surgery, so its presence should not delay or contraindicate surgery. A comprehensive approach of the associated psychosocial factors remains necessary in painful patients, but outcomes and satisfaction are not negatively impacted.

Significance: Level of evidence—III—retrospective cohort study.

The postoperative decline in health-related quality of life for adolescents with idiopathic scoliosis undergoing spinal fusion

Adam A. Jamnik, Emily E. Lachmann, Anne-Marie D. Datcu, David C. Thornberg, Chan-Hee Jo, Karl E. Rathjen, Megan Johnson, Brandon A. Ramo

Scottish Rite for Children, Dallas, TX, USA

LOE-Not Applicable-Not Applicable

Purpose: The daily lives of adolescents have changed dramatically since the initial creation of the Scoliosis Research Society (SRS) questionnaire, with more time spent on social media and engaging in sedentary behaviors. We have previously demonstrated a decline in SRS scores over the last 2 decades for adolescent idiopathic scoliosis (AIS) patients prior to surgery. Meanwhile, one would reasonably conclude that improvement of techniques in surgical scoliosis management should improve postop outcomes over this same time. The purpose of this study is to evaluate how 2-year postoperative SRS scores for surgical AIS patients have changed over 2 decades.

Methods: Retrospective review of consecutively enrolled AIS patients undergoing definitive fusion (DF) at a single institution between 2002 and 2022. Patients were included if they completed an SRS questionnaire between 21 and 49 months postoperatively, classified as 2-year follow-up. A multivariate linear regression was performed with the outcome as the 2-year SRS domain and total scores and the variable of interest as the year the survey was completed. Variables included to minimize confounding were the corresponding preoperative SRS domain score, gender, race, body mass index (BMI), and age at surgery.

Results: A total of 788 patients met the inclusion criteria. Patients were 83.8% female, had a mean age at DF of 14.3 ± 2.0 years, and a mean preoperative major Cobb angle of 61.7° ± 11.1°. Of the potential confounding variables included, preoperative SRS score, body mass index (BMI), gender, and race contributed significantly to the model. Later (more recent) SRS completion years were associated with worse health-related quality of life (HRQoL) in the Mental Health domain (N = 782, coefficient = −0.009, p = 0.0432) and the Activity domain (N = 782, coefficient = −0.007, p = 0.0340) (Figure 1). The relationship between SRS completion year and SRS scores was not significant for Pain (N = 785, coefficient = −0.003, p = 0.4680), Appearance (N = 784, coefficient = 0.005, p = 0.2645), or Satisfaction (N = 689, coefficient = −0.003, p = 0.4939).

Conclusions: Patients with AIS who underwent DF in more recent years are reporting worse postoperative HRQoL for both Activity and Mental Health. Extrinsic causes likely play a role. Whereas we previously showed declining preoperative SRS Pain and Appearance scores over time, the effects of surgery may mitigate this, as the 2-year postoperative scores have not worsened over time.

Significance: Healthcare providers should be cognizant of these downward trends and should consider using adjunct treatments to support postoperative patients with AIS, such as physical and psychological therapy. Similarly, when comparing postoperative patient-reported outcomes of scoliosis patients from different time periods, clinicians and researchers should be cautious of the confounding role that external, societal factors may play in adolescents’ lives.

The CoCo (Core Outcome ClubfOot) study: recurrence, with poorer clinical and quality of life outcomes, affects 37% of patients—an international multicenter observational study

Yael Gelfer, Sean Cavanagh, Anna Bridgens, Maryse Bouchard, Elizabeth Ashby, Deborah Eastwood

St George’s Hospital, London, UK

LOE-Therapeutic-Level II

Purpose: The Ponseti method is the global gold standard treatment for clubfoot. However, recurrence rates, treatment of recurrence, and the frequency of further interventions vary widely. All increase with duration of follow-up. There is a lack of high-quality research investigating clinical and quality of life (QoL) outcomes in children with clubfoot and how they correlate with deformity recurrence. The aim of this study was to assess clinical and QoL outcomes in children with clubfoot treated by the Ponseti method using a standardized Core Outcome Set (COS) and compare these in children with and without relapse and in those requiring further treatment.

Methods: Eleven international clubfoot centers participated in this institutional review board (IRB)–approved observational study. Data were collected from consecutive clinic patients from 1 June 2022 to 30 June 2023, with a minimum 5-year follow-up. A standardized proforma was used for data collection including demographics, information regarding primary presentation and treatment, and details of subsequent relapse and further treatment. The clubfoot COS incorporating 31 parameters of clinical and QoL outcomes was used. Data were pooled across hospital centers for analysis. A regression model was used to assess relationships between baseline variables and outcome measures and between the clinical and QoL outcomes.

Results: A total of 293 patients (432 feet), median age 89 months (range, 72–113), were included. The relapse rate was 37%, with more than one relapse in 13%. Treatment for relapse considered a standard part of the Ponseti journey (recasting, repeat tenotomy, and tibialis anterior tendon transfer (TATT)) was documented in 35% of cases. Soft tissue release and osteotomies were documented in 5.4% and 2.4% of cases, respectively. Predictors of relapse included higher age at follow-up, higher initial Pirani score, and poor evertor muscle activity. Relapse and further treatment were associated with poorer outcomes. The total score and the physical component of the Oxford Foot and Ankle Score were higher and the clinical examination better in children who did not experience relapse. Thirty-seven percent of the children were unable to squat. This was related to reduced ankle dorsiflexion.

Conclusions: Patients who were treated according to the Ponseti method experienced 37% relapse. Patients with no relapse had superior core outcomes.

Significance: This is the first multi-center study using a standardized COS following Ponseti clubfoot treatment. It distinguishes patients with and without relapse in terms of clinical outcomes and QoL. This tool can serve to compare treatment methods and outcomes, can facilitate information sharing, and sets expectations with families.

Comparison of clinical outcomes, parental anxiety, and surgeon satisfaction during outpatient clinic versus operating room setting for Achilles tenotomy during Ponseti method of clubfoot correction—a randomized controlled trial

Karthick Sengoda Gounder Rangasamy, Premkumar Rajakumar, Nirmal Raj Gopinathan

Post Graduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

LOE-Prognostic-Level II

Purpose: Achilles tenotomy (AT) during idiopathic clubfoot correction can be performed either in an outpatient clinic (OPC) or operating room (OR) setting. However, there is no clear consensus regarding where to perform AT to produce better outcomes without major complications. To the best of our knowledge, no randomized controlled trial (RCT) was conducted to compare these two settings. Research question: Is AT done in OR under general anesthesia (GA) better than in OPC under local anesthesia (LA)?

Methods: Through block randomization, 40 idiopathic clubfoot cases were allotted equally to the OPC and OR groups. After satisfactory correction of midfoot cavus, forefoot adduction, and hindfoot varus using Ponseti casting, percutaneous scalpel AT was performed for correcting ankle equinus either in OPC or OR. Parental anxiety through the STAI questionnaire, self-designed surgeon satisfaction questionnaire, and ankle dorsiflexion at the final follow-up were assessed.

Results: The mean age in the OPC and OR groups were 4.18 and 5.03 months, respectively. The mean number of casts was comparable between groups. The mean duration of the procedure was 19.80 and 53.50 minutes in the OPC and OR groups, respectively (p < 0.001). The mean infant–mother separation time was 59.75 minutes in the OR group. Intra-op bleeding was seen in three (15%) patients in the OPC group and one (5%) patient in the OR group; statistically insignificant (p = 0.605) between groups. Parental anxiety during the initial presentation at the clinic and pre- and post-operative periods were comparable in both groups. The mean total surgeon satisfaction score favors OR (25.10) as their preferred setting in comparison to OPC (23.05) (p = 0.009). The mean ankle dorsiflexion on 1-year follow-up in the OPC and OR groups was 17.80° and 17.95°, respectively.

Conclusions: AT done in OPC under LA is a safe, cost-effective procedure with less infant–mother separation time and decreased overall procedural duration. Also, this setting has comparable clinical outcomes and similar parental anxiety during the pre- and post-procedural period to that done under GA in the OR.

Significance: AT during the Ponseti correction method in idiopathic clubfoot can be done safely, cost-effectively, and with less infant–mother separation time in OPCs. OR setting shows no significant additional advantage for doing AT in idiopathic clubfoot except for surgeon satisfaction.

Effect of the abduction bracing on the contralateral foot in patients with unilateral clubfoot deformity: a longitudinal study

Kelly Jeans, Victoria Blackwood, Anthony Riccio

Scottish Rite for Children, Dallas, TX, USA

LOE-Therapeutic-Level II

Purpose: Although numerous studies have assessed function of idiopathic clubfeet (CF) treated by the Ponseti method, few studies have assessed alterations in the normal foot of patients with unilateral CF imparted by Ponseti bracing. The purpose of this study was to identify pedobarographic differences in the unaffected foot of unilateral CF patients treated with the Ponseti abduction brace, and if those differences persist over time.

Methods: Patients with unilateral idiopathic clubfoot treated during infancy by either the Ponseti method (CF group) or the French Physical Therapy Method (PT group) were prospectively followed until 16 years of age. Those in the CF group used abduction braces 23 h a day for a minimum of 3 months followed by nighttime brace wear until a minimum of 2 years of age. Those in the PT group used a custom unilateral orthosis on only the affected foot during the maintenance phase of treatment. Pedobarographic assessment of the unaffected foot was collected longitudinally at 2, 5, 10, and 16 years of age. Non-parametric Mann–Whitney U tests were used to compare plantar contact area (CA%), mean force (MF%), and contact time (CT%) between groups.

Results: Forty-four patients (19 in the CF group and 25 in the PT group) underwent pedobarographic assessment at all time intervals. The CF group demonstrated significantly higher MF% and CT% as a percentage of the rollover process in the medial and lateral midfoot at 2 years of age. These differences, however, resolved by age 5, with no difference noted in plantar pressures between groups at age 16 (Table 1).

Conclusions: Abduction bracing of normal feet in patients with unilateral CF deformity results in measurable increases in midfoot plantar pressure and contact time during the brace phase of treatment indicating that bracing does impart an increased pes planus posture of the unaffected foot. These changes, however, appear to resolve following cessation of brace use.

Significance: Although abduction bracing of the unaffected foot in children with unilateral CF deformity following Ponseti casting does result in increased midfoot pressures, these changes are transient and are expected to resolve after brace discontinuation. This information should be reassuring to parents of unilateral CF patients concerned about the effect of abduction bracing on the normal foot.

Clubfoot relapse rates in idiopathic clubfoot using the Ponseti method: 65 years of data from a single institution

Jose A. Morcuende, Arianna Dalamaggas, Malynda Wynn

University of Iowa, Iowa City, IA, USA

LOE-Therapeutic-Level IV

Purpose: Clubfoot is the most common musculoskeletal birth defect in the United States, and it is widely agreed upon that conservative management using the Ponseti method of casting is the most effective treatment, consisting of correction by manipulation and casting, Achilles tenotomy, and prevention of relapse by bracing. However, relapses can still occur and may result in significant long-term morbidity for patients if not managed appropriately. No current literature discusses long-term rate of relapses with most studies limited by smaller cohorts, short follow-up, and deviations from the original method making clinical conclusions difficult. The purpose of this study is to determine relapse rate of patients with congenital idiopathic clubfoot spanning 65 years of practice.

Methods: This is a retrospective cohort study of patients treated for clubfoot at a tertiary center between 1948 and 2013. A Kaplan–Meier survivorship curve for relapse rate was used to evaluate relapse rate in patients with and without prior treatment, as well as for those adherent and non-adherents with brace treatment. Multivariate Cox models were used to assess which factors were significantly associated with relapse rate.

Results: Of the 1580 feet in the population, we found that non-compliance with bracing during the maintenance phase is the largest risk factor for relapse. The relapse rate without prior treatment was 8.9% and with prior treatment, 18.9%. Non-adherence without prior treatment increased the relapse rate to 93.6%, and with prior treatment, 95.7%. Patient age, gender, family history, unilateral versus bilateral clubfoot, prior treatment, and indication for tenotomy were not significantly associated with relapse rate. However, a higher number of casts required for correction (>7) and brace non-adherence were significant associated with relapse rate (p < 0.0001).

Conclusions: This study supports that proper technique and brace adherence lead to lower rates of relapse and further demonstrates that after the age of 5 years, relapse occurs as low as 5%. Other studies provide highly variable relapse rates with shorter follow-up. This study provides information on risk for relapses, with the greatest risk factor being number of casts needed for correction and brace adherence over 65 years of practice.

Significance: Many factors are described in the current literature to be associated with clubfoot relapse; however, this study with 65 years of data shows that number of casts for correction and brace adherence are most significant factors for preventing relapse, with decreasing likelihood after the age of 5.

Comparative analysis of three anterior tibialis tendon transfer techniques in idiopathic clubfoot

Gregory Firth, Javier Masquijo, Alexandre Arkader, Victoria Allende Nores, Manoj Ramachandran

Royal London Hospital, London, UK

LOE-Therapeutic-Level III

Purpose: Recurrence in clubfoot after the Ponseti technique is a well-recognized phenomenon. The use of the anterior tibial tendon transfer (ATTT) is a well-documented technique for addressing dynamic supination following treatment. Despite its widespread use, there is a lack of comparative studies evaluating different fixation techniques for ATTT. Consequently, we conducted a multicenter study to assess and compare the outcomes of three techniques.

Methods: Three centers from three different continents collaborated in this comparative analysis. Children with dynamic supination after initial Ponseti treatment who underwent ATTT between 2008 and 2023 were included. Exclusion criteria were neuromuscular disease, associated syndromes, partial tendon transfer, and follow-up <6 months. A pull-through technique with suture over a button (Group A) was used in 52 patients (72 feet), a pull-through technique using an interference screw fixation (Group B) was used in 23 patients (26 feet), and a novel suture anchor technique (Group C) was used in 55 patients (65 feet). A retrospective review included demographics, surgical procedures, and outcomes (active contraction of the ATTT, recurrence, and postoperative complications). The comparison between the three groups was made with analysis of variance (ANOVA) and Kruskal–Wallis tests. p-values of <0.05 were considered statistically significant.

Results: A total of 130 children (163 feet), were included. There were no statistically significant differences in age, sex, side, and isolated ATTT versus ATTT + additional surgery (soft tissue or bony) between groups. At final follow-up (mean, 42.0 months), 28 patients (38 feet) experienced some form of relapse of any aspect of the clubfoot: Group A, 27.8% (20/72 feet), Group B, 27.6% (8/29 feet) and Group C, 15.4% (10/65 feet), p = 0.41. There were five postoperative complications directly related to the ATTT, comprising four cases in Group A (5.6%) and one in Group B (1.5%), p = 0.88. Of these complications, only one case from Group A included pull out of the transferred tendon. There were no cases of overcorrection.

Conclusions: All three fixation techniques effectively secure the tibialis anterior tendon transfer in the pediatric population. The suture anchor technique may present a slightly lower recurrence and complication rate. Future large-scale studies are needed to validate these findings and provide further insights into optimal ATTT techniques.

Significance: The suture anchor and screw techniques have demonstrated their utility in addressing recurrent dynamic supination. In situations where the tendon length may be insufficient or the bone dimensions are inadequate for a pull-through approach, the suture anchor technique stands as a safe and effective alternative.

Rebound of the radiological effect of anterior distal tibia epiphysiodesis in relapsed Ponseti-treated clubfoot patients after implant removal

Arnold T. Besselaar, Maria Christine Van Der Steen, Isabelle Van Tilburg

Máxima Medisch Centrum, Veldhoven, The Netherlands

LOE-Therapeutic-Level IV

Purpose: Despite good initial results, 1.9% and 67% of the clubfoot patients treated with the Ponseti method might experience a relapse. In case repeated Ponseti treatment yields insufficient results, additional treatment might be necessary. Anterior distal tibial hemi-epiphysiodesis (ADTE) with eight plates has been proposed as treatment for recurrent equinus deformity. So far clinical results were inconsistent and mainly focused on correction during treatment. In the current project, we also evaluate outcome at least 1 year after 8-plate removal as a resumption on the project earlier presented at EPOS.

Methods: We included idiopathic clubfoot patients aged between 4 and 12 years, who are initially treated with the Ponseti method and treated with hemi-epiphysiodesis between 2015 and 2022. Patients were assessed preoperatively at indication for 8-plate removal and if available at least 1 year after 8-plate removal. A senior orthopedic surgeon specialized in clubfoot treatment determined passive ankle dorsiflexion. Furthermore, on standardized lateral radiographs, the anterior distal tibial angle (ADTA) was measured.

Results: The median treatment duration of 45 feet (31 patients, 42% female) was 22.7 (range: 9.4–37.0) months. Both ankle dorsiflexion and ADTA improved during hemi-epiphysiodesis. From 23 of these feet also a radiological assessment at least 1-year follow-up after removal of hardware was available. The ADTA showed significant increase (p < 0.001) after 8-plate removal, by modeling back to preoperative values. In the subgroup with 1-year follow-up available, no change in dorsiflexion was found. One patient undergoing bilateral hemi-epiphysiodesis experienced a wound infection and screw migration.

Conclusions: ADTE with eight plates is an effective method to achieve slope correction. However, after 8-plate removal, regression of ADTA occurs. Physicians, parents, and patients should be informed about this mechanism which might result in different timing of the treatment and possibly the need for renewed hemi-epiphysiodesis. To what extent the degree of relapse is related to age and therefore potential for residual growth is still unclear. More clarity on this would provide insight into the timing of epiphysiodesis as relapse treatment for clubfeet patients with a recurrent equinus deformity.

Significance: As far as we are aware, this is the first study showing a rebound effect of the ADTA after implant removal of an anterior distal tibial hemi-epiphysiodesis as a treatment for recurrent equinus deformity in Ponseti-treated clubfoot patients.

Is there a benefit to rigid fixation in calcaneal lengthening osteotomy in painful pediatric idiopathic flatfoot deformity? Comparing results of Kirschner wire versus plate fixation

Abhishek Tippabhatla, Beltran Torres-Izquierdo, Daniel Pereira, Zachary Meyer, Pooya Hosseinzadeh

Washington University School of Medicine, Saint Louis, MO, USA

LOE-Therapeutic-Level III

Purpose: Flexible flatfoot (FF) is a common pediatric condition that is mostly asymptomatic, and surgical intervention is only considered when painful FF is refractory to conservative treatment. Calcaneal lengthening osteotomy (CLO) is one of the most used procedures to address painful FF. Traditionally, Kirschner wires were used for fixation, but there has been a recent increase in the use of plates. We compared the clinical and radiographic outcomes of these two fixation methods.

Methods: This single-center retrospective study included children aged 8–18 years with symptomatic FF that received CLO using K-wire or plate fixation. Primary outcomes include weight-bearing radiographic measurements and complications following surgery. Secondary outcomes included patient-reported outcomes. Statistical significance was held at 0.05.

Results: Among 102 feet (65 patients), 42 feet (41.2%) underwent K-wire and 60 feet (58.8%) underwent plate fixation. No differences in casting duration (p = 0.525) and time to radiographic healing (p = 0.17) were noted. Total complications were higher in the plate cohort (12 versus 2, p = 0.04) due to a higher rate of reoperations (16.7%) for hardware-related pain (10 versus 0; odds ratio (OR) = 17.74, 95% confidence interval (CI) = 1.01–310.54, p < 0.05), and infection rates were similar. Both interventions significantly improved (p ≤ 0.001) anteroposterior (AP) talo-first metatarsal and calcaneal pitch angles. Irrespective of intervention, CLO significantly improved pain at 6 months and mobility scores at 12 months. Neither modality demonstrated superior pain or mobility scores at final follow-up.

Conclusions: Both K-wire and plate fixations lead to similar radiographic and functional outcomes after CLO in painful, pediatric flatfeet. Compared to K-wire fixation, plates cause a 17.7-fold increased risk of reoperations for painful hardware, with 16.7% of plated cases requiring reoperation.

Significance: Noting this, along with the higher costs associated with using plates, our study advocates for K-wire fixation for children undergoing CLO.

Nonunion rate of Evans osteotomy without fixation in pediatric flatfoot

Smitha Mathew, Brian Gallagher, Megan Miles, Gregory Paul Guyton

MedStar Union Memorial Hospital, Baltimore, MD, USA

LOE-Therapeutic-Level IV

Purpose: The Evans osteotomy is a lateral column lengthening procedure of the calcaneus that is commonly used to correct flatfoot deformities. Currently, there is no consensus on whether fixation is needed when performing this osteotomy. Furthermore, the use of large sized grafts for the Evans osteotomy has been associated with degeneration of the calcaneocuboid joint. The aim of this study was to determine the union rate of an unfixed Evans osteotomy with the use of a relatively small allograft wedge performed in pediatric patients undergoing flexible flatfoot reconstruction.

Methods: We retrospectively reviewed 39 pediatric patients with idiopathic symptomatic flatfoot deformity who underwent 46 unfixed Evans osteotomies with allogenic bone graft for flatfoot reconstruction between March 2013 and September 2017, with a mean follow-up of 49 (range, 9.9–243.9) weeks. Hospital record, preoperative and follow-up radiographs, and complications were reviewed.

Results: Of the 46 feet, 42 (91.3%) underwent a double calcaneal osteotomy with an associated medial displacement calcaneal osteotomy (MDCO). Mean graft wedge size was 7.2 mm (SD = 1.6 mm). The mean time to union was 10.3 (range, 6.7–13.9; SD = 1.5) weeks (Table 1). There were no nonunions. One patient (one foot) had persistent sinus tarsi pain requiring arthroscopic debridement of fibrosis and graft prominence. Significant improvement was observed in all radiographic parameters at final follow-up, including calcaneal pitch, talonavicular uncoverage, anteroposterior and lateral talo-first metatarsal angle, and lateral column length (p < 0.05) (Table 2). Postoperative calcaneocuboid subluxation occurred in 70% of feet, with no correlation with wedge size (r = 0.01, p = 0.53). Mean change in calcaneocuboid subluxation at final follow-up was 1.07 mm (SD = 2.18), with postoperative calcaneocuboid subluxation diminishing over time (Table 3). There was also no correlation between wedge size and change in lateral column length (r = 0.01, p = 0.45).

Conclusions: An unfixed Evans osteotomy for symptomatic flatfoot deformity resulted in a significant improvement in the radiographic alignment of the foot with no nonunion at final follow-up. Although calcaneocuboid subluxation was found in 70% of feet, it was small in magnitude and diminished with time. Mean graft size in our cohort was smaller than that reported by other studies.

Significance: No fixation of the Evans osteotomy was required to achieve a 100% union rate with minimal calcaneocuboid subluxation in this pediatric population. Performing adjunctive procedures may allow for less than 10 mm of lengthening with this procedure.

Improving detection of underlying neurologic etiology for pediatric cavovarus foot deformity: we can do better

Mike O’Sullivan, Michelle Christie, Rusty Lynn Hartman, Jacob Zide, Anthony Riccio

Scottish Rite for Children, Dallas, TX, USA

LOE-Diagnostic-Level IV

Purpose: While evaluation of an underlying neurologic disorder in children with cavovarus foot deformity (CVD) is of great importance, no standardized method of neurologic assessment exists. Moreover, the yield of commonly used diagnostic measures remains unstudied. This study aims to compare the diagnostic yield of traditional methods of neurologic assessment for patients with CVD of unknown etiology to a more advanced diagnostic algorithm (ADA).

Methods: An institutional review board (IRB)-approved retrospective review of patients presenting to a single pediatric tertiary care center for bilateral or unilateral CVD was performed over a 19-year period. Patients with a known etiology for their deformity or pre-existing neurologic or syndromic diagnoses were excluded. Neurologic evaluation of all remaining patients was conducted by a pediatric neurologist using one of two diagnostic algorithms. The traditional diagnostic algorithm (TDA) consisted of clinical examination, magnetic resonance imaging (MRI) of the brain and spinal cord, and/or electromyography (EMG)/nerve conduction velocity (NCV). The ADA included all components of the TDA in addition to genetic testing, and/or muscle/nerve biopsy and/or repeat EMG/NCV testing when initial workup remained negative. These diagnostic algorithms were compared regarding determination of an underlying etiology for CVD.

Results: A total of 108 patients (average age 9.7 years) were included. Ninety-six patients were assessed via the TDA which detected an underlying neurologic diagnosis in 56 (58%) patients. Of these, 15 had central neurologic disease, 35 had peripheral neuropathies, and 6 had combined central and peripheral neuropathology. Of the 40 patients in whom no diagnosis was made using the TDA, 21 were further assessed using the ADA which revealed a diagnosis in 15 (71%) patients, thereby increasing the diagnostic yield to 71/77 patients (92%) when the ADA was incorporated and 71/96 (74%) overall. Moreover, 23 of 41 patients (56%) diagnosed with an unspecified polyneuropathy by TDA received a more specific diagnosis when TDA was combined with ADA. Definitive diagnosis was achieved solely by genetic testing without utilization of the TDA in 12 patients, 75% of whom had variants of Charcot-Marie-Tooth disease.

Conclusions: Neurologic etiology remains undetected in 42% of children with CVD using solely neuroaxis imaging and electrodiagnostic testing. Determination of an underlying neurologic cause for deformity can be increased by over 30% through the incorporation of genetic testing and other components of the ADA.

Significance: When used in lieu of or in addition to traditional diagnostic tools for neuropathology, the addition of genetic testing and, as needed, tissue biopsy or repeat EMG/NCV markedly increases the ability to detect or further specify a neurologic etiology for CVD.

Efficacy of Ponseti casting in arthrogryposis

Theresa A. Hennessey, Ferran Romero, Amanda Purcell, Angielyn San Juan, Bruce MacWilliams

Shiners Children’s Salt Lake City, Salt Lake City, UT, USA

LOE-Therapeutic-Level III

Purpose: Clubfeet associated with a diagnosis of arthrogryposis multiplex congenita (AMC) are more difficult to treat than idiopathic classifications. The amount of initial treatment, additional treatment, and relative success of casting protocols, however, are not well known, particularly with moderate to long-term follow-up.

Methods: A total of 624 consecutive patients undergoing Ponseti casting for clubfoot between 2000 and 2022 were retrospectively reviewed and grouped by diagnosis with inclusion criteria being treatment initiated within the first 6 months of age, a diagnosis of arthrogryposis or an exclusionary diagnosis of idiopathic club foot (ICF), and a minimum of 5-year follow-up. Groups were propensity matched on gender, laterality, and age at initial cast. Casting data, recurrence incidence, posterior medial release (PMR) surgery, salvage procedures, and additional surgeries such as repeat TALs and other tenotomies were compared between groups. Differences reported here met statistical significance by Student t-test or Fisher’s exact test unless otherwise indicated.

Results: Twenty-three individuals with AMC and 185 with ICF met the inclusion criteria. All AMC subjects were bilateral. Propensity matching determined a subset of 23 ICF subjects for statistical comparison. No differences in time or casts applied were found in initial casting treatment. Children with AMC were 30% more likely to have at least one recurrence (87% versus 50%, odds ratio = 5.0), recurred at a younger age (1.4 versus 3.0 years), and required more casts for the first recurrence (4.0 versus 2.3). Ten AMC children (43%) went on to PMR surgery compared to one ICF (4%), an odds ratio of 17. Five AMC individuals further went on to a salvage procedure, primarily talectomy. Rates of additional surgeries were not different between groups (61% AMC versus 48% ICF).

Conclusions: Some important differences reflecting the challenges of treating club foot in the arthrogryposis population were determined in this study. Most significantly, children with AMC were 5 times more likely to have recurrence and 17 times more likely to require PMR. Still, more than half of the AMC group were successfully treated without PMR or salvage operations and did not require more frequent additional surgeries to augment casting.

Significance: Understanding the treatment course and outcomes of Ponseti treatment for clubfoot can help inform the expectations of both the provider and the patient and family. We compare these findings in children with arthrogryposis to the more commonly managed idiopathic clubfoot group to add perspective for this population.

Is the proximal lateral epiphysiodesis of the first metatarsal effective in the correction of hallux valgus in the pediatric population?

Emanuel Seiça, Teresa Clode Araújo, Afonso Cardoso, Susana Norte, Monika Thüsing, Manuel Cassiano Neves

Hospital CUF Descobertas, Lisbon, Portugal

LOE-Diagnostic-Level III

Purpose: Juvenile hallux valgus (JHV) is a forefoot deformity of the skeletally immature population with an increase in the hallux valgus and intermetatarsal angle, impacting the quality of adult life. Correction of angular deformities by epiphysiodesis was first described by Blount and become popular with “guided growth surgery.” Our objective study was to analyze the effect of guided growth of the first metatarsal in the correction of these parameters at the end of growth.

Methods: We retrospectively reviewed 64 feet with JHV with a severe family history, submitted to proximal lateral epiphysiodesis of the first metatarsal (2012–2022). Radiological measurements were performed on weight-bearing X-rays. Seven parameters were measured. The American-Orthopedic-Foot-and-Ankle-Society-Hallux-Scale (AOFAS) was applied pre/post-operatively. All patients were operated by the same technique (percutaneous drilling/curettage of the lateral quart of the proximal growth plate of the first metatarsal). One-way analysis of variance (ANOVA), Mann–Whitney U test, and chi-square test were used in statistical analysis.

Results: The mean follow-up was 2.6 years. There were no major complications. HVA improved in 48/64 feet (75%). The average HVA change was 4.2° (SD = 4.3°). IMA improved in 50/64 feet (78%). The average IMA change was 1.6° (SD = 1.9°). Both IMA and HVA improved in 47/64 feet (73.4%). Correction to reference values for HVA occurred in 38/64 feet (60%) and for IMA occurred in 37 feet (58%). These patients had a lower pre-operative HVA (20.8° versus 25.3°), IMA (11.0° versus 14.0°), PMAA (2.0° versus 3.1°), and DMAA (3.7° versus 9.1°). The AOFAS improved in 62/64 feet (97%). The average change in AOFAS was 23.6 (SD = 10.09). The average PMAA change was 1.4 valgus (SD = 2.6). There was no considerable effect in DMAA. No evidence of bone growth arrest was documented.

Conclusions: Hemiepiphysiodesis is a valid management option, with an overall good outcome in 60% of patients. However, if we consider the mild/moderate cases, this number rises to 84%, being effective in halting disease progression and improvement in patient’s symptoms and aesthetic concerns.

Significance: Few studies reported on the results of hemiepiphysiodesis in correction of JHV, however with mixed results. This is the first study that tries to identify predictive X-ray parameters for good results.

The creation and validation of an ankle bone age atlas and data predicting remaining ankle growth

Andrew Pennock, James David Bomar, Jason Pedowitz, Stephen Carveth

Rady Children’s Hospital, San Diego, CA, USA

LOE-Prognostic-Level IV

Purpose: Premature physeal closure is frequently encountered in the management of pediatric distal tibia physeal fractures. An accurate bone age assessment is invaluable in optimizing decision making with respect to prognosis and treatment. The purpose of the current study was twofold: (1) to create and validate an ankle bone age (BA) atlas spanning the pediatric and adolescent years and (2) to utilize this atlas in conjunction with distal tibia/fibular growth data as measured on serial radiographs to develop a predictive growth model for the lower extremity.

Methods: Radiographs were surveyed to identify distinguishable and reproducible features of the tibia, fibula, hindfoot, and midfoot. In a similar manner to the creation of the Greulich and Pyle atlas, a “standard” for each age/sex was selected to create an ankle BA atlas. A separate cohort of 90 patients was selected to validate the atlas. A sub-cohort of 41 patients with left-hand radiographs within 3 months of ankle imaging was used to compare the two BA approaches. Furthermore, Harris growth lines were evaluated using 304 serial images of the distal tibia to determine remaining growth.

Results: The distal tibia/fibula ossification centers provided the best age assessment for early childhood (male age: 1–6 years; female age: 1–5 years). The ossification/fusion of the calcaneal apophysis provided the best age assessment in the preadolescent stage (male age: 6–14 years; female age: 5–12 years). The closure of the distal tibia/fibula physes best determined skeletal maturity (male age: 14–16; female age: 12–14 years). The ankle atlas had excellent inter- and intra-observer reliability (intraclass correlation coefficient (ICC) = 0.993, p < 0.001 and 0.998, p < 0.001), respectively. We found excellent correlation between the patient’s chronologic age and ankle BA (r = 0.984; p < 0.001). Ankle BA assessment and G&P were correlated (rs = 0.822, p < 0.001). We found that males with a BA of ≥15 years and females with a BA of ≥13 years had ≤2 mm of residual growth of the distal tibia/fibula physes (Table 1).

Conclusions: BA can be determined using ankle films ordered to assess/treat ankle injuries. This tool, along with our growth remaining table, may have important clinical implications when managing ankle trauma patients with premature physeal closure.

Significance: This atlas can be utilized to help guide the surgeon as to the potential need and timing of surgical intervention after skeletal trauma associated with a premature physeal closure.

Hip progression after triradiate cartilage closure in ambulatory cerebral palsy: who needs continued surveillance?

Amelia M. Lindgren, Ali Asma, Kenneth Rogers, Freeman Miller, M. Wade Shrader, Jason Howard

Nemours Children’s Hospital, Delaware, Wilmington, DE, USA

LOE-Prognostic-Level III

Purpose: Hip surveillance in cerebral palsy (CP) is an accepted practice with consensus-based guidelines adopted worldwide. For the skeletally immature, with open triradiate cartilage (TRC), recommendations for radiographic surveillance stemmed from population-based studies. For non-ambulatory CP, progression of hip displacement (HD) after skeletal maturity has recently been reported, with risk factors including a migration percentage (MP) >35% and pelvic obliquity at TRC closure. Less is known, however, for ambulatory CP. The objective was to determine the prevalence and risk factors associated with progressive HD after TRC closure, a proxy for skeletal maturity, for persons with ambulatory CP.

Methods: Retrospective cohort study. Patients with ambulatory CP (Gross Motor Function Classification System (GMFCS) I-III), hypertonic motor type, regular hip surveillance (3+ X-rays after age 10 years, 1 prior to TRC closure, 1+ after age 16 years), and 2-year follow-up post TRC closure were included. The primary outcome was MP. Secondary outcomes included previous preventive/reconstructive surgery, topographic pattern, gender, scoliosis (>40° or spinal fusion after TRC closure), epilepsy, ventriculoperitoneal shunt, and the presence of G-tube. An “unsuccessful hip” was defined by MP ≥30%, MP progression by ≥10%, and/or requiring reconstructive surgery after TRC closure. Statistical analyses included chi-square and multivariate Cox regression analyses. Kaplan–Meier survivorship was also determined. Receiver-operating characteristic (ROC) curve analysis was used to identify the MP threshold for progression to an unsuccessful hip after TRC closure.

Results: Seventy-six patients (39.5% female) met the inclusion criteria, with mean follow-up 4.7 ± 2.1 years after TRC closure. The age at TRC closure for females/males was 13.5(± 1.7)/14.2(± 1.7) years (p = 0.12). Sixteen (21.1%) patients had an unsuccessful hip outcome at final follow-up. By chi-square analysis, diplegia (p = 0.002) and the presence of epilepsy (p = 0.04) were risk factors for an unsuccessful hip. By multivariate analysis, only “first MP at TRC closure” (p < 0.001) was a significant risk factor for progression to an unsuccessful hip. The mean survival time for progression to an unsuccessful hip was higher for females (p < 0.02). A first MP at TRC closure of ≥28% was associated with an unsuccessful hip at final follow-up (ROC analysis, area under the curve (AUC): 0.845, p < 0.02).

Conclusions: The risk of MP progression after skeletal maturity was still relatively high (21%), similar to non-ambulatory CP. Annual hip surveillance X-rays after TRC closure should continue for GMFCS levels I–III with an MP ≥28% after TRC closure, especially for bilateral CP, males, and for those with epilepsy.

Significance: The prevalence and risk factors for progressive HD after skeletal maturity can be used to inform hip surveillance programs for ambulatory CP.

How well does physical examination predict radiographic hip displacement in children with cerebral palsy?

Unni G. Narayanan, N. Susan Stott, Darcy Fehlings, H. Graham, Kishore Mulpuri, Benjamin J. Shore, M. Wade Shrader, Moon Seok Park, Tim Theologis, Marek Jozwiak, Jon R. Davids, Eva M. Ponten, Gunnar Hagglund, Bjarne Moeller-Madsen, Uri Givon, Deborah Eastwood, Tom F. Novacheck, Cerebral Palsy Hip Outcomes Project (CHOP)

The Hospital for Sick Children, Toronto, ON, Canada

LOE-Diagnostic-Level I

Purpose: Hip surveillance programs for cerebral palsy (CP) recommend standardized radiographs to identify hip displacement because physical examination is believed to be an unreliable predictor of hip displacement. The purpose of this study is to evaluate whether physical examination of hip range of motion can rule out or predict a clinically significant hip displacement (migration percentage > 30%).

Methods: The CP Hip Outcomes Project (CHOP) is a long-term prospective cohort study of patients (2–18 years old) enrolled from 28 sites in 10 countries evaluating the comparative effectiveness of different management strategies for hip displacement in non-ambulant children with CP (Gross Motor Function Classification System (GMFCS) IV and V). A total of 622 children enrolled prospectively in CHOP were evaluated at baseline for range of motion (ROM: R1 and R2) of hip abduction with hips and knees flexed; hip abduction with hips and knees extended; hip flexion contracture; and popliteal angles. ROM assessments were correlated with the respective MPs of each hip measured from the standardized AP radiograph of the pelvis, using linear regression, with analysis conducted for each side. Receiver operating curves were created for each specific ROM with area under the curve (AUC) reported to quantify the power of each test to predict MP ≥30%.

Results: Mean (range) age of the cohort was 6.9 (2–18) years. Mean MP was 58% (0%–100%) on each side. There was a non-linear relationship between hip abduction (R1 and R2) and MP. Ranges of motion > 30° had little correlation with the MP. When abduction was less than 25°, there was a stronger negative correlation between ROM and increasing MP (see Figure). However, the predictive value of the ROM was poor with an AUC of only 54%–59%, respectively, for R1 and R2 of hip abduction both with knees flexed or extended (see Figure). The knee popliteal angle fared worse with AUC of just over 50%.

Conclusions: This study confirms empirically in a large prospectively collected data set that the physical examination based on the ROM of the hip is a poor predictor of hip displacement, justifying the rationale of radiographs for hip surveillance programs in CP.

Significance: Evaluation of the range of motion during hip surveillance might still have clinical utility to inform indications for, and decision making about, specific interventions such as spasticity management or muscle contracture releases based on the type and magnitude of these contractures, but should not be used to predict hip displacement in CP.

Femoral head shaft angle changes based on severity of neurologic impairment in children with cerebral palsy and spinal muscle atrophy

Luiz Carlos Almeida Da Silva, Yusuke Hori, Burak Kaymaz, Kenneth Rogers, Arianna Trionfo, Jason Howard, J. Richard Bowen, M. Wade Shrader, Freeman Miller

Nemours Alfred I. Dupont Children’s Hospital, Wilmington, DE, USA

LOE-Not Applicable-Level III

Purpose: Guided growth of the proximal femur (PFGG) is a treatment option for coxa valga deformity in children with cerebral palsy (CP). Understanding proximal femoral geometric features by functional level is important to provide a rationale for the use of guided growth as a potential intervention for hip displacement. This study evaluates the neck shaft angle (NSA) and the head shaft angle (HSA) in children with different levels of neurological disability.

Methods: Medical records and pelvic radiographs of children aged 1–12 with a diagnosis of spastic CP, spinal muscular atrophy (SMA) I-II, or typical development from 2006 to 2021 were reviewed to evaluate the NSA and HSA. Patients were divided into five groups: Gross Motor Function Classification System (GMFCS) I-II, GMFCS III, GMFCS IV-V, SMA 1-2, and typical development. A linear mixed model (LMM) was utilized to evaluate NSA and HSA within the groups during the growth.

Results: In total, 196 consecutive children with multiple visits were included. The distribution and measurement of children were: GMFCS I–II 8 children (33 hip evaluations, NSA 143.7 ± 7.4, HSA 160.0 ± 7.1), GMFCS III 22 children (130 hip evaluations, NSA 153.1 ± 4.3, HSA 163.4 ± 4.2), GMFCS IV–V 30 children (137 hip evaluations, NSA 156.4 ± 5.6, HSA 167.9 ± 6.8), SMA 1–2 32 children (83 hip evaluations, NSA 161.9 ± 9.7, HSA 173.4 ± 7.4), and typical development 104 children (222 hip evaluations, NSA 138.6 ± 7.0, HSA 156.4 ± 5.9). Mean age of all children was 4.8 ± 4.5 years (range: 0.4–12.78 years). Typical development group had NSA lower than SMA 1–2, GMFCS III, and GMFCS IV–V groups. GMFCS I–II group had lower HSA than GMFCS IV–V. Typical development group had NSA lower than SMA 1–2, GMFCS III, and GMFCS IV–V groups (Figure). GMFCS I–II had lower NSA than GMFCS III and GMFCS IV–V. LMM shows that there are significant statistical differences when comparing NSA and HSA among the groups.

Conclusions: As children grow, NSA and HSA tend to decrease in typical development and GMFCS I–II groups. However, in low-tone groups (SMA 1–2) and high-tone groups (GMFCS IV–V), NSA and HSA tend to increase with age. In both low-tone and high-tone groups, coxa valga is observed. Using PFGG may help change the course of HSA and NSA in both low- and high-tone groups.

Significance: PFGG relies on anatomical parameters to correct the coxa valga deformity. Understanding NSA and HSA for different levels of neurological impairment may assist with diagnostic tools, aid in treatment planning, and help monitor the progress of interventions.

Proximal femur guided growth for spastic hip displacement in cerebral palsy children—long-term follow-up

Wei-Chun Lee, Szu-Yao Wang, Hsuan Kai Kao, Wen-E Yang, Chia-Hsieh Chang

Chang Gung Memorial Hospital, Taipei

LOE-Therapeutic-Level IV

Purpose: Proximal femur guided growth by an eccentric transphyseal screw has been used to treat spastic hip displacement in cerebral palsy children. However, there was no long-term follow-up outcome for this inventive treatment. The purpose of this study was to report postoperative long-term results.

Methods: This case series study included consecutive patients who received soft-tissue release and guided growth at the proximal femur from 2005 to 2014 with more than 5-year follow-up. Surgical indications were children with spastic cerebral palsy aged 4–12 years, gross motor function classification system level III, IV, or V, and hip displacement noted on X-ray image. Parameters include Reimer’s migration percentage (MP) which was compared between preoperative and postoperative 1 year, 2 years, 5 years, and last follow-up pelvis X-ray image.

Results: Sixteen children with 30 spastic hip displacements received guided growth surgery and soft tissue release in the study period. The mean operation age was 7.4 years, and the mean follow-up duration was 9.4 years. The mean pre-operative MP of the 30 hips was 46.5%. The mean MP in the post-operative 1, 2, 5 years and last follow-up was 37.8%, 36.4%, 26.1%, and 26.6%. The changes of MP were statistically significant between preoperative and postoperative 5-year and last follow-up. After guided growth, only two patients received proximal femur varus osteotomy to treat residual spastic hip displacement.

Conclusions: Although not as acute and definite as varus osteotomy, the correction effect by the guided growth on proximal femur was encouraging in our study. Less surgical dissection and hospital stay, less complication, and faster recovery of motion make the guided growth surgery a treatment option for cerebral palsy children with spastic hip displacement.

Significance: This study reported the long-term follow-up results of guided growth in the capital femoral physis for spastic hip displacement in cerebral palsy children.

Does the addition of proximal femoral epiphysiodesis in neuromuscular hips improve caput valgum?

M. Bryant Transtrum, Katelyn S. Rourk, Julia Todderud, Christina Regan, Anthony A. Stans, William J. Shaughnessy, A. Noelle Larson, Todd A. Milbrandt, Emmanouil (Manos) Grigoriou

Mayo Clinic, Rochester, MN, USA

LOE-Therapeutic-Level III

Purpose: Children with cerebral palsy (CP) are at an increased risk of developing progressive neuromuscular hip subluxation and potential dislocation due to altered muscle forces across the hip joint. Progressive neuromuscular hip displacement can lead to altered spino-pelvic mechanics and eventual hip pain and arthritis, imposing challenges for both patients and caregivers. This current study aims to evaluate the potential additive protective role of proximal femoral epiphysiodesis (PFE) after varus derotation femoral osteotomy (VDRO) in preventing recurrent coxa valga in CP patients with progressive neuromuscular hip disease.

Methods: We conducted a retrospective study of pediatric patients with CP who underwent a combined staged treatment approach involving VDRO at index surgery and PFE at the time of hardware removal at a tertiary pediatric orthopedic hospital between 2007 and 2022. We evaluated radiographic hip progression at four distinct time points: initial presentation, post-VDRO, post-PFE, and latest follow-up. Radiographic progression of the included hips was evaluated by measuring the acetabular index (AI), lateral center-edge angle (LCEA), neck-shaft angle (NSA), physeal tilt (PT), head-shaft angle (HAS), and migration percentage (MP) at all time points.

Results: Twenty-one hips in 12 patients were included with a mean follow-up of 6.5 years (2.1–10.5, ± 2.5). The average age at VDRO was 6.2 (3.6–12.7, ± 2.3) and PFE was 7.2 (4.2–12.7, ± 2.0). Significant radiographic improvements in all measurements were observed following VDRO. Similar improvements were not observed after PFE at latest follow-up of 6.5 years (2.1–10.5, ± 2.5) (see Table 1).

Conclusions: Our findings suggest that while VDRO is effective in improving radiographic measurements in CP patients with neuromuscular hip dysplasia, the addition of PFE does not provide further radiographic benefits at an average follow-up of 6.5 years (2.1–10.5, ± 2.5 years). Large-scale prospective studies are needed to further delineate the role of PFE and help us refine our treatment strategies for this patient population and their families.

Significance: Previous studies have investigated the benefits of both procedures individually, but there is limited knowledge regarding the additive effects of PFE following VDRO. We sought to address this gap in the literature, and we found that while VDRO led to significantly improved radiological outcomes for every measurement, PFE yielded no additional significant improvements at latest follow-up. These results can help guide caregiver counseling and can aid providers in surgical planning.

Medialization at the osteotomy site may reduce relapse after varus de-rotational osteotomy (VDRO) of the proximal femur in cerebral palsy

Frederico Vallim, Eduardo Duarte Pinto Godoy, Juliana Lyra, Joao Antonio Matheus Guimaraes, Marcello Henrique Nogueira-Barbosa, H. Graham

Hospital Estadual da Criança, Rio de Janeiro, Brazil

LOE-Therapeutic-Level II

Purpose: Non-ambulatory children with cerebral palsy have a high prevalence of hip dislocation, which may cause pain, seating, and perineal hygiene difficulty, contributing to scoliosis and pelvic obliquity. Surgical treatment includes procedures like proximal femoral reorientation by varus de-rotational osteotomy (VDRO). However, recurrent hip instability can be as high as 77% between index surgery and skeletal maturity. We evaluated the association between femoral diaphyseal medialization at the VDRO site and the recurrence of hip displacement. We hypothesized that medialization modifies the resultant forces around the hip, reducing the femoral remodeling that leads to recurrent instability.

Methods: We retrospectively evaluated the medical and radiographic records of 138 patients (276 hips) with cerebral palsy (CP), GMFCS IV or V, and followed until skeletal maturity (average 11.3 years). All patients had been registered on the Statewide CP Register (SCPR) and managed by bilateral femoral VDRO as index surgery, between 1998 and 2012. Radiographic measurement of medialization was performed using the medialization index (MeI) described by Davids et al. (2013), preoperatively, at 6 weeks, at 12 months post-surgery, and at skeletal maturity. Failure or relapse of instability was defined as the need for revision surgery before skeletal maturity or final migration percentage (MP) > 40%. The influence of MeI was determined by Poisson regression with multiple variances. Inter and intra-observer reliability of MeI was assessed by 4 different observers in 72 hip radiographs using Cohen’s D test. Finally, medialization was bench tested in synthetic models to evaluate its stability, with system rigidity and deformation compared by analysis of variance (ANOVA) and logistic regression.

Results: Groups with and without relapse were comparable pre-operatively regarding femoral morphological parameters and acetabular index. Baseline MP was higher in the relapse group (p < 0.001). MeI at 6 weeks post-surgery differed significantly between groups, being lower in the relapse group (p = 0.004, risk ratio (RR) = 0.07; confidence interval (CI) = 0.01–0.42), considering confounding factors, including pre-operative MP. MeI showed good inter and intra-observer reliability with D inferior to 0.5 in all tests. Finally, the medialization of the diaphysis did not cause system failure clinically or instability in mechanical testing.

Conclusions: Patients with more femoral diaphysis medialization had reduced recurrence of hip instability in long-term follow-up. MeI proved to be radiographically reliable, and medialization did not increase mechanical instability.

Significance: Osteotomy medialization at VDRO site can be controlled by implant selection and surgical technique. This study suggests that medialization may reduce recurrence of hip instability in non-ambulatory CP patients, after VDRO.

Combined pelvic osteotomy and proximal femur guided growth for serious hip subluxation in cerebral palsy children

Kuan-wen Wu, Hsiang Chieh Hsieh, Chia-Che Lee, Ting-ming Wang, Ken N. Kuo

National Taiwan University Hospital, Taipei

LOE-Therapeutic-Level III

Purpose: Guided growth of proximal femur has been shown to be feasible to stabilize the hip subluxation in cerebral palsy (CP) children. However, in those with higher migration percentage (MP), guided growth alone might not suffice. This study addresses whether additional pelvic osteotomy can provide early hip stability and prevent further subluxation in CP children with serious hip displacement.

Methods: We retrospectively studied data on CP children who underwent combined guided growth and pelvic osteotomy from 2016 to 2020 at a single institution. The indications for surgery were MP >40% and head-shaft angle (HSA) >155° with acetabular dysplasia (acetabular index (AI) > 22°). There were 26 consecutive CP children (12 boys and 14 girls; 42 hips) who underwent index procedures with a minimum of 2-year follow-up. The mean age at surgery was 6.3 years. Radiographic parameters including the HSA, AI, center-edge angle (CEA), Hilgenreiner’s epiphyseal angle (HEA), and Reimer’s MP were assessed before surgery and at 6 months, 1 year, 2 years, and final visit after surgery.

Results: All radiographic measurements had statistical improvement in terms of the mean reduction of HSA by 9° ± 8° (p < 0.001), the AI by 9° ± 6° (p < 0.001), and the MP by 28% (p < 0.001) at final visit. The HSA continued to steady decline after surgery, with lower HSA in patients with more follow-up times (p < 0.001), older age at surgery (p = 0.007), and milder GMFCS level (p < 0.001). However, the AI and MP had an immediate reduction at 6-month follow-up and continue to slow decline until the latest visit. We also found that the 19 of 42 hips required a longer screw change due to physis grew off the screw tip. The four of 42 of hips who had a latest MP > 50% were considered poor outcome. They were GMFCS IV and V and three hips had a severe baseline MP ≥80%.

Conclusions: The combined guided growth and pelvic osteotomy provide hip stability with immediate improvement of AI and MP, while guided growth takes time to correct coxa valgus. It is an effective strategy in CP children with avoidance of multiple osteotomies.

Significance: The study may be the first one to demonstrate the encouraging results of combined procedures in CP children with serious subluxation. Further study is necessary to determine the optimal timing of intervention and develop an algorithm for unstable CP hips.

Medium-term results after femoral head resection and subtrochanteric valgus osteotomy in children and adolescents with cerebral palsy

Ralf D. Stuecker, Madeleine Marowsky, Oliver Jungesblut, André Strahl, Martin Rupprecht

Childrens Hospital Hamburg-Altona, Hamburg, Germany

LOE-Therapeutic-Level IV

Purpose: Various salvage surgical procedures for painful hip dislocation in adolescent patients with cerebral palsy (CP) exist. In our institution, the McHale procedure is the standard of care for painful chronically dislocated hips with or without deformity of the femoral head in non-ambulatory patients with CP. This study focuses on mid-term results after surgical treatment.

Methods: Surgical reports and patient charts were analyzed retrospectively. All X-rays were evaluated and migration of the proximal femur and heterotopic ossification (HTO) according to Brooker were recorded. In addition, we conducted a telephone interview with the caretakers with special reference to pain pre- and postoperatively and after implant removal, sitting tolerance, range of motion, mobility, quality of life, and personal hygiene.

Results: Fifty-two patients (65 hips) with a mean age of 13.5 ± 3.6 years (range: 4–20 years) were included. Mean surgery time was 178.4 ± 63.4 min (range: 45–380 min) and mean follow-up was 45.17 ± 30.6 months (range: 12–204 months). A significant difference between pre- and post-operative pain levels was found, p < 0.001. Personal hygiene (p = 0.02) and quality of life (p = 0.013) improved significantly. Eighty-five percent of the caregivers would have the surgery performed on their child again and 81% of the caregivers would recommend the surgery to others. The removal of implants leads to a significant improvement in pain (p = 0.011). A total of 22 complications in 65 McHale procedures (33.9 %) were related to the procedures.

Conclusions: A significant reduction in pain and a significant improvement in hygiene as well as quality of life can be achieved with the McHale procedure in painful chronically dislocated hips in patients with cerebral palsy. Overall, the procedure is predominantly experienced as helpful by the caregivers and recommended to others. The removal of the implants improves pain significantly, but complications may occur in one third of the patients.

Significance: Medium-term follow-up data after McHale procedures are very encouraging. However, further studies to assess long-term outcomes after such procedures are needed.

The association between hip displacement, scoliosis, and pelvic obliquity in 106 non-ambulatory patients with cerebral palsy: a longitudinal, population-based study

Terje Terjesen, Svend Vinje, Thomas Kibsgård

Oslo University Hospital, Rikshospitalet, Oslo, Norway

LOE-Not Applicable-Level II

Purpose: The relationship between hip displacement (HD), pelvic obliquity (PO), and scoliosis in non-ambulatory children with cerebral palsy (CP) has not been clearly elucidated. The aims of this longitudinal, population-based study were to examine the prevalence and temporal sequence of these deformities and to evaluate how probable it is that severe unilateral HD could be the cause of scoliosis.

Methods: The study comprised 106 non-ambulatory children (37 girls and 69 boys) with bilateral CP, born between 2002 and 2006, at Gross Motor Function Classification System (GMFCS) level IV or V, and enrolled in a surveillance program. Pelvic radiographs for measurements of migration percentage (MP) and PO were taken once a year from the diagnosis of HD. Spinal radiographs were taken in patients with clinical scoliosis. Only clinically significant deformities, defined as scoliosis with Cobb angle ≥40°, HD with MP ≥40%, and supine position PO ≥5°, were included in the analyses.

Results: Scoliosis occurred in 60 patients (57%) at a mean age of 11.8 years (5.3–16.3 years). Sixty-five patients (61%) developed HD at a mean age of 4.8 years (range, 0.7–12.5 years). Thirty-three children (31%) had PO in the supine position ≥5° at the latest follow-up; the mean PO was 10.4° (5°–30°). Nineteen patients had no deformities, 35 had 1 deformity, 33 had 2 deformities, and 19 patients had 3 deformities. The temporal sequence showed that HD was diagnosed before scoliosis in all except 2 of the 43 patients with both deformities. Thirteen of 17 patients (74%) with severe unilateral HD (MP ≥ 60%) had scoliosis convexity to the opposite side of the displaced hip.

Conclusions: The combination of clinically significant scoliosis and HD occurred in 41% of non-ambulatory children, and HD was diagnosed first in most of these patients. In children with severe, unilateral HD and PO, HD was probably a contributing cause of scoliosis.

Significance: The surveillance of scoliosis in children with CP needs improvement to detect the deformity earlier. In children at GMFCS levels IV/V, a routine spine radiograph should be taken at an age of 5–6 years. If scoliosis occurs in young children, a spinal orthosis could be tried temporarily, providing improved sitting balance, head/neck control, and hand function until the child gets old enough for surgical correction.

The evaluation of total hip replacement in management of spastic painful hip dislocation in cerebral palsy

Andrzej Sionek, Bartosz Babik, Jaroslaw Czubak

Department of Orthopedic, Pediatric Orthopedic and Traumatology, Gruca Teaching Hospital CMKP, Warsaw-Otwock, Masovia, Poland

LOE-Therapeutic-Level III

Purpose: Spasticity used to be considered as a contraindication for total hip replacement (THR). Due to the development of implants as well as surgical skills, THR became an option of treatment of painful dislocation of the hip joint in patients with spastic plegia. The aim of this study was an evaluation of mid-term results of THR in spastic cerebral palsy (CP) adolescent patients with painful hip with hip joint subluxation or dislocation.

Methods: In 2014–2022, 18 patients (19 hips) with CP between the ages of 16 and 20 years underwent THR in our department. The mean follow-up was 4 years (range, 1–9 years). Results were examined with Gross Motor Function Scale, VAS scale in accordance with Ashworth scale, types of implants used (dual mobility cup and not dual mobility cup), and radiological factors (Crowe scale). Complications were noted.

Results: In case of 10 patients, there was improvement in Gross Motor Function Classification System (GMFCS) scale average from 1 to 2 points after surgery. In all patients, improvement was observed in VAS scale average 8 points (from 10 to 7) According to Crowe scale, reposition of preoperative dislocation to primary acetabulum was achieved in all cases. Complications occurred in four cases: dislocation of two THR with dual mobility cup and two THR with non-dual mobility cup requiring revision surgery with good final result. No statistical significance was noted according to the type of cup (Mann–Whitney U test).

Conclusions: Increased risk factors of complications are severe spasticity. We believe that CP patients with painful hip should be treated using THR. We did not observe any differences between the type of implant.

Significance: These findings may serve as a basis for prediction of outcomes of THR treatment in this group of patients.

Inter-rater reliability of a photo-based modified foot posture index (MFPI) in identifying severity of foot deformity in children with cerebral palsy

Beltran Torres-Izquierdo, Jason Howard, Sean Tabaie, Mara S. Karamitopoulos, Benjamin J. Shore, Monica Payares-Lizano, Robert Lane Wimberly, M. Wade Shrader, Kristan A. Pierz, Andrew Gregory Georgiadis, Jason Rhodes, Jon R. Davids, Rachel Mednick Thompson, Pooya Hosseinzadeh

Washington University School of Medicine, Saint Louis, MO, USA

LOE-Not Applicable-Level V

Purpose: Children with cerebral palsy (CP) have high rates of foot deformity. Accurate assessment of these deformities is crucial for therapeutic planning and outcome evaluation. This study aims to evaluate the reliability of a novel photo-based Modified Foot Posture Index (MFPI) in the neuromuscular patient population.

Methods: In this multicenter study, 12 orthopedic surgeons from 12 institutions evaluated standardized photographs of feet from 20 children with CP and scored their deformities using the MFPI. Raters scored photographs based on five parameters: for the hindfoot, curvature above and below the malleoli and calcaneal inversion/eversion were assessed and scored; for the forefoot, talonavicular congruence, medial arch height, and forefoot abduction/adduction were similarly evaluated and scored. Summary of MFPI scores range from −10 to +10, where a positive number connotes a tendency toward planovalgus while a negative number connotes a tendency toward cavovarus. Inter-rater reliability was determined using a two-way mixed model of the intraclass correlation coefficient (ICC) set to absolute agreement.

Results: According to the MFPI, feet examined spanned the spectrum of potential pathology, including no deformity, mild, and moderate deformities. The inter-rater reliability for various components of the MFPI was evaluated, and the findings are summarized in Table 1. All scored variables showed high inter-rater reliability with ICC ranging from 0.965 to 0.984. Rear foot total score had a mean of 0.97, an ICC of 0.979, and a confidence interval (CI) of 0.968–0.988 (p < 0.001). The forefoot total score was 2.12 with an ICC of 0.984 and a CI of 0.976–0.991 (p < 0.001). Finally, the overall total for the MFPI was mean 3.67 with an ICC of 0.982 and a CI of 0.972–0.990 (p < 0.001).

Conclusions: The photo-based MFPI demonstrates high inter-reliability in assessing foot deformities in children with CP, with specific high reliability in both rear foot and forefoot assessments. Its non-invasive nature and ease of use make it a promising tool for both clinical and research settings, especially in prospective studies evaluating the outcomes of foot reconstruction.

Significance: In addition to radiographic and patient-reported outcomes, MFPI should be adopted as part of standard outcomes scores in studies regarding the treatment of CP-associated foot deformities.

Impact of femoral derotation osteotomy and equinus varus foot correction on transverse plane asymmetry in patients with hemiplegic cerebral palsy

Mauro Cesar Morais Filho, Marcelo Hideki Fujino, Catia Miyuki Kawamura, Jose Augusto Fernandes Lopes, Fernanda Piumbini Azevedo

AACD, São Paulo, Brazil

LOE-Therapeutic-Level III

Purpose: Our objective was to evaluate the impact of external femoral derotation osteotomy (FDO) and equinus varus foot surgical correction (EVFSC) on transverse plane asymmetry in patients with spastic hemiplegic cerebral palsy (SHCP).

Methods: Patients with SHCP, with two or more gait analyses and with pelvic asymmetry at baseline were included. A total of 453 participants were identified; however, those with mixed tone (n = 45), other pathologies combined with hemiplegia (n = 11), previous orthopedic surgeries (n = 248), pre-operative gait analysis performed more than 2 years before surgery (n = 12), and those who underwent other surgical procedures (n = 47) were excluded. The remaining 90 patients were divided into 4 groups according to the surgical procedure: Control Group (CG)—no surgery (n = 21); FDO (n = 7); EVFSC (n = 40); and FDO + EVFSC (n = 22). Age at surgery, follow-up time, and baseline Gait Deviation Index (GDI) were compared. GDI, pelvic asymmetry (PA) in the transverse plane, internal hip rotation (IHR), and foot progression angle (FPA) in kinematics were analyzed at baseline and at the final follow-up, and the intra- and inter-group results were compared.

Results: The four groups matched in age at surgery (12.4–16.7 years) and follow-up time (34.5–66 months). At the final follow-up, a GDI reduction in the CG (3.6) and an increase in the other groups (7.7 in FDO, 10 in EVFSC and 21.8 in FDO + EVFSC) were observed. GDI improvement was greater in the FDO + EVFSC group than in the other groups (p < 0.001). The reduction in PA in the transverse plane was higher in the FDO + EVFSC group (15.3°) than in the EVFSC (5°, p = 0.007) and CG (1.2°, p = 0.001) groups. Regarding IHR, we observed a decrease in the FDO + EVFSC (17°), EVFSC (4.6°), and FDO (3.1°) groups, whereas the CG group exhibited a mild increase (1.1°). The reduction in IHR in the FDO + EVFSC group was greater than that of the other groups (p < 0.001). Finally, the reduction in internal FPA was similar in the FDO + EVFSC (18.6°, p = 0.002) and EVFSC (12.7°, p = 0.021) groups.

Conclusions: The combination of external FDO and EVFSC produced the greatest reduction in IHR and increase in GDI. Groups receiving EVFSC exhibited a greater reduction in internal FPA. Finally, the reduction in PA in the transverse plane was higher in the FDO + EVFSC groups than in the EVFSC or CG group, and it was similar to that of the FDO group.

Significance: Asymmetry in transverse plane is frequent in patients with SHCP and increased IHR and EVF deformity have been implicated as possible causes.

Aspirations dashed: serum neutrophil-to-lymphocyte ratio is not a good predictor of septic arthritis of the hip and knee in pediatric patients

Christopher John DeFrancesco, David Peter VanEenenaam, Carter Hall, Vineet Desai, Kevin Jossue Orellana, Wudbhav N. Sankar

Children’s Hospital of Philadelphia, Philadelphia, PA, USA

LOE-Diagnostic-Level II

Purpose: Recent research in adults suggests that serum neutrophil-to-lymphocyte ratio (S-NLR) is a superior diagnostic criterion for pyogenic septic arthritis (SA) compared to synovial fluid white blood cell count (SF-WBC) >50,000 cells/mm³ or >90% polymorphonuclear leukocytes (%PMN). However, it is not known whether this finding extends to pediatric patients.

Methods: Medical records at a large urban tertiary care children’s hospital were queried for emergency department visits between 2012 and 2022 where an aspiration was performed to evaluate for SA of the hip or knee. Patients ≥18 years old were excluded. Patients were considered “positive” for SA if synovial fluid analysis showed any of the following: (1) SF-WBC > 50,000 cells/mm³, (2) > 90% PMNs, or (3) organisms reported on gram stain. This was termed the “composite test.” Cases where aspirate and/or operating room (OR) cultures revealed an offending organism were considered culture-positive septic arthritis (CPSA). Serum and synovial fluid test data were compared to assess their diagnostic utility. Receiver-operating characteristic (ROC) curves were examined to compare the predictive value of S-NLR compared with traditional indicators of SA.

Results: In total, 392 cases met the inclusion criteria (Figure 1A). Those with CPSA had higher ESR and CRP compared to culture negative patients (p = 0.01 and p < 0.01, respectively). Univariate testing showed no difference in S-NLR between those with CPSA and those without CPSA (p = 0.97). ROC analysis for the diagnosis of CPSA showed no difference between S-NLR and SF-WBC as diagnostics (area under the curve (AUC) = 0.68 versus 0.63, respectively, p = 0.42, Figure 1B). S-NLR was also not found to be a predictor of CPSA (p = 0.90) in the 126 cases where the composite test was negative. Among patients with negative cultures, 38% had positive Lyme serologies.

Conclusions: In contrast to adults, S-NLR is not a good indicator of SA of the hip or knee in children. This may be because competing diagnoses in children (e.g. Lyme) come with systemic humoral inflammatory responses—and increased S-NLR—similar to that seen in pyogenic SA, while noninfectious/degenerative conditions that might represent the major alternate diagnosis in adults do not increase S-NLR significantly. Given regional variations in the incidence of Lyme disease, this topic should be further studied at centers outside Lyme-endemic areas.

Significance: Despite excitement regarding S-NLR as a diagnostic for adult SA, this criterion appears less useful in the diagnosis of pyogenic SA in pediatric patients.

Severity of osteomyelitis—the bug is the problem

Haemish A. Crawford, Anna McDonald, Simon Swift, Jillian Cornish, Reece Joseph, Sophia Huiyao Hamada-Zhu, Christina Straub, Brya Matthews

Starship Children’s Hospital, Auckland, New Zealand

LOE-Prognostic-Level II

Purpose: Pediatric acute hematogenous osteomyelitis (PAHO) infections in New Zealand are often caused by methicillin-susceptible Staphylococcus aureus (MSSA), and flucloxacillin is frequently used as a first-line treatment antibiotic. While most PAHO cases can be successfully treated with antibiotics, some patients respond poorly and/or require prolonged treatment despite their infecting bacteria having shown susceptibility in a laboratory setting. Other factors may contribute to antibiotic failure, such as the ability of S. aureus to form biofilms. Currently, there is a knowledge gap in how biofilm formation may be linked with antibiotic effectivity in PAHO. However, it is hypothesized that biofilms may influence antibiotic failure and poor clinical outcomes by shielding bacteria from antibiotics and giving bacteria the potential to become antibiotic-tolerant persister cells. Or is it due to increased virulence of the bacteria itself?

Methods: We sequenced 85 PAHO isolates from patients at Starship Hospital collected between 2008 and 2017 using short-read technology. Our institution is unique in that all bacteria causing PAHO over this period are stored in our “bio bank.” These bacteria were available for further analysis. The patient’s clinical course was classified as complicated or uncomplicated. A. Complicated: pediatric intensive care unit (PICU) admission, Chronic or recurrent osteomyelitis, >8 weeks antibiotics. B. Uncomplicated: resolved with no major complication and maximum 1 operation, <6.5 weeks antibiotics. The microbiologists performing the bacterial analysis were blinded to the classification of the patient’s clinical outcome. Genomes were sequenced using multiplexed SeqWell libraries run on two HiSeq lanes and assembled with SKESA.

Results: Analysis to explore whether there are bacterial gene markers beyond antibiotic resistance that correlate with treatment failure found that increased carriage of LukS Panton/Valentine Leukocidin was associated with complications in treatment (84% complicated, 19% uncomplicated isolates). All nine PICU patients had increased LukS Panton/Valentine Leukocidin. In addition, there appears to be a correlation with a shift in the carriage of the type 8 capsule serotype-specific genes over the type 5 capsule serotype-specific genes (cap8HIJK versus cap5HIJK). Complicated 7% Type 5: 93% Type 8 Uncomplicated 38% Type 5: 62% Type 8.

Conclusions: MSSA that causes complicated PAHO have a significantly increased carriage of LukS Panton/Valentine Leukocidin and a higher carriage of type 8 capsule serotype-specific genes when compared to MSSA bacteria in children with uncomplicated PAHO.

Significance: Osteomyelitis remains a significant issue among children in New Zealand. Identifying bacterial genes associated with poor outcomes in PAHO could help inform novel strategies for treatment in the future.

Featherweight versus heavyweight of pediatric musculoskeletal infections: Kingella versus the titans of Staphylococcus and Streptococcus

Brian Quincey Hou, Malini Anand, William Franklin Hefley, Katherine Sara Hajdu, Stephen Chenard, Anoop Chandrashekar, Naadir Jamal, Michael Joseph Greenberg, Courtney Baker, Stephanie N. Moore-Lotridge, Jonathan G. Schoenecker

Vanderbilt University Medical Center, Nashville, TN, USA

LOE-Prognostic-Level III

Purpose: Pediatric musculoskeletal infections (MSKIs) are common invasive infections that account for, on average, 1 in every 10 consultations to a pediatric orthopedic provider at a tertiary care center. MSKIs may cause rapid decompensation and severe complications if not triaged and treated appropriately. Complications from these infections can be severe and often involve thromboembolic processes that lead to multiorgan dysfunction and death. The epidemiological landscape of causative pathogens has evolved over time, with Kingella kingae rising in incidence alongside the mainstays of Staphylococcus aureus and Streptococcus pyogenes. However, many septic arthritis cases, regardless of pathogen, are managed similarly. We hypothesize that there are different characteristics in Kingella infections compared to Staphylococcus or Streptococcus infections, with less severe presentations of Kingella MSKIs that warrant alternative treatment.

Methods: A retrospective review of patients at a single academic tertiary care center identified 712 pediatric MSKIs from January 2013 to July 2022. Each record was reviewed to collect patient and disease characteristics as well as blood and tissue culture data. Patients without a confirmed infection resulting from Kingella, Staphylococcus, or Streptococcus were excluded.

Results: Patients with confirmed Kingella infection were significantly younger than those with confirmed Staphylococcus or Streptococcus infection (1.23 versus 7.98 versus 6.77 years, respectively, p < 0.0001). Kingella had a non-statistically significant increased propensity for the lower extremity (95.2%) compared to Staphylococcus (85.7%) or Streptococcus (77.3%). Moreover, Kingella infections induced a significantly milder inflammatory response compared to Staphylococcus or Streptococcus, with a lower admission C-reactive protein (mean, 41.3 versus 119.9 versus 100.0 mg/L, respectively, p < 0.0001) and higher admission platelet count (402.4 versus 304.0 versus 318.3, respectively, p = .0036). 92.3% of patients with disseminated disease and complications had a diagnosed Staphylococcus infection, while 3.8% had Streptococcus and none had a Kingella infection. As such, unlike those with Staphylococcus infections, no patients with Kingella or Streptococcus infections required supplemental oxygen.

Conclusions: Cases of diagnosed Kingella-caused pediatric MSKIs differ from cases of Staphylococcus or Streptococcus MSKIs. Patients with a MSKI resulting from Kingella are often younger and have a milder course of disease from MSKIs compared to patients with a Staphylococcus or Streptococcus MSKI.

Significance: Understanding the epidemiological trends in MSKI is critical for developing appropriate treatment pathways for pediatric MSKI. Differences in Kingella infection severity compared to other pathogens may indicate a need to reevaluate contemporary approaches to their treatments. Use of molecular detection methods may allow for expedited diagnosis of Kingella infections, setting the stage for new focused treatment algorithms.

Low prevalence of anaerobic bacteria in pediatric septic arthritis makes obtaining anaerobic cultures of questionable value

Maia Regan, David A. Spiegel, Kenneth Smith, Keith D. Baldwin

Children’s Hospital of Philadelphia, Philadelphia, PA, USA

LOE-Diagnostic-Level III

Purpose: Pediatric acute septic arthritis necessitates urgent identification and treatment to avoid irreversible joint damage if not recognized and treated in a timely fashion. Many centers routinely send both aerobic and anaerobic cultures for the purpose of pathogen identification; however, the yield of anaerobic culture results has been called into question. The goal of this study was to determine the positivity rate of anaerobic cultures collected intraoperatively in pediatric patients with a clinical diagnosis of septic arthritis.

Methods: Patients with a clinical diagnosis of septic arthritis were extracted from a search of musculoskeletal infections at a large tertiary care pediatric hospital from 2007 to 2021. We examined cultures obtained from the emergency room or operating room through arthrocentesis or intraoperative culture. These cultures are routinely obtained in both aerobic and anaerobic culture media at our institution. We assessed the cultures and the clinical scenario surrounding them to assess the utility of anaerobic cultures in the diagnosis and management of pediatric septic arthritis.

Results: We identified 466 cases of pediatric septic arthritis of which only 3 cultures were positive (0.6%) for anaerobic organisms, two involving a hip and one involving a shoulder (Table 1). All cases involved a single isolate. The anaerobic bacteria that were detected included one facultative anaerobe, one strict anaerobe, and one relatively aerotolerant anaerobe (Table 1). Four aerobic organisms grew on the anaerobic cultures. These included methicillin-resistant Staphylococcus aureus (2) and Staphylococcus (2). These four false positives also grew on aerobic culture media. Neither the facultative anaerobe nor the aerotolerant anaerobe grew on aerobic culture media.

Conclusions: Over a 14-year period, only 0.6% (3 cases) of septic arthritis cases yielded positive anaerobic cultures that were not able to be cultured on aerobic media. As such, an anaerobic culture was more likely to culture a bacterium that would have also been also cultured on aerobic media. In addition, a true positive anaerobic culture that would not have been diagnosed on aerobic cultures only occurs about once every 5 years at our institution.

Significance: These results suggest that routinely obtaining anaerobic cultures may be of limited value in pediatric septic arthritis. In our sample, a positive culture of strict or facultative anaerobe was exceedingly unusual. The false-positive rate of anaerobic labs exceeds that of true-positive cases. These results provide actionable opportunity to help guide clinician decision making in a more cost-effective and efficient management of pediatric septic arthritis.

Tiny humans versus a deadly disease: an epidemiologic review of necrotizing fasciitis in pediatric patients

Stephanie N. Moore-Lotridge, Samuel Johnson, Wendy Ramalingam, Jonathan G. Schoenecker

Vanderbilt University Medical Center, Nashville, TN, USA

LOE-Prognostic-Level III

Purpose: Necrotizing fasciitis (NF) is a rare but potentially life-threatening infection that can occur in both pediatric and adult patients. Urgent identification and differentiation of NF from other soft tissue infections is vital to prevent associated morbidity and mortality. To date, few studies have examined the diagnosis and treatment of necrotizing fasciitis in pediatric populations. Thus, most of the diagnosis, prognosis, and management recommendations have been based on adult data. The goal of this study was to examine the variance in epidemiology and outcomes of pediatric and adult patients with confirmed NF.

Methods: A retrospective review from a single tertiary center identified 354 patients with confirmed NF, 23 children (<17 years of age) and 331 adults. Records were individually verified for the presence of disease prior to inclusion. Demographics including age, gender, infection origin, comorbidities, causative microbe, patient presenting symptoms, and mortality rate were compared.

Results: Between children and adults with confirmed NF, similar rates of gender and ethnic distribution were observed. Upon admission, pain, erythema, and swelling were present at similar rates between cohorts. The origin and mechanism of infection were significantly different between cohorts with children experiencing extremity infections at double the rate of adults (73.9% versus 36.0%). Children presented with lower rates of NF from infected wounds (0.0% versus 25.8%) and higher rates of NF from puncture wounds (17.4% versus 8.1%) compared to adults. Tissue culture results were significantly different between cohorts with children developing culture-negative NF at more than five times the rate of adults (23.8% versus 4.6%, p < 0.01). Finally, children, along with having fewer comorbidities (p < 0.001), had more favorable outcomes with shorter length of stay and lower rates of amputation and multi-organ failure. Importantly, children also had lower mortality rates compared to adults (4.3% versus 19.8%, p < 0.01).

Conclusions: Cases of pediatric NF differ from those of adults with lower rates of comorbidities, different mechanisms of contracting NF, and higher rates of culture negativity. Future multi-center clinical studies are necessary to develop pediatric-specific (1) diagnostic criteria to allow us to detect NF and distinguish NF from other severe soft tissue infections, (2) prognostic algorithms to predict patient outcomes, and (3) clinical practice guidelines to direct optimal treatment.

Significance: The unique features of pediatric NF, coupled with the rapidly progressive nature and potentially devastating sequelae, should prompt high clinical suspicion even when a child lacks traditional risk factors, disease features, and/or culture results commonly found in adults with NF.

Awake biopsy in pediatric patients with suspected musculoskeletal malignancy is safe, feasible, cost-effective and reduces time to tissue diagnosis

Huw R.F. Walters, Alpesh Kothari, Max Mifsud, Andrew Wainwright, Karen Partington

Oxford University Hospitals NHS Foundation Trust, Oxford, UK

LOE-Diagnostic-Level III

Purpose: Biopsy under general anesthetic (BGA) has been the standard of care in pediatric patients with suspected musculoskeletal malignancy. Scheduling and undertaking BGA may lead to delays in histological diagnosis and increase patient and parental anxiety. In our UK-based tertiary sarcoma center, we have increasingly performed awake percutaneous ultrasound-guided biopsies (AUPB) in children in cases of suspected musculoskeletal bone and soft tissue tumors. The aim of this study was to evaluate the safety, feasibility, and cost-effectiveness of AUPB in children with suspected musculoskeletal tumors compared to conventional BGA.

Methods: A retrospective, institutional review board (IRB)-approved database review was undertaken on all consecutive pediatric biopsies (<16 years old) performed in a UK tertiary sarcoma center between March 2016 and March 2023. Data extracted included patient demographics, biopsy type, time to biopsy (TTB) (from decision being made), as well as duration of biopsy and associated complications. Procedure costings were also obtained. Comparison was made between patients undergoing AUPB versus BGA. The primary outcome measure was TTB. Secondary outcomes measures included diagnostic specimen rates, complications, and costs. Standard parametric and non-parametric statistical methods were undertaken (alpha < 0.05).

Results: Eighty-six patients were included in this study with 53 undergoing BGA versus 33 AUPB. AUPB was associated with a significant reduction in median TTB compared to BGA (3 days (interquartile range (IQR), 2–7) versus 5 days (IQR, 2–18 days), p = 0.037). There was no significant difference in diagnostic specimen rate in AUPB compared to BGA (96% versus 97.5% p = 0.86) There were no complications in either group and no procedural abandonment in the AUPB group. Fifty-eight percent of patients undergoing AUPB had their biopsy on the same day as their initial assessment in clinic. AUPB conferred a reduction in financial cost of £647 per patient (£1177 versus £1824, p < 0.001). The mean age of patients undergoing AUPB was older than BGA (13.2 years (range, 5–16 years) versus 8.5 years (range, 3 months–15 years), p < 0.001).

Conclusions: Awake ultrasound-guided biopsy in suspected pediatric musculoskeletal tumors is an efficient, safe, well-tolerated, and cost-minimizing diagnostic tool. Its use reduces the time to obtain a histological diagnosis compared to biopsy under general anesthetic. While it may not be suitable for all children, this method is feasible and an important part of the diagnostic armamentarium.

Significance: Collaboration between clinicians and radiologists in identifying suitable patients for AUPB can facilitate efficient one-stop clinics, timely diagnosis, and reduce costs incurred in pediatric oncological services.

Comparison of diaphyseal reconstruction techniques of the lower limbs in childhood malignant tumors: long-term results

Edouard Haumont, Lys Budiartha, Manon Pigeolet, Marine De Tienda, Frank Fitoussi, Stephanie Pannier, Eric Mascard

Necker—Enfants Malades, Paris, France

LOE-Therapeutic-Level III

Purpose: Osteosarcomas or Ewing’s sarcomas represent 90% of malignant bone tumors in children. The improvement of oncological resections and oncological treatments pushes orthopedists to improve their limb reconstructions. Endoprostheses seems to provide a significant complication rate, around 60% of cumulative complication. In biologic intercalary reconstruction, four techniques are described: vascularized fibula (VF), allograft and vascularized fibula (AVF), allograft alone (A), or induced membrane (IM).

Methods: This is a multicenter retrospective study. It concerns 93 (mean age, 12.3) patients with resection of primary malignant diaphyseal tumors between June 1986 and December 2017 and were included in French or European-chemotherapy protocols. Healing index (HI), functional score, and complications were investigated. The number of interventions was counted and compared between the different techniques. The minimal follow-up is 5 years.

Results: There were 53 osteosarcomas, 28 Ewing, and 12 other tumors, in 63 femurs and 30 tibias. Length of resection ranged from 18% to 88% of the bone length (mean 51%), with a mean length of resections was 19 cm (± 5.2 cm). Thirty-two patients had a VF, 13 had an AVF, 20 had an A, and 28 were treated with IM. Demographic data were comparable. Mean follow-up was 12 (5–25) years. Per primam consolidation was obtained in 57% of patients. Nine patients (9.5%) required amputation, seven due to local recurrence, one due to infection, and one due to mechanical failure. HI were comparable, but the average number of interventions per patient was 1.7 for AVF, 3.3 for VF, 3.3 for A, and 2.9 for IM. Four allograft resorptions were reported. In VF, 10 patients had donor-site complications (15%).

Conclusions: To our knowledge, no recent study offers a decision tree for choosing which technique for which resection. Functional results are good in all groups (MSTS mean: 85), comparison takes place on number of reoperations, complications, and time to consolidation. In femoral resections larger than 50% of bone length, AVFs provide good results. In tibias, VF and IM are comparable. However, IM leads to fewer reoperations and avoids donor-site complications. In IM non-unions, usually at proximal junction, fibula graft remained an option.

Significance: The significance of this study is its comprehensive patient cohort and long-term follow-up data to help inform decision making for individual malignant tumors patients.

Survivorship of custom-made non-invasive extendable implants in pediatric sarcoma patients

Max Mifsud, Ruben Thumbadoo, Tim Theologis

Oxford University Hospitals NHS Foundation Trust, Oxford, UK

LOE-Therapeutic-Level IV

Purpose: Osteosarcoma and Ewing’s sarcoma are the most common primary malignant bone tumors of adolescents. Unlike in adult patients, where non-custom non-extendable modular implants are typically used, in pediatric patients often custom-made non-invasive extendable implants (CMNIEI) are used. The survivorship of CMNIEI is not well documented. Previous research suggests that revision rates ranged from 39% to 47%, including revisions due to marked leg length discrepancies, and that on average patients underwent 2.7 further operations at a mean follow-up of 29.4 years. Complication and revision rates are important factors to consider in the management of these skeletally immature patients and this study seeks to add to that body of knowledge.

Methods: Retrospective case series between 2006 and 2023 of all skeletally immature patients with bone sarcoma treated by resection and reconstruction with CMNIEI by three surgeons within one nationally commissioned sarcoma unit. Patients with biological reconstructions were excluded. All prostheses were custom-made for each patient by Stanmore Implants Worldwide JTS®. Data were collected on type of tumor, type and length of bone resection performed, implant design characteristics, cementation techniques, complication rate, and revision surgery rate.

Results: Eighteen (9 male, 9 female) patients aged ≤16 years at the time of diagnosis were included in the study with minimum 2-year follow-up (mean = 56 months, range = 12–132). Fifteen had osteosarcoma and three had Ewing’s sarcoma. Mean age at the time of diagnosis was 11.6 years (range, 8–15 years). The mean resection length was 192 mm (range, 150–243 mm, femur n = 13, tibia n = 5). 77.8% of CMNIEI implants survived at least 56 months. Six of 18 patients (33.3%) suffered complications (fracture n = 1, arthrofibrosis, scar revision n = 3, other n = 1) and 4/18 (22.2%) required implant revision (revision of growing component n = 2, aseptic loosening n = 2). There were no cases of deep infection. Of those who needed implant revision, the mean implant survival was 50.1 months (range, 4.9–135.0 months), with the best implant survival in fully cemented implants. Six of 18 patients died of metastatic disease at a mean of 31.7 months after presentation.

Conclusions: 77.8% of implants survive at least 56 months. Although the rate of complications is high (33.3%), only four implants were revised (22.2%), at a mean of 50.1 months (range, 4.9–135.0 months). The lack of deep infections in our series (as opposed to the published rate of 6%–18%) likely contributed to better implant survivorship.

Significance: This information helps guide pre-operative discussions with patients/families. There is a high incidence of complications (33.3%) but overall low revision rate (22.2%).

Sufficiency of isolated vascularized fibula for intercalary reconstruction

Laura Saenz, Sevan Hopyan

Hospital for Sick Children, Toronto, ON, Canada

LOE-Therapeutic-Level III

Purpose: Since survivors of childhood sarcoma may reach their elder years, the durability of limb reconstruction is important. Several methods for intercalary femoral and tibial defect reconstruction have been used. Allografts and endoprostheses are associated with a fairly high incidence of major revision and lifelong activity restrictions. Bone transport is a resurgent method of achieving living bone reconstruction that involves several additional procedures. Emerging evidence suggests that “single barrel” vascularized fibular free flaps (VFFF) are an alternative option for limb salvage. However, this approach remains uncommon because of uncertainty in the field about the complication profile, donor-site morbidity, and whether that reconstruction is mechanically sufficient for massive defects. Our main objective is to evaluate sufficiency and complications of intercalary reconstructions with a single-barrel VFFF with and without allograft.

Methods: Retrospective cohort study of 34 pediatric patients who underwent intercalary reconstruction for primary bone sarcoma of the lower extremity from 2005 to 2021 with minimum 2-year follow-up in one institution. To compare isolated vascularized fibular grafts to allograft or composite grafts (Capanna), we used descriptive and inferential statistical analyses using SPSS and classified complications according to the Clavien–Dindo System.

Results: Group No. 1 underwent reconstructions with allograft alone or combined with VFFF (n = 13) while Group No. 2 underwent reconstructions with a single-barrel VFFF alone (n = 21). No significant differences were found in gender, age (10.5 years versus 11.4 years), or diagnosis (osteosarcoma, Ewing sarcoma). Both groups had similar reconstruction length (21 cm), incidence of unplanned surgeries (4.2 versus 2.4) for complications (soft tissue, infection, fracture, hardware failure) except for nonunion which was significantly lower in Group No. 2 (19%) compared to Group No. 1 (38%, p = 0.025). Most complications in both groups required surgical intervention. All patients resumed full weight-bearing. Donor-site morbidity was not a substantive source of complications.

Conclusions: A single-barrel VFFF is not inferior and has at least one short/medium-term advantage regarding union compared to intercalary reconstruction using allograft or composite graft. Both groups had statistically similar rates of complications and unplanned surgeries. Full weight-bearing is readily achievable with a single-barrel VFFF even after massive reconstructions.

Significance: Since a VFFF restores living bone, the approach may allow for greater activity level, resistance to infection, and healing potential over several decades. Moreover, in regions worldwide where allografts are not available, intercalary reconstruction with VFFF is a feasible option.

Femoral head cartilage window approach combined with artificial bone implantation for treatment of epiphyseal chondroblastoma in children

Xuemin Lyu, Zheng Yang

Beijing Jishuitan Hospital, Beijing, People’s Republic of China

LOE-Therapeutic-Level IV

Purpose: Chondroblastoma of the femoral head epiphysis (CBFHE) is not an uncommon occurrence in children. Due to the vulnerability of the femoral head to necrosis and the potential for secondary deformity induced by the lesion, treatment poses a significant challenge. To address this concern, we sought to evaluate the impact of a simple and efficient method—the femoral head cartilage window approach combined with artificial bone implantation—on the growth of the femoral head.

Methods: Twenty-five cases of CBFHE were reviewed in our hospital. All patients were treated by the same surgical technique: anterior exposure of the femoral head was achieved using the Bikini approach. A cartilage window was opened in the weakest area of the lesion, which measured approximately 1.5 cm × 1 cm. The lesion was then curetted, followed by the implantation of artificial bone. The average age was 10.5 years (range: 6–14 years); 14 cases were male and 11 were female. Preoperative symptom duration was on average 6 months (range: 1–24 months). The Lodwick classification was used, showing two cases of class IA, 14 cases of class IB, and 9 cases of class IC. Nine cases presented invasion of the epiphyseal plate.

Results: Follow-up postoperative evaluation was conducted at 37 months (range: 22–84 months) and revealed no evidence of avascular necrosis or lesion recurrence. Femoral head deformity was observed in three cases, and two cases exhibited leg length discrepancies with shortening of 1.6 and 2 cm, respectively. Preoperative MSTS scores averaged at 21.5 (range: 16–24) and it improved to 28.7 (27–30) postoperatively (p < 0.05).

Conclusions: The combined approach of the femoral head cartilage window and artificial bone implantation represents a safe and effective method for treating CBFHE in children.

Significance: The combined approach of the femoral head cartilage window and artificial bone implantation could be an option for CBFHE in children.

Evidence-based recommendations for treating pediatric desmoid tumors: consensus of the Desmoid Tumor Working Group*

Benjamin A. Alman, The Desmoid Tumor Working Group

Duke University, Durham, NC, USA

LOE-Economic-Not Applicable

Purpose: Desmoid tumors are locally invasive soft tissue lesions that frequently occur in the pediatric population. There have been substantial advances in diagnosis, natural history, and treatments in the past 5 years that have altered management. An international team used a formal process to develop evidence-based treatment recommendation for desmoid tumors with a sub-focus on the pediatric population.

Methods: Eighty-nine clinicians or researchers with expertise in desmoid tumors and leaders of patient advocacy groups were invited to participate. Data from a prior consensus paper was used as a baseline. New articles were assessed in several subtopics, and a meta-analysis was undertaken when appropriate data were available. A modified Delphi process was used with over 60 experts actively participating to develop recommendations.

Results: The diagnosis can be reliably by needle biopsy. Mutational analysis of the CTNNB1 gene, which encodes beta-catenin, should be undertaken to confirm the diagnosis. In mutation-negative cases, consideration should be given to genetic counseling or germ line APC mutational analysis, as these patients may have a form of familial adenomatous polyposis and routine colonoscopy is needed for surveillance. Almost a quarter of tumors will regress over time, and as such, active surveillance should be the first line of therapy unless a tumor is causing significant morbidity or has the potential for mortality. Surgery should be avoided, even in the smallest tumors, as recurrence rates are quite high, and overall outcome, as shown in meta-analysis, is worse than without surgery. In case of progressive or life-threatening situations, first-line therapy should be a methotrexate-vinblastine regimen. Tyrosine kinase inhibitors can be considered in the case of progression under this treatment. Studies on gamma secretase inhibitors in children are ongoing; however, this is associated with ovarian dysfunction, and should this be used with caution in girls. Radiation therapy should be avoided in children due to long-term side effects. The role of local therapy with other modalities, such as cryotherapy, is not well defined but could be discussed in small progressive lesion and/or refractory disease after several lines of therapy.

Conclusions: A Delphi process resulted in treatment recommendations for pediatric patients with desmoid tumors.

Significance: There was consensus around a shift away from surgery and toward active surveillance as a first-line treatment for pediatric desmoid tumors.

*Indicates a presentation in which the Food and Drug Administration (FDA) has not cleared the drug and/or medical device for the use described (i.e. the drug or medical device is being discussed for an “off label” use.)

Retrospective analysis and characterization of avascular necrosis in pediatric leukemia/lymphoma patients using BLAST classification

Amin Alayleh, Hiba Naz, Vanessa Taylor, Taylor Renee Johnson, Saima Farook, Grady Harrison Hofmann, Chiamaka Nneka Obilo, Katie Harbacheck, Tara Anne Laureano, Stephanie M. Smith, Karen Chao, Stuart B. Goodman, Kevin G. Shea

Stanford University, Palo Alto, CA, USA

LOE-Prognostic-Level III

Purpose: Avascular necrosis (AVN) is a serious complication of high-dose corticosteroid therapy for pediatric patients with leukemia/lymphoma. AVN primarily affects weight-bearing joints and long bones and may lead to significant osteoarthritis and early joint replacement. Early identification and intervention for AVN may prevent progressive joint collapse. The purpose of this study is to evaluate and characterize the occurrence and location of AVN using the newly developed Bone Location for AVN from STeroids (BLAST) classification system that considers specific bone locations and its impact on future joint collapse.

Methods: An imaging database was queried for patients 25 years of age and younger with a diagnosis of AVN and leukemia/lymphoma who required corticosteroid treatment. Magnetic resonance imaging (MRI) of the patients were reviewed and AVN sites classified using the BLAST system. AVN locations were described using descriptive statistics. Multivariable logistic regression analysis was used to assess odds of AVN bilaterality based on location.

Results: A total of 84 patients (49/35 males/females) with acute lymphoblastic leukemia (B-cell 74%, T-cell 21%) or acute myeloid leukemia (5%) were included in this cohort. The median age was 14.8 years at leukemia diagnosis and 16.5 years at AVN diagnosis. The majority of AVN locations include femur (87%), tibia (68%), and humerus (25%). Based on the Stanford-BLAST classification system, the most common sites of AVN overall include the proximal tibial metaphysis (61%), distal femoral metaphysis (60%) and epiphysis (60%), and femoral head epiphysis (50%). The most common sites of AVN in the tibia, humerus, and femur are proximal tibial metaphysis (89%), humeral head epiphysis (86%), and distal femoral metaphysis and epiphysis (68%), respectively. Patients with AVN in their tibia and humerus are 7.22 times and 9.11 times more likely to develop AVN bilaterally (p = 0.002) than patients with AVN elsewhere. Epiphyseal location was considered high risk for subsequent joint collapse in all locations.

Conclusions: This retrospective analysis demonstrates that AVN in leukemia/lymphoma patients on corticosteroid therapy has a clear predilection for specific locations in long bones. Using the BLAST classification system, practitioners are better equipped to characterize the incidence and location of AVN, monitor high-risk locations for joint collapse, and track early outcomes of preventive treatment.

Significance: A standardized classification system for AVN that identifies high-risk locations for joint collapse will facilitate timely intervention for joint preservation in this population. Development of screening protocols for early detection and prospective multi-center AVN study groups will be critical to improve functional outcomes for leukemia/lymphoma survivors.

Survival of telescoping rods decreases with successive surgeries in patients with osteogenesis imperfecta

Cynthia Nguyen, Chris Makarewich, Selina Poon, Robert Hyun Cho, Theresa A. Hennessey

Shriners for Children Medical Center, Pasadena, CA, USA

LOE-Therapeutic-Level IV

Purpose: Intramedullary rodding of lower extremity long bones in patients with osteogenesis imperfecta (OI) is a widely accepted technique for fracture treatment and prevention. Previous smaller studies have shown that telescoping rods have longer survival times compared to static rods. However, there are concerns about telescopic rod performance with potential complications such as failure to lengthen, rod tip migration, and breakage requiring revision. The purpose of this study was to examine risk factors that affect rod survival length in a large cohort of patients with OI.

Methods: This was a multi-center retrospective comparative study of patients with OI who underwent intramedullary nailing of the femur and/or tibia. Each bone included had 2-year minimum follow-up from initial implantation, no rod replacement surgeries outside of the primary hospital system, and had full survival history of each rod implanted available (followed until removal or replacement). Patient demographics, clinical data, and details of each procedure were obtained. Multivariate regression analysis was used to examine risk factors for rod survival.

Results: In total, 597 rod surgeries in 171 patients met the inclusion criteria. Average patient follow-up was 10.3 (range, 2.1–18.3) years. There were 405 static rods and 190 telescoping rods. Overall, telescoping nails had longer survival times compared to solid rods (p < 0.01). Within each rod type, the number of previous rod surgeries had a significant effect on longevity for telescoping rods (p = 0.04) but not for static nails (p = 0.9), with increasing number of surgeries leading to a decrease in rod survival. The type of bone (tibia versus femur) was significant for static nails (p = 0.04) but not for telescoping nails (p = 0.13), with femur rods being a risk factor for decreasing survival. The surgeon, number of osteotomies, and the use of bisphosphonates did not have a significant effect on survival for either group. Figure 1 shows the average rod survival time versus number of previous rod surgeries per bone for telescoping and static nails.

Conclusions: For telescoping rods, the higher the number of previous surgeries, the shorter the survival time of the rod. After three rod surgeries in a bone, average survival of a telescoping rod was no longer than a static rod. For static nails, tibia nails had a longer survival than femur nails.

Significance: Surgeons should be aware that although telescoping rods do have longer survival compared to static rods, these advantages may decrease with increasing number of rod surgeries in the same bone.

Augmentation of submuscular plates in addition to telescopic rodding in the management of long bone fractures in patients with osteogenesis imperfecta

Baris Gorgun, Onur Oto, Sema Ertan Birsel, Ozan Ali Erdal, Muharrem Inan

Ortopediatri Istanbul, Academy of Pediatric Orthopedics, Istanbul, Turkey

LOE-Therapeutic-Level III

Purpose: The use of telescopic nails is a frequently used treatment method in the treatment of osteogenesis imperfecta (OI). However, providing rotational stability in telescopic nail systems is not always possible. In this respect, we believe that a long plate used in addition to the telescopic nail may be an ideal solution for this problem. Submuscular plates not only provide rotational stability but also protect the integrity of the long bones by providing additional stability and decreasing the rate of metaphyseal fractures. Our aim is to analyze the results and characteristics of the patients diagnosed with OI and underwent deformity correction surgery of the lower extremities with submuscular plates and telescopic nails simultaneously.

Methods: The study included 48 OI patients who were operated between 2019 and 2022. The deformities were in the femur, tibia, or both. In the surgical procedure, the deformity was corrected by one or two osteotomies. After that, a telescopic nail was inserted anterogradely, retrogradely, or retropatellarly. Afterward, a submuscular bridging plate was applied through the metaphyseal parts of the bone. At least six cortices were fixated at both ends of the plate. Patients were splinted for 1 week postoperatively. Partial weight-bearing was provided at the postoperative third week with thermoplastic braces. At the end of the 6 weeks, full weight was allowed. Complication and refracture rates were recorded after the surgery.

Results: In this retrospective study, patients with OI and long bone fractures of the lower extremities with a follow-up period of at least 12 months were included. In total, 48 OI patients (mean age = 8.7) and 73 bones were included in the study. The average follow-up time is 26 months (12–36 months). In all patients, the bony union was achieved without pseudoarthrosis and infections. There was a significant decrease in the rates of nail bending and fracture in patients with submuscular plate. It was also found that the immobilization period was shortened due to additional stabilization and load bearing was started earlier.

Conclusions: Telescopic nailing is an optimal surgical method in the treatment of long bone fractures in patients of OI. To decrease the fracture rates and the need for revision surgery, submuscular plates may be added simultaneously during the surgery to provide rotational stability. Further research is needed to provide clear evidence for this hypothesis.

Significance: Based on the results, a long plate used in addition to the telescopic nail could be an ideal solution for providing rotational stability.

Long-term outcomes of intramedullary nails in osteogenesis imperfecta: fewer surgeries and longer survival times with telescoping rods in patients with over 10-year follow-up

Cynthia Nguyen, Chris Makarewich, Selina Poon, Robert Hyun Cho, Theresa A. Hennessey

Shriners for Children Medical Center, Pasadena, CA, USA

LOE-Therapeutic-Level IV

Purpose: Intramedullary rodding of lower extremity long bones in patients with osteogenesis imperfecta (OI) is a widely accepted technique for fracture treatment and prevention. Previous small studies with short- to medium-term follow-up have shown that telescoping rods have longer survival times compared to static rods. However, there are no long-term studies (10 or more years) evaluating the results of these procedures. The purpose of this study was to compare the rod survival duration and number of procedures in patients treated with static versus telescoping rods with 10-year minimum follow-up.

Methods: This was a multi-center retrospective comparative study of patients with a diagnosis of OI who had intramedullary nailing of the femur and/or tibia. Each bone included in the study had 10-year minimum follow-up, only one type of rod utilized over the follow-up period, and no rod replacement surgeries outside the primary hospital system. Patient demographics, clinical data, and details of each procedure were obtained. Bones treated with static rods were compared to those treated with telescoping rods using independent-samples t-test for continuous variables and chi-square test for categorical variables.

Results: In total, 119 bones in 52 patients met the inclusion criteria. Average follow-up was 13.1 (range, 10.2–18.4) years. Average age at implantation of first rod was 4.5 years for static rods and 4.2 years for telescopic rods. There were 69 femurs and 50 tibias. There were no differences between groups in follow-up length, Sillence type, bisphosphonate use, or age at first rod placement. There were significantly more tibias treated in the static rod group compared to the telescopic rod group. Average survival length was significantly longer for telescopic rods (Table 1). Bones treated with telescopic rods had significantly fewer surgeries compared to static rods (Table 1).

Conclusions: At 10-year minimum follow-up, bone segments that received telescoping rods as the initial treatment required significantly fewer surgeries than those treated with static nails. The average survival time of telescopic rods was significantly higher.

Significance: Over the long term, bone segments in patients with OI treated with telescopic rods required fewer surgeries and had longer implant survival times compared to those treated with static rods. When available, surgeons should consider telescopic rods as the initial implant choice in this patient population.

Prophylactic intramedullary rodding following femoral lengthening in patients with achondroplasia and hypochondroplasia

Cesar G. Fontecha, Pilar Rovira Martí

Sant Joan de Déu Children Hospital Barcelona, Esplugues de Llobregat (Barcelona), Spain

LOE-Therapeutic-Level III

Purpose: Femoral fracture after femoral lengthening in patients with achondroplasia and hypochondroplasia is a frequent complication, occurring in up to 30%. The purpose of this study is to demonstrate the effectiveness of prophylactic intramedullary rodding in preventing this complication.

Methods: Multicenter retrospective study involving 86 femoral lengthening procedures in 43 patients with achondroplasia or hypochondroplasia. Forty-two femora (21 patients) were prophylactically managed with intramedullary Rush rodding following external fixation removal (11 females and 10 males, mean age 14.53 years) compared with 44 femora (22 patients) without prophylactic intramedullary rodding (13 females and 9 males, mean age 15.14 years). The mean amount of lengthening in the rodding group was 13.3 cm (52.6%), with an External Fixation Index of 25.79 days/cm; inpatients without rodding was 14.32 cm (61.5%) and 24.46 days/cm, respectively.

Results: Seven cases (15.9%) without rodding developed fracture. Four of them required surgical correction due to displacement or shortening. Only one patient (2.38%) had fracture of the femur after prophylactic rodding and surgery was not required. The incidence of femur fracture was significantly lower in the prophylactic rodding group compared to the non-rodding group (2.38% versus 15.9%, respectively; p = 0.034). There were no cases of infection or avascular necrosis.

Conclusions: Prophylactic intramedullary rodding is a safe and effective method for preventing femoral fractures following femoral lengthening in patient with achondroplasia or hypochondroplasia.

Significance: Level III—a retrospective comparative study. Mandatory treatment after femur lengthening when removing the external fixator.

Spinal surgery in achondroplasia: causes of reoperation and reduction of risks

Arun R. Hariharan, Hans K. Nugraha, Aaron Huser, David S. Feldman

Paley Orthopedic & Spine Institute, West Palm Beach, FL, USA

LOE-Therapeutic-Level IV

Purpose: Children with achondroplasia are prone to developing symptomatic spinal stenosis requiring surgery. Revision rates are thought to be high; however, the causes and rates of re-operation are unknown. The primary aim of this study was to investigate the causes that necessitate re-operation. In addition, we report on surgical techniques aimed at reducing the risks of these re-operations.

Methods: Retrospective review over an 8-year period of all patients with achondroplasia at a single institution. Demographics and surgical/revision details were recorded. Type of surgery was placed into four categories: decompression only, decompression with a short fusion (T10), decompression with a midlevel fusion (T7–T9), and decompression with a long fusion (T2–T4). The use of interbody cage was documented. Descriptive statistics and Fisher’s exact test were performed.

Results: A total of 148 patients were identified, 33 underwent spinal surgery (22.2%) at a mean age of 17.6 years. Twenty-one patients were included, 12 were excluded for follow-up. Sixteen revisions were performed on 9 patients (43%) and 4 required multiple revisions. Fourteen (67%) primary surgeries were done at our institution and 4 (29%) required revision. On average, the time from initial surgery to revision was 1.9 years. Some revision surgeries were performed for multiple reasons: 8 pseudarthroses, 7 proximal junctional kyphosis (PJK), and 7 new neurologic findings. The mean follow-up from the index procedure was 5.8 years. Short fusions were more likely to develop PJK with an odds ratio (OR) of 31.2 (p = 0.007). Short fusions without a caudal interbody were also more likely to develop a caudal pseudarthrosis when compared with long and mid-fusions without a caudal interbody (p = 0.044). To date, none of the initial long fusions with interbody have required revision.

Conclusions: This is the largest study of re-operative spinal deformity in patients with achondroplasia. The rate of surgery is 21.5% and the risk of revision is 32.1%. This is primarily due to pseudarthrosis, PJK, and recurrent neurologic symptoms. Surgeons should consider discussing spinal surgery as part of the patient’s life plan and should consider wide decompression of the stenotic levels and fusion from T4 with the use of interbody cage at the caudal level in all patients to reduce risks of revision.

Significance: Children with achondroplasia have a high risk of developing symptomatic spinal stenosis requiring surgery. Revision rates of surgery are high and are most often due to pseudarthrosis and PJK. Longer fusions and use of an interbody cage can help mitigate these risks.

Collagen-type 2 skeletal dysplasias: key clinical, radiographic, and MRI findings guide cervical stabilization decision-making

Bryan Menapace, Colleen P. Ditro, Kenneth Rogers, Jeffrey Campbell, William G. Mackenzie, Stuart Mackenzie

A.I. duPont/Nemours Children’s Hospital, Wilmington, DE, USA

LOE-Economic-Level IV

Purpose: Cervical spine (CS) pathology is frequently encountered in pediatric skeletal dysplasias (SD). Collagen type 2 (COL2) patients are at risk for pathologic atlantoaxial instability (AAI). The value of associated clinical and imaging findings has yet to be compared. This study assesses these modalities and identifies significant values for surgical decision-making.

Methods: An institutional review board (IRB)-approved retrospective case series via Epic query (2007–2023) for SD with CS imaging. Inclusion criteria: COL2 diagnosis, pediatric orthopedist’s history and examination, CS flexion and extension (F:E) radiographs, and CS F:E magnetic resonance imaging (MRI). Patient demographics and clinical findings were collected. Radiographic measurements included dens diameter, anterior atlanto-dens interval (AADI), and posterior ADI (PADI). MRI measurements included cord diameter, space available for the cord (SAC), and myelomalacia. Surgical records were reviewed. Statistical analysis involved t-tests, significance p ≤ 0.05.

Results: In total, 547 SD patients with CS imaging included 78 COL2 patients. Fifty-one of 78 met all inclusion criteria. Patients included five COL2 diagnoses, most commonly spondyloepiphyseal dysplasia (64.7%, 33). Patients were majority 53.9% female (n = 27) and 80.4% White (n = 41). 17.6% underwent surgery (n = 9). Clinically, ≥1 AAI symptom was documented in 33.3% (3/9) of surgical and 19.0% (8/42) of nonsurgical patients (p = 0.18). Physical examination revealed ≥1 upper motor neuron sign(s) in 55.5% (5/9) of surgical and 23.8% (10/42) of nonsurgical patients (p = 0.030). Surgical versus nonsurgical comparison found radiographic dens diameter as a percentage of C1 (51.3 versus 39.3%, p < 0.0001), change in AADI F:E (6.46 versus 3.26 mm, p = 0.004), and change in PADI F:E (−7.23 versus −2.90 mm, p < 0.0001). MRI findings included SAC at neutral position (7.4 versus 9.2 mm, p = 0.016), change in SAC F:E in mm (1.9 mm versus 0.6, p = 0.002) and relative to cord diameter (31.1% versus 8.2%, p = 0.0008), and incidence of myelomalacia (77.8% versus 16.7%, p < 0.0001). Surgical management included 77.8% (7) C1 decompression with immediate (6) or delayed (1) occipital-cervical fusion. Two patients (22.2%) underwent C1-2 fusion.

Conclusions: Clinical history did not provide a significant association, while the presence of a myelopathic examination did. Since both COL2 cohorts demonstrated AAI on imaging, the key value is identifying appropriate surgical thresholds. The most valuable measures include F:E PADI, change in MRI SAC F:E relative to cord diameter, and myelomalacia.

Significance: This study, the largest case series on COL2 conditions, shows the value of various clinical and radiology studies in the management of COL2 AAI, and it provides insight into the critical measurements that can aid surgical decision-making.

Screening and early management of hips in children with spina bifida following prenatal surgical closure

Domenic Grisch, Aurelia Hof, Britta Krautwurst, Thomas Dreher

University Children’s Hospital Zurich, Zurich, Switzerland

LOE-Prognostic-Level II

Purpose: Spina bifida (SB) patients are more likely to suffer from hip dysplasia and developmental delays. At the Children’s Hospital Zurich, a study to assess the prevalence of hip dysplasia in a group of SB patients with fetal closure of the spinal lesion was conducted and described this population’s characteristics.

Methods: Hip ultrasound images of 144 children with SB following fetal SB repair surgery who underwent the examination during their first 6 months of life were analyzed. The sonographic angle measurements were classified into hip maturity types according to the Graf method, divided into five types. Descriptive statistics were applied to summarize and present the data collected during the assessment of hip dysplasia prevalence in the group of SB patients following prenatal spinal closure.

Results: The prevalence hip dysplasia in this SB patient group was 3.8%, which was in the middle of the range reported in the general population (0.24%–7.15%) and did not reach the expected percentages (30%–50%) of previous SB studies. Eleven percent of immature hips were found, and in the subpopulation, 90% matured within 6 months.

Conclusions: This study may prove that fetal SB repair also benefits hip development. However, future studies with larger patient groups and control sonographies are needed to provide more significant statements and determine the long-term outcomes of SB patients with hip dysplasia who have undergone fetal surgery.

Significance: SB patients are more likely to suffer from hip dysplasia and developmental delays. Overall, these notable findings, that there may not be a significant difference in hip dysplasia rates between fetal-operated SB children and the general population, represent a significant contribution to the field, providing a solid foundation for further research on this topic. As this study is one of the first to examine the hip development of fetal-operated SB patients, the pleasing results further highlight the importance of studying hip development in children with SB and emphasize the need for continued investigation.

Does open reduction of arthrogrypotic hips cause stiffness?

Harold J.P. Van Bosse, David Teytelbaum, Solomon Samuel, Vinieth Bijanki, Stephen Silva

St. Louis University, St. Louis, MO, USA

LOE-Therapeutic-Level IV

Purpose: Congenital hip dislocation in arthrogryposis multiplex congenita (AMC) occurs in 15%–30% of patients. Although open reduction yields positive outcomes, concerns about potential iatrogenic hip stiffness persist. Our study compares pre- and postoperative hip ranges of motion (ROMs) and assesses ambulatory abilities in patients after open reduction. We hypothesize that the pre-existing motion limitations are minimally affected by relocation.

Methods: From 2008 to 2018, 52 consecutive patients with AMC underwent 75 open reductions of congenitally dislocated hips (23 bilateral), most via a medial approach with femoral shortenings osteotomies as needed. All had ≥2-year follow-up. Hip ROM was recorded preoperatively, post-operatively, and at last follow-up, as was ambulatory ability. Simultaneous and subsequent hip operations were recorded.

Results: Average age at surgery was 23 months (range, 10–58 months), with 68-month (24–152 months) follow-up. Thirty-four hips had flexion contractures >20° preoperatively (average 33°), improving by 22° at follow-up; 61 hips had <45° frog leg abduction (average 27°), improving by 11°; 41 hips had <30° abduction preoperatively (average 20°) which improved by 11°, all p-values < 0.001 (Table 1). Hips that had <90° of flexion pre-operatively showed no improvement or worsening at follow-up (p = 0.867). Flexion-extension (Flx-Ext) total arc of motion (TAM) for all hips decreased by only 2° from pre-operative to follow-up (p = 0.507), and the internal-external rotational (IR-ER) TAM in extension decreased 13° (p < 0.001). Pre-operatively, the 29 unilateral dislocated hips had an average TAM flx-ext of 7° less than the contralateral hip (85° versus 92°); at final follow-up, that discrepancy only increased 3° (78° versus 88°). All other TAMs were statistically unchanged at follow-up. Of the 52 patients, 30 were independently ambulatory at follow-up, most with braces, while 16 were walker-dependent but still progressing; six remained non-ambulatory.

Conclusions: Open reduction for arthrogrypotic hip dislocations maintains mobility without causing stiffness. While pre-existing hip motion limitations slightly worsened, lower limb positioning improved, enhancing ambulation—especially in hip extension and abduction. Most achieved independent ambulation.

Significance: The results of this study challenge the prevailing belief that performing an open reduction on arthrogrypotic hips leads to significant functional stiffness. Our findings suggest that open reduction can offer the patient a more functional hip, contrary to the perception of limited mobility. Children with AMC often have ambulatory potential if their limb positioning can be optimized.

Burosomab reduces the need for hemiepiphysiodesis in hypophosphatemic rickets

Christopher James Marusza, Zakir Haider, Kelvin Miu, Daniel Thomas Fontannaz, Deborah Eastwood

Great Ormond Street Hospital, London, UK

LOE-Therapeutic-Level III

Purpose: X-linked hypophosphatemic (XLH) rickets is the commonest inherited form of rickets. Children often present with lower limb deformities. Corrective osteotomies in early childhood are associated with high recurrence rates and the risks of surgery. Burosumab is a monoclonal antibody approved in 2018 for the treatment of XLH. Administered fortnightly by subcutaneous injection, it is now the preferred medical treatment. We report the use of hemi-epiphysiodesis in a cohort of XLH patients pre and post the introduction of Burosumab.

Methods: XLH patients referred to our orthopedic unit since 2005 were reviewed. Hemi-epiphysiodesis was undertaken if there was progressive mechanical axis deviation after 1 year of maximized medical treatment. The rate of correction of deformity per month (RoC) was calculated from standardized long-leg radiographs using Traumacad software. Statistical analysis was performed using SPSS. The Mann–Whitney U test was used for analysis of continuous data. Statistical significance was set at p < 0.05.

Results: Since 2005, 28 patients were seen for orthopedic assessment. Sixteen patients required 18 episodes of guided growth (32 limbs, 40 segments). Mean age was 10.2 (4.3–14.7) years. Neutral mechanical axis was restored in 13/32 (41%) limbs: a further 7 (22%) improved. Mean follow-up was 42.7 (6–116) months. Three patients (four limbs) underwent an osteotomy. The mean RoC with hemi-epiphysiodesis was 0.35° for the distal femur and 0.20° for the proximal tibia. Diaphyseal deformity RoC was 0.12° per month in both the femur and the tibia. Younger patients (≥3 years of growth remaining) corrected at the same rate as older patients. The RoC was higher for the distal femoral deformity and the femoral diaphyseal bow in females (p = 0.043 and 0.047, respectively). Since 2018, 28 patients commenced Burosumab treatment, mean age 7.8 (0.5–14.5) years. Mean follow-up was 31 (5–53) months. The mean RoC per month with Burosumab was 0.21° for the distal femur and 0.15° for the proximal tibia, and 0.18° for both femoral and tibial diaphyseal deformity. No statistical significance was found in RoC between hemi-epiphysiodesis and Burosomab treatment. No patient treated with Burosumab has required hemi-epiphysiodesis.

Conclusions: The data confirm that, in conjunction with medical therapy, hemi-epiphysiodesis is effective in correcting lower limb deformity. Burosumab alone demonstrates a similar improvement and no hemi-epiphysiodesis has yet been required.

Significance: This study suggests that Burosomab treatment of XLH improves lower limb deformity and reduces the need for hemi-epiphysiodesis.

Ethnicity is a risk factor for permanent brachial plexus birth injury

Petra Grahn, Aarno Yrjana Nietosvaara, Mika Gissler, Marja Kaijomaa

Helsinki University Hospital, New Children’s Hospital, Helsinki, Finland

LOE-Prognostic-Level II

Purpose: We aimed to assess the incidence and risk factors related to permanent brachial plexus birth injury (BPBI) in Southern Finland.

Methods: Helsinki University Hospital (HUS) New Children’s Hospital is the primary and only care center for BPBI in Southern Finland, while HUS Women’s Hospital oversees and trains all birth units in the region which hosts a population of 1.7 million. Birth information of all mothers and their children born 2006–2022 were gathered from the national database and compared to prospectively gathered data from mothers and children with a permanent BPBI. Permanent injury was defined as limited active or passive ROM or decreased strength of the affected limb detected at 1 year of age. Severity of the injury was assessed using the 3-month Toronto test score. In Finland, all women are subjected to the same prenatal care, and no private delivery hospitals exist.

Results: A total of 298,428 children were born within the study period. Six hundred acquired a BPBI, 100 of which were permanent. The incidence of permanent BPBI was 0.4 per 1000 vagin*l live births (0.34 all births) with a declining trend. Children to immigrant mothers, and women of Black ethnicity had a higher incidence of a permanent injury (0.85 and 1.52). Non-White background and immigrant status correlated with a more severe injury (β = −1.12 range, −2.17 to −0.07, p = 0.004) with children of Black mothers having the least favorable outcome (β = −1.64 range, −2.79 to −0.49, p = 0.005). In the multivariate regression model, gestational diabetes in a Black mother further strengthened the negative effect (β = −0.99, range, −1.98 to −0.00, p = 0.049).

Conclusions: The overall incidence for a permanent BPBI is on a decline. However, care needs to be taken especially in non-White women to prevent and in a timely manner recognize and treat risk factors leading to severe birth complications.

Significance: Black ethnicity and immigrant status are risk factors for a permanent and more severe BPBI.

Sprengel deformity: what is the functional outcome and quality of life after surgery according to the EQ-5D-Y and the short version of Disabilites of the Arm, Shoulder, and Hand Questionnaire (quickDASH)?

Carina Antfang, Adrien Frommer, Georg Gosheger, Robert Roedl, Andrea Marira Laufer, Gregor Toporowski, Henning Tretow, Jan Duedal Rölfing, Bjoern Vogt

Department of Pediatric Orthopedics, Deformity correction and Foot surgery, University Hospital, Muenster, Germany

LOE-Therapeutic-Level IV

Purpose: Sprengel deformity is a rare skeletal deformity due to a maldescent of the scapula. In addition to cosmetic issues, abduction limitation of the affected shoulder can be functionally impairing and thus the primary indication for surgical treatment. This study is designed to evaluate the function of the shoulder of patients who underwent surgery for deformity correction and those treated conservatively.

Methods: The longitudinally maintained database of our orthopedic teaching hospital was retrospectively analyzed to identify all patients with Sprengel deformity who were treated from 2016 to 2022. Nineteen patients with a mean age of 6 years (min./max. 0/16 y), with 21 affected shoulders (right = 10; left = 7; bilateral = 2) were included. Of these, 8/19 patients with a median Cavendish 4/Rigault 3 score and severe functional impairment were treated surgically according to a modified Green procedure and 11/19 patients (Cavendish 2/Rigault 2) were treated conservatively by means of physiotherapy. Shoulder function was assessed by clinical examination and the short version of Disabilities of the Arm, Shoulder, and Hand Questionnaire (quickDASH). Quality of life was evaluated using the EQ-5D-Y.

Results: In the surgical group, the Cavendish 4 could be reduced to Cavendish 1 (min./max. 1/3). Shoulder abduction improved by 45° (interquartile range, 27.5) to a postoperative abduction of 135° (median conservative cohort: 105°). Five of those eight patients were satisfied with the outcome, one unsure and one unsatisfied. On average, patients needed 3 months to be able to return to daily life activity. Nine of 19 questionnaires were complete. The median quickDASH score was 23 in the surgical cohort and 9 in the conservative cohort. Average follow-up was 66 months (min./max. 4/145). The quality of life according to the EQ-5D-Y was 72/100 in the surgical cohort and 77/100 in the conservative group.

Conclusions: Although surgical intervention in Sprengel deformity implies intense surgery, it can improve the abduction of the affected shoulder and may reduce disability in daily life at a moderate level. Yet, mild forms can still have a sufficient function and should therefore not be considered for surgery. Quality of life seems to be almost equal in both groups. Thus, when advising about performing surgery on patients with Sprengel deformity, the quickDASH can help to evaluate the current limb function and lead together with the clinical evaluation to a better decision making.

Significance: Sprengel deformity is a rare congenital disorder and surgical correction means high impact surgery. It is therefore important for clinicians to know the gain by surgery for better patient counseling.

Medium- and long-term clinical and functional outcomes of modified Green’s procedure for Sprengel shoulder in children

Giovanni Trisolino, Marco Todisco, Paola Zarantonello, Giovanni Di Gennaro, Alessandro Depaoli, Gino Rocca

IRCCS Istituto Ortopedico Rizzoli, Bologna, Italy

LOE-Therapeutic-Level IV

Purpose: Sprengel shoulder is a rare congenital deformity caused by failure in the descent of the scapula during early fetal growth. This condition leads to hypoplastic, elevated, and malrotated scapula. The present study described clinical and functional outcomes in pediatric patients treated with surgery for Sprengel shoulder.

Methods: From June 2010 to April 2021, 32 scapulae in 31 pediatric patients were surgically treated at our institution. Cavendish classification for aesthetics and Rigault classification for radiological appearance were used to assess deformity. Abduction and elevation movements were quantified. The following patient-reported outcome measures (PROMs) were administered: QuickDASH, Simple Shoulder Test (SST), the Shoulder Pain Index and Shoulder Disability Index (SPADI), and UCLA Shoulder Scale. Complications were evaluated according to the modified Clavien–Dindo–Sink classification. Clavicle osteotomy and omovertebral bone resection were also performed in some cases.

Results: With a mean follow-up of 6.4 years (range, 1–11.2 years), 2/32 major complications were found. In 18/32 patients, a Cavendish value of 1 was found, 18/32 achieved a degree of abduction greater than 120°, and 21/32 achieved a degree of elevation greater than 135°. The scores indicate that a good functional and aesthetic result was achieved in most cases. Few studies have used validated scores, often after treatment with Woodward technique. Moreover, this case series represents the largest collection of SD patients treated surgically.

Conclusions: In our experience, the modified Green’s technique proved to be a safe procedure with a low rate of complications and recurrences. Despite the prolonged follow-up, we believe that the functional results should be evaluated even after the full maturation of the musculoskeletal system to confirm the long-term efficacy of this treatment.

Significance: This case series confirmed the good results of the modified Green’s procedure in one of the largest cohort of patients ever reported.

Osteot-OH MY! contemporary surgical techniques may reduce revision rates following preaxial polydactyly reconstruction

Eliza Buttrick, Sarah L. Struble, Shaun Mendenhall, Benjamin Chang, Sulagna Sarkar, Apurva S. Shah

Children’s Hospital of Philadelphia, Philadelphia, PA, USA

LOE-Therapeutic-Level IV

Purpose: Previous investigations into preaxial polydactyly reconstruction established revision surgery rates ranging from 7% to 19%. This study aimed to define the rate and type of post-operative complications following surgical treatment of preaxial polydactyly.

Methods: Demographics, modified Wassel-Flatt classification, and surgical and post-operative data were abstracted from the electronic medical records of patients with preaxial polydactyly at a single children’s hospital between 2011 and 2022. Patients were contacted to complete an aesthetic survey adapted from the Australian Hand Difference Register, Wong-Baker pain, and PROMIS Upper Extremity Function and Peer Relations modules. Data were analyzed via descriptive statistics and non-parametric tests.

Results: A total of 232 children with 248 cases of preaxial polydactyly (6.5% bilateral, 43.5% female) underwent surgical treatment at a mean age of 1.5 ± 1.4 years, with a mean 1.3 ± 2.0 years of follow-up. Patients mostly presented with Wassel-Flatt type IV duplications (87 thumbs, 35.1%), and notably, an additional 31 (12.5%) were the hypoplastic subtype. The majority of cases (181/248, 73.0%) underwent resection with reconstruction at a mean age of 1.6 ± 1.5 years, with reconstruction including elements of ligament reconstruction, corrective osteotomy (n = 97, 53.4%), and/or tendon realignment. Early post-operative complications occurred in 23/248 (9.3%) patients. Long-term complications occurred in 87/249 cases (34.9%), significantly more likely in triphalangeal thumbs and in cases needing reconstructive surgery (p < 0.001, Table 1). Only nine patients underwent revision surgery, all requiring corrective osteotomy for coronal plane deformity (mean pre-revision IPJ angular deformity 29.7° and mean MPJ angular deformity 20.0°). Thirty-nine patients completed prospective questionnaires at an average age of 7.3 ± 3.5 years, with normal PROMIS UE and Peer Relations scores (mean 46.4 ± 9.2 and 52.4 ± 9.2, respectively). Nearly all (31/37) respondents reported negligible pain (0–1). Regarding aesthetics, 27 patients (69.2%) reported they “strongly agreed” or “agreed” with the statement “I am happy with how my hands look.” Similarly, 35 respondents (89.7%) “strongly agreed” or “agreed” that their operation(s) made their hands look better.

Conclusions: The revision surgery rate was 3.6%, reflecting technical emphasis on corrective osteotomy, ligament reconstruction, and/or tendon realignment at index surgery. In fact, more than half of the cases were treated with osteotomy at the time of the initial procedure. Normal post-operative function and excellent cosmesis can be expected.

Significance: Our series demonstrated a lower rate of revision surgery than previously reported in children with preaxial polydactyly, indicating the importance of corrective osteotomy at index surgery.

Surgical versus nonsurgical management of pediatric ganglia—a cost and outcomes analysis

Bryce Bell, Umar Ghilzai, Zuhair Jameel Mohammed, Christine Yin, Abdullah Ghali, Qianzi Zhang

Baylor College of Medicine, Houston, TX, USA

LOE-Therapeutic-Level III

Purpose: Pediatric ganglia are an understudied phenomenon, with an estimated 10% of all ganglia occurring in pediatric patients. Data on treatment modalities, recurrence rates, and cost effectiveness remains sparse. This study aims to characterize recurrence rates of surgical versus conservative management of pediatric ganglia, along with cost-effectiveness analysis of treatment modalities.

Methods: This was a retrospective cohort study that included all patients from 1 December 2011 to 15 March 2023 that were treated at a single institution for primary or recurrent ganglia of the wrist. Patient data was gathered regarding ICD-10 code, CPT code, date of procedure, primary payer, and total cost of procedure associated. These data were then stratified into two cohorts: surgery and percutaneous ultrasound-guided fenestration of ganglia (PUGG). Patient data was analyzed for rates of recurrence, treatment methods for recurrence, and total costs of treatment.

Results: Five hundred and ninety-five patients comprised our cohort, with a total of 710 procedures performed—296 underwent PUGG and 414 underwent surgery. We noted a recurrence rate of 14.7% after PUGG and 21.2% after surgery. Mean time to recurrence was 302.66 days after the PUGG cohort and 1619.13 days after surgery (p < 0.0001). PUGG recurrence was treated with surgery in 76.3% of cases and surgery recurrence was treated with re-operation in 59.2% of cases. Average payer charge was $2224.01 for PUGG and $11,645.71 for surgery (p < 0.0001). Significant differences were seen in charges among private insurers (BCBS, p < 0.0001) and Medicaid/CHIP (p = 0.0004). No significant differences were seen in cost of re-operation and repeat procedure for both surgery and PUGG groups, respectively, even after controlling for insurer status.

Conclusions: Ganglia are a challenging entity to definitively treat. PUGG can present as an alternative option for pediatric ganglia given its cost-effectiveness and comparable recurrence rates. One factor for practitioners to note is the faster time to recurrence for PUGG-treated ganglia. Both cohorts saw a preference for recurrence treatment with surgical management. Overall, PUGG can serve as an effective management strategy for pediatric ganglia.

Significance: This study is one of a few to examine pediatric ganglia recurrence rates after surgery and PUGG along with associated costs. This study shows that PUGG can effectively manage pediatric ganglia, with lower associated charges and recurrence rates.

Tendon transfer in spastic cerebral palsy upper limb

Mahzad Javid, G. Hossain Shahcheraghi, Hadi Gerami

Shiraz Medical University, Shiraz, Iran

LOE-Therapeutic-Level III

Purpose: Cerebral palsy (CP) in upper limb produces functional, aesthetic, and hygienic issues and is not always amenable to surgical procedures. We are reporting a single-center, long follow-up experience with tendon transfer in wrist and forearm CP.

Methods: The CP cases who had undergone tendon transfer in hand, wrist, and forearm in a 14-year period were evaluated for change in motion, function, and cosmetic appearance and also assessed by MACS (Manual Ability Classification System) and DASH (Disability of Arm, Shoulder, Hand) scores.

Results: Forty-two spastic CP patients with a mean age of 19.81 (10–34 years, SD: 5.36) years, with a mean follow-up of 5.5 (2–14) years, entered the study. Twenty-four cases were Gross Motor Function Classification System (GMFCS) I or II and 18 were III or IV. Supination beyond neutral was seen in 48.5%, and improvement in MACS scores in all the cases. Improved “grasp” and “release” and keyboard use was seen in 50%, 71%, and 87% respectively. The satisfaction from appearance and improved function of 83%–96% correlated positively with GMFCS, MACS, and DASH scores. Noticeable improvement in personal hygienic care was seen in only 52% of cases.

Conclusions: Tendon transfer in well-selected spastic upper extremity CP cases results in long-time improved function and limb appearance—correlating with initial GMFCS and MACS scores.

Significance: A long-term results of function improvement in pure spastic CP upper limb.

Factors influencing return of elbow motion following pinning of displaced supracondylar humeral fractures

Akbar Nawaz Syed, Pooja Nilesh Balar, Margaret Bowen, J. Todd Lawrence

Children’s Hospital of Philadelphia, Philadelphia, PA, USA

LOE-Therapeutic-Level IV

Purpose: Various operative and post-operative protocols exist for the management of displaced supracondylar humeral fractures with most providing excellent functional outcomes by 1 year. However, factors influencing the early return of elbow motion have not been evaluated. This study aims to identify factors influencing the time to return of elbow motion following operative fixation of displaced supracondylar humeral fractures.

Methods: An institutional review board (IRB)-approved retrospective study was conducted of patients <18 years old treated at a tertiary children’s hospital from January 2013 to January 2020 for a displaced supracondylar humerus fracture (Gartland type III/IV). Time to functional elbow motion was recorded for each patient and defined as elbow extension <30 degrees and flexion >130 degrees. Patients with screw fixation, inadequate documentation, insufficient follow-up, and incomplete recovery <1 year were excluded. Demographic data, injury characteristics, operative technique, and post-operative factors were recorded. Analysis was performed using Kruskal–Wallis/Mann–Whitney U tests after ensuring normality using the Shapiro–Wilk test.

Results: A total of 433 displaced supracondylar humeral fractures were identified with an average age of 6.2 years. Patient characteristics are outlined in Table 1. Median time to recovery of elbow motion was 40 days. Flexion injuries (n = 17/433, 3.9%, p = 0.012), fractures associated with nerve injuries (n = 86/433, 19.9%, p < 0.001), fractures undergoing open reduction (n = 26/433, 6%, p = 0.012), and those fixed using >3 pins (n = 63/433, 14.5%, p = 0.018) took twice as long to regain motion. Referral to physical therapy was not beneficial for return of motion (p < 0.001). Immobilization >21 days following pin removal (<0.001) delayed the return of motion by ~4 weeks. Age, gender, additional ipsilateral fracture, days to pin removal, pin construct, and management of additional ipsilateral fracture (Op versus non-Op) did not delay the return of elbow motion (p > 0.05).

Conclusions: To our knowledge, this is the first study to describe factors influencing the return of elbow motion. Delay in return of motion is seen with flexion-type injuries, concomitant nerve injuries, fractures requiring open reduction, and >3 pins. Splinting more than 3 weeks following pin removal also delays recovery.

Significance: At-risk populations for delay in return to elbow function include flexion-type injuries or nerve injuries and those requiring open reduction or more than 3 pins for fixation. Once pins are removed, immobilization in a splint for more than three additional weeks, even if motion is permitted when the splint is removed, can delay recovery as well.

Development of a new classification for forearm involvement in patients with multiple hereditary exostosis (MHE) using the Delphi process

Carley Vuillermin, Maria F. Canizares, CoULD Study Group

Boston Children’s Hospital, Boston, MA, USA

LOE-Diagnostic-Level V

Purpose: Classification schemes of forearm deformity in multiple hereditary exostosis (MHE) were developed to guide surgical care in retrospective case series and have been shown to have low reliability in a prospective population. The purpose of this study is to develop a widely applicable classification system using the Delphi consensus technique.

Methods: Fifteen experienced pediatric orthopedic hand surgeons from the Congenital Upper Limb Registry (CoULD) Study Group undertook seven rounds of online surveys to establish consensus. Investigators (CV, MC) analyzed results and provided structured feedback between rounds, without direct contact with experts. A research associate administered expert surveys and anonymized results to investigators. Round 1 began with an open-ended questionnaire to identify the clinical features of forearm MHE experts believe may predict outcomes. Round 2 and beyond aimed to achieve consensus. Consensus agreement was defined as ≥70% responses of either “strongly agree” or “agree” for a 7-point Likert-type or “agree” in a dichotomous question.

Results: In Round 1, a total of 10 patients and radiological characteristics were identified as themes for inclusion: (1) age at diagnosis; (2) family history; (3) genetic diagnosis; (4) symptoms (pain, stiffness, activity limitations); (5) clinical exam findings (visible deformity, wrist deviation, ROM limitation); (6) location of the osteochondroma; (7) number of osteochondroma; (8) bone alignment and shortening; (9) joint alignment; and (10) progression of the deformity. Round 2 consisted of a 7-point Likert-type scale. At this round, radial head dislocation, bone segment, ulnar shortening, radial alignment, distal radial inclination, distal radio ulnar joint (DRUJ), radial shortening, and ulnar variance were the items that reached the agreement threshold. In Round 3, experts were asked to choose between 3 or 4 categorical levels for the classification. Rounds 4 and 5 proposed a mild/moderate/severe classification and asked about the placement of each characteristic. In Round 6, experts asked to classify nine radiographs with the proposed classification and indicate whether this version fully classified the deformity. Closing criteria were met in the seventh and final round with 93% agreement. The proposed classification is presented in Table 1. Response rate was 100% in all rounds, with a range between 1 and 35 days to responses.

Conclusions: The new CoULD MHE forearm classification developed by experts in the field of pediatric hand surgery incorporates three categories of forearm involvement: Type 3-Severe forearm involvement, Type 2-Moderate forearm involvement, and Type 1-Mild forearm involvement.

Significance: This work represents the first step in the validation process of a clinically applicable and inclusive classification for patients with forearm MHE.

Flippin’ out over gymnast wrist: presentation and treatment of distal radial physeal stress syndrome in young gymnasts

David Peter VanEenenaam, Scott J. Mahon, Naomi Brown, Joseph Yellin, Apurva S. Shah

The Children’s Hospital of Philadelphia, Philadelphia, PA, USA

LOE-Not Applicable-Level IV

Purpose: Gymnast wrist is a niche injury experienced primarily by young gymnasts secondary to repetitive load bearing on the distal radius. Patients report insidious-onset wrist pain with physeal tenderness on examination and characteristic physeal widening on radiographs. Treatment typically involves prolonged immobilization, posing challenges for an otherwise highly active population. Current literature is limited when describing the overall treatment outcomes of gymnast wrist. This study aimed to understand the demographics, treatment outcomes, and return-to-sport timelines in a large patient sample.

Methods: Pediatric patients from 2016 to 2023 were identified using ICD-10 codes for “gymnast wrist” and wrist pain. Adults, acute fractures, non-gymnasts, and congenital etiologies were excluded. We collected patient data, Child Opportunity Index (COI), activity, treatment, radiographs, and outcomes. Statistical tests included chi-square, t-test, and Wilcoxon rank-sum analysis.

Results: A total of 78 wrists in 62 patients (60, 97% female) with an average age of 11.8 ± 1.7 (range, 6.9–15.4) were included. Average time to presentation was 12.2 weeks after symptom onset. All patients presented with wrist pain and tenderness over the physis, including 35.5% with symptoms on the dominant side, 38.7% on the nondominant side, and 25.8% with bilateral symptoms. The affected individuals averaged 15.9 ± 9.1 h of gymnastics participation per week. On initial wrist radiographs, 43 (69%) had distal radial physeal widening, 45 (73%) had juxtaphyseal sclerosis, and 12 (19%) had positive ulnar variance. For initial treatment, 55 (89%) were treated with splinting/rest/avoidance of weight-bearing with or without PT, 5 (8%) with short arm cast, and 2 (3%) underwent surgery. Thirty-two (53%) patients had resolved tenderness and return to sport after initial non-operative treatment. Among these 32 patients, average time to resolved tenderness was 7.4 weeks and average time to return to sport was 8.6 weeks. Approximately 23% of patients who initially improved with non-operative treatment later developed recurrent symptoms requiring a new period of rest. No significant differences were found in age, height, weight, body mass index (BMI), activity frequency, or prescribed treatment duration between patients whose symptoms resolved versus those whose symptoms persistent/recurred. Seven (11%) patients progressed to growth arrest, all of whom were eventually treated with ulnar shortening osteotomy.

Conclusions: Only about half of the patients with gymnast wrist respond well to conservative treatment. Recurrence rates are high, necessitating extended breaks from competitive athletics. More than 10% of patients progress to growth arrest requiring surgery.

Significance: Further prospective research is needed to identify risk factors for persistent or recurrent wrist pain in gymnasts.

Cerebral palsy in the British Orthopedic Surgery Surveillance Study (CPinBOSS)

Marie-Caroline Nogaro, Julie Stebbins, Daniel Christopher Perry, Tim Theologis

University of Oxford, Oxford, UK

LOE-Not Applicable-Level IV

Purpose: Musculoskeletal deformity in ambulant children with cerebral palsy (CP) is often addressed with single-event multi-level surgery (SEMLS). Despite its wide use, there is weak evidence on the effectiveness of SEMLS in improving gait and function, and variation in surgical practices and clinical indications. The aim of this study was to undertake national surveillance of current SEMLS practice in children with CP in the United Kingdom.

Methods: A national prospective cohort study of diplegic children (5–16 years) with Gross Motor Function Classification System (GMFCS) level I–III was conducted. Between April 2019 and March 2022, children who attended any hospital providing SEMLS in the United Kingdom, and who were deemed eligible to undergo SEMLS were recruited. Routine clinical data including GMFCS level, gait analysis data as summarized by the Gait Profile Score (GPS), motor function (FMS), and details of the surgical interventions were collected. Patient-reported outcome measures (PROMs) were assessed at baseline and 1 year post recruitment or surgery in a subgroup of consented participants.

Results: A total of 202 children were recruited over a 36-month period. In total, 152 children (75%) underwent surgery within the study period. The mean GMFCS, GPS, and FMS (5 m, 50 m, 500 m) scores at baseline for these children were 2.47 (0.68), 15.7 (4.8), and 4.01 (1.64), 3.09 (1.75), and 2.38 (1.76), respectively. In surgical patients, there was an overall improvement of GPS and FMS scores at 1 year (12.6 (3.9) and 4.14 (1.39), 3.50 (1.57) and 2.68 (1.84)), and the GMFCS level remained stable. Similarly, in non-surgical patients, there was no change in GMFCS level (2.56 (0.55) versus 2.71 (0.49)), but contrary to the surgical group, there was no overall improvement to GPS and FMS scores at 1 year. Of the 72 patients who completed the baseline PROM (in the consented cohort), 53% completed this at the 1-year follow-up. In surgical patients, there was overall improvement of GOAL at 1 year, but no change in the non-surgical patients.

Conclusions: The results from this national study in the United Kingdom indicate that SEMLS is effective in improving gait pattern, as measured by the GPS and patient-reported functional mobility as measured by the GOAL at 1 year from SEMLS.

Significance: This study has assessed current SEMLS practice in the United Kingdom and laid the foundation for the development of large-scale randomized trials on the management of this complex patient population.

Health-related quality of life in ambulatory children with physical disabilities

Chris Church, Sana Patil, Stephanie Butler, Freeman Miller, Jose De Jesus Salazar-Torres, Nancy Lennon, M. Wade Shrader, Maureen Donohoe, Faithe Rassias Kalisperis, Stuart Mackenzie, L. Reid Boyce Nichols, Nemours Gait Lab

Nemours A.I. Dupont Hospital for Children, Wilmington, DE, USA

LOE-Prognostic-Level III

Purpose: Health-related quality of life (HRQOL) is defined as “perceived wellbeing in physical, mental, and social domains of health.” This study aimed to evaluate HRQOL in children with the most common physical disabilities and examine its relationship with gross motor ability.

Methods: In this institutional review board (IRB)-approved retrospective study, the parent-reported Pediatric Outcomes Data Collection Instrument (PODCI) and Section D of the Gross Motor Function Measure (GMFM-D) were administered to ambulatory children aged 2–18 with a diagnosis of cerebral palsy (CP; GMFCS II), arthrogryposis, achondroplasia, or Morquio syndrome during clinical visits to the Gait Laboratory. The PODCI assesses perceptions across six domains associated with mobility, pain, and happiness. It is validated for children (age 2–10 years) and adolescents (age 11–18 years). PODCI results were compared to published norms (TDY) using T-tests. The relationship between PODCI and GMFM-D scores was analyzed with Pearson correlations. To examine differences in HRQOL between diagnoses, pairwise comparisons of PODCI scores using the Wilcoxon rank-sum test with Bonferroni corrections was conducted.

Results: Children and adolescent groups within all four diagnoses demonstrated limited mobility and higher pain compared to TDY (p < 0.015). Happiness was lower in children and adolescents with CP, arthrogryposis multiplex congenital (AMC), and children with Morquio syndrome compared to TDY (p < 0.002). Among diagnostic groups, global function was higher (p < 0.0001) in people with achondroplasia (83(2)), compared to people with AMC (73(3)), CP (72(2)), and Morquio syndrome (66(4)). Global functioning was higher in people with AMC (p < 0.05) and CP (p < 0.01) compared to Morquio. Despite functional differences, there were no significant differences between the four diagnoses in pain (p > 0.10). Happiness was lower in people with CP (80(3)) compared to achondroplasia (90(3); p = 0.01). GMFM-D was associated with the PODCI mobility domains for all diagnoses (r = 0.31 to 0.79, p < 0.03), but was not correlated with the happiness domain for any group (r = −0.16 to 0.092; p > 0.14). GMFM-D and PODCI pain scores were correlated only in children with achondroplasia (r = 0.355; p < 0.001).

Conclusions: Significant limitations in HRQOL are seen in children with the most common physical disabilities of childhood. Pain is present at higher levels than non-disabled peers and tends not to be related to low motor function. Happiness was also not associated with gross motor function, suggesting the need to examine other contributors when mental health concerns exist in children with physical disabilities.

Significance: It is essential to utilize patient-reported outcomes to best understand and assist in the management of HRQOL in children and adolescents with lifelong physical disabilities.

What is the prevalence of depressive symptoms and antidepressant use among adult patients with cerebral palsy?

Michael G. Vitale, Chun Wai Hung, Daniel Linhares, Afrain Z. Boby, Hiroko Matsumoto, Joshua E. Hyman, David P. Roye

Columbia University Medical Center, NY, USA

LOE-Prognostic-Level III

Purpose: Depressive symptoms are reported to be higher for individuals with cerebral palsy (CP); however, as patients with CP often lack effective means of communication, the prevalence of depression may be underestimated. The objective of this study was to better understand the prevalence of depression in the CP population as determined by self-reported depressed mood and/or prescribed antidepressants (as a surrogate measure in patients unable to provide self-report).

Methods: This retrospective cohort study examined adults (≥18 years) with CP seen in an academic medical center from 2006 to 2016. The associations between depression and patient characteristics (demographics, comorbidities, gross motor function, medication) were analyzed to determine any association with depression. We also identified commonly used antidepressants.

Results: In total, 101 of 501(20.2%) CP patients (31+ 13 years, 53% F) met the criteria for depression. Of the patients without CP, 31/240 (13.1%) met the criteria for depression. The depressed and non-depressed groups did not differ in age, weight, height, body mass index (BMI), or gender. There was no significant difference in depression between ambulatory and non-ambulatory patients (21% versus 19%, p = 0.7). GI comorbidity, stimulants, antipsychotics, analgesics, benzodiazepines, and antispasmodics use were associated with higher rates of depression (p < 0.05). Selective serotonin reuptake inhibitors (SSRIs) were the most prescribed antidepressants.

Conclusions: One of five adult CP patients in our sample met the criteria for depression, a substantially greater rate than that identified through self-report alone. Although ambulatory status was not found to be associated with depression, non-ambulatory patients are more likely to have difficulty communicating their symptoms, which could confound estimates of depression prevalence in this population.

Significance: The significance of this study lies in uncovering a potentially underestimated prevalence of depression among individuals with CP. Our study reveals that 20.2% of adult CP patients met the criteria for depression—significantly higher than estimates based solely on self-report. The findings emphasize the limitations of relying on self-disclosure in a population where effective communication may be challenging. In addition, the study identifies associations between depression and various patient characteristics, shedding light on factors that may contribute to depressive symptoms in this population. The results underscore the critical importance of employing comprehensive screening measures beyond self-report to accurately assess and address depression in individuals with CP.

Accuracy and reliability of mobile app–enhanced observational gait analysis in children with cerebral palsy

Donald T. Kephart, Jon R. Davids, Vedant A. Kulkarni

Shriners Children’s Northern California, Sacramento, CA, USA

LOE-Diagnostic-Level I

Purpose: Three-dimensional gait analysis (3DGA) is the gold standard for quantifying gait deviation in children with cerebral palsy (CP). Where 3DGA is unavailable, observational gait analysis using the Edinburgh Visual Gait Score (EVGS) has shown moderate reliability and accuracy. The addition of mobile app slow-motion video analysis may improve the accuracy and reliability of this technique. This study prospectively evaluates the accuracy and reliability of mobile app–enhanced observational gait analysis (mOGA) when compared to the gold standard of 3DGA in children with CP.

Methods: Slow-motion gait video was captured using two Apple iPhone 8S units while simultaneous 3DGA was acquired using a 12-camera infrared system. Using a low-cost commercially available slow-motion video analysis mobile app (Dartfish Express) on an iPad Pro, two observers (orthopedic surgery resident and attending pediatric orthopedic surgeon) made 11 quantitative mOGA measurements per limb. Inter-class correlation coefficient (ICC) was used to compare reliability between observers and between methods. Pearson’s correlation coefficient and Spearman’s rank correlation coefficient were used to determine correlation of error magnitude with Gait Profile Score and Gross Motor Function Classification System (GMFCS), respectively.

Results: Thirty-five subjects with CP were recruited for the study (26 M, 9 F; GMFCS I = 7; GMFCS II = 21, GMFCS III = 7; mean = 12 years, range = 6–18 years). Subjects had significant gait deviation as measured by the Gait Profile Score (GPS), with a mean GPS of 12.55 (SD = 3.97). There was excellent overall reliability between raters for the mOGA technique (ICC = 0.964). Of the 11 mOGA measurements, inter-rater reliability was good to excellent (ICC > 0.75) for nine measurements and moderate (ICC = 0.5–0.75) for two measurements. When compared with the gold standard of 3DGA, mOGA had excellent reliability for both the less experienced (ICC = 0.927) and more experienced (ICC = 0.947) rater. Reliability between mOGA and 3DGA was good to excellent for six measurements, with no measurement having poor reliability (ICC < 0.5) and all measurements having acceptable error magnitude. The error magnitude was not correlated to GPS or GMFCS (p > 0.05), indicating that the technique could be used in patients with increased gait deviation and functional impairment.

Conclusions: mOGA has good reliability between raters and acceptable accuracy compared with 3DGA.

Significance: mOGA in children with CP can allow for enhanced communication between providers using low-cost technology. Further studies should clarify the role of mOGA in assisting with treatment decision-making and outcome assessment in environments where 3DGA is not possible or available.

The risk factors associated with increased anterior pelvic tilt in ambulatory children with cerebral palsy

Chris Church, Jose De Jesus Salazar-Torres, Tanmayee Joshi, Nancy Lennon, Thomas Shields, John Henley, Freeman Miller, M. Wade Shrader, Jason Howard

Nemours Children’s Health, Wilmington, DE, USA

LOE-Not Applicable-Level III

Purpose: Anterior pelvic tilt (APT) is common in children with cerebral palsy (CP) and may be associated with excessive lumbar lordosis, back pain, and functional limitations. Factors related to increased APT are unclear but may be associated with hamstring length, proximal muscle strength, sagittal plane gait deviations, or orthopedic surgery. This study aims to determine risk factors for pathologic change in APT in ambulatory children with CP.

Methods: In this institutional review board (IRB)-approved retrospective study, ambulatory children with bilateral spastic CP were identified from our institutional gait lab database from 2002 to 2022. Pairs of gait analyses were used to measure change in APT with and without surgical interventions. A mixed-effects linear regression analysis was conducted to predict which patient and surgical factors influenced change in APT between gait analyses. A logistic mixed-effects generalized linear regression was also performed to identify factors associated with an APT increase of 5° or 10°.

Results: In total, 528 children with CP were analyzed, 422 in the surgical group (GMFCS level I (n = 35), II (n = 281), III (n = 106)) and 106 in the non-surgical group (GMFCS level I (n = 17), II (n = 64), III (n = 25)), with mean ages of 11.1 ± 3.8 and 10.0 ± 3.6 years at first gait analysis and of 13.2 ± 6.2 and 12.3 ± 3.8 years at the follow-up gait analysis, respectively. Both the surgical and non-surgical groups exhibited APT at last follow-up, at 20° ± 9° and 20° ± 9°, respectively (p = 0.63). Prior rhizotomy and hamstring lengthening (medial, medial + lateral) were significant factors for increasing APT, while Achilles lengthening, jump gait pattern, increased stance hip flexion, and high initial APT, were related to decreasing APT (Table 1). The mean change in APT after medial and medial + lateral hamstring lengthening was 1° ± 9° (p = 0.046) and 4° ± 9° (p < 0.0001), respectively. The mean change in APT in children with history of rhizotomy was 2° ± 11° (p = 0.44).

Conclusions: Rather than being solely related to hamstring lengthening, factors associated with changes in APT were multifactorial, with prior rhizotomy having the most profound effect. Contrary to conventional thinking, the presence of high initial APT was not a risk factor for further progression. Factors associated with ankle equinus—jump gait and Achilles lengthening—also had a protective effect.

Significance: Identified risk factors for change in APT should be considered when developing treatment programs for children with bilateral CP, particularly for those with a normal or retroverted pelvis given their inherent susceptibility to worsening pelvic tilt.

The association between increase in knee range of motion and patient satisfaction after rectus femoris transfer in cerebral palsy

Mauro Cesar Morais Filho, Marcelo Hideki Fujino, Catia Miyuki Kawamura, Jose Augusto Fernandes Lopes

Ageu Saraiva AACD, São Paulo, Brazil

LOE-Prognostic-Level III

Purpose: The purpose of this study was to analyze the association between patient satisfaction and increase in knee range of motion (KRM) after rectus femoris transfer (RFT) in cerebral palsy (CP).

Methods: Ninety-eight patients with spastic diplegic CP, Gross Motor Function Classification System (GMFCS) I–III, and who underwent RFT, with pre- and postoperative gait analyses, were included in the study. They were divided into three groups according to the change in KRM after RFT: (1) Group reduction (GR): reduction > 1 standard deviation (SD); (2) Group unchanged (GU): reduction or increase ≤ 1 SD; and (3) Group improvement (GI): increase > 1 SD. Demographic data, kinematics, Gait Deviation Index (GDI), and self-reported questionnaire were analyzed, and the results compared among groups.

Results: The mean age at surgery and the follow-up time were 14.2 and 3.04 years, respectively. After surgical intervention, the KRM was reduced in 6 (6.1%), unchanged in 69 (70.4%), and increased in 23 (23.5%) patients. No comparisons were made with GR due to the small number of patients in this group (6). Age at surgery (GU 14.3 years/GI 14.3 years), follow-up time (GU 3.2 years/GI 2.7 years), GMFCS distribution, GDI increase (GU 8.1/GI 11), increase in peak knee flexion in the swing phase (GU 6.40/GI 7.30), and change in Gillette Functional Assessment Questionnaire walking ability score (GU + 0.8/GI −0.1) showed no difference when comparing GU and GI. The minimal knee flexion in stance phase was reduced (−14.30) in GI and increased (+4.80) in GU (p < 0.001). The prevalence of other surgical procedures performed concomitant to RFT was also similar between GU and GI, except for hip adductor surgical lengthening, which was observed in 13% of the patients in GU and 34.8% in GI (p = 0.02). Finally, 24.6% of the patients from GU and 13% from GI were extremely satisfied with the treatment results (p = 0.243), whereas 34.8% of the patients in both groups were satisfied. According to the parents’ perspective, 23.2% from GU and 13% from GI were extremely satisfied (p = 0.298), whereas 36.2% from GU and 39.1% from GI were satisfied with the outcome (p = 0.803).

Conclusions: In the present study, the increase in KRM after RFT was related to reduction of knee flexion in stance phase. The improvement in KRM did not generate higher levels of satisfaction with treatment.

Significance: The outcomes after RFT in CP have been frequently analyzed using tridimensional gait analysis. The evidence regarding patient satisfaction regarding this procedure remains scarce.

Is the CPCHILD questionnaire responsive—assessing HRQoL changes and performance of the CPCHILD after hip and spine surgery in children with severe cerebral palsy

Lennert Plasschaert, Patricia E. Miller, Rachel DiFazio, Brian D. Snyder, Colyn Watkins, Travis Matheney, Benjamin J. Shore

Boston Children’s Hospital, Boston, MA, USA

LOE-Not Applicable-Not Applicable

Purpose: Non-ambulant (Gross Motor Function Classification System (GMFCS) IV and V) children with cerebral palsy (CP) undergo surgical interventions designed to manage pain, facilitate seating, and improve their health-related quality of life (HRQoL). The CPCHILD questionnaire was specifically designed to measure HRQoL in children with severe non-ambulatory CP and has been shown valid and reliable. However, the responsiveness and sensitivity to change of all CPCHILD domains has yet to be demonstrated in this population. Therefore, the purpose of this study was to report the responsiveness of the CPCHILD in non-ambulant children with CP undergoing hip and spine surgery.

Methods: Parents or caregivers of non-ambulant children with CP undergoing spine or hip surgery at our hospital between 2011 and 2019 completed the CPCHILD questionnaire at baseline and 6, 12, and 24 months post-operative. This prospectively collected cohort was used to assess the responsiveness of the CPCHILD domains, as measured by the standard response mean (SRM) and the effect size (ES). Sensitivity to change was assessed by the standard error of measurement (SEM) and the minimal detectable change at the 90% confidence level (MDC90) from preoperative measurement to each follow-up measurement.

Results: One-hundred and eight patients with CP (38% GMFCS level IV; 62% level V) underwent spine or hip surgery at a mean age of 11.5 years (SD = 4.2). There were no statistically significant differences in cohort characteristics at each follow-up measurement. At 24 months of follow-up, significant score improvements were seen in the personal (6.5, SD = 21.2; 95% confidence interval (CI) = 1.6–11.5; p = 0.01), positioning (7.8, SD = 15.7; 95% CI = 4.1–11.5); p < 0.001), and health (4.9, SD = 20.0; 95% CI = 0.3–9.5; p = 0.04) domains. A small ES and SRM were detected in the personal domain (ES, 0.39; SRM, 0.44), a moderate ES and SRM were detected in the positioning domain (ES, 0.54; SRM, 0.70), and a small ES and SRM were detected in the health domain (ES, 0.28; SRM, 0.32). The SEM and MDC90 for the personal domain were, respectively, 4.7 and 9.2 points, for the positioning domain 4.1 points and 8.1 points, and for the health domain 4.9 points and 9.6 points.

Conclusions: The CPCHILD questionnaire is currently our best tool to assess HRQoL in children with severe CP; however, we found only the Personal, Positioning, and Health domains to be responsive to change after spine and lower extremity orthopedic surgery.

Significance: Surgeons should interpret CPCHILD scores with caution as not all domains are responsive to change after orthopedic surgery.

Disease-modifying therapy changed the natural course of spinal muscular atrophy type 1: what about spine and hip?

Niyazi Erdem Yasar, Guzelali Ozdemir, Elif Uzun Ata, Naim Ata, Mahir Mustafa Ülgü, Ebru Dumlupinar, Suayip Birinci, Izzet Bingöl, Senol Bekmez

Ankara Bilkent Children’s Hospital, Ankara, Turkey

LOE-Therapeutic-Level IV

Purpose: Spinal muscular atrophy (SMA) type 1 has a devastating natural course. Scoliosis and hip subluxation are common in non-ambulatory patients with SMA. Nusinersen, the first Food and Drug Administration-approved therapy for SMA, has not only prolonged survival, but has also improved motor function in SMA type 1. However, the impact of Nusinersen treatment on the development of spine and hip deformities remains unclear.

Methods: We conducted a retrospective electronic health record database review. We included patients with SMA type 1 born between 2017 and 2021, the diagnosis confirmed by genetic testing and received intrathecal Nusinersen (Spinraza®) therapy. Patients having <2 years radiological follow-up, inadequate clinical or radiographic data, first Nusinersen dose >6 months old, and SMA treatment other than Nusinersen were excluded. We obtained demographic parameters, age at first Nusinersen dose, total number of intrathecal Nusinersen administrations, and Children’s Hospital of Philadelphia Infant Test of Neuromuscular Disorders (CHOP-INTEND) assessment scores of motor function. Radiological evaluation was also performed to assess parasol rib deformity, scoliosis, pelvic obliquity, and hip subluxation.

Results: We included 29 patients, mean age 3.7 ± 1.1 years (range, 2–6 years), male to female ratio 1.07. Mean number of intrathecal Nusinersen administrations was 8.9 ± 2.9 (range, 4–19). There was a significant correlation between CHOP-INTEND score and number of Nusinersen administrations (r = 0.539, p = 0.05). Correlations between CHOP-INTEND score and patient age (r = 0.361) or age at first Nusinersen dose (r = 0.39) were not significant (p = 0.076 and p = 0.054, respectively). 93.1% of the patients had scoliosis, 69% had pelvic obliquity, and 60.7% had hip subluxation, after a mean 3.3 ± 1.1 years of follow-up (range, 2–6). Distribution of age, total number of Nusinersen administrations, age at first Nusinersen dose, and CHOP-INTEND scores were the same across patients with or without scoliosis, pelvic obliquity, and hip subluxation. There was also no significant correlation between these parameters and progression rate of scoliosis, hip subluxation, or pelvic obliquity.

Conclusions: Prior to the introduction of disease-modifying therapies, management of musculoskeletal deformities was rarely discussed in SMA type 1 due to limited survival. Although Nusinersen treatment changed the natural course of SMA type 1 in terms of overall survival and motor function, progressive scoliosis and hip subluxation remain significant problems in most cases.

Significance: Significant improvements in SMA type 1 regarding overall survival and motor function in the disease-modifying therapy era should encourage us to reconsider our management algorithms for treating spine deformity and hip subluxation, from only observation toward surgical reconstruction.

Increased knee range of motion in patients with arthrogryposis: minimum 2-year follow-up

Aaron Huser, Michael William Brown, Arun R. Hariharan, Hans K. Nugraha, David S. Feldman

Paley Orthopedic and Spine Institute, West Palm Beach, FL, USA

LOE-Therapeutic-Level IV

Purpose: Arthrogryposis describes several conditions characterized by multiple congenital joint contractures. Surgical interventions for knee flexion contractures have achieved extension to redirect the arc of motion and improve ambulation but have not demonstrated maintained increases in total range of motion (ROM) over time. The purpose of this study is to review our patients’ latest ROM and ambulation status that underwent posterior knee release, proximal femoral shortening, and peroneal nerve decompression.

Methods: A retrospective chart and radiographic review was performed on patients presenting with arthrogryposis from 2016 through September of 2021 who underwent the above procedure. Patients were included if they had a minimum 2-year follow-up. Charts were reviewed for preoperative and latest follow-up ROM, ambulation status, and any complications. An immediate postoperative ROM was also reviewed to determine if there were any decreases over time. Group comparisons were performed with Friedman’s test and pairwise comparisons were performed using Dunn’s test. Fisher’s exact test was used to compare preoperative and most recent follow-up ambulation.

Results: Twenty-nine patients with 51 limbs were included in the final analysis. The mean age at surgery was 6.0 years (± 3.4 years). The mean follow-up was 40.5 months (± 15.1 months). The median preoperative flexion deformity was 45° (interquartile range (IQR) = 20°) and this improved to 3° (IQR = 12°) (p < 0.0001). The mean preoperative arc of motion was 50° (IQR 30°) and this improved to 85° (IQR 30°) (p < 0.0001). No changes in ROM occurred between immediate postoperative ROM and most recent follow-up (Figure 1). Thirty-eight percent of patients were able to ambulate preoperatively and this improved to 93% at most recent follow-up (p < 0.0001). Sixteen of 29 patients had at least limited ability to ambulate in the community, and 11/29 were home ambulators. Ten patients experienced a complication during the follow-up period. Five patients had postoperative neuritis which was treated conservatively with gabapentin and eventually weaned off. Five patients sustained fractures: three intraoperatively, requiring pinning of the distal femoral physis, one in the acute postoperative period, which was treated conservatively, and one patient sustained a femur fracture at 29 months and required fixation in the operative theater.

Conclusions: Posterior knee release with proximal femoral shortening and peroneal nerve decompression achieved increased and maintained ROM in patients with knee flexion deformities at a minimum 2-year follow-up. In addition, over 90% were at least home ambulators.

Significance: This study demonstrates maintained knee ROM and ambulation improvements at a minimum 2-year follow-up using a novel technique.

Obesity-related alterations in capital femoral epiphysis morphology: an extensive analysis of 8717 hips utilizing automated 3D-CT imaging

Eduardo Novais, Mohammadreza Movahhedi, Mallika Singh, Nazgol Tavabi, Shanika De Silva, Sarah D. Bixby, Ata M. Kiapour

Boston Children’s Hospital, Boston, MA, USA

LOE-Diagnostic-Level IV

Purpose: The epiphyseal tubercle provides stability to the capital femoral epiphysis and acts as a rotational pivot point in slipped capital femoral epiphysis (SCFE). A posterior tilt of the epiphysis (retroversion) is recognized as a crucial factor in SCFE development. This study investigates the impact of childhood obesity on the morphology of the capital femoral epiphysis in children without pre-existing hip disorders.

Methods: After institutional review board (IRB) approval, we included patients aged 7–19 years who had undergone pelvic computed tomography (CT) from 2012 to 2022. Inclusion criteria were no documented bone/joint disorders and quality for three-dimensional (3D) segmentation. We applied validated custom software (VirtualHip) for automated femoral head and neck segmentation. We measured epiphyseal tubercle height, peripheral cupping (12 o’clock position), and epiphyseal tilt. Bilateral measurements were averaged. We modeled the association between body mass index (BMI) and each measurement using linear models. We compared the mean differences in the measurements for each BMI group at each age using t-tests and corrected the p-values for multiple testing using the Bonferroni method.

Results: Our study encompassed 8717 patients, equally distributed across the sexes. The average BMI percentile was 61% ± 33%, with 24% classified as obese (BMI >= 95th percentile). After adjusting for age and sex, obese subjects had reduced epiphyseal tubercle height (estimated differences: −0.6% (95% confidence interval (CI): −0.8 to −0.4; p < 0.001)), a more posteriorly tilted epiphysis (estimated difference: −1.6% (95% CI: −2.2 to −1.0)), and decreased peripheral cupping (estimated difference: −0.3% (95% CI: −0.6 to −0.1)) compared to non-obese counterparts. Figure 1 shows trends concerning the measurements across all age groups. A smaller epiphyseal tubercle and a more posteriorly tilted epiphysis in obese children were consistent findings across all age groups until approximately age 14 (Table 1). Peripheral cupping was statistically lower in obese compared to non-obese individuals after age 13.

Conclusions: In this comprehensive automated 3D analysis of pelvic CT scans encompassing 8717 subjects, our observations underline that obese children exhibit a smaller epiphyseal tubercle, greater posterior tilt of the epiphysis, and reduced peripheral cupping compared to their non-obese counterparts.

Significance: Our study describes the impact of obesity on capital femoral epiphysis development in children. The reduced epiphyseal tubercle and diminished peripheral cupping observed in obese children are morphological features known to compromise epiphyseal stability, while posteriorly tilted epiphysis increases shear forces on the growth plate, thereby increasing the risk of slip. These findings contribute to our understanding of the pathomechanics underlying the association between obesity and SCFE.

Intraoperative perfusion monitoring does not reliably predict osteonecrosis following treatment of unstable SCFE

Bridget Ellsworth, Julianna Lee, Wudbhav N. Sankar

Children’s Hospital of Philadelphia, Philadelphia, PA, USA

LOE-Prognostic-Level IV

Purpose: Avascular necrosis (AVN) remains the most dreaded complication of unstable slipped capital femoral epiphysis (SCFE) treatment. Newer closed reduction techniques (with perfusion monitoring) have emerged as a technically straightforward means to address residual SCFE deformity while still minimizing the risk of osteonecrosis. However, limited data exist regarding the reliability of perfusion monitoring to predict the development of AVN. The purpose of this study was to evaluate this reliability.

Methods: We retrospectively reviewed all patients with unstable SCFE who underwent closed or open reduction with epiphyseal perfusion monitoring using an intracranial pressure (ICP) probe from 2015 to 2023 at a single institution with minimum 6-month radiographic follow-up. Demographic, clinical, and radiographic data were recorded, including duration of symptoms, type of reduction, capsulotomy performed, presence of a waveform on ICP monitoring after epiphyseal fixation, and development of AVN on follow-up radiographs.

Results: Our cohort included 33 hips (32 patients). The average age was 12.5 ± 1.8 years, with a median follow-up of 13.9 months. Eleven hips were treated with open reduction using the modified Dunn technique (10 hips) or anterior approach (1 hip), and 22 hips were treated with inadvertent (5 hips) or purposeful closed reduction using the Leadbetter technique (17 hips). Overall, 7 of 33 hips (21.2%) developed AVN, 5 of which (16.7%) had a pulsatile waveform intraoperatively on perfusion monitoring. The incidence of AVN after closed reduction with a detectable waveform was 25% (5 of 20 hips). There was no significant association between time to surgery (p = 0.273) or type of reduction (p = 0.378) and incidence of AVN.

Conclusions: In this series, the presence of a pulsatile waveform on intraoperative epiphyseal perfusion monitoring during reduction of unstable slips did not preclude the development of AVN.

Significance: To our knowledge, this is the first study to report development of AVN after demonstrable intraoperative epiphyseal perfusion following closed reduction of unstable slips.

Bone scintigraphy can predict post-operative femoral head avascular necrosis in children with hip trauma and slipped capital femoral epiphysis

Patrick Curran, Katharine Hollnagel, James David Bomar, V. Salil Upasani

Rady Children’s Hospital, San Diego, CA, USA

LOE-Diagnostic-Level III

Purpose: Femoral head avascular necrosis (AVN) is a devastating complication that can occur in the setting of trauma to the hip. Bone scintigraphy (BoS) can be used to evaluate femoral head perfusion, but current evidence in support of its use in the pediatric population is lacking. The purpose of this study was to test the hypothesis that BoS would have high sensitivity and specificity to accurately diagnose femoral head AVN.

Methods: We performed a retrospective review of patients who underwent BoS to assess femoral head perfusion following treatment of traumatic femoral neck fracture, hip dislocation, or slipped capital femoral epiphysis (SCFE) at a single pediatric tertiary referral center. All patients had minimal clinical follow-up of 1 year. Assessment of intra-operative femoral head perfusion was also recorded if performed. Results of BoS and intra-op assessments were compared to radiographic findings of AVN at final follow-up.

Results: Thirty patients (47 hips) (23 M, 7 F; age: 12.8 ± 1.6 years; follow-up: 24.5 ± 15.3 months) were included. Radiographic AVN (RAVN) developed in 13% of hips. Seven hips had an abnormal BoS: 6 hips had absent perfusion and 1 hip had diminished perfusion. All six hips with absent perfusion went on to develop RAVN with collapse. The hip with diminished perfusion did not go on to femoral head collapse. No hip with normal BoS developed radiographic evidence of AVN at final follow up. Twenty-one hips underwent intra-op perfusion assessment, 15/21 were evaluated with femoral head drilling with 3/15 indicating absent perfusion. Fourteen of 21 were evaluated with intraosseous pressure (IOP) monitoring with 2/14 indicating absent perfusion. Eight of 21 hips were evaluated with both techniques with concordant absent perfusion in 1/8 hips, and this finding of absent perfusion was not noted on BoS and this hip did not develop RAVN. BoS demonstrating absent perfusion had a sensitivity of 100% and a specificity of 98% in predicting RAVN. Evidence of femoral head perfusion at the time of surgery (drilling or IOP) had a sensitivity of 17% and a specificity of 79% in predicting RAVN.

Conclusions: BoS demonstrates high sensitivity and specificity for predicting RAVN and femoral head collapse postoperatively in the setting of trauma and SCFE compared to intra-operative assessments.

Significance: Clinicians should have increased suspicion for impending femoral head collapse if there is absent perfusion on BoS at 6–12 weeks post-op, even with evidence of preserved femoral head perfusion at the time of surgery. Early diagnosis of AVN can allow for earlier interventions to improve femoral head perfusion prior to femoral head collapse.

Epiphyseal stability increases specificity of the Loder classification system in prognosticating AVN after slipped capital femoral epiphysis

Katherine Sara Hajdu, Emilie Amaro Zoldos, Courtney Baker, Simone Herzberg, Benjamin Asbury, Stephanie N. Moore-Lotridge, Kevin Michael Dale, David Ebenezer, Nathaniel Lempert, Craig R. Louer, Jeffrey E. Martus, Gregory A. Mencio, Jonathan G. Schoenecker, Vanderbilt SCFE Study Group

Vanderbilt University Medical Center, Nashville, TN, USA

LOE-Not Applicable-Level III

Purpose: Avascular necrosis (AVN) of the femoral head is the leading cause of morbidity following slipped capital femoral epiphysis (SCFE). To identify patients at risk for AVN, Loder developed a classification system based on weight-bearing status (WBS). While sensitive in predicting which patients are at risk for AVN, determining the Loder classification can be subjective and lacks specificity for AVN. We hypothesized that only Loder-unstable patients with epiphyseal instability are at risk of AVN.

Methods: Retrospective chart review of SCFE patients who underwent closed operative treatment at a single institution from 2007 to 2022. Patient data, surgical details, and all imaging were analyzed. Loder status was determined by whether the patient could independently weight-bear (WB) at presentation. Epiphyseal instability was defined as a change from preoperative to postoperative Southwick angle (SWA) over 15 degrees with an operative note documenting a successful reduction, indicating there was relative motion between the epiphysis and metaphysis. The primary outcome was postoperative AVN rate. Statistical analysis was performed using R (version 4.3.1).

Results: There were 458 hips in 397 patients identified. Three hundred sixty-six (80%) hips were Loder-stable with a 0% AVN rate at median follow-up of 15 months (range, 7–35 months). Ninety-two (20%) hips were Loder-unstable, with an 11% (10/92) AVN rate at median follow-up of 16 months (range, 4–43 months). Of the Loder-unstable patients, 34% (31/92) were epiphyseal-stable and 66% (61/92) were epiphyseal-unstable. The AVN rate of the Loder-unstable/epiphyseal-stable patients was 0% (0/31) at a median follow-up of 12 months (range, 12–32 months). The AVN rate of the Loder-unstable/epiphyseal-unstable was 16% (10/61) at a median follow-up of 19 months (range, 4–48 months) (Table A). The addition of epiphyseal stability as a classifier significantly increased the specificity of predicting AVN (p = 0.005) (Table B).

Conclusions: The Loder classification is a sensitive preoperative screening tool for predicting risk of AVN following SCFE resulting in no false negatives. Epiphyseal stability, diagnosed intraoperatively, is an additional classifier that increases the specificity of predicting AVN risk without sacrificing sensitivity. Patients with confirmed epiphyseal instability are at the greatest risk of AVN and may benefit from perioperative and postoperative imaging or interventions.

Significance: The addition of epiphyseal stability to the Loder classification increases physicians’ ability to identify patients at risk for developing AVN. Intraoperative assessment of epiphyseal stability is essential for treatment algorithms and counseling families. The results help explain the wide range of AVN rates for Loder-unstable patients in the literature as epiphyseal stability has not been previously accounted for.

Rate and risk factors for contralateral slippage in adolescents treated for slipped capital femoral epiphysis: a comprehensive analysis of 3528 cases

David Momtaz, Rishi Gonuguntla, Aaron Singh, Mehul Mittal, Beltran Torres-Izquierdo, Pooya Hosseinzadeh

Washington University School of Medicine, Saint Louis, MO, USA

LOE-Prognostic-Level III

Purpose: After a unilateral slipped capital femoral epiphysis (SCFE), the contralateral hip is at risk for a subsequent SCFE. However, further information regarding risk factors involved in the development of contralateral hip SCFE must be investigated. The purpose of the study was to report the rate and risk factors for subsequent contralateral SCFE in adolescents treated for unilateral SCFE.

Methods: A case-control study utilizing aggregated multi-institutional electronic medical record (EMR) data between January 2003 and March 2023 was conducted. Skeletally immature children diagnosed with SCFE who underwent surgical management were included. Variables associated with contralateral SCFE were identified using multivariate logistic regression models that adjusted for patient characteristics and time of surgery, providing adjusted odds ratios. False discovery rate was accounted for via the Benjamini–Hochberg method.

Results: 15.3% of patients developed contralateral SCFE with a mean of 296.53 (± 17.23 SE) days and a median of 190 days following their initial SCFE. Increased thyrotropin (odds ratio (OR) = 1.43, p = 0.036), diabetes mellitus (OR = 1.67, p = 0.005), severe obesity (OR = 1.81, p < 0.001), history of human growth hormone (HGH) use (OR = 1.85, p < 0.001), low vitamin D (OR = 5.75, p < 0.001), younger age (boys under 12 years of age (OR = 1.85, p < 0.001), and girls under 11 years of age (OR = 1.47, p = 0.026)), and tobacco exposure (OR = 2.43, p < 0.001) were significantly associated with an increased odds of developing contralateral disease.

Conclusions: In the largest study on this topic, we identified the rate, odds, and risk factors associated with development of contralateral SCFE. We found younger age, hypothyroidism, severe obesity, low Vitamin D, diabetes mellitus, and history of HGH use as independent risk factors.

Significance: Our findings can aid clinical decision making in at-risk patients.

Temporary in situ pinning with subsequent modified Dunn is a safe alternative to primary modified Dunn

Jordyn Adams, Graham Whiting, Jordan Archer, Courtney Selberg

Children’s Hospital Colorado, Aurora, CO, USA

LOE-Therapeutic-Level II

Purpose: Surgical hip dislocation with modified Dunn (MD) is a viable treatment option for moderate to severe unstable slipped capital femoral epiphysis (SCFE) at an experienced tertiary care center. In the acute setting, an experienced surgeon may not be immediately available for MD. However, in situ pinning (ISP) is a well-accepted treatment in the acute setting and is more familiar to the on-call orthopedic surgeon. To date, no literature exists on the impact of temporary ISP prior to MD to compare its impact on outcome. Our primary aim was to compare complication rate and radiographic outcome at 2 years in patients with moderate/severe unstable SCFE undergoing primary MD versus ISP with subsequent MD.

Methods: Thirty-one patients underwent treatment for moderate/severe SCFE at a single institution. Demographics, surgical characteristics, and radiographic measurements were compared. Investigation of treatment outcomes was performed using linear regression models, adjusted for age and sex, to analyze continuous variables across ISP + MD and MD only treatment groups. Mean and standard deviation were calculated to summarize continuous variables, and the mean difference between treatment groups with a 95% confidence interval was calculated from the linear regression model for slip angle. Binary outcomes were explored through logistic regression models again adjusted for age and sex. The Wilcoxon rank-sum test assessed the significance of Merle d’Aubigné scores between the two groups.

Results: Fifteen patients underwent ISP with subsequent MD; 16 patients underwent primary MD. Average time between ISP and MD was 47.86 days. There was no difference in demographic characteristics, preoperative Southwick angle, alpha angle, or body mass index (BMI). Implant failure rate was 6.67% (ISP + MD) and 6.25% (MD), p = 0.094; both occurred at 3 months postoperatively and required upsizing of femoral neck implants without loss of initial correction. The incidence of AVN was not significantly different between groups; 6.67% ISP + MD versus 12.5% MD, p = 0.443. Merle d’Aubigné scores at final follow-up were 18 (ISP + MD) and 17 (MD), p = 0.581. There was no significant difference between groups in Tonnis grade, Southwick, or alpha angles at final follow-up.

Conclusions: We demonstrate no difference in complication rate in patients with moderate/severe SCFE treated with primary MD compared to temporary ISP with subsequent conversion to MD.

Significance: Temporary ISP for moderate/severe unstable SCFE may be a safe alternative to primary MD if an experienced hip surgeon is not immediately available. This also allows additional time for patients and their family to understand the risk/benefit ratio for the MD surgery prior to conversion.

Risk factors of vitamin D deficiencies on SCFE development

David Momtaz, Abhishek Tippabhatla, Rishi Gonuguntla, Mehul Mittal, Beltran Torres-Izquierdo, Pooya Hosseinzadeh, Zachary Meyer

Washington University School of Medicine, Saint Louis, MO, USA

LOE-Prognostic-Level III

Purpose: Slipped capital femoral epiphysis (SCFE) is one of the most common hip disorders in adolescents. Multiple metabolic disorders have been associated with secondary SCFE, and the association of obesity with primary SCFE has been well studied. While the incidence of vitamin D deficiency in childhood has been increasing, few studies have examined a potential link with SCFE. This study investigates the impact of vitamin D on SCFE development and related complications.

Methods: A large comprehensive national database was queried for patients under age 9 years who have records of calcidiol lab drawn between 9 and 18 years of age. These patients are followed up until SCFE occurrences or until the patient turns 18 years. Patients were divided into vitamin D adequate (calcidiol >= 30 ng/mL) and vitamin D-deficient (calcidiol < 30 ng/mL) groups. Propensity score matching was performed adjusting for demographics and risk factors, including body mass index (BMI), medications, medical comorbidities, and laboratory values. Temporal analysis was performed comparing risks of SCFE development between the two cohorts. Statistical significance is held at 0.05.

Results: On preliminary analysis, 98,045 patients met the inclusion criteria. After matching, 34,552 in the vitamin D-deficient and 34,552 in the vitamin D-adequate groups were included. The average patient ages were 11.4 years and 50% were female. In total, 136 (0.39%) and 48 (0.14%) patients developed SCFE in vitamin D-deficient and vitamin D-adequate groups, respectively (p < 0.0001). In this study, 64.7% of SCFE development was attributed to vitamin D deficiency (risk ratio (RR) = 2.833, 95% confidence interval (CI) = 2.040–3.936; hazard ratio (HR) = 1.558, 95% CI (1.119–2.168), p < 0.0001).

Conclusions: This is one of the largest studies to date demonstrating the association between vitamin D deficiency and SCFE development. Vitamin D-deficient children are more susceptible to developing SCFE than those with adequate supplementation.

Significance: These findings highlight the importance of managing adequate vitamin D supplementation in at-risk adolescents, especially among the obese, nutrient-deficient, and populations living further from the equator to lower the risk of future SCFE development.

Intertrochanteric Imhauser’s osteotomy combined with osteochondroplasty in management of slipped capital femoral epiphysis

Mostafa Baraka

Ain Shams University, Cairo, Egypt

LOE-Therapeutic-Level IV

Purpose: Treatment of moderate to severe stable slipped capital femoral epiphysis (SCFE) remains a challenging problem. Open reduction by modified Dunn procedure carries a considerable risk of osteonecrosis (ON). Imhauser osteotomy can realign the deformity without the risk of ON, but the remaining metaphyseal bump is implicated with significant chondro-labral lesions and accelerated osteoarthritis. We conducted this study to evaluate the efficacy and safety of Imhauser osteotomy combined with osteochondroplasty (OCP) through the surgical hip dislocation (SHD) approach.

Methods: A prospective series of 23 patients with moderate-severe stable SCFE underwent Imhauser osteotomy and OCP through SHD. The mean age was 14.4 years (13–20 years) and the mean follow-up period was 45 months (24–66 months). The outcome measures included clinical and radiological parameters, and Harris hip score (HHS) was used as a functional score.

Results: The mean HHS improved significantly from 65.39 to 93.3. The limb length discrepancy improved by a mean of 1.72 cm. The mean flexion and abduction arcs showed a significant improvement (mean increase of 37.5° and 18.5°, respectively). The mean internal rotation demonstrated the most significant improvement (mean increase of 38.5°). All the radiographic parameters improved significantly, including anterior and lateral slip angles (mean improvement 37.52° and 44.37°, respectively). The mean alpha angle decreased by 39.19°. The articulo-trochanteric distance significantly increased to a mean of 23.26 mm. No cases of ON or chondrolysis were identified.

Conclusions: Combined Imhauser osteotomy and OCP through the surgical dislocation approach provides a comprehensive and safe management of moderate to severe stable SCFE.

Significance: Additional OCP should be considered with corrective osteotomies in moderate and severe stable slips.

Long-term outcomes for total joint arthroplasties in pediatric and young adult populations

Andrea Rogers, Gabrielle J. Patin, Carson L. Keeter, Nathan Donaldson

Children’s Hospital Colorado, Aurora, CO, USA

LOE-Therapeutic-Level II

Purpose: A growing number of pediatric and young adult patients are undergoing total joint arthroplasty (TJA) due to rheumatic, congenital, degenerative, and traumatic joint diseases as well as treatment for malignant blood disorders that lead to avascular necrosis (AVN). Surgical considerations differ from adult populations, contributing to patient and provider hesitancy. Data are now available to examine short- and medium-term clinical outcomes, but little is known about long-term functional or quality of life (QoL) outcomes.

Methods: A retrospective cohort study was performed on patients who have undergone total hip arthroplasty (THA) and total knee arthroplasty (TKA) at our institution between 2000 and 2022. Clinical data were collected from patient records. Functional outcomes were assessed using the modified Harris Hip score (HHS) and Oxford Knee score (OKS), and QoL outcomes were assessed using the 36-Item Short Form Survey (SF-36v2). Scores were averaged for each patient preoperatively, and between 0 and 6 months, 6–12 months, 1–2 years, 2–5 years, 5–10 years, and 10+ years postoperatively. Exclusion criteria included less than 5 years of follow-up, patients over age 25 at the time of surgery, TJA for oncologic treatment, and patients with mucopolysaccharidoses and other progressive degenerative diseases.

Results: In total, 112 TJAs in 95 patients were identified and consented to analysis (43% female, 55% male, 2% non-binary, average 17.3 years of age at surgery). Eighty-eight THA patients (102 TJA) provided 189 HHS responses. Scores increased by a mean of 30.3 and 38.4 points, at 6 months (n = 28) and 10+ years (n = 16), respectively, from preoperative surveys (n = 30). Seven TKA patients (10 TJA) had insufficient preoperative OKS data (n = 1) and were not included in the final analysis. Ninety-three patients (110 TJA) provided 154 SF-36vs responses. Scores increased by a mean of 16.9 and 16.7 points at 6 months (n = 27) and 10+ years (n = 13), respectively, from preoperative surveys (n = 21). Five revisions were necessary due to manufacturer implant recall (n = 3), deep infection (n = 1), and femur fracture (n = 1).

Conclusions: Study data suggests that patients under the age of 25 experience significant improvement in both QoL and functional outcomes by 6 months following TJA procedures and that these improvements are sustained for up to 10 years. Non-recall-related revision rates are also like adult population TJA revision rates within the first 10 years postoperatively.

Significance: This is one of the largest single institution samples to systemically track long-term functional and implant survivorship outcomes and one of the first to track QoL for pediatric and young adult TJA procedures.

Universal ultrasound screening for DDH may be cost effective: a Markov decision analysis model incorporating the entire lifespan

Joshua Bram, Drake Glenn LeBrun, Peter Cirrincione, Erikson Nichols, Bridget Ellsworth, Ernest Sink, Emily Dodwell

Hospital for Special Surgery, New York, NY, USA

LOE-Economic-Not Applicable

Purpose: Developmental hip dysplasia (DDH) is a common pediatric orthopedic condition with long-term implications for quality of life and is now considered the primary cause of hip arthritis requiring joint replacement. Complications of missed or delayed DDH may be minimized when diagnosed early when non-operative interventions (e.g. Pavlik harness) are highly successful. Current standard of care in the United States is selective ultrasound screening, performed only with abnormal physical examination, family history, or presence of risk factors. An updated cost-utility analysis, considering health utilities and costs, is warranted.

Methods: A Markov simulation model was designed representing six DDH screening strategies: (1) Universal ultrasound screening within first 2 weeks (USB), (2) Selective screening within first 2 weeks (SSB, high-risk patients screened with ultrasound, low-risk patients not screened), (3) Universal ultrasound screening at 6 weeks (US6), (4) Universal hybrid screening (UHS, clinical instability receives ultrasound within first 2 weeks while all others undergo ultrasound at 6 weeks), (5) German model (ultrasound for high-risk patients within first 2 weeks and all others at 6 weeks), and (6) a no screening model (i.e. natural history of untreated DDH). A standardized DDH treatment algorithm was used, and a comprehensive literature review provided relevant probabilities and incidences, Medicare costs, and utilities. A hypothetical patient cohort was modeled from birth to death capped at age 100, with sensitivity analyses performed.

Results: From a societal perspective, the US6 strategy was more effective (77.06 versus 77.00 quality-adjusted life years (QALYs)) and more costly ($5677.36 versus $5606.07) than SSB. Utilizing a willingness-to-pay (WTP) threshold of <$100,000/QALY, US6 was more cost-effective than SSB (incremental cost-effectiveness ratio (ICER) = $1111.78/QALY). From the payer perspective, a US6 approach was more costly ($4355.83 versus $4262.12) than SSB, but more cost-effective (ICER = $1461.27/QALY). The US6 strategy was overall cost-effective (<$100,000/QALY) in 59.2% of 10,000 probabilistic sensitivity analysis (Figure 1) iterations assessing societal costs (versus 27.7% UHS, 7.2% USB, and 5.9% SSB) and 58.8% of payer iterations (versus 27.5% UHS, 7.6% USB, and 6.1% SSB). The second most cost-effective model was UHS (77.04 QALYs, societal = $5721.20, payer = $4391.79).

Conclusions: Universal screening at 6 weeks of age or a universal hybrid screening model was cost-effective in screening for DDH compared to traditional selective screening models in the United States.

Significance: These results may justify updated guidelines in the United States regarding appropriate screening for DDH, bringing our standards in line with many other countries where all children are screened for DDH with ultrasound.

Deep-learning algorithm accurately measures migration percentage on hip surveillance radiographs

Vedant A. Kulkarni, Chun-Hsiao Yeh, Anna Kay, Michael Eli Firtha, Marie Villalba, Patrick Donohue, H. Graham, Unni G. Narayanan, Stella X. Yu

Shriners Children’s Northern California, Sacramento, CA, USA

LOE-Diagnostic-Level I

Purpose: All hip surveillance programs for children with cerebral palsy (CP) rely on accurate quantification of the hip migration percentage (MP) to determine referral and treatment. Accurate reporting and measurement of the MP has been shown as a substantial barrier to hip surveillance implementation. This study validates a deep-learning algorithm to automatically measure MP on hip surveillance radiographs, with the goal of allowing broader access to “specialist-level” radiographic measurement.

Methods: Pelvic and femoral landmarks relevant for the calculation for MP were labeled by pediatric orthopedic surgeons on a de-identified data set of anteroposterior (AP) pelvis radiographs of children with CP. A deep-learning algorithm to automatically detect hip landmarks and calculate the MP was then created using a ResNet 18 and Cascaded Pyramid Network architecture. The algorithm was tested for accuracy on a prospectively collected set of AP pelvis radiographs, with “ground truth” MP measurements performed by pediatric orthopedic surgeons in their actual clinical practice. The mean absolute difference (MAD) between algorithm and surgeon was calculated with linear mixed modeling, and multi-variate analysis was used to determine sources of error. The sensitivity, specificity, and area under the receiver-operating characteristic curve (AUC) were calculated with an MP > 30% representing a “positive” case.

Results: A total of 3354 pelvis radiographs from 1509 subjects with CP were included in the study (56% M, 44% F; mean age at X-ray, 8 years; 14% Gross Motor Function Classification System (GMFCS) I, 15% GMFCS II, 6% GMFCS III, 24% GMFCS IV, and 36% GMFCS V). A total of 2500 X-rays from two centers were used to develop and optimize the algorithm. In total, 854 prospectively collected X-rays from 24 different international centers were utilized as a diverse external test set. When compared to the MP measurements calculated by each center’s orthopedic surgeon, the algorithm had an MAD of 9.3% (95% confidence interval (CI): 8.7%–9.9%), sensitivity of 84% , specificity of 85%, and AUC of 0.915. Increased error was correlated most strongly with increasing MP (b = 0.117, p < 0.001), and weakly with GMFCS 5 (b = 0.04, p = 0.047), presence of femoral implant (b = 0.03, p = 0.005), and increasing age (b = 0.005, p < 0.001).

Conclusions: A deep-learning algorithm can automatically measure MP on hip surveillance radiographs with excellent discriminatory characteristics in a large and diverse set of “real world” radiographs. The measurement error of the algorithm falls within the published range of specialist error (MAD = 4%–13%).

Significance: Incorporation of this deep-learning algorithm into the clinical workflow of community and non-specialist providers may remove a substantial barrier to the implementation of hip surveillance.

Suprainguinal fascia iliaca nerve blocks outperform epidural analgesia in patients undergoing periacetabular osteotomy

David Peter VanEenenaam, Stefano Cardin, Wallis Muhly, Wudbhav N. Sankar

Children’s Hospital of Philadelphia, Philadelphia, PA, USA

LOE-Therapeutic-Level III

Purpose: Adequate pain control and early mobilization are critical in the postoperative period following periacetabular osteotomy (PAO). Regional anesthesia can reduce postoperative pain, but certain techniques can increase the risk of postoperative motor block and delayed functional recovery. Continuous lumbar epidurals can not only provide excellent analgesia but also create challenges with early ambulation. Recently, suprainguinal fascia iliaca (SIFI) single-shot blocks have been shown to provide effective analgesia in PAO patients. The goal of this study was to compare opioid use, time to achieve inpatient physical therapy (PT) goals, and length of stay (LOS) between a cohort of patients receiving SIFI blocks and a cohort of patients receiving epidural analgesia (EA).

Methods: This retrospective single-surgeon comparative cohort study included all patients who underwent a PAO between 2012 and 2022. Regional anesthetic technique (SIFI versus EA), length of hospital stay, intra- and post-operative opioid use, pain scores, and time to achievement of PT milestones prior to discharge were recorded. Patients were excluded if they had any preexisting neuromuscular syndrome or neurosensory deficit. All opioid use was converted to morphine milligram equivalents (MME) using standard conversions.

Results: A total of 204 patients met the inclusion criteria; 164 patients received EA and 32 received a SIFI block. The average age of our cohort was 19.5 years (range, 10–35 years). There were no significant differences in age, gender, race, or body mass index (BMI) between the two groups. Compared to EA patients, the SIFI block cohort had shorter mean LOS (2.9 days versus 4.1 days) (p < 0.001) and lower total opioid use on post-operative days 0–2 (p < 0.05). Time to sit on edge of bed, time to ambulate > 10 feet, and time to negotiate > 3 stairs were achieved sooner in those who received a SIFI block compared to EA (Figure 1) (p < 0.001). There were no differences in pain scores reported at inpatient PT visits.

Conclusions: After PAO surgery, the SIFI block is associated with shorter hospital LOS, reduced postoperative opioid use, and earlier mobilization when compared with those who were managed with an epidural.

Significance: This is the largest direct comparison of patients receiving SIFI blocks with those receiving lumbar epidurals for PAO surgery. Our results support the use of SIFI blocks given the equivalent pain scores, lower opiate use, faster time to ambulation, and shorter LOS.

Virtual children’s fracture clinic—a prospective study of 5536 patients confirming that efficiency and cost saving does not compromise safety

Anish P. Sanghrajka, Kareem Edres, Aly Pathan, Matthew Edward Kenneth Goodbun, Joe Hwong Pang, Graeme Carlile, Rajiv Merchant, Helen Chase

Norfolk & Norwich University Hospitals, Norwich, UK

LOE-Economic-Level II

Purpose: Virtual fracture clinics, in which patient’s notes and radiographs are reviewed after referral from the Emergency Department (ED), to plan and implement further management, have become increasingly popular in clinical practice. They reduce burden of travel for unnecessary appointments to the hospital, saving patients time and money, as well as having environmental benefits. The main concern about this method of delivery of clinical care is safety, with the potential for missed injuries as the patient is not seen or examined in person. The purpose of this study was to determine the safety of Virtual Children’s Fracture Clinics (VCFC) specifically with reference to time to review, missed/incorrect diagnoses, and return rate after discharge.

Methods: This was a prospective consecutive case series study performed at a University Teaching Hospital. The VCFC database was used to identify all patients seen between August 2017 and May 2021. Records were reviewed and data recorded about the ED diagnosis, the orthopedic diagnosis and final outcome, unplanned return to clinic for a face-to-face (F2F) review, and later referral back for the same injury.

Results: A total of 5536 consecutive patients were seen in the VCFC over the study period (cost saving £465,496). Mean time from ED visit to VCFC appointment was 2.25 days. Median follow-up at the time of this study was 31 months (range: 19–63 months). Thirty-eight patients (0.5%) required admission for urgent surgical intervention following their virtual clinic review. A total of 2325 patients (42%) were discharged from VCFC without further F2F follow-up. Sixty-three of these patients (3%) returned for unplanned F2F clinic. None were found to have missed injuries or required further interventions. The remaining 2262 patients never attended the hospital again during the entire study period, suggesting that there were no significant missed injuries. The diagnosis made in VCFC differed from ED diagnosis in 1921 cases (34%) (downgraded to a lesser injury in 1460, upgraded to more severe fracture type in 220).

Conclusions: This is the largest reported series of patients from a VCFC, with the longest follow-up, providing a solid basis for making conclusions about safety and missed/incorrect diagnoses. The VCFC model saved 110 F2F appointments per month. Our data demonstrate that the system allows prompt identification of the small number of patients requiring urgent attention and allows safe modification of the ED diagnosis without further F2F review.

Significance: A 3% return rate after discharge from a VCFC, with no missed injuries at a minimum follow-up of 19 months, confirms the safety of VCFCs. We advise those units that do not employ a VCFC model to strongly consider doing so.

Prevalence of osteochondromas in the spine in patients with multiple hereditary exostoses

Carlos Monroig-Rivera, Lauren Bockhorn, Brenda Santillan, David C. Thornberg, Karl E. Rathjen

Scottish Rite for Children, Dallas, TX, USA

Purpose: Multiple hereditary exostoses (MHE), an autosomal dominant musculoskeletal disorder, is characterized by the development of multiple cartilage-capped exostoses originating from the physis, known as osteochondromas. The potential for these osteochondromas to impinge on the spinal cord is a clinical concern. Our study aims to determine the prevalence of osteochondromas in the spine in individuals with MHE. In addition, we aim to identify any risk factors for osteochondromas impinging on the spinal cord.

Methods: We prospectively enrolled a cohort of MHE patients/families at a single institution from 2010 to 2022. Demographics, osteochondroma location, and clinical outcomes were documented. Magnetic resonance imagings (MRIs) were obtained and interpreted by musculoskeletal pediatric radiologists. Patients were categorized based on osteochondroma location: no spinal involvement, on the spinal column, in the spinal canal, or neural impinging. We also noted when osteochondromas were present on the scapula, ribs, or pelvis (aka “harbinger lesions”).

Results: Ninety-four MHE patients received an MRI. Almost 44 (46.8%) had osteochondromas on their spine. Around 22 (23.4%) had osteochondromas located only on their spinal column, 18 (19.1%) had an osteochondroma encroaching in their spinal canal, and 4 (4.3%) were found to have one causing neural impingement (Table 1). Of the four with neural impingement, and two displayed paraparesis requiring immediate surgical intervention (Figure 1). The remaining two patients were observed clinically and monitored with serial MRIs. One patient began to exhibit worsening paretic symptoms and had excision. The remaining patient remained stable throughout follow-up. Age, gender, and presence of harbinger lesions were not associated with neural impingement although logistic regression showed Hispanic individuals were ~32.8 times more likely to have neural impingement compared to white individuals (p = 0.0419).

Conclusions: Although nearly half of patients have osteochondromas on the spine, neural impingement is rare (4.3%) and resulted in operative intervention in 75% of cases. Age, gender, and harbinger lesions were note associated with neural impingement. Significance: Although the prevalence of osteochondromas of the spine approaches 50%, neural impingement is rare. This information can be used to inform clinical decision-making regarding screening MRIs.

Predictors of complication in pediatric hardware removal

Pablo Coello, David A. Hsiou, Luke Austin Nordstrom, Todd Phillips, Rachel Silverstein, Scott B. Rosenfeld

Texas Children’s Hospital, Houston, TX, USA

LOE-Therapeutic-Level III

Purpose: Hardware removal (HWR) is a common procedure in the pediatric population. Yet, current literature is mixed on the risks and benefits especially in the elective setting. Identification of appropriate indications for HWR is essential to guiding treatment. The objective of this study was to assess the incidence of complications as well as identifying significant predictors of complications based on patient risk factors.

Methods: A retrospective cohort study was conducted using institutional CPT code 20680 database query. Patient demographics, comorbidities, peri-procedural data, and complications were recorded. Complications were subcategorized as infection, pain, and re-fracture. Data were analyzed for patient-specific and complication-specific factors.

Results: A total of 1452 patients met the inclusion criteria. The average age was 11.51 years (range, 0 + 3 to 21 + 0 years). There were 107 (7.4%) complications within the cohort, of which 44.9% were infections (75% superficial and 25% deep), 19.6% were refractures, 17.7% had pain (68.4% had pain symptoms unresolved with HWR and 31.6% had new-onset pain after HWR), and 17.8% had various other complications. The overall rate of superficial infection was 2.5%, deep infection was 0.83%, refracture was 1.44%, pain symptoms unresolved with HWR was 0.9%, and new-onset pain was 0.4%. A total of 464 (32%) patients had a comorbidity. 9.5% were diagnosed with a neuromuscular syndrome such as cerebral palsy, 2.5% had osteogenesis imperfecta (OI), 1.3% had rickets, 0.55% had Neurofibromatosis Type-1, 1.3% had a skeletal dysplasia, 2.6% had a growth plate disorder such as Blount’s disease, 4.6% had another syndrome or growth disorder such as Morquio syndrome, 0.83% had a bone disease such as fibrous dysplasia, 3.9% were obese, and 4.5% had a systemic illness/disease such as sickle cell disease. Of the 107 complications, 42% were associated with comorbidities (p = 0.019), OI having the highest at 16.2% and growth plate disorders at 15.8%. There was no difference in complications with elective versus trauma cases. For refractures after HWR, there was no difference among type of hardware used and which bone the HWR occurred. For infections, the proximal femur and the use of plates and screws had the highest infections for both superficial and deep infections.

Conclusions: The incidence of HWR-related complications is low (7.6%). Given HWR is often left to surgeon discretion without evidence-based decision making, these results can help surgeons and families to make evidence-based decisions on whether or not to remove hardware.

Significance: Help surgeons determine whether HWR is appropriate, especially in the elective setting, given the risks and benefits found in this cohort.

Significant improvement in health-related quality of life following surgical treatment of congenital muscular torticollis among a 2-year follow-up cohort of children, adolescents, and young adults

Per Reidar Hoiness, Anja Medbø

Oslo University Hospital, Oslo, Norway

LOE-Prognostic-Level II

Purpose: Surgical treatment of congenital muscular torticollis (CMT) has demonstrated clinical efficacy in cases unresponsive to conservative therapies. Nevertheless, there is a paucity of reports on self-reported outcomes and health-related quality of life (HRQoL). This study aims to assess HRQoL in children, adolescents, and young adults undergoing surgical treatment for CMT, comparing HRQoL with clinical outcomes.

Methods: We conducted a Level 2 prospective observational study on a cohort of CMT patients treated surgically at a single tertiary center. Surgical intervention was guided by clinical indicators, including sternocleidomastoid (SCM) tightness, pain, restricted neck motion, craniofacial asymmetry, and insufficient response to physical therapy. The surgical procedures focused on complete release of tightness, with routine elongation of the medial head to preserve cosmetic function. Patients underwent rigorous pre-operative and post-operative assessments, followed by long-term follow-up. A strict postoperative protocol, including stretching, physiotherapy, and positional exercises, was followed for a minimum of 6 months. Clinical outcomes were assessed using the Cheng and Tang torticollis scoring system, while HRQoL was evaluated using the PedsQL™ 4.0 generic core scales.

Results: The study comprised 31 patients with an average age of 11.4 years. Significant improvements in range of motion, deformities, and overall subjective satisfaction were observed at the 2-year follow-up. The Cheng and Tang score improved significantly from fair (9.6 points) to excellent (17.9 points) after 2 years (p < 0.001). Older patients exhibited less improvement than their younger counterparts, particularly in terms of craniofacial asymmetry (p = 0.004). Patient PedsQL scores also significantly improved at the 2-year mark (p = 0.040), with no discernible age-related differences.

Conclusions: This study demonstrates that surgical treatment significantly enhances PedsQL scores in children, adolescents, and young adults with CMT. Notably, clinical parameters such as shoulder, neck, and plagiofacial deformities showed significant improvement, particularly in younger patients. Utilizing the Cheng and Tang system, all patients achieved good or excellent outcomes, suggesting a correlation between improved clinical results and higher HRQoL post-CMT surgery. Importantly, PedsQL scores detected surgical changes, underscoring their sensitivity. Our findings underscore substantial HRQoL and clinical improvements following CMT surgery, with age-related differences, particularly in craniofacial asymmetry improvement. We recommend further research, including the development of a specific CMT HRQoL questionnaire.

Significance: This study contributes to our understanding of the effectiveness of surgical treatment for CMT and its impact on the quality of life of affected individuals. It provides valuable insights for healthcare professionals and researchers in this field.

The hidden consequences of advanced operative spine imaging in children: do the suggested benefits of intraoperative computed tomography and navigation in posterior spinal fusion for adolescent idiopathic scoliosis outweigh the possible lifetime oncological risks of increased radiation exposure?

Bram Verhofste, Brendan M. Striano, Alex Crawford, Andrew M. Hresko, Andrew Schoenfeld, Andrew K. Simpson, Daniel Hedequist

Boston Children’s Hospital, Boston, MA, USA

LOE-Therapeutic-Level II

Purpose: Adolescent idiopathic scoliosis (AIS) develops in 1%–3% of adolescents, with recent trends demonstrating increased rates of posterior spinal fusion (PSF). Advances in intraoperative computed tomography (iCT), navigation, and robotics suggest improved outcomes and safer instrumentation. However, imaging techniques expose children to varying yet often significant amounts of ionizing radiation. Furthermore, scoliosis patients accumulate exposure due to recurrent surveillance imaging. Despite the well-established dose-dependent relationship of radiation and oncogenesis, minimal data exist on the true oncological risks of advanced operative spine imaging in children. Our goal was to compare the lifetime cancer risk of iCT-navigation compared to traditional imaging techniques (non-iCT) in AIS PSF.

Methods: Retrospective analysis of AIS patients (0–18 years) treated with PSF at a quaternary pediatric center was performed (2014–2019). Demographic, surgical, deformity, and radiation characteristics were compared between groups based on intraoperative imaging technique (iCT versus non-iCT). Cumulative radiation exposure and carcinogenesis risk were calculated as total effective dose (ED) in millisieverts (mSv) based on established conversion factors utilizing age, gender, and body region. Pediatric low-dose iCT protocols were used.

Results: A total of 245 patients (mean = 14.4 years; 83% female) were included: 119 iCT (49%) were compared to 126 non-iCT (51%) (Table 1). After accounting for clinical/radiographic variations, radiation exposure was statistically different between imaging groups, with iCT patients receiving the highest degree of ionization (p < 0.001). Total ED (median; interquartile range (IQR)) per group included fluoroscopy (0.05 mSv; 0.03–0.07), fluoroscopy/radiography (4.14 mSv; 0.23–4.4), and iCT group (8.19 mSv; 5.36–11.7). Overall, iCT-navigated PSF theoretically resulted in 0.9 iatrogenic malignancies per 1000 patients (p < 0.001; IQR, 0.59–1.29), while 2.37/1000 patients (p < 0.001; IQR, 1.47–3.27) are projected to develop in patients with 3 iCT spins/acquisitions.

Conclusions: Compared to traditional non-iCT PSF, an additional 1/1000 AIS patients are expected to develop cancer directly related to iCT usage. In addition, the true incidence of carcinogenesis after a complete treatment course is likely underestimated due to cumulative radiation exposure from surveillance imaging. Further research is necessary to evaluate long-term population risks of iatrogenic imaging-induced malignancies.

Significance: Advanced operative imaging modalities are being increasingly applied in spine surgery. Despite rising trends in iCT, a paucity of data exists on the potential risks of neoplastic transformation secondary to iatrogenic, imaging-induced radiation exposure. We found an alarming incidence of anticipated cancers due to iCT which exposes patients to significantly higher ED compared to alternative techniques. The dose-dependent risk of cancer necessitates limited radiation exposure in developing children and a discussion of iCT indications/regulations is required to promote development of radiation-free imaging modalities in the biomedical sector.

Radiation shielding during bedside fluoroscopy reduces radiation exposure to pediatric patients

Steven Zhang, William Huffman, Caroline Fay, Margaret Bowen, Divya Talwar, J. Todd Lawrence

Children’s Hospital of Philadelphia, Philadelphia, PA, USA

LOE-Diagnostic-Level II

Purpose: Pediatric patients are sensitive to the effects of radiation. There are no established standard protocols or society guidelines supporting the use of radiation shielding during bedside orthopedic procedures requiring fluoroscopic guidance despite frequent use. Consequently, there is a wide variation among institutions. Some position papers have suggested that radiation shields may actually increase radiation exposure to the patient due to scatter of the X-ray beams. In this study, we sought to quantify the amount of radiation at three critical anatomic locations using a mannequin model undergoing a simulated bedside upper extremity fracture reduction with and without lead shielding.

Methods: A mannequin (Laerdal Medical) approximating the size of an average 5-year-old was placed supine on a standard hospital stretcher. The upper extremity was extended over the image intensifier of a standard fluoroscopy machine (GE Healthcare). Exposures were performed at 110 kVp and 4.5 mA for 10 s. A dosimeter (Mirion Technologies) was used to assess scattered radiation in three locations: neck/thyroid, chest/breast, and groin. Readings were assessed with and without a standard rectangular radiation shield that spanned from the chin to the upper thighs. Each configuration was replicated five times.

Results: With no shielding, the neck/thyroid received 61.5 mSv, breast/chest received 58.2 mSv, and the groin received 39.7 mSv. These values were significantly higher (p < 0.0001) compared to the values measured at the thyroid, chest, and groin when in the presence of an anterior shield (4.1, 5.3, and 1.2 mSv, respectively) or in the presence of an anterior and posterior shield (1.4, 4.7, and 1.8 mSv, respectively.) There was no statistical difference between anterior shielding and anterior and posterior shielding in all three locations.

Conclusions: Radiation shields placed anteriorly over a pediatric patient significantly reduce the radiation exposure of pediatric patients in the neck/thyroid, chest/breast, and groin regions by an average of 20-fold. The use of a posterior shield confers no additional protection.

Significance: Although fluoroscopy is commonly used in pediatric orthopedic procedures, there remains no standard guideline for the use of radiation shielding, leading to wide variation in clinical practice. This study not only offers the first of its kind data supporting the use of anterior radiation shielding during reduction procedures where the radiation source is above the patient but also suggests that additional shielding techniques may need to be considered to reduce the exposure even further. Providers should strongly consider placing an anterior shield on the patient when performing a bedside procedure requiring fluoroscopy.

Suicidal ideation in pediatric orthopedic patients

Taylor Zak, Whitney Meghan Herge, Chan-Hee Jo, Anthony Riccio

Scottish Rite for Children, Dallas, TX, USA

LOE-Prognostic-Level III

Purpose: Suicide is the second leading cause of death among American children aged 10–18 years, and those with physical health conditions have a 20% increase in suicidal ideation (SI) relative to their healthy peers. Despite this, suicidality in children with musculoskeletal problems is largely unstudied. This study therefore aims to determine the prevalence of SI in pediatric orthopedic patients and identify risk factors for endorsem*nt of suicidality.

Methods: A retrospective review of all patients aged 10 years and older presenting for outpatient orthopedic evaluation to a single tertiary pediatric orthopedic institution over a 1-year period was conducted. Demographic information, chief complaint, pain symptomatology, treatment history, primary orthopedic diagnosis, and responses on the Ask Suicide Screening Questionnaire were compared between patients who endorsed suicidality and those that did not.

Results: Of 11,590 individual patient encounters, 201 children (2.0%) endorsed SI with 9 (4.3%) experiencing active and imminent suicidal thoughts. Patients endorsing suicidality were predominantly female (67%), Caucasian (75%), and had a mean age of 14.3 years. Patients who endorsed SI were significantly more likely to present with a chief complaint of pain (46.4% versus 33.3%, p = 0.00008) and ultimately more likely to receive non-structural, non-mechanical pain diagnosis (23.7% versus 16.9%, p = 0.010). Those presenting for pain who screened positive for SI were more likely to characterize their pain as persistent or chronic (81.6%), which trended toward but did not reach statistical significance (p = 0.05). Adjusting for age and sex, multivariate analysis showed that patients taking a psychotropic medication (odds ratio (OR) = 2.45, 95% confidence interval (CI): 1.51–3.84) and carrying a documented preexisting mental health diagnosis (OR = 3.95, 95% CI: 2.85–5.43) were more likely to report SI. Moreover, patients with an underlying mental health disorder presenting with a chief complaint of pain without an identifiable structural or mechanical etiology were 12 times more likely to screen positive for SI (OR = 12.1, 95% CI: 6.36–22.83).

Conclusions: SI is not uncommon among adolescent pediatric orthopedic patients presenting for outpatient care. White females with chronic pain complaints and an existing diagnosis of a mental health disorder appear to be at particularly high risk.

Significance: This is the first comprehensive study to assess SI in a large, outpatient, pediatric orthopedic population over this period. These data suggest that mental health screening and the provision of integrated clinical care with psychologists and other behavioral health specialists may be important to provide comprehensive care to adolescent orthopedic patients.

From bytes to bones: assessing the ability of ChatGPT to educate patients and families in pediatric orthopedic surgery

Alex Gornitzky, Raghav Badrinath, Joseph Yellin, Brett R. Lullo

Lurie Children’s Hospital of Chicago, Chicago, IL, USA

LOE-Not Applicable-Level V

Purpose: Generative artificial intelligence (GenAI) applications such as ChatGPT are increasingly being utilized to interpret, summarize, and deliver information across many disciplines. It is imperative that we understand the ability and limitations of GenAI within healthcare. The aim of this study was to assess the level and accuracy of ChatGPT in communicating information about pediatric orthopedic procedures to patients and families.

Methods: We prompted ChatGPT to explain the indications, risks, benefits, and alternatives to five common surgeries: closed reduction percutaneous pinning for supracondylar humerus fractures, flexible intramedullary nailing for femoral shaft fractures, posterior spinal fusion for adolescent idiopathic scoliosis, open hip reduction for developmental hip dysplasia, and physeal-sparing anterior cruciate ligament reconstruction. The unedited responses were distributed to practicing pediatric orthopedists across four academic institutions. Surgeons were asked to rate the author’s level of experience, assess the accuracy and completeness of the information, and determine whether it was sufficient to provide education to families without modification. Respondents were blinded to the fact that the information was written by GenAI.

Results: There were 18 respondents from 4 institutions, with the majority (n = 15; 83%) having ≥11 years of experience. Respondents most attributed the level at which the information was written to that of a fellow (32%) (Figure 1A). However, 35% of respondents felt the information was written by an author at either the junior or senior attending level. There were no differences in perceived level of experience by subsections (indications, risks, benefits, alternatives) (Figure 1B). Overall, respondents felt that the medical information provided was accurate and complete (Figure 1C), with 63% of total explanations receiving a score ≥ 4/5. Across all subsections, 50% of respondents felt the explanations were sufficient to provide education to families without modification (Figure 1D). When answering questions concordant with their own self-selected subspecialty, respondents were equally likely to grade the explanations as accurate and complete (60% with a score ≥ 4/5), but less likely to feel they were sufficient for family consumption without modification (43%).

Conclusions: ChatGPT can provide high-level information about common pediatric orthopedic procedures. However, while most respondents felt the information was accurate and complete, only half felt the information was sufficient to provide education to families without modification.

Significance: As GenAI applications such as ChatGPT become increasingly available, families will utilize them more often to obtain medical information. This study is the first to evaluate the ability of GenAI to communicate accurate information about pediatric orthopedic surgery to patients.

Greater obstetric barriers for female orthopedic surgeons compared to the general population and peer physicians

Emily Reeson, Gwen Grimsby, Melissa Esparza, Heather Menzer

Phoenix Children’s Hospital, Phoenix, AZ, USA

LOE-Not Applicable-Not Applicable

Purpose: Medical training occurs during optimal childbearing years. While unique family planning challenges for surgeons are becoming more widely reported, a gap in knowledge remains regarding fertility and pregnancy risks for each subspecialty. Establishing contemporary trends of infertility risks and pregnancy barriers specific to orthopedic surgeons is imperative to defining steps to improve maternal support at all levels of training and practice. The goal of this study was to determine the prevalence of pregnancy complications, infertility, and maternal support for female orthopedic surgeons in comparison to the general population and other female physicians.

Methods: An anonymous, voluntary survey was distributed to female physicians via private physician social media groups from June to August 2021. The survey queried pregnancy demographics and complications, infertility diagnosis and treatment, workplace environment, and prior education on these topics. Results were compared between orthopedic surgeons and the general population using data from the Centers for Disease Control and Prevention as well as between orthopedic surgeons and other female physicians. Statistical analysis was performed using Fisher’s exact test, chi-square with Yates’ correction, or Student’s t-tests as indicated.

Results: A total of 4638 female physicians completed the survey, including 141 (3%) orthopedic surgeons. Compared with the general population, orthopedic surgeons had children later in life (34.1 versus 23.6 years; p < 0.0001), were more likely to have had a miscarriage (40% versus 19.1%; p < 0.0001), to have undergone infertility evaluation (40.1% versus 8.8%; p < 0.0001) or infertility treatment (31.9% versus 12.7%; p < 0.0001), and to have had a pre-term birth (19.9% versus 10.2%; p < 0.0001). Forty-six percent of orthopedic surgeons reported a pregnancy complication and only 6% received education on risks of delaying pregnancy. Compared to other physicians, orthopedic surgeons were older at first pregnancy (34.1 versus 31.7, p < 0.0001), had fewer children (1.8 versus 2.0, p = 0.0094), were more often discouraged from starting a family during training and practice (56% versus 42%, p = 0.0007), and worked significantly more hours per week while pregnant (59.1 versus 54.1, p = 0.0002; Table 1).

Conclusions: Female orthopedic surgeons have increased risk of miscarriage, infertility, and preterm birth compared to the general population. In addition, orthopedic surgeons experience more negative workplace attitudes and longer working hours while pregnant compared to physician peers. The culture of orthopedic surgery must continue to evolve to better support family planning and childbearing for young physicians during their schooling, training, and early career stages.

Significance: These data will foster steps forward in overcoming the well-described obstetric barriers for orthopedic surgeons in efforts to better support, attract, and retain women in orthopedics.

Full-thickness skin graft versus hyaluronic acid skin graft substitute in syndactyly release: a randomized trial

Ann Van Heest, Deborah Bohn, Jamie N. Price, Susan A. Novotny, Tonye Sylvanus

Gillette Children’s Specialty Healthcare, St. Paul, MN, USA

LOE-Therapeutic-Level II

Purpose: This study assesses scar healing over a 24-month period following syndactyly release surgery using full-thickness autograft (FTSG) versus hyaluronic acid (HA) matrix skin graft substitute. We hypothesized that scar healing would be non-inferior for HA compared with FTSG.

Methods: A randomized, single-blinded, within-subject controlled study was used to compare the effectiveness of HA versus FTSG for 40 webs in 18 patients (aged 1.2 ± 0.8 years) undergoing syndactyly release. HA was used on one side of the web and FTSG was used on the other side of the web in each case. At 12- and 24- month post-surgery scar healing was assessed using Patient and Observer Scar Assessment Scale (POSAS), Vancouver Scar Score (VSS), Web Creep, and the family and surgeon’s preferred scar (FTSG, HA, or equivalent). FTSG source was groin (12 webs) and antecubital fossa (28 webs).

Results: Families rated the syndactyly scar with lower/better scores for FTSG compared to HA at the 12- and 24-month visits. Surgeons’ VSS scores showed preference for FTSG in thickness and vascularity without clear superiority. At the final 24-month visit, FTSG and HA were preferred equally by 42% of families with 16% of families rating HA and FTSG as equivalent; HA was the preferred scar in 46%, FTSG preferred in 31%, and equivalent in 23% according to surgeons. Major complications (CD Level III) occurred in three patients (17%) and included infection, graft failure, and scar revision with return to the OR during the study period.

Conclusions: Neither HA nor FTSG demonstrated superiority as defect coverage for syndactyly release surgery at final follow-up. When compared with FTSG, HA provides diminished surgical time and no donor-site morbidity; however, HA has an open wound for a longer period postoperatively, with greater wound care requirements by families to ensure appropriate wound healing.

Significance: Level of Evidence: II.

Outcomes following operative versus non-operative treatment of completely displaced midshaft clavicle fractures in adolescent baseball players and other overhead athletes

Eric W. Edmonds, David D. Spence, Michael Quinn, Benton E. Heyworth, FACTS Study Group

Boston Children’s Hospital, Boston, MA, USA

LOE-Therapeutic-Level II

Purpose: Recent evidence has demonstrated equivalent or superior outcomes following conservative treatment of completely displaced midshaft clavicle fractures in adolescents. However, specific sub-populations, such as throwers, who may have unique relative benefits from surgical intervention, remain under-investigated. The current study therefore assessed complications and patient-reported outcomes (PROs) in baseball players and other overhead athletes, with specific comparisons of dominant versus non-dominant shoulders and operative versus non-operative treatment.

Methods: Ten 18-year-old patients with completely displaced midshaft clavicle fractures treated at eight participating institutions from 2013 to 2022 were filtered for those who self-reported participation in baseball, softball, racquet sports, water polo, lacrosse, and football quarterbacks. Athletes were divided into operative (open reduction internal fixation (ORIF)) or nonoperative (NonOp) treatment cohorts and sub-divided into dominant versus non-dominant injured laterality. Demographics, fracture characteristics, outcomes, complications, and PROs (ASES, QuickDASH, Marx Shoulder Activity, EQ5D, EQVAS) were analyzed for patients with a minimum of 1-year follow-up.

Results: Out of a total of 788 adolescents with completely displaced clavicle fractures, 238 overhead athletes (30.2%) included 123 baseball players (51.7%, 15.6% overall), fewer of which were the dominant (46, 37.4%) versus non-dominant shoulder (77, 62.6%). Of the 46 dominant baseball player sub-population, 18 (39.1%) were in the ORIF cohort, compared to 28 (60.9%) in the NonOp cohort. When complications, secondary surgeries, or PROs were compared, the only difference was a superior mean QuickDASH score in the NonOp cohort (p = 0.01), a finding that was replicated when treatment groups in the dominant shoulder of all overhead athletes were compared as well (p = 0.04). When dominant versus non-dominant shoulders were analyzed among baseball players, both within ORIF and NonOp cohorts, there were no differences in complications, secondary surgeries, or PROs between sides. Complications in the dominant side of all overhead athletes were rare, including delayed union (2%), symptomatic malunion (1%), and refracture (4%), and were not statistically different between ORIF and Nonop cohorts. Only one case of nonunion (on the non-dominant side of a baseball player) was identified, despite most patients (77%) undergoing non-operative treatment.

Conclusions: Overhead throwers, including baseball players, who sustained completely displaced midshaft clavicle fractures appear to have equivalent, or perhaps superior results from non-operative treatment, when compared to surgical fixation. Outcomes appear similar between the dominant and non-dominant side, regardless of treatment approach.

Significance: Unlike completely displaced midshaft clavicle fractures in adult patients, similar fractures in adolescent overhead athletes are associated with low rates of complications, such as non-union and symptomatic malunion, regardless of treatment approach.

Two-year patient-reported outcomes and graft rupture following ACL reconstruction in skeletally immature athletes: results from the PLUTO (pediatric ACL: understanding treatment options) prospective cohort study

Mininder S. Kocher, Lauren E. Hutchinson, Danielle Cook, Jeffrey Kay, Benton E. Heyworth, PLUTO Study Group

Boston Children’s Hospital, Boston, MA, USA

LOE-Therapeutic-Level II

Purpose: The purpose of this study was to compare patient-reported outcomes (PROs) and graft rupture in a multicenter prospective cohort study of skeletally immature patients who underwent anterior cruciate ligament reconstruction (ACLR) using growth preservation techniques. The hypothesis was that PROs across sub-cohorts would be similar, but anterior cruciate ligament (ACL) graft rupture rates would be higher in the early adolescent/pubescent group than in the pediatric/pre-pubescent group.

Methods: Skeletally immature patients who underwent ACLR by one of 23 PLUTO surgeon-investigators at one of 10 participating academic medical centers across the United States over a 5-year period (2016–2020) were included. Surgical techniques were categorized as one of three different pediatric/prepubescent physeal-sparing techniques (all-epiphyseal, AE; partial transphyseal, PTP; combined intra-articular/extra-articular, extraphyseal using ilitiobial band, ITB) or an early adolescent/pubescent transphyseal (TP) physeal-respecting technique, and by autograft type (hamstring, HS; soft tissue quadriceps, Q; iliotibial band, ITB). Demographics, surgical characteristics, pedi-IKDC scores, and ACL graft rupture (re-tear) rates were analyzed with comparative statistics.

Results: A total of 742 patients (mean age: 12.9 years (SD, 1.9), 62% male) were included. Two-year follow-up pedi-IKDC was available in 553 (74%) patients at a median 24 (interquartile range (IQR), 24.0–26.7) months post-ACLR. The median pedi-IKDC score for the full cohort at 2-year follow-up was 94.6 (range, 21.7–98.9). Median pedi-IKDC was not found to be different across surgical technique groups (p = 0.22, Figure 1) or graft types (p = 0.51) at 2-year follow-up. Of 665 (665/742, 90%) patients with adequate 2-year re-tear data, 48 (7%) experienced a re-tear at a median 16 months post-operatively (IQR, 10–22 months). Significant differences were detected in re-tear across surgical techniques (p = 0.008), with pairwise comparisons revealing higher re-tear in TP (10%) than ITB (3%; p = 0.02). No differences were otherwise detected in re-tear across graft types (p = 0.12). The rates of re-tear were significantly higher in the pubescent group (10%) than in the prepubescent group (3%; p = 0.001).

Conclusions: Among skeletally immature patients undergoing ACLR, pre-pubescent children undergoing physeal-sparing techniques have superior 2-year re-tear rates than pubescent adolescents undergoing transphyseal techniques. The physeal-sparing ITB technique has superior re-tear rates compared to the transphyseal technique, but similar outcomes to other pediatric physeal-sparing (PTP, AE) techniques.

Significance: The frequency of ACLR is increasing at a faster rate in pediatric patients than in any other sub-population. There remains a dearth of prospective comparative studies designed to elucidate the optimal techniques for this active, high-risk sub-population.

Fabrication of a biomimetic 3D-printed scaffold for the treatment of large osteochondral defects in an adolescent porcine model: outcomes at 6 months

Sanjoy Kumar Ghorai, Patrick William Whitlock, Sumit Murab, Anish Gangavaram, Chia-Ying James Lin; Jenna Hall

Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, USA

LOE-Not Applicable-Not Applicable

Purpose: Repair of large osteochondral (OC) defects represents a significant challenge in pediatric orthopedics and often leads to progressive and early osteoarthritis. Current treatments for large OC defects are limited in their clinical success by the defect size and geometry of the recipient, donor tissue availability, and inferior mechanical properties. Differences in structure, architecture, metabolic activity, timing of bone and cartilage regeneration, and the differing moduli of cartilage and subchondral bone require a comprehensive strategy to regenerate OC tissue. Thus, this study aimed to assess the integration of a novel, biomimetic scaffold, and its concomitant regeneration of OC tissue 6 months after treatment of an OC defect. Our hypothesis was that the regenerated OC tissue would be comparable to OATS autograft.

Methods: The study was approved by the IACUC (IACUC2020-0081). Regeneration of OC tissue within a distal femoral OC defect was studied in an adolescent porcine model (Yucatan minipig) after implantation of a novel, biomimetic scaffold. Polycaprolactone (PCL) scaffolds containing thermally stable polylactic acid (PLA) microspheres encapsulating bioinductive, decellularized cartilage or bone matrix were three-dimensional (3D) printed in a biphasic fashion (DCM+ scaffold). The PLA microspheres containing bone DCM were printed in the “subchondral layer” and PLA microspheres containing cartilage DCM within the “cartilage layer” of the scaffolds, mimicking the hierarchical micro-architecture of OC tissue. To assess and compare OC regeneration and integration, PCL scaffolds containing PLA microspheres without DCM (DCM−) and OATS autografts were included as negative and positive controls, respectively. Six months after implantation, OC regeneration was assessed by micro-computed tomography (CT), magnetic resonance imaging (MRI), and biochemical assays.

Results: Six-month implants showed higher percentage of new bone formation in the outer annulus bone of DCM+ scaffolds as compared to DCM− and OATS (p < 0.05) indicating excellent integration of the DCM+ scaffolds within host tissue (Figure 1). Two-dimensional (2D) Magnetic Resonance Observation of Cartilage Repair Tissue (MOCART) score by MRI study showed new OC regeneration in DCM+ scaffolds comparable to OATS autografts (Figure 2). Histologic staining showed excellent OC regeneration within the DCM+ scaffolds at least comparable to OATS (Figures 3 and 4).

Conclusions: The biomimetic 3D-printed scaffolds exhibited excellent OC regeneration and scaffold integration at 6 months. The tidemark layer was also maintained. Scoring by histology and MRI was comparable to OATS.

Significance: Biomimetic 3D-printed scaffolds recapitulate normal OC tissue architecture at 6 months post-implantation providing evidence for their further development as an alternative to current treatment strategies for large OC defects in pediatric, adolescent, and young adult patients.

Long-term outcome of nonoperative treatment of Perthes disease—244 hips with a mean follow-up of 48 years

Anders Wensaas, Chiara Blatti, Terje Terjesen, Stefan Huhnstock

Department for Children’s Orthopaedics and Reconstructive Surgery, Division of Orthopedic Surgery, Oslo University Hospital, Oslo, Norway

LOE-Prognostic-Level III

Purpose: In Perthes disease, there is an increased risk of osteoarthritis later in life and thus long-term studies are necessary to analyze the outcome. The purpose of this study was to evaluate the prevalence of total hip arthroplasty (THA) after Perthes disease and to define risk factors for poor outcome.

Methods: Patients were recruited from a search of the radiographic archive at our hospital. Inclusion criteria were patients with nonoperative treatment for Perthes disease and a minimum of 25-year follow-up. The femoral head at the healing stage was classified with the modified Stulberg method (a 3-group classification based on the shape of the femoral head: spherical, ovoid, or flat). Information regarding THA was provided by the Norwegian Arthroplasty Register. Kaplan–Meier survival analysis with conversion to THA as the endpoint was used to find the percentage of hips that had not undergone THA.

Results: A total of 229 patients (244 hips) were included in the study, 184 boys and 45 girls. The mean age at diagnosis was 6.2 years (2.1–13.7 years). Treatment was nonoperative with relief of weight-bearing and/or abduction orthosis. At the healing stage, 105 hips (43%) were classified as spherical, 93 (38%) as ovoid, and 46 (19%) as flat. The mean time from diagnosis to follow-up was 48 years (27–72 years). At follow-up, 47 hips (19%) had undergone THA at a mean patient age of 46 years (22–72 years). Age ≥ 6 years at diagnosis was associated with significantly worse outcome compared with that of age < 6 years; the rate of THA in the two groups were 28% and 10%, respectively (p < 0.001). The frequency of THA was 3% in hips with spherical femoral heads, 25% in hips with ovoid heads, and 46% in hips with flat heads (p < 0.001). Survival analysis showed a mean survival rate of 100% in spherical hips, 99% in ovoid hips, and 98% in flat hips at 20-year follow-up and 99%, 76%, and 48%, respectively, at 50-year follow-up.

Conclusions: After a mean follow-up of 48 years, 19% of non-operatively treated hips with Perthes disease had undergone THA. The negative prognostic factors were age ≥ 6 years at diagnosis and deformed (especially flat) femoral head at skeletal maturity.

Significance: The study demonstrates the long-term outcome of non-operatively treated Perthes disease, which should represent a valuable basis for comparison with other treatment methods.

In situ fixation of slipped capital femoral epiphysis carries an over 40% risk for later total hip replacement during a long-term follow-up

Thomas Schlenzka, Joni Serlo, Timo Juhani Viljakka, Kaj Tallroth, Ilkka J. Helenius

Helsinki University Hospital, Helsinki, Finland

LOE-Therapeutic-Level IV

Purpose: The elevated long-term risk for arthrosis after treatment of slipped capital femoral epiphysis (SCFE) has been described. Previous studies have observed the need for later total hip replacement (THR) to range from 12% to 24% during a mean follow-up varying between 16 and 38 years. It is possible that the incidence increases with the patients age due to progression of degenerative changes. To our knowledge, no study has described the risk for THR in a more than 40-year follow-up.

Methods: In this study, 138 patients with 172 affected hips treated with in situ fixation were evaluated retrospectively. Ninety-seven patients (70%) were male (mean age 14.0 years for males and 12.7 years for females at surgery) and 35 patients (25%) had a bilateral disease. The median follow-up time was 49 years (range, 37–64 years). Basic demographics, stability, and surgical details were obtained from patient records, preoperative radiographs (slip angle) were measured, and data on THR was gathered from the National Arthroplasty Register.

Results: The preoperative slip angle averaged 39° (SD, 19°). After a median follow-up of 49 years, 56 (41%) patients had undergone THR of a hip previously fixed in situ for SCFE and 64 (37%) of all affected hips had been replaced. Kaplan–Meier analysis gave a median prosthesis-free postoperative survival of 55 years (95% confidence interval, (CI) = 45–64 years) for the affected hips. In a multivariate analysis, females had a twofold (hazard ratio (HR) = 2.42, 95% CI = 1.16–5.07) risk for THR and a greater preoperative slip angle (HR = 1.03 for every increment of 1°, 95% CI = 1.01–1.05) increased the risk for replacement surgery while patient age at surgery, slip laterality, stability of slip, or diagnostic delay did not have a statistically significant effect on risk of THR.

Conclusions: SCFE treated primarily with in situ fixation may lead to THR in more than 40% of the affected hips, when follow-up reaches 50 years. This risk is approximately 15 times the reported life-time risk in our general population. Female sex and increasing preoperative slip angle significantly predicted higher risk of total hip replacement.

Significance: With a median follow-up time of 49 years, our study group of 138 patients represents one of the largest patient series with very long-term follow-up data.

Mid-term outcomes following vertebral body tethering: a single-center cohort with 5+ years of follow-up

Daniel G. ho*rnschemeyer, Sam Hawkins, Nicole Tweedy, Melanie E. Boeyer

University of Missouri, Columbia, MO, USA

LOE-Therapeutic-Level III

Purpose: Vertebral body tethering (VBT) is a new, non-fusion alternative for adolescent idiopathic scoliosis. To date, there are only several published outcome studies with no more than 2 years of postoperative follow-up and almost no outcome data beyond 2 years. We aimed to fill this gap in the VBT literature by evaluating mid-term outcomes in our first 31 consecutive patients. We hypothesized that (1) the postoperative success would decrease, (2) the surgical revision and suspected broken tether rate would increase, and (3) additional patients would convert to a posterior spinal fusion (PSF).

Methods: We retrospectively assessed additional clinical and radiographic data (mean follow-up: 5.7 ± 0.7) from the original 31 consecutive VBT patients previously included in ho*rnschemeyer et al. (2020). This cohort included patients with various curve patterns (e.g. thoracolumbar only, bilateral). Radiographic measurements included standard deformity measures and skeletal maturity status at latest follow-up. Using the same definition of success (i.e. cobb ≤ 30°; no PSF), we revisited the success rate, revision rate, broken tether rate, and conversion to PSF. To assess differences between the values presented in ho*rnschemeyer et al. and those derived from additional follow-up, we performed a two-sided t-test with an alpha of ≤0.05 indicating statistical significance.

Results: Of our first 31 patients treated with VBT, 29 patients returned for additional follow-up after publication. At 5 years, the success rate dropped to 64% as deformity measures increased and the revision rate increased to 28% following two additional surgical revisions. Four additional suspected broken tethers were identified for a rate of 55%, with only one occurring beyond 4 years. Ninety-three percent of this cohort continues to avoid PSF. We observed an average cobb increase of 4° and 8° in the thoracic and lumbar spine, respectively. Statistically significant progression of both thoracic (p < 0.001) and lumbar curves (p = 0.047) were seen only in Lenke 1B/1C patients where we tethered only the main thoracic curve.

Conclusions: With more than 5 years of follow-up on our VBT cohort, we observed: (1) a decrease in the postoperative success rate due to deformity progression in most subgroups and (2) an increase in the revision and suspected broken tether rate. No additional patients received a PSF.

Significance: With more than 5 years of follow-up, 93% of VBT patients continue to avoid a PSF and 89% exhibited deformities that are ≤40°, which may indicate long-term survivorship.

A CNP analog as adjuvant treatment for moderate-to-severe osteogenesis imperfecta in the growing mouse: a pilot study

Jack Mulcrone, Ketsia Seide, Erin Carter, Nancy Pleshko, Cathleen L. Raggio

Hospital for Special Surgery, New York, NY, USA

LOE-Therapeutic-Level I

Purpose: Osteogenesis imperfecta (OI), a heterogeneous type 1 collagenopathy, results in fragile bones. Bisphosphonates reduce fractures and are used to manage moderate-to-severe OI in children. C-type natriuretic peptide (CNP) is produced in the growth plate and positively regulates linear bone growth. This study aims to evaluate whether the addition of CNP analog to standard bisphosphonate (alendronate (ALN)) therapy will reduce fracture incidence, improve growth, increase bone mineral density (BMD), and/or improve bone strength in the growing oim/oim mouse.

Methods: At 2 weeks, growing oim/oim mice (N = 17) were divided into four groups: control mice (N = 8) receiving saline treatment and treated oim/oim (oim) mice receiving weekly ALN along with one of three CNP dosages: 20 µg/kg 5 days/week (20x5) (N = 3), 20 µg/kg 3 days/week (20x3) (N = 3), or 10 µg/kg 3 days/week (10x3) (N = 3). Faxitron images were taken at 2 and 14 weeks (sacrifice) to assess fracture incidence, femoral length, and vertebral height. Microcomputed tomography (micro-CT) was used to assess bone microstructural parameters.

Results: This IACUC-approved study found that the 20x5 group had no new fractures at sacrifice, while the 20x3 and 10x3 groups each had one fracture, and untreated oim mice had an average of 1.57 ± 1.33 fractures. All treatment groups showed increased femoral lengths compared to untreated oim mice (Figure 1), with the 20x5 group showing the greatest increase and the 20x3 and 10x3 groups having similar increases. In addition, all treated groups exhibited increased vertebral height; the 20x5 and 10x3 groups had higher vertebral heights than the 20x3 group but were not different from each other. All three dosage groups showed increased cortical bone tissue mineral density (TMD), cortical BMD, cortical bone thickness, trabecular bone volume fraction (BVF), trabecular TMD, trabecular BMD, and trabecular bone number. All three dosage groups had reduced trabecular bone separation (Figure 2). Due to small sample sizes and variability, these differences are reported qualitatively.

Conclusions: Adjuvant CNP analog treatment increased femoral length and vertebral height without compromising fracture reduction and showed added benefits for both trabecular and cortical bone, which was not observed with bisphosphonates alone. Continued enrollment of oim mice will enhance the significance and help determine the optimal dosage for maximizing these positive bone effects.

Significance: This research project aims to discern the optimal dose of the CNP analog as a promising adjuvant treatment with ALN therapy to reduce fracture incidence and improve bone growth, quality, and strength in pediatric patients suffering from OI.

Bi-lateral and bi-level erector spinae plane block in pediatric idiopathic scoliosis surgery: a randomized, double-blind, controlled trial

Malgorzata Domagalska, Piotr Janusz, Tomssz Reysner, Grzegorz Kowalski, Juliusz Huber, Przemyslaw Daroszewski, Tomasz Kotwicki

University of Medical Sciences, Poznan, Poland

LOE-Therapeutic-Level I

Purpose: This study aimed to compare the effect of ultrasound-guided bi-lateral and bi-level erector spinae plane block (ESPB) on pain scores, opioid requirements, neuromonitoring parameters, and surgery-induced stress response, compared with standard analgesia following posterior spinal correction and fusion surgery for idiopathic scoliosis in children.

Methods: This was a prospective, double-blind, randomized controlled trial. Sixty patients aged 10–18 years and with physical status ASA 1 or 2 were randomized into two groups, each receiving either ESPB (n = 30) or a sham block (n = 30). The primary outcomes were pain scores (Numeric Rating Scale, NRS) within 48 h of posterior spinal surgery. The secondary outcomes were total opioid consumption, intraoperative motor-evoked potentials’ (MEPs) amplitude in µV recorded from the tibialis anterior muscle following transcranial magnetic stimulation at the threshold measured in mA, as well as the surgery-induced stress response as expressed by the neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) at 12- and 24-h post-operation.

Results: The ESPB group had lower NRS scores at all time points (30, 60, 90, 120 min, and 6, 12, 24, and 48 h after surgery), all at p < 0.0001, compared with the control group. Total opioid consumption, incidence of nausea and vomiting, and the need for remifentanil or propofol during surgery were significantly lower in the ESPB group. The MEPs showed higher amplitudes and lower stimulus strengths for evoking the threshold responses in the ESPB group. The surgery-induced stress response indicators (NLR and PLR) were significantly lower in the ESPB group.

Conclusions: ESPB performed in adolescents during idiopathic scoliosis surgery provided more effective analgesia and reduced post-operative opioid requirements. Intra-operatively, ESPB reduced the demand for remifentanil or propofol, thus also reducing the anesthesia influence on MEPs. Post-operatively, ESPB inhibited the stress response as expressed by NLR and PLR ratios.

Significance: The study presents better pain control after spinal correction and fusion surgery for idiopathic scoliosis in children with ultrasound-guided ESPB. Ultrasound-guided ESPB does not affect intraoperative MEP’s amplitude.

Local wound infiltration reduces acute postoperative opioid requirements in AIS: a prospective double-blind randomized controlled trial

Craig Munro Birch, Sydney Lee, Kelsey Mikayla Flowers Zachos, Shanika De Silva, Grant Douglas Hogue, Michael T. Hresko, Daniel Hedequist

Boston Children’s Hospital, Boston, MA, USA

LOE-Therapeutic-Level I

Purpose: Local wound infiltration is a non-narcotic method of acute pain management following surgical intervention. This double-blind randomized controlled trial (RCT) aimed to determine the impact of wound infiltration with 0.25% bupivacaine with epinephrine, compared to a placebo of equal volume injectable saline, on pain and opioid consumption during the first 24 h postoperatively in patients with adolescent idiopathic scoliosis (AIS) undergoing posterior spinal fusion (PSF).

Methods: AIS patients, ages 10–17 years old, undergoing PSF were randomized to receive intraoperative wound infiltration with 0.25% bupivacaine + epinephrine (treatment) or saline (placebo). Providers, study staff, and patients were blinded to randomization. Postoperatively, patients rated their pain using an 11-point numeric rating scale, 0 to 10. Inpatient pain scores and opioid administration were extracted from medical records. Opioids were converted to morphine milligram equivalents per kilogram (MME/kg). Four 6-h intervals were created to assess average pain scores and total opioids administered during the first 24 h. Linear mixed models were used to analyze differences between treatment groups in pain scores and opioid consumption over time.

Results: Fifty-six patients were included (mean age = 14.7 years, mean body mass index (BMI) = 21.5): 26 randomized to the treatment group and 30 to placebo. Patients were predominantly female (75%), White (82%), and non-Hispanic (83%). On average, patients in the treatment group had slightly higher preoperative major Cobb angles compared to the placebo group (63° versus 55°). Patients in the treatment group consumed significantly less opioids during the first 24 h, with the placebo group receiving an average estimated 6 MME/kg more compared to the treatment group (Figure B, p = 0.049). Difference at interval 1 was most notable with a 27.7% reduction, average 36.1 MME/kg (placebo) compared to 26.1 MME/kg (treatment). Opioid consumption decreased significantly over time with interval 4 (18–24 h) having estimated mean decrease of 13 MME/kg (p < 0.001) compared to interval 1 (0–6 h). Despite reduced opioid usage, there was no statistically significant difference in pain scores between treatment groups during the 24-h postoperative period. However, pain scores decreased across time for the entire cohort by approximately 1 point per interval (Figure A, p < 0.001).

Conclusions: Local anesthetic injection of 0.25% bupivacaine with epinephrine can effectively decrease postoperative opioid consumption in AIS patients in the first 24 h without compromising pain control.

Significance: This is the first RCT to assess the impact of wound infiltration on postoperative pain following PSF for AIS. The results suggest the potential of incorporating local wound infiltration as standard practice.

Efficacy of a multimodal surgical site injection in pediatric patients with cerebral palsy undergoing hip reconstruction: a randomized controlled trial

Danielle Brown, Christina-Angèle Kaulueloa’ainalani Sun, Daniel McBride, Bailey Young, Vineeta Swaroop, Rachel Mednick Thompson

UCLA, Los Angeles, CA, USA

LOE-Therapeutic-Level I

Purpose: One in four children with cerebral palsy (CP) will undergo orthopedic surgery during their childhood. Despite its ubiquity, postoperative pain control has been poorly studied in this patient population. Moreover, poor pain management has been associated with adverse surgical outcomes. Multimodal analgesic injections have been well studied in the adult population, demonstrating safety and efficacy at reducing postoperative pain and narcotic consumption, but this modality has not been studied in pediatric patients undergoing similarly complex procedures. The objective of this study was to evaluate the efficacy of a multimodal surgical site injection for postoperative pain control following operative management of hip dysplasia in patients with CP.

Methods: After obtaining institutional review board (IRB) approval, a multi-center, randomized double-blind placebo control trial was completed. Patients < 18 years old with a diagnosis of CP who were scheduled for varus derotation osteotomy (VDRO) of the proximal femur were randomized to receive a surgical-site injection with either a combination of ropivacaine (3 mg/kg), epinephrine (0.5 mg), and ketorolac (0.5 mg/kg) (experimental group) or normal saline (control). All included patients had identical post-operative care, including immobilization, physical therapy, and standardized, multi-modal postoperative pain control. Pain scores and narcotic consumption were recorded at regular intervals and compared between groups utilizing two-tailed t-test or a non-parametric Mann–Whitney test for quantitative variables and a Fisher’s exact test for categorical variables.

Results: Thirty-four patients were included, evenly divided between study arms. There were no significant differences in demographic variables, Gross Motor Function Classification System (GMFCS), comorbidities, preoperative radiographic parameters, or concomitant surgeries between groups. Patients in the experimental group required significantly lower narcotic medications at all postoperative time points from post-anesthesia care unit (PACU) until hospital discharge compared to controls (0.41 ± 0.42 versus 1.87 ± 2.05 total morphine mEQ/kg, p = 0.01). Similarly, patients in the experimental group were found to have significantly lower pain scores throughout their hospital stays compared to controls (1.0 ± 0.6 versus 2.4 ± 1.1 mean pain score, p < 0.001). There were no significant differences in operative time, OR time, blood transfusion requirements, or hospital length of stay between groups. There were no adverse medication reactions or injection site complications in either group.

Conclusions: In patients with CP undergoing hip reconstruction, surgical-site injection with a multimodal analgesic combination improves pain control and reduces narcotic consumption in the early postoperative period with no observed adverse effects.

Significance: Local multimodal analgesic injections should be adopted as part of standard multi-modal pain control in this patient population for all osseous surgeries.

Analysis of regenerate bone formation using internal lengthening nails in a large animal model: a pilot study

Christopher A. Iobst, Anirejuoritse Bafor, Aidan Gene Isler, Sara McBride-Gagyi, Kell Sprangel

Nationwide Children’s Hospital, Columbus, OH, USA

LOE-Not Applicable-Not Applicable

Purpose: Internal lengthening nails have become the preferred method of long bone lengthening. As the experience with these nails increases, it is clear they behave differently than traditional external fixators. For example, radiographically, internal lengthening nail regenerate bone formation is visually quite different than traditional external fixator bone. While the biology and mechanics of regenerate bone formation with external fixators has been extensively studied, very little is known about the same basic science of regenerate bone formation with internal lengthening nails. This study is the first to attempt an evaluation of internal lengthening nail regenerate bone formation while simultaneously attempting to develop the first feasible large animal model.

Methods: After obtaining IACUC approval, nine male goats had successful implantation of a tibial internal lengthening nail. The surgical technique and implants were identical to normal human surgery. After a 6-day latency, the goats were lengthened three times at 0.25 mm for a total of 0.75 mm per day. Once 2 cm of lengthening was completed, the animals entered the consolidation phase with necropsy at 4 or 8 weeks. Weekly radiographs during distraction and every 2 weeks during consolidation were obtained. Micro-computed tomography (CT) analysis and histology were also performed on each of the specimens.

Results: Although feasible to insert human internal lengthening nails into the goat tibia, the goat anatomy consistently had a deformity apex in the diaphysis. This required the osteotomy site to be more distal than the typical human scenario. The radiographic timeline of regenerate bone formation visually mimicked the human pattern of progressive outside to inside healing. The micro-CT imaging identified a unique characteristic not previously noted in external fixator lengthenings. A fibrous interzone in the center of the regenerate like external fixation was noted, but there was an additional layer of the same tissue on the nail surface within the regenerate bone. This new finding is called the “fibrous innerzone” (see Figure 1). Histology confirmed healing of the regenerate from the outside to inside with a distinct demarcation between the outer more mature bone and the inner fibro-cartilaginous layer.

Conclusions: This pilot study is the first to demonstrate a feasible large animal model for studying internal lengthening nails. We discovered a unique “fibrous innerzone” not previously identified in external fixator regenerate bone.

Significance: This pilot study confirms that regenerate bone formation and healing appears to occur in a different pattern than external fixator lengthenings and warrants further investigation.

Long-term results of epiphyseal distraction prior to resection (Cañadell’s technique) in 169 patients with metaphyseal pediatric bone sarcomas

Jorge Gómez-Álvarez, José María Lamo-Espinosa, Rocío López, Mikel San-Julián

Clínica Universidad de Navarra, Pamplona, Navarra, Spain

LOE-Therapeutic-Level II

Purpose: Cañadell’s technique consists of a Type I epiphysiolysis performed using continuous distraction (1 mm/day) in the last 10–12 days of neoadjuvant chemotherapy by an external fixator. It is performed, prior to resection, in metaphyseal pediatric bone sarcomas. It was designed to achieve a safe margin due to the ability of the physeal cartilage to be a barrier to tumor spread to the epiphysis, avoiding the need for articular reconstruction, and preserving the growth capacity most of the times.

Methods: We retrospectively reviewed patients who had a metaphyseal pediatric osteosarcoma or Ewing’s sarcoma treated at our institution with Cañadell’s technique between 1984 and 2021 with a minimum 2-year follow-up. Demographic data, metastases at diagnosis, type of reconstruction, functional results according to the Musculoskeletal Tumor Society scale, infection rate, local recurrence, and overall survival were obtained. All outcome data were collected prospectively and compared to patients in whom the epiphysis could not be preserved.

Results: A total of 169 patients with pediatric metaphyseal sarcomas (100 osteosarcomas and 69 Ewing’s sarcomas) were treated in our institution with Cañadell’s technique between 1984 and 2021. We found 74% of disease-free survival in osteosarcoma and 68% in Ewing’s sarcoma at a mean follow-up of 15 years versus 59% in those cases in which the epiphysis could not be preserved (p = 0.03). Also, patients whose epiphyses could not be preserved had more metastases at diagnosis. The global infection rate was 9%, the same as in other tumor operations in which chemotherapy and/or radiotherapy were used. In three cases (1.7%), the distraction occurred unexpectedly through the tumor.

Conclusions: Epiphysiolysis prior to resection of metaphyseal pediatric bone sarcomas provides better clinical outcomes in terms of survival and function than patients in whom the epiphysis could not be preserved.

Significance: This study shows the largest series of patients treated with epiphysiolysis prior to resection (Cañadell’s technique) of metaphyseal pediatric bone sarcomas during the last 40 years.

Use of serum biomarkers and cytokines to differentiate septic arthritis, osteomyelitis, and transient synovitis in pediatric and adolescent patients

Nichelle Enata, Kirsten Brouillet, Ling Chen, Kim Quayle, Scott J. Luhmann

Washington University School of Medicine, St Louis, MO, USA

LOE-Diagnostic-Level II

Purpose: Differentiating transient synovitis (TS), osteomyelitis (OM), and septic arthritis (SA) in the pediatric population remains a clinical challenge, often resulting in the use of painful, invasive procedures and advanced imaging, which can delay diagnosis and management. The purpose of this study is to develop a serum-based panel of biomarkers and cytokines to accurately diagnose/differentiate TS, OM, and SA at initial presentation.

Methods: A prospective study was completed over an 8-year period, collecting serum samples at the initial evaluation at a tertiary care children’s hospital in patients with the working diagnosis of a possible musculoskeletal infection. Each sample was analyzed for 103 distinct biomarkers and cytokines using enzyme-linked immunosorbent assay (ELISA) testing. Final diagnoses were classified based on clinical scenario, arthrocentesis results, and post-operative synovial fluid cultures. Linear discriminant analysis was performed to identify limited sets of predictive biomarkers and cytokines to accurately diagnose SA, OM, and TS.

Results: A total of 164 pediatric patients whose working diagnosis included a musculoskeletal infection were identified at a single institution over an 8-year period. Average age of SA patients was 5.2 years (0–11 years) with 68% male. Average age of OM patients was 8.1 years (0–18 years) with 59% male. Average age of TS patients was 5.4 years (0–17 years) with 56% male. Twenty-six biomarkers were identified that demonstrated significant differences between SA, OM, and TS. To differentiate SA and TS, a panel of three biomarkers and cytokines was compiled (COMP, TIMP-1, CTACK), which was able to classify 61/69 cases properly (88% accuracy). To differentiate SA and OM, a separate three-panel combination was compiled (COMP, Fractalikine, EGF), which was able to classify 49/64 cases properly (77% accuracy).

Conclusions: The creation of serum-based biomarker/cytokines panels aimed to improve diagnostic accuracy and decrease delays in management in the diagnostic process, in the pediatric population. This study identified two panels of three serum-based biomarkers/cytokines that can accurately differentiate between SA, OM, and TS at initial presentation. Differentiation between TS and SA was 88% accurate, and OM-SA was 77%.

Significance: Serum-based biomarker/cytokine panels can differentiate TS-SA with 88% accuracy and SA-OM with 77% accuracy.

Kicking the can in DDH: the impact of age on outcomes following secondary reconstructive surgery for residual dysplasia

Shamrez Haider, Laura M. Mayfield, Corey Gill, Harry K.W. Kim, Daniel J. Sucato, David A. Podeszwa, William Zachary Morris

Scottish Rite for Children, Dallas, TX, USA

LOE-Therapeutic-Level III

Purpose: Following closed reduction of developmental hip dislocations, residual dysplasia is common with rates of secondary reconstructive surgery approaching 50%–60%. The determination and timing of when to proceed with surgery is difficult as acetabular remodeling occurs gradually over the first years following closed reduction. The purpose of this study was to evaluate how age at secondary reconstructive surgery influences the clinical and radiographic outcomes following pelvic osteotomy for residual dysplasia.

Methods: Following institutional review board (IRB) approval, we retrospectively reviewed all Salter or Pemberton pelvic osteotomies performed for residual dysplasia following index closed reduction at a single institution between 1983 and 2020. Patients with concurrent femoral osteotomy, femoral head deformity on follow-up radiographs, or patients with <2 years radiographic follow-up were excluded. Acetabular index (AI) and migration index (MI) were measured on pre-operative, immediate post-operative, and 2-year follow-up radiographs. Following triradiate cartilage closure, final follow-up (FFU) anterior-posterior pelvis radiographs were measured for lateral center-edge angle (LCEA), Tönnis angle, MI, and lateralization ratio (LR). Univariate and multivariate analyses and spearman’s correlation coefficients were used to evaluate outcomes based on age at the time of surgery. Significance was set at p < 0.05.

Results: Eighty-four hips from 77 patients were included, including 52 Salter osteotomies (62%) and 32 Pemberton osteotomies (38%), with mean age of surgery 5.19 ± 1.57 years. There was no significant difference in markers of acetabular dysplasia at 2 years postop or after triradiate closure between hips treated before or after age 5 (all p > 0.05). A higher-than-expected number of patients (41/84, 49%) across both treatment groups demonstrated residual dysplasia at FFU with LCEA < 25 degrees (Figure). However, there was no difference in the rate of residual dysplasia at triradiate closure between hips treated before or after age 5 (48% versus 51%, p = 0.19). There was a weak negative linear correlation between age at pelvic osteotomy and FFU LCEA (−0.23, p = 0.04) across both treatment groups. However, there was no correlation between age at surgery and Tönnis angle, MI, or LR (all p > 0.05). In addition, age at the time of surgery was not a significant predictor of LCEA on multivariate analysis (p = 0.16).

Conclusions: We demonstrated high rates of residual dysplasia at triradiate closure following closed reduction and subsequent pelvic osteotomy. However, the age at secondary reconstructive surgery for residual dysplasia does not significantly influence the radiographic markers of dysplasia at triradiate closure.

Significance: These findings may guide counseling for families and suggest there is no opportunity cost to short-term continued observation while monitoring for acetabular remodeling.

Late-diagnosed DDH is rare in Finland with universal clinical screening program complemented with selective ultrasonography

Emma Luoto, Jenni Katariina Jalkanen, Ilari Kuitunen, Reijo Sund, Aarno Yrjana Nietosvaara

Kuopio University Hospital, Kuopio, Finland

LOE-Not Applicable-Level IV

Purpose: There is an ongoing debate regarding whether universal ultrasound screening should be used to detect developmental dysplasia of the hip (DDH) in newborn. The aim of our study was to assess the effectiveness of universal clinical screening complemented with selective ultrasound used in our country.

Methods: In Finland, pediatricians examine all newborns’ hips in the maternity hospital, and refer infants with suspected DDH to pediatric orthopedic outpatient clinic for a re-examination at 2 weeks of age. General practitioners conduct hip examinations on all infants at 4–6 weeks, at 4 months and at 1.5 years of age in child welfare clinics and refer the children with suspected DDH expectantly to a pediatric orthopedic service. Hip ultrasound is performed at 8 weeks of age to children with suspected or diagnosed DDH. For this retrospective cohort study, we collected the number of children under the age of 15 years that were given DDH (ICD-10 codes Q65.0-6) as the major diagnosis in three or more visits. The data were obtained from The Finnish Care Register for Health Care (HILMO), which collects ICD-10 codes of every medical appointment. We calculated the annual incidence per 1000 newborns of a given DDH diagnosis between 2002 and 2021. Late detected DDH was defined as children aged between 6 months and 15 years at the initial diagnosis who received invasive treatment.

Results: During the 20-year-long study period, altogether 1,103,269 babies were born (median per year = 57,214, range per year = 45,346–60,694). A total of 6421 children were given DDH as a diagnosis (mean per year = 321, range per year = 193–405), giving a mean calculated incidence of 5.8 per 1000 (95% confidence interval (CI) = 5.7–6.0) newborns. The female-to-male incidence rate of DDH was 4.4 (95% CI, 4.3–4.5). Altogether 120 (0.1 per 1000 newborns (95% CI = 0.09–0.1)) children aged between 6 months and 15 years were treated for DDH, with little yearly variation (0.04–0.2 per 1000 newborns). The incidences of late-detected cases varied between 0.07 and 0.2 per 1000 newborns in the five University Hospital Specific Catchment Areas.

Conclusions: The risk of late diagnosis of DDH leading to invasive treatment was 0.01% in Finland without a universal ultrasound screening program.

Significance: Finland’s universal clinical DDH screening of neonates complemented with selective ultrasonography is effective, resulting in one of the lowest reported nationwide incidences of late diagnosed DDH globally suggesting that universal ultrasound screening is unnecessary in our country.

Navigation versus fluoroscopy for anterior VBT screw placement, analysis of 530 screws with confirmatory 3D imaging

Chunho Chen, Jimmy Daher, A. Noelle Larson, Todd A. Milbrandt, Lawrence L. Haber

Ochsner Hospital for Children, New Orleans, LA, USA

LOE-Therapeutic-Level III

Purpose: Vertebral body tethering (VBT) is used to treat scoliosis in skeletally immature patients. Especially with MIS techniques, screw placement can be challenging, and ramifications of screw malposition are significant. The techniques of intraoperative three-dimensional (3D)-imaging/navigation and fluoroscopic screw placement with confirmatory 3D fluoro spine can possibly reduce the rate of pedicle screw malposition. We compare 3D screw accuracy for anterior VBT screws placed with fluoroscopy/check spin versus computed tomography (CT)-guided navigation.

Methods: Retrospective review of the two experienced VBT centers that routinely used either intraoperative CT-guided navigation or fluoroscopic screw insertion followed by check 3D fluoro scan (similar to CT), prior to cord placement, to confirm the screw positions after MIS surgery for VBT. Twenty-two of 133 patients who underwent CT-guided navigation had a postoperative CT scan. Fifty-five patients underwent fluoroscopic-guided screw insertions followed by intraoperative 3D fluoroscopic scan. Images of screws were assessed for appropriate length (no more than 2.5 mm short or beyond the far cortex) and appropriate placement within the vertebral body. Screw accuracy and rate of intraoperative screw revision rate were evaluated.

Results: A total of 530 screws had axial imaging available, including 385 fluoroscopically guided screws and 145 CT-navigated screws. The rate of ideal length screws (no more than 2.5 mm short or beyond the far cortex) was 125/144 (87%) in navigated group and 356/385 (92%) in the check spin group (p = 0.07), respectively. The mean screw length protruding from the cortex were 0.91 mm in navigated group and 0.62 mm in fluoroscopic group (p = 0.02). Ten out of 380 screws were revised intraoperatively after 3D spinsin, prior to cord placement in the fluoroscopic-guided group while no screw was revised in the navigated group. There were no vascular injuries, neurologic monitoring events, or cerebral spinal fluid leaks in either group.

Conclusions: Both CT navigation and use of a check spin are effective strategies to ensure accurate screw position and length. Fluoroscopic guidance might increase the radiation exposure to the surgeons but gives results while still in the OR, prior to cord placement, so revision of screws is simple and during the initial surgery. It also eliminates the need for an incision for the reference Array.

Significance: This is the first study that analyzes the accuracy of navigated and fluoroscopic VBT screws with confirmatory axial imaging. Both fluoroscopy- and CT-guided navigation have high accuracy in VBT surgery that may help lessen the learning curve and improve safety in VBT surgery.

Growth modulation response in thoracic VBT depends primarily on magnitude of concave vertebral body growth

Craig R. Louer, V. Salil Upasani, Jennifer Hurry, Hui Nian, Christine L. Farnsworth, Peter O. Newton, Stefan Parent, Pediatric Spine Study Group, Ron El-Hawary

Vanderbilt University Medical Center, Nashville, TN, USA

LOE-Therapeutic-Level III

Purpose: There is variability in clinical outcomes with vertebral body tethering (VBT) partly due to a limited understanding of the growth modulation (GM) response. We used the largest sample of patients with three-dimensional (3D) spine reconstructions to characterize the changes in disk and vertebra morphology that accompany GM during the first 2 years following VBT.

Methods: A multicenter registry was used to identify adolescent idiopathic scoliosis (AIS) patients who underwent VBT with ≥2 years of follow-up. Calibrated biplanar X-rays obtained at longitudinal time points (pre-op, post-op, and 2 years) underwent 3D reconstruction and subsequent analysis with custom MATLAB software to obtain precision measurements of apical 3 vertebra and 2 disks. GM was defined as change in instrumented Cobb from post-op to 2 years. Groupings based on GM magnitude were compared to determine which morphological changes are responsible for the GM response.

Results: Fifty patients (mean age: 12.5 ± 1.3 years) were analyzed over mean 27.7 months. GM ranged from 46.9° correction to 28.2° loss of correction. Improvement in instrumented Cobb from post-op to 2 years was positively correlated with concave vertebra height growth (r = 0.57, p < 0.001), 3D spine length growth (r = 0.36, p = 0.008), and decreased convex disk height (r = −0.42, p = 0.002). Fifty-four percent of patients were Neutral Modulators (“NM”; 10°> GM > −10°), 36% were High Modulators (“+M”; GM > 10°), and 10% were Poor Modulators (“−M”; −10°> GM[LC1]°). +M patients experienced an additional 1.6 mm (229% increase) of mean concave vertebra growth during study period compared to the −M group, (2.3 versus 0.7 mm, p = 0.039), while convex vertebra height growth was similar (1.3 versus 1.4 mm, p = 0.91, Figure 1).

Conclusions: The mechanism for initial correction and subsequent GM has been further elucidated. If GM occurs, vertebra body heights increase asymmetrically during the GM phase. A strong GM response is correlated with concave vertebral body growth and overall instrumented spine growth. A poor GM response is associated with increase in convex disk height (suspected tether rupture). Future aims will investigate the patient and technique-specific factors which influence increased growth remodeling.

Significance: This is the largest such study on the growth modulation phenomenon of VBT using 3D reconstructions. VBT induces vertebra remodeling by fostering concave vertebra growth, not by inhibiting convex vertebra growth as often hypothesized.

Spontaneous correction of the thoracic curve in Lenke 5 patients: lumbar vertebral body tether (VBT) versus posterior fusion

Jennifer Marie Bauer, Suken A. Shah, Jaysson T. Brooks, Baron S. Lonner, Amer F. Samdani, Firoz Miyanji, Peter O. Newton, Burt Yaszay, Harms Study Group

Seattle Children’s Hospital, Seattle, WA, USA

LOE-Therapeutic-Level II

Purpose: Vertebral body tethering (VBT) is a non-fusion option for skeletally immature patients with idiopathic scoliosis. As with posterior spinal fusion (PSF), compensatory curves are not commonly included in the construct. Prior studies demonstrated spontaneous correction of the compensatory thoracic curve after selective lumbar fusion, which can guide decision making of instrumented levels. However, no prior studies have examined thoracic curve correction after lumbar VBT. We hypothesize a smaller spontaneous correction of the unoperated thoracic curve from lumbar VBT compared to lumbar fusion because of decreased three-dimensional (3D) correction with VBT.

Methods: Retrospectively compared prospectively collected multicenter cohort study.

Results: Twenty-four VBT and 24 fusion patients were matched 1:1 for upper instrumented vertebra (UIV; T8-11), lowest instrumented vertebra (LIV; L2-3), and thoracic Cobb with no differences between pre-operative thoracic or lumbar curves. Fusion patients were 1.4 years older (p = 0.008). There were no significant differences between VBT and PSF for average pre-op or 2 years post-op major T or L curves, T1 tilt, or coronal balance. VBT had 4° less PJK and 10° less L lordosis at 2 years; PSF had 12 mm better coronal balance (Table 1). Thoracic curves worsened an insignificant amount from first erect to 2 years (2° in VBT, <1° in PSF). Compared to pre-operative flexibility radiographs, 2-year post-op thoracic curves were 6.2° (VBT) and 7.0° (PSF) larger (p = 0.83). There were 7 (24%) reoperations in the VBT group: 2 overcorrections relaxed, 2T adding-on (extended to T by PSF-1, VBT-1), 1 broken tether converted to PSF, and 1 (4%) reoperation in the fusion group (pseudarthrosis/broken screw). No other major complications.

Conclusions: Spontaneous thoracic correction is achieved to a similar degree for lumbar VBT and PSF patients operatively treated for their lumbar curves. There is little change in thoracic curve magnitude over time, and, on average, the correction does not reach the pre-operative flexibility curve measurement. There is a higher rate of revision in VBT patients to achieve these outcomes.

Significance: The uninstrumented thoracic curve spontaneously corrects to a similar degree in either tethered or fused lumbar spines, but to less than on the pre-operative flexibility imaging. This should be considered when deciding on whether to include the thoracic curve in a main lumbar VBT or PSF.

Are outcomes improving for AIS following FDA HDE approval?

Lawrence L. Haber, Melanie E. Boeyer, Daniel G. ho*rnschemeyer, Samantha C. Ahrens, Julia Todderud, Todd A. Milbrandt, Susan Scariano, Nicole Tweedy, A. Noelle Larson

Ochsner Hospital for Children, New Orleans, LA, USA

LOE-Therapeutic-Level III

Purpose: The US Food and Drug Administration (FDA) first approved vertebral body tethering (VBT) for AIS in August 2019. Published data thus far are from cohorts treated prior to approval 9/2019 with off-label products (first-generation data (1G)). We sought to evaluate second-generation outcomes (2G) from three experienced centers using a consecutive series of adolescent idiopathic scoliosis (AIS) patients, within FDA indications, treated with VBT for AIS. We hypothesized that 2G results would be superior to 1G due to the use of an on-label device and instrumentation and experience at these centers.

Methods: Multicenter retrospective review of perioperative and postoperative outcomes following VBT. Patient series were consecutive and from three experienced centers. Inclusion criteria were curves between 35 and 65 degrees, skeletally immature patients, and minimum follow-up (fu) of 2 years.

Results: A total of 92 subjects with 104 treated curves were included. Mean initial Cobb 52 (38–65) with mean fu 28 months (21–43). Mean coronal Cobb angles for the first post op visit (FV), 1- and 2-year time points were 26 (9–46), 21 (0–41), and 24 (3–49). Thirty-three patients > 3-year fu with a mean Cobb of 26 (1–40). At most recent fu 88% < 35 degrees. There was 1 fusion (1%) and 9 additional interventions 10%, including 4 reoperations for overcorrection, 4 surgeries for instrumentation complications, and 1 aspiration for pulmonary effusion. The overall additional intervention rate was 11%. Tether rupture 27%. There were 0 neurologic injuries or infections.

Conclusions: For this series of 92 AIS patients treated with an FDA-approved VBT, 2- to 4-year post op outcomes were improved with only 1% fusion rate and 88% of curves < 35. Complications occurred in 11% of patients, which included four reoperations for overcorrection. Good correction on first erect radiograph appears to be a predictor for success. Compared to historical series, second-generation data show promising improvements in outcomes and decreasing rate of complications/reoperations to the original first-generation cohorts.

Significance: This cohort of 92 patients treated with an FDA-approved VBT for AIS from three experienced centers shows improved outcomes, higher success rates, one fusion, and lower reoperation rates than earlier studies using off-label devices. First erect Cobb is likely a predictor for success. Cord rupture continues to be a concern. Larger cohorts and longer follow-up is critical to fully evaluate VBT for AIS.

Outcomes in patients with tether rupture after anterior vertebral tethering (AVT) for adolescent idiopathic scoliosis: the good, the bad, and the ugly

John T. Braun, Sofia Federico, David Lawlor, Brian E. Grottkau