Osteotomies At The Time of Graduation Surgery: How Much Do We Get from Them?

STUDY DESIGN: Retrospective, Multicenter. OBJECTIVE: Determine if posterior column osteotomies (PCO) at time of conversion from growth friendly instrumentation (GFI) to definitive fusion in early onset scoliosis (EOS) graduates impacts outcomes. SUMMARY OF BACKGROUND DATA: Increasing spinal rigidity following treatment of EOS with GFI can limit curve correction at time of conversion to definitive spinal fusion. PCO are often employed at the time of fusion to improve flexibility. This technique's efficacy has not been studied. METHODS: EOS patients with GFI undergoing conversion to fusion were grouped by those that did or did not have PCO. Patients with inadequate radiographs, <2 years follow-up, or three-column osteotomies at time of fusion were excluded. RESULTS: 832 patients met inclusion criteria. 175 (21%) patients had PCO. Age at index surgery was younger (6.6 vs. 7.4 y, P =0.0009) and the mean duration of GFI was greater (6.2 vs. 5.5 y, 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). On average 4.4 osteotomies (range: 1 - 12) were performed for the PCO group and EBL (PCO=820 cc vs no PCO=752 cc, P<0.01) and surgical time (PCO=403 min vs no PCO=349 min, P<0.01) were greater. Postoperatively, mean curve correction (PCO=16.6, no PCO=14.4 degrees, P=0.18) was similar. Accounting for preoperative curve magnitude, there was a relationship between number of PCOs and curve correction (P=0.04). There was no relationship between degrees of correction per osteotomy and duration of GFI (P=0.12). Postoperative complications at 2 years were similar (PCO=25% vs no PCO=27%, P=0.63). CONCLUSION: EOS graduates achieve minimal correction at time of conversion regardless of whether PCOs are performed. PCOs increase EBL and operative time but have a similar complication rate. More PCOs resulted in more correction, though less than that anticipated in a previously uninstrumented spine. LEVEL OF EVIDENCE: III.

Proximal foundation anchor variations and their correlation with unplanned return to the operating room (UPROR) in children with EOS treated with magnetically controlled growing rods (MCGR).

INTRODUCTION: The evolution of MCGR technique has led to modifications in the configuration of the proximal construct to decrease the incidence of implant-related complications (IRC) and revision surgeries. However, there is no data characterizing the performance of the most used configurations reducing the risk of complications. METHODS: 487 patients were identified from an international multicenter EOS database. INCLUSION CRITERIA: EOS patients, primary dual MCGR, complete radiographs, and minimum of 2-year follow-up. 76 patients had incomplete X-rays, 5 had apical fusions, and 18 had inconclusive complications, leaving 388 patients for review. A digital spine template was created to document UIV; number of levels; number, type, and location of anchors; as well as implant configuration. First available postoperative and latest follow-up radiographs were reviewed by two senior surgeons and two spine fellows. UPROR due to IRC was defined as any change in proximal anchors between the postoperative and final follow-up radiographs. RESULTS: The most common proximal construct configuration: UIV at T2 (50.0%) with 17.5% UPROR, followed by T3 (34.0%) with 12.1% UPROR; number of levels was three (57.1%) with 16.8% UPROR and two (26.0%) with 17.0% UPROR; number of proximal anchors was six (49.9%) with 14.1% UPROR and four (27.0%) with 18.3% UPROR. The most common anchors were all screws (42.0%) with 9.9% UPROR, and all hooks (26.4%) with 31.4% UPROR (P < 0.001). The construct with the lowest rate of UPROR was a UIV at T2, with six anchors (all screws) across three levels (42 cases), with 0% UPROR. Other construct combinations that yielded 0% UPROR rates were UIV of T3, six anchors (all screws) across three levels (25 cases), and a UIV of T3 with six anchors (screws and hooks) across three3 levels (9 cases). CONCLUSION: Proximal anchor configuration impacts the incidence of UPROR due to IRC in MCGR. UIV at T2 and T3 compared to T4, and the use of all screws or combination of screws and hooks compared to all hooks were associated with a lower UPROR rate. The most common construct configuration was T2 UIV, three levels, six anchors, and all screws. The use of a combination of six anchors (screws or screws and hooks) across three levels with a UIV at T2 or T3 was associated with a lower UPROR rate. Additional research is needed to further evaluate the variables contributing to configuration selection and their association with IRC.

Association between structural rib autograft and the rate of arthrodesis in children undergoing occiput-C2 instrumentation and fusion.

OBJECTIVE: The purpose of this study was to identify factors associated with fusion success among pediatric patients undergoing occiput-C2 rigid instrumentation and fusion. METHODS: The Pediatric Spine Study Group registry was queried to identify patients ≤ 21 years of age who underwent occiput-C2 posterior spinal rigid instrumentation and fusion and had a 2-year minimum clinical and radiographic (postoperative lateral cervical radiograph or CT scan) follow-up. Fusion failure was defined clinically if a patient underwent hardware revision surgery > 30 days after the index procedure or radiographically by the presence of hardware failure or screw haloing on the most recent follow-up imaging study. Univariate comparisons and multivariable logistic regression analyses were subsequently performed. RESULTS: Seventy-six patients met inclusion criteria. The median age at surgery was 9 years (range 1.5-17.2 years), and 51% of the cohort was male. Overall, 75% of patients had syndromic (n = 41) or congenital (n = 15) etiologies, with the most frequent diagnoses of Down syndrome (28%), Chiari malformation (13%), and Klippel-Feil syndrome (12%). Data were available to determine if there was a fusion failure in 97% (74/76) of patients. Overall, 38% (28/74) of patients had fusion failure (95% CI 27%-50%). Univariate analysis demonstrated that use of a rigid cervical collar postoperatively (p = 0.04) and structural rib autograft (p = 0.02) were associated with successful fusion. Multivariable logistic regression analysis determined that patients who had rib autograft used in surgery had a 73% decrease in the odds of fusion failure (OR 0.27, 95% CI 0.09-0.82; p = 0.02). Age, etiology including Down syndrome, instrumentation type, unilateral instrumentation, use of recombinant human bone morphogenetic protein, and other variables did not influence the risk for fusion failure. CONCLUSIONS: In this multicenter, multidisciplinary, international registry of children undergoing occiput-C2 instrumentation and fusion, fusion failure was seen in 38% of patients, a higher rate than previously reported in the literature. The authors' data suggest that postoperative immobilization in a rigid cervical collar may be beneficial, and the use of structural rib autograft should be considered, as rib autograft was associated with a 75% higher chance of successful fusion.

Predicting radiographic outcomes of vertebral body tethering in adolescent idiopathic scoliosis patients using machine learning.

Anterior Vertebral Body Tethering (AVBT) is a growing alternative treatment for adolescent idiopathic scoliosis (AIS), offering an option besides spinal fusion. While AVBT aims to correct spinal deformity through growth correction, its outcomes have been mixed. To improve surgical outcomes, this study aimed to develop a machine learning-based tool to predict short- and midterm spinal curve correction in AIS patients who underwent AVBT surgery, using the most predictive clinical, radiographic, and surgical parameters. After institutional review board approval and based on inclusion criteria, 91 AIS patients who underwent AVBT surgery were selected from the Shriners Hospitals for Children, Philadelphia. For all patients, longitudinal standing (PA or AP, and lateral) and side bending spinal Radiographs were retrospectively obtained at six visits: preop and first standing, one year, two years, five years postop, and at the most recent follow-up. Demographic, radiographic, and surgical features associated with curve correction were collected. The sequential backward feature selection method was used to eliminate correlated features and to provide a rank-ordered list of the most predictive features of the AVBT correction. A Gradient Boosting Regressor (GBR) model was trained and tested using the selected features to predict the final correction of the curve in AIS patients. Eleven most predictive features were identified. The GBR model predicted the final Cobb angle with an average error of 6.3 ± 5.6 degrees. The model also provided a prediction interval, where 84% of the actual values were within the 90% prediction interval. A list of the most predictive features for AVBT curve correction was provided. The GBR model, trained on these features, predicted the final curve magnitude with a clinically acceptable margin of error. This model can be used as a clinical tool to plan AVBT surgical parameters and improve outcomes.

Intraoperative neuromonitoring predicts postoperative deficits in severe pediatric spinal deformity patients.

PURPOSE: To evaluate intraoperative monitoring (IOM) alerts and neurologic deficits during severe pediatric spinal deformity surgery. METHODS: Patients with a minimum Cobb angle of 100° in any plane or a scheduled vertebral column resection (VCR) with minimum 2-year follow-up were prospectively evaluated (n = 243). Preoperative, immediate postoperative, and 2-year postoperative neurologic status were reported. Radiographic data included preoperative and 2-year postoperative coronal and sagittal Cobb angles and deformity angular ratios (DAR). IOM alert type and triggering event were recorded. SRS-22r scores were collected preoperatively and 2-years postoperatively. RESULTS: IOM alerts occurred in 37% of procedures with three-column osteotomy (n = 36) and correction maneuver (n = 32) as most common triggering events. Patients with IOM alerts had greater maximum kyphosis (101.4° vs. 87.5°) and sagittal DAR (16.8 vs. 12.7) (p < 0.01). Multivariate regression demonstrated that sagittal DAR independently predicted IOM alerts (OR 1.05, 95% CI 1.02-1.08) with moderate sensitivity (60.2%) and specificity (64.8%) using a threshold value of 14.3 (p < 0.01). IOM alerts occurred more frequently in procedures with new postoperative neurologic deficits (17/24), and alerts with both SSEP and TCeMEP signals were associated with new postoperative deficits (p < 0.01). Most patients with new deficits experienced resolution at 2 years (16/20) and had equivalent postoperative SRS-22r scores. However, patients with persistent deficits had worse SRS-22r total score (3.8 vs. 4.2), self-image subscore (3.5 vs. 4.1), and function subscore (3.8 vs. 4.3) (p ≤ 0.04). CONCLUSION: Multimodal IOM alerts are associated with sagittal kyphosis, and predict postoperative neurologic deficits. Most patients with new deficits experience resolution of their symptoms and have equivalent 2-year outcomes. LEVEL OF EVIDENCE: II.

Baclofen pumps do not increase risk of complications following spinal fusion.

PURPOSE: To assess the complication risks associated with intrathecal baclofen (ITB) pumps in cerebral palsy (CP) patients undergoing posterior spinal fusion (PSF) and to determine if timing of pump implantation before or during PSF impacts the risk of complications. METHODS: A prospectively collected multicenter database was retrospectively reviewed to identify CP patients undergoing PSF from 2008 to 2023. Patients were divided into 2 cohorts: those with an ITB pump (ITB cohort) and those without (non-ITB cohort). The ITB cohort was further categorized by placement of the pump prior to or during PSF. Cohorts were then compared in terms of postoperative complications, perioperative complications, and need for revision surgery. RESULTS: Four hundred six patients (ITB n = 79 [53 prior to, 26 during PSF], non-ITB n = 326) were included in this analysis. At an average follow-up of 4.0 years (range 2-10 years), there were no significant differences between the ITB and non-ITB cohorts in the rate of perioperative complications (5.0% vs 6.5%, p = 0.80), revision surgeries (2.5% vs 4.6%, p = 0.54), or any complication type, regardless of whether pumps were placed prior to or during PSF, aside from longer surgical times in the latter group. CONCLUSION: Complication rates are similar for ITBs placed prior to and during PSF. Patients with spastic CP may safely be treated with ITB pumps without increased risks of complication or further reoperation/revision following PSF. LEVEL OF EVIDENCE: Level III.

Fifteen Years of Spinal Fusion Outcomes in Children With Cerebral Palsy: Are We Getting Better?

STUDY DESIGN: Retrospective multicenter study. OBJECTIVE: We reviewed 15-year trends in operative factors, radiographic and quality of life outcomes, and complication rates in children with cerebral palsy (CP)-related scoliosis who underwent spinal fusion. SUMMARY OF BACKGROUND DATA: Over the past two decades, significant efforts have been made to decrease complications and improve outcomes of this population. MATERIALS AND METHODS: We retrospectively reviewed a multicenter registry of pediatric CP patients who underwent spinal fusion from 2008 to 2020. We evaluated baseline and operative, hospitalization, and complication data as well as radiographic and quality of life outcomes at a minimum 2-year follow-up. RESULTS: Mean estimated blood loss and transfusion volume declined from 2.7±2.0 L in 2008 to 0.71±0.34 L in 2020 and 1.0±0.5 L in 2008 to 0.5±0.2 L in 2020, respectively, with a concomitant increase in antifibrinolytic use from 58% to 97% (all, P <0.01). Unit rod and pelvic fusion use declined from 33% in 2008 to 0% in 2020 and 96% in 2008 to 79% in 2020, respectively (both, P <0.05). Mean postoperative intubation time declined from 2.5±2.6 to 0.42±0.63 days ( P< 0.01). No changes were observed in preoperative and postoperative coronal angle and pelvic obliquity, operative time, frequency of anterior/anterior-posterior approach, and durations of hospital and intensive care unit stays. Improvements in the Caregiver Priorities and Child Health Index of Life with Disabilities postoperatively did not change significantly over the study period. Complication rates, including reoperation, superficial and deep surgical site infection, and gastrointestinal and medical complications remained stable over the study period. CONCLUSIONS: Over the past 15 years of CP scoliosis surgery, surgical blood loss, transfusion volumes, duration of postoperative intubation, and pelvic fusion rates have decreased. However, the degree of radiographic correction, the rates of surgical and medical complications (including infection), and health-related quality of life measures have broadly remained constant.

Results of posterior spinal fusion after failed anterior vertebral body tethering.

PURPOSE: In patients with adolescent idiopathic scoliosis (AIS) undergoing anterior vertebral tethering (AVBT), some will subsequently require posterior spinal fusion (PSF). Limited data exist on clinical and radiographic outcomes of fusion after tether failure. METHODS: 490 patients who underwent AVBT were retrospectively analyzed. Twenty patients (4.1%) subsequently underwent conversion to PSF. A control group of patients with primary PSF (no previous AVBT) was matched for comparison. Data were compared using paired t-tests and Fisher Exact Tests. RESULTS: There was a significant increase in estimated blood loss (EBL) (p = 0.002), percent estimated blood volume (%EBV) (p = 0.013), operative time (p = 0.002), and increased amount of fluoroscopy (mGy) (p = 0.04) as well as number of levels fused (p = 0.02) in the AVBT conversion group compared to primary fusion. However, no difference was found in implant density (p = 0.37), blood transfusions (p = 0.11), or intraoperative neuromonitoring events (p > 0.99). Both groups attained similar thoracic and lumbar percent correction (major coronal curve angle) from pre-op to the latest follow-up (thoracic p = 0.507, lumbar p = 0.952). CONCLUSION: A subset of patients with AVBT will require conversion to PSF. Although technically more challenging, revision surgery can be safely performed with similar clinical and radiographic outcomes to primary PSF.

Pain After Anterior Vertebral Body Tethering: Incidence, Risk Factors, and Timing.

STUDY DESIGN: Retrospective case series. OBJECTIVE: To examine the incidence and risk factors for postoperative pain following anterior vertebral body tethering (AVBT) for adolescent idiopathic scoliosis (AIS). SUMMARY OF BACKGROUND DATA: Up to 78% of patients with AIS report preoperative pain; it is the greatest patient concern surrounding surgery. Pain significantly decreases following posterior spinal fusion, but pain following AVBT is poorly understood. MATERIALS AND METHODS: We retrospectively reviewed 279 patients with a two-year follow-up after AVBT for AIS. We collected demographic, radiographic, and clinical data pertinent to postoperative pain at each time interval of preoperative and postoperative visits (6 wk, 6 mo, 1 y, and annually thereafter). RESULTS: Within our cohort, 68.1% of patients reported preoperative pain. Older age ( P =0.014) and greater proximal thoracic ( P =0.013) and main thoracic ( P =0.002) coronal curve magnitudes were associated with preoperative pain. Pain at any time point > 6 weeks postoperatively was reported in 41.6% of patients; it was associated with the female sex ( P =0.032), need for revision surgery ( P =0.019), and greater lateral displacement of the apical lumbar vertebrae ( P =0.028). The association between preoperative and postoperative pain trended toward significance ( P =0.07). At 6 months postoperatively, 91.8% had pain resolution; the same number remained pain-free at the time of last follow-up. The presence of a postoperative complication was associated with new-onset postoperative pain that resolved ( P =0.009). Only 8.2% had persistent pain, although no risk factors were found to be associated with persistent pain. CONCLUSION: In our cohort of 279 patients with a minimum 2-year follow-up after AVBT, 68.1% reported preoperative pain. Nearly 42% reported postoperative pain at any time point, but only 8.2% had persistent pain. Postoperative pain after AVBT was associated with female sex, revision surgery, and Lenke lumbar modifier. AVBT is associated with a significant reduction in pain, and few patients report long-term postoperative pain.

Is spinal height gain associated with rod orientation and the use of cross-links in magnetically controlled growing rods in early-onset scoliosis?

Optimal orientation for magnetically controlled growing rods (MCGRs) is unclear. The objective of this study was to investigate associations of rod orientation with implant-related complications (IRCs) and spinal height gains. Using an international early-onset scoliosis (EOS) database, we retrospectively reviewed 57 patients treated with dual MCGRs from May 2013 to July 2015 with minimum 2-year follow-up. Outcomes of interest were IRCs and left/right rod length gains and thoracic (T1-T12) and spinal (T1-S1) heights. We compared patients with two rods lengthened in the cephalad ( standard; n  = 18) versus opposite ( offset; n  = 39) directions. Groups did not differ in age, sex, BMI, duration of follow-up, EOS cause, ambulatory status, primary curve magnitude, baseline thoracic height, or number of distractions/year. We compared patients whose constructs used ≥1 cross-link (CL group; n  = 22) versus no CLs (NCL group; n  = 35), analyzing thoracic height gains per distraction ( α  = 0.05). Offset and standard groups did not differ in left or right rod length gains overall or per year or in thoracic or spinal height gain. Per distraction, the CL and NCL groups did not differ significantly in left or right rod length or thoracic or spinal height gain. Complications did not differ significantly between rod orientation groups or between CL groups. MCGR orientation and presence of cross-links were not associated with differences in rod length gain, thoracic height, spinal height, or IRCs at 2-year follow-up. Surgeons should feel comfortable using either MCGR orientation. Level of evidence: 3, retrospective.

Perioperative Complications and Health-related Quality of Life Outcomes in Severe Pediatric Spinal Deformity.

STUDY DESIGN: Prospective multicenter cohort study. OBJECTIVE: To evaluate perioperative complications and mid-term outcomes for severe pediatric spinal deformity. SUMMARY OF BACKGROUND DATA: Few studies have evaluated the impact of complications on health-related quality of life (HRQoL) outcomes in severe pediatric spinal deformity. METHODS: Patients from a prospective, multicenter database with severe pediatric spinal deformity (minimum of 100 degree curve in any plane or planned vertebral column resection (VCR)) with a minimum of 2-years follow-up were evaluated (n=231). SRS-22r scores were collected preoperatively and at 2-years postoperatively. Complications were categorized as intraoperative, early postoperative (within 90-days of surgery), major, or minor. Perioperative complication rate was evaluated between patients with and without VCR. Additionally, SRS-22r scores were compared between patients with and without complications. RESULTS: Perioperative complications occurred in 135 (58%) patients, and major complications occurred in 53 (23%) patients. Patients that underwent VCR had a higher incidence of early postoperative complications than patients without VCR (28.9% vs. 16.2%, P =0.02). Complications resolved in 126/135 (93.3%) patients with a mean time to resolution of 91.63 days. Unresolved major complications included motor deficit (n=4), spinal cord deficit (n=1), nerve root deficit (n=1), compartment syndrome (n=1), and motor weakness due to recurrent intradural tumor (n=1). Patients with complications, major complications, or multiple complications had equivalent postoperative SRS-22r scores. Patients with motor deficits had lower postoperative satisfaction subscore (4.32 vs. 4.51, P =0.03), but patients with resolved motor deficits had equivalent postoperative scores in all domains. Patients with unresolved complications had lower postoperative satisfaction subscore (3.94 vs. 4.47, P =0.03) and less postoperative improvement in self-image subscore (0.64 vs. 1.42, P =0.03) as compared to patients with resolved complications. CONCLUSION: Most perioperative complications for severe pediatric spinal deformity resolve within 2-years postoperatively and do not result in adverse HRQoL outcomes. However, patients with unresolved complications have decreased HRQoL outcomes.

What Happens After a Vertebral Body Tether Break? Incidence, Location, and Progression With Five-year Follow-up.

STUDY DESIGN: Retrospective review. OBJECTIVE: To study risk factors for anterior vertebral body tether (VBT) breakage. SUMMARY OF BACKGROUND DATA: VBT is used to treat adolescent idiopathic scoliosis in skeletally immature patients. However, tethers break in up to 48% of cases. MATERIALS AND METHODS: We reviewed 63 patients who underwent thoracic and/or lumbar VBT with a minimum five-year follow-up. We radiographically characterized suspected tether breaks as a change in interscrew angle >5°. Demographic, radiographic, and clinical risk factors for presumed VBT breaks were evaluated. RESULTS: In confirmed VBT breaks, the average interscrew angle change was 8.1°, and segmental coronal curve change was 13.6°, with a high correlation ( r =0.82). Our presumed VBT break cohort constituted 50 thoracic tethers, four lumbar tethers, and nine combined thoracic/lumbar tethers; the average age was 12.1±1.2 years and the mean follow-up was 73.1±11.7 months. Of 59 patients with thoracic VBTs, 12 patients (20.3%) had a total of 18 breaks. Eleven thoracic breaks (61.1%) occurred between two and five years postoperatively, and 15 (83.3%) occurred below the curve apex ( P <0.05). The timing of thoracic VBT breakage moderately correlated with more distal breaks ( r =0.35). Of 13 patients who underwent lumbar VBT, eight patients (61.5%) had a total of 12 presumed breaks. Six lumbar breaks (50%) occurred between one and two years postoperatively, and seven (58.3%) occurred at or distal to the apex. Age, sex, body mass index, Risser score, and curve flexibility were not associated with VBT breaks, but the association between percent curve correction and thoracic VBT breakage trended toward significance ( P =0.054). Lumbar VBTs were more likely to break than thoracic VBTs ( P =0.016). Seven of the patients with presumed VBT breaks (35%) underwent revision surgery. CONCLUSIONS: Lumbar VBTs broke with greater frequency than thoracic VBTs, and VBT breaks typically occurred at levels distal to the curve apex. Only 15% of all patients required revision. LEVEL OF EVIDENCE: 3.

Surgical Outcomes of Cerebral Palsy Patients With Scoliosis and Lumbar Hyperlordosis: A Comparative Analysis With 2-year Minimum Follow-up.

STUDY DESIGN: Retrospective review of a prospectively collected multicenter database. OBJECTIVE: To compare outcomes of patients with cerebral palsy (CP) who undergo surgery for scoliosis with normal lordosis (NL) versus hyperlordosis. SUMMARY OF BACKGROUND DATA: Surgical correction of scoliosis with lumbar hyperlordosis is challenging. Hyperlordosis may confer higher perioperative morbidity, but this is not well understood. MATERIALS AND METHODS: A multicenter database was queried for CP patients who underwent surgery from 2008 to 2017. The minimum follow-up was 2 years. Two groups were identified: lumbar lordosis <75° (NL) versus ≥ 75° hyperlordosis (HL). Perioperative, radiographic, and clinical outcomes were compared. RESULTS: Two hundred seventy-five patients were studied: 236 NL and 39 HL (-75 to -125°). The mean age was 14.1 years, and 52.4% were male. Patients with hyperlordosis had less cognitive impairment (76.9% vs. 94.0%, P =0.008) and higher CPCHILD scores (59.4 vs. 51.0, P =0.003). Other demographics were similar between the groups. Patients with hyperlordosis had greater lumbar lordosis (-90.5 vs. -31.5°, P <0.001) and smaller sagittal vertical axis (-4.0 vs. 2.6 cm, P <0.001). Patients with hyperlordosis had greater estimated blood loss (2222.0 vs. 1460.7 mL, P <0.001) but a similar perioperative complication rate (20.5% vs. 22.5%, P =0.787). Significant correction of all radiographic parameters was achieved in both groups. The HL group had postoperative lumbar lordosis of -68.2° and sagittal vertical axis of -1.0 cm. At a 2-year follow-up, patients with hyperlordosis continued to have higher CPCHILD scores and gained the greatest benefit in overall quality of life measures (20.0 vs. 6.1, P =0.008). The reoperation rate was 10.2%: implant failure (3.6%), pseudarthrosis (0.7%), and wound complications (7.3%). There were no differences in the reoperation rate between the groups. CONCLUSION: Surgical correction of scoliosis with hyperlordosis is associated with greater estimated blood loss but similar radiographic results, perioperative morbidity, and reoperation rate as normal lordosis. Patients with hyperlordosis gained greater overall health benefits. Correction of ≥25% of hyperlordosis seems satisfactory. LEVEL OF EVIDENCE: 3.

Risk factors for proximal junctional kyphosis after pediatric spinal deformity surgery with halo gravity traction.

PURPOSE: To determine risk factors for proximal junctional kyphosis (PJK) in pediatric patients with scoliosis undergoing halo gravity traction (HGT) prior to posterior spinal fusion (PSF). METHODS: Data from consecutive patients who underwent PSF after HGT with 2-year follow-up were retrospectively collected from a single center. Patients were divided into two groups: PJK vs. no PJK. RESULTS: Twenty-five patients (age 13.6 ± 3.1 years) underwent HGT for a mean of 42 ± 37 days. Eight patients (32%) developed radiographic PJK and 1 (4%) developed proximal junctional failure. PJK patients had greater loss of cervical lordosis (-17.4 ± 23.2 vs. 2.7 ± 16.2°, p < 0.05), greater increase in lumbar lordosis (9.3 ± 19.5 vs. -2.8 ± 12.8°; p = 0.034) during traction, and smaller overall major coronal curve angle correction (-16.8 ± 30.6 vs. -36.6 ± 16.4°; p = 0.026). From postoperative to last follow-up, PJK patients had a greater increase in upper end vertebrae (UEV) slope (3.3 ± 7.8 vs. -4.0 ± 7.7°; p = 0.004). Selection of UIV based on which vertebra was most level either pre-traction or in-traction was not associated with PJK development (p > 0.05). CONCLUSION: In-traction radiographic changes of decreased cervical lordosis, decreased major coronal curve, increased lumbar lordosis, and disruption of cervical sagittal balance at last follow-up may have implications for level selection and risk of PJK.

Trunk Range of Motion and Patient Outcomes After Anterior Vertebral Body Tethering Versus Posterior Spinal Fusion: Comparison Using Computerized 3D Motion Capture Technology.

BACKGROUND: Anterior vertebral body tethering (AVBT) for adolescent idiopathic scoliosis (AIS) is postulated to preserve motion compared with traditional posterior spinal fusion (PSF), but few studies exist to date. We used a validated computerized 3D model to compare trunk motion between patients treated with PSF and AVBT, and analyzed trunk motion in relation to the lowest instrumented vertebra (LIV). METHODS: This was a single-center retrospective review of a consecutive series of skeletally immature patients with AIS who underwent motion analysis prior to PSF (n = 47) or AVBT (n = 65) and 2 years postoperatively. Patients were divided into 4 groups on the basis of the LIV (≤L1, L2, L3, L4). Computerized 3D kinematic evaluations included thoracic and lumbar flexion, extension, side-bending, and rotation. Patient outcomes were assessed using the Scoliosis Research Society (SRS)-22 questionnaire. RESULTS: The LIV was ≤L1 in 48 patients treated with AVBT and 23 treated with PSF, L2 in 4 AVBT and 8 PSF patients, L3 in 10 AVBT and 8 PSF patients, and L4 in 3 AVBT and 8 PSF patients. PSF patients had a significant loss of motion in all 4 directions at 2 years postoperatively (e.g., flexion loss was 11° for ≤L1 to 30° for L4; p < 0.001). This equated to a 7° loss of trunk flexion per additional LIV level included in the fusion. AVBT patients only demonstrated loss of flexion and side-bending at 2 years postoperatively (e.g., flexion loss of 11° for L1 to 17° for L4; p < 0.001). Preoperative curve size and flexibility did not have any significant impact on differences in trunk motion between AVBT and PSF. SRS-22 scores were predominantly similar for AVBT versus PSF preoperatively and at 2 years postoperatively. CONCLUSIONS: Patients treated with AVBT experienced predominantly less motion loss compared with PSF patients at 2 years postoperatively. Patients treated with PSF demonstrated loss of motion in all planes that increased with each additional LIV from ≤L1 to L4, with 7° loss of flexion per additional LIV. However, the differences in total trunk motions were relatively modest for PSF and AVBT with an LIV of ≤L1. Preoperative curve magnitude and flexibility had no significant impact on trunk motion in either group. SRS-22 scores were similar for both groups at 2 years postoperatively. LEVEL OF EVIDENCE: Therapeutic Level III . See Instructions for Authors for a complete description of levels of evidence.

Posterior-only Resection of Single Hemivertebrae With 2-Level Versus >2-Level Fusion: Can We Improve Outcomes?

BACKGROUND: The outcomes of congenital scoliosis (CS) patients undergoing hemivertebra (HV) resection surgery with a 2-level fusion versus a >2-level fusion are unclear. We hypothesized that CS patients undergoing HV resection and a >2-level fusion have decreased curve progression and reoperation rates compared with 2-level fusions. METHODS: Retrospective review of prospectively collected data from a multicenter scoliosis database. Fifty-three CS patients (average age 4.5, range 1.2 to 10.9 y) at index surgery were included. Radiographic and surgical parameters, complications, as well as revision surgery rates were tracked at a minimum of 2-year follow-up. RESULTS: Twenty-six patients had a 2-level fusion while 27 patients had a >2-level fusion with similar age and body mass index between groups. The HV was located in the lumbar spine for 69% (18/26) 2-level fusions and 30% (8/27) >2-level fusions ( P =0.006). Segmental HV scoliosis curve was smaller in 2-level fusions compared to >2-level fusions preoperatively (38 vs. 50 degrees, P =0.016) and at follow-up (25 vs. 34 degrees, P =0.038). Preoperative T2-T12 (28 vs. 41 degrees, P =0.013) and segmental kyphosis (11 vs. 23 degrees, P =0.046) were smaller in 2-level fusions, but did not differ significantly at postoperative follow-up (32 vs. 39 degrees, P =0.22; 13 vs. 11 degrees, P =0.64, respectively). Furthermore, the 2 groups did not significantly differ in terms of surgical complications (27% vs. 22%, P =0.69; 2-level fusion vs. >2-level fusion, respectively), unplanned revision surgery rate (23% vs. 22%, 0.94), growing rod placement or extension of spinal fusion (15% vs. 15%, P =0.95), or health-related quality of life per the EOS-Questionnaire 24 (EOSQ-24). Comparison of patients with or without the need for growing rod placement or posterior spinal fusion revealed no significant differences in all parameters analyzed. CONCLUSIONS: Two-level and >2-level fusions can control congenital curves successfully. No differences existed in curve correction, proximal junctional kyphosis or complications between short and long-level fusion after HV resection. Both short and long level fusions are viable options and generate similar risk of revision. The decision should be individualized by patient and surgeon.

Complications following surgical treatment of adolescent idiopathic scoliosis: a 10-year prospective follow-up study.

BACKGROUND: Accurate reporting of long-term complications of surgical treatment of adolescent idiopathic scoliosis (AIS) is critical, but incomplete. This study aimed to report on the rate of complications following surgical treatment of AIS among patients with at least 10 years of follow-up. METHODS: This was a retrospective review of prospectively collected data from a multicenter registry of patients who underwent surgical treatment for AIS with minimum 10-year follow-up. Previously published complications were defined as major if they resulted in reoperation, prolonged hospital stay/readmission, neurological deficits, or were considered life-threatening. Rates and causes of reoperations were also reviewed. RESULTS: Two hundred and eighty-two patients were identified with mean age at surgery of 14.6 ± 2.1 years. Mean follow-up was 10.6 (range 9.5-14) years. Eighty-seven patients had anterior spinal fusion (ASF); 195 had posterior spinal fusion (PSF). The overall major complication rate was 9.9% (n = 28) in 27 patients. Among PSF patients, the complication rate was 9.7% (n = 19) in 18 patients. The complications were surgical site infection (37%), adding-on (26%), pulmonary (16%), neurologic (11%), instrumentation (5%), and gastrointestinal (5%). In ASF patients, the complication rate was 10.3% (n = 9) among nine patients. The complications were pulmonary (44%), pseudoarthrosis (22%), neurologic (11%), adding-on (11%), and gastrointestinal (11%). The reoperation rate was 6.0% (n = 17) among 17 patients. Although most of the complications presented within the first 2 years (60.7%), surgical site infection and adding-on were also seen late into the 10-year period. CONCLUSION: This is the largest prospective study with at least a 10-year follow-up of complications following spinal fusion for AIS, the overall major complication rate was 9.9% with a reoperation rate of 6.0%. Complications presented throughout the 10-year period, making long-term follow-up very important for surveillance. LEVEL OF EVIDENCE: Therapeutic II.

Development of consensus-based best practice guidelines for response to intraoperative neuromonitoring events in high-risk spinal deformity surgery.

PURPOSE: To expand on previously described intraoperative aids by developing consensus-based best practice guidelines to optimize the approach to intraoperative neuromonitoring (IONM) events associated with "high-risk" spinal deformity surgery. METHODS: Consensus was established among a group of experienced spinal deformity surgeons by way of the Delphi method. Through a series of iterative surveys and a final virtual consensus meeting, participants expressed their agreement (strongly agree, agree, disagree, and strongly disagree) with various items. Consensus was defined as ≥ 80% agreement ("strongly agree" or "agree"). Near-consensus was defined as ≥ 60% but < 80%. Equipoise was ≥ 20% but < 60%, and consensus to exclude was < 20%. RESULTS: 15 out of 15 (100%) invited surgeons agreed to participate. Final consensus supported inclusion of 105 items (53 in Response Algorithm, 13 in Ongoing Consideration of Etiology, 31 in Real-Time Data Scenarios, 8 in Patterns of IONM Loss), which were organized into a final set of best practice guidelines. CONCLUSION: Detailed consensus-based best practice guidelines and aids were successfully created with the intention to help organize and direct the surgical team in exploring and responding to neurological complications during high-risk spinal deformity surgery. LEVEL OF EVIDENCE: Level V.

Surgical outcomes of severe spinal deformities exceeding 100° or treated by vertebral column resection (VCR). Does implant density matter?: an observational study of deformity groupings.

STUDY DESIGN: Prospective multicenter international observational study. OBJECTIVE: To investigate the effect of implant density on clinical outcomes in complex pediatric spine reconstruction. Implant density in spine deformity surgery has been a subject of much debate with some authors advocating higher density for better correction. Few studies have looked at the effect of implant density on severe curves > 100 deg or treated with vertebral column resection (VCR). METHODS: 250/311 pts with 2-year f/u enrolled in the FOX pediatric database from 17 international sites were queried for the impact of implant density and surgical outcomes. Patients were grouped into three implant density categories for comparative analysis Group 1 (density ≤ 1), Group 2 (1 < density < 1.5) and Group 3 (density; 1.5-2). RESULTS: 250 pts: 47 (Grp1)/99 (Grp2) /104 (Grp3); Pre-op age and etiology and curve types were similar in all groups, but body mass index (BMI) was higher in Grp3. Grps 1 and 2 had significantly higher sagittal deformity angular ratio (S-DAR) compared to Grp 3 (p < 0.001). Pre-op Halo Gravity Traction (HGT) was used in 55.3%/44.4%/31.7%, p = 0.017; Grp1/Grp2/Grp3, respectively. Average duration of surgery (min) was higher in Grp3 relative to Grp1 only: 352.5/456.5/515.0, p = 0.0029. Blood loss was similar in all Grps. Rate of VCR, PSO and SPO was similar in all Grps. Pre-op Coronal Cobb avg 96.1/83.6/88.6, p = 0.2342, attained similar correction after HGT (24.6%/27.2%/23.2%, p = 0.4864. Coronal Cobb corrections at 2-year follow-up (FU) were (37.1%/40.3%/53.5%, p = 0.0004). Pre-op sagittal Cobb was (105.4/101.9/75.9, p < 0.01.), achieved similar %correction in HGT (19.1%/22.3%/22.5%, p = 0.6851) and at 2-year FU (39.6%/41.4%/29.8%, p = 0.1916). After adjusting for C-DAR, S-DAR, pre-op coronal and sagittal Cobb, etiology, curve types, age, BMI and number of rods in multivariate analysis, the odds of developing post-operative implant complication was 11 times greater in group 1 compared to group 3 (OR = 11.17,95% CI 2.34-53.32). There was significant improvement in SRS scores in all Grps at 2-year FU. CONCLUSION: Although higher implant density was observed to be associated with greater curve correction and lower rates of post-operative implant-related complication and revision in heterogeneous case groups, the results may not imply causality of implant density on the outcomes in severe pediatric spine reconstruction.

Establishing consensus: determinants of high-risk and preventative strategies for neurological events in complex spinal deformity surgery.

PURPOSE: To establish expert consensus on various parameters that constitute elevated risk during spinal deformity surgery and potential preventative strategies that may minimize the risk of intraoperative neuromonitoring (IONM) events and postoperative neurological deficits. METHODS: Through a series of surveys and a final virtual consensus meeting, the Delphi method was utilized to establish consensus among a group of expert spinal deformity surgeons. During iterative rounds of voting, participants were asked to express their agreement (strongly agree, agree, disagree, strongly disagree) to include items in a final set of guidelines. Consensus was defined as ≥ 80% agreement among participants. Near-consensus was ≥ 60% but < 80% agreement, equipoise was ≥ 20% but < 60%, and consensus to exclude was < 20%. RESULTS: Fifteen of the 15 (100%) invited expert spinal deformity surgeons agreed to participate. There was consensus to include 22 determinants of high-risk (8 patient factors, 8 curve and spinal cord factors, and 6 surgical factors) and 21 preventative strategies (4 preoperative, 14 intraoperative, and 3 postoperative) in the final set of best practice guidelines. CONCLUSION: A resource highlighting several salient clinical factors found in high-risk spinal deformity patients as well as strategies to prevent neurological events was successfully created through expert consensus. This is intended to serve as a reference for surgeons and other clinicians involved in the care of spinal deformity patients. LEVEL OF EVIDENCE: Level V.