Citations

OBJECTIVE Chordomas are rare tumors of the skull base and spine believed to arise from the vestiges of the embryonic notochord. These tumors are locally aggressive and frequently recur following resection and adjuvant radiotherapy. Proton therapy has been introduced as a tissue-sparing option because of the higher level of precision that proton-beam techniques offer compared with traditional photon radiotherapy. This study aimed to compare recurrence in patients with chordomas receiving proton versus photon radiotherapy following resection by applying tree-based machine learning models. METHODS The clinical records of all patients treated with resection followed by adjuvant proton or photon radiotherapy for chordoma at Mayo Clinic were reviewed. Patient demographics, type of surgery and radiotherapy, tumor recurrence, and other variables were extracted. Decision tree classifiers were trained and tested to predict long-term recurrence based on unseen data using an 80/20 split. RESULTS Fifty-three patients with a mean ± SD age of 55.2 ± 13.4 years receiving surgery and adjuvant proton or photon therapy to treat chordoma were identified; most patients were male. Gross-total resection was achieved in 54.7% of cases. Proton therapy was the most common adjuvant radiotherapy (84.9%), followed by conventional or external-beam radiation therapy (9.4%) and stereotactic radiosurgery (5.7%). Patients receiving proton therapy exhibited a 40% likelihood of having recurrence, significantly lower than the 88% likelihood observed in those treated with nonproton therapy. This was confirmed on logistic regression analysis adjusted for extent of tumor resection and tumor location, which revealed that proton adjuvant radiotherapy was associated with a decreased risk of recurrence (OR 0.1, 95% CI 0.01–0.71; p = 0.047) compared with photon therapy. The decision tree algorithm predicted recurrence with an accuracy of 90% (95% CI 55.5%–99.8%), with the lowest risk of recurrence observed in patients receiving gross-total resection with adjuvant proton therapy (23%). CONCLUSIONS Following resection, adjuvant proton therapy was associated with a lower risk of chordoma recurrence compared with photon therapy. The described machine learning models were able to predict tumor progression based on the extent of tumor resection and adjuvant radiotherapy modality used.

OBJECTIVE Use of bone morphogenetic protein (BMP)—an osteoinductive agent commonly used in lumbar arthrodesis—is off-label for cervical arthrodesis. This study aimed to identify the effect of BMP use on clinical and radiological outcomes in instrumented cervical arthrodesis. METHODS A comprehensive systematic review of the literature was performed to identify studies directly comparing outcomes between cervical arthrodeses with and without using BMP. Outcomes were analyzed separately for cases of anterior cervical discectomy and fusion (ACDF) and posterior cervical fusion (PCF). RESULTS A total of 20 studies with 5828 patients (1948 with BMP and 3880 without BMP) were included. In the ACDF cases, BMP use was associated with higher fusion rates (98.9% vs 93.6%, risk difference [RD] 8%; risk ratio [RR] 1.12, p = 0.02), lower reoperation rates (2.2% vs 3.1%, RD 3%; RR 0.48, p = 0.04), and higher risk of dysphagia (24.7% vs 8.1%, RD 11%; RR 1.93, p = 0.02). No significant differences in the Neck Disability Index, neck pain, or arm pain scores were associated with the use of BMP. On subgroup meta-analysis of ACDF cases, older age (≥ 50 years) and higher BMP dose (≥ 0.9 mg/level) were associated with significantly higher fusion rates and relatively lower risk for dysphagia, whereas arthrodesis of fewer segments (< 2 levels) showed significantly higher dysphagia rates without a significant increase in fusion rates. In the PCF cases, the use of BMP was not associated with significant differences in fusion (p = 0.38) or reoperation (p = 0.61) rates but was associated with significantly higher blood loss during surgery (mean difference 146.7 ml, p ≤ 0.01). CONCLUSIONS Use of BMP in ACDF offers higher rates of augmented fusion and lower rates of all-cause reoperation but with an increased risk of dysphagia. The benefit of fusion outweighs the risk of dysphagia with a higher BMP dose in older patients being operated on for < 2 levels. The use of BMP in PCF seems to have a less important effect on clinical and radiological outcomes.

OBJECTIVE The objective of this study was to compare clinical and patient-reported outcomes (PROs) between posterior foraminotomy and anterior cervical discectomy and fusion (ACDF) in patients presenting with cervical radiculopathy. METHODS The Quality Outcomes Database was queried for patients who had undergone ACDF or posterior foraminotomy for radiculopathy. To create two highly homogeneous groups, optimal individual matching was performed at a 5:1 ratio between the two groups on 29 baseline variables (including demographic characteristics, comorbidities, symptoms, patient-reported scores, underlying pathologies, and levels treated). Outcomes of interest were length of stay, reoperations, patient-reported satisfaction, increase in EQ-5D score, and decrease in Neck Disability Index (NDI) scores for arm and neck pain as long as 1 year after surgery. Noninferiority analysis of achieving patient satisfaction and minimal clinically important difference (MCID) in PROs was performed with an accepted risk difference of 5%. RESULTS A total of 7805 eligible patients were identified: 216 of these underwent posterior foraminotomy and were matched to 1080 patients who underwent ACDF. The patients who underwent ACDF had more underlying pathologies, lower EQ-5D scores, and higher NDI and neck pain scores at baseline. Posterior foraminotomy was associated with shorter hospitalization (0.5 vs 0.9 days, p < 0.001). Reoperations within 12 months were significantly more common among the posterior foraminotomy group (4.2% vs 1.9%, p = 0.04). The two groups performed similarly in PROs, with posterior foraminotomy being noninferior to ACDF in achieving MCID in EQ-5D and neck pain scores but also having lower rates of maximal satisfaction at 12 months (North American Spine Society score of 1 achieved by 65.2% posterior foraminotomy patients vs 74.6% of ACDF patients, p = 0.02). CONCLUSIONS The two procedures were found to be offered to different populations, with ACDF being selected for patients with more complicated pathologies and symptoms. After individual matching, posterior foraminotomy was associated with a higher reoperation risk within 1 year after surgery compared to ACDF (4.2% vs 1.9%). In terms of 12-month PROs, posterior foraminotomy was noninferior to ACDF in improving quality of life and neck pain. The two procedures also performed similarly in improving NDI scores and arm pain, but ACDF patients had higher maximal satisfaction rates.

Study Design: Retrospective cohort analysis. Objective: To determine whether the C2 exposure technique was a predictor of change in cervical alignment and patient-reported outcomes measures (PROMs) after posterior cervical decompression and fusion (PCDF) for degenerative indications. Background: In PCDF handling of the C2 posterior paraspinal musculature during the operative approach varies by surgeon technique. To date, no studies have investigated whether maintenance of the upper cervical semispinalis cervicis attachments as compared with complete reflection of upper cervical paraspinal musculature from the posterior bony elements is associated with superior radiographic and clinical outcomes after PCDF. Patients and Methods: All adult patients who underwent C2–T2 PCDF for myelopathy or myeloradiculopathy at multi-institutional academic centers between 2013 and 2020 were retrospectively identified. Patients were dichotomized by the C2 exposure technique into semispinalis preservation or midline muscular reflection groups. Preoperative and short and long-term postoperative radiographic outcomes (upper cervical alignment, global alignment, and fusion status) and PROMs (Visual Analog Scale–Neck, Neck Disability Index, and Short Form-12) were collected. Univariate analysis compared patient factors, radiographic measures, and PROMs across C2 exposure groups. Results: A total of 129 patients met the inclusion/exclusion criteria (73 muscle preservation and 56 muscle reflection). Patients in the muscular preservation group were on average younger (P = 0.005) and more likely to have bone morphogenic protein (P < 0.001) and C2 pars screws (P = 0.006) used during surgery. Preoperative to postoperative changes in C2 slope, C2 tilt, C2–C3 segmental lordosis, C2–C3 listhesis, C0–C2 Cobb angle, proximal junctional kyphosis, ADI, C1 lamina-occiput distance, C2 sagittal vertical axis, C2–C7 lordosis, and PROMs at all follow-up intervals did not vary significantly by C2 exposure technique. Likewise, there were no significant differences in fusion status, C2–C3 pseudoarthrosis, C2 screw loosening, and complication and revision rates between C2 exposure groups. Conclusions: Preservation of C2 semispinalis attachments versus muscular reflection did not significantly impact cervical alignment, clinical outcomes, or proximal junction complications in long-segment PCDF. Level of Evidence: Level III.

OBJECTIVE Cervical spondylotic myelopathy (CSM) can cause significant difficulty with driving and a subsequent reduction in an individual’s quality of life due to neurological deterioration. The positive impact of surgery on postoperative patient-reported driving capabilities has been seldom explored. METHODS The CSM module of the Quality Outcomes Database was utilized. Patient-reported driving ability was assessed via the driving section of the Neck Disability Index (NDI) questionnaire. This is an ordinal scale in which 0 represents the absence of symptoms while driving and 5 represents a complete inability to drive due to symptoms. Patients were considered to have an impairment in their driving ability if they reported an NDI driving score of 3 or higher (signifying impairment in driving duration due to symptoms). Multivariable logistic regression models were fitted to evaluate mediators of baseline impairment and improvement at 24 months after surgery, which was defined as an NDI driving score < 3. RESULTS A total of 1128 patients who underwent surgical intervention for CSM were included, of whom 354 (31.4%) had baseline driving impairment due to CSM. Moderate (OR 2.3) and severe (OR 6.3) neck pain, severe arm pain (OR 1.6), mild-moderate (OR 2.1) and severe (OR 2.5) impairment in hand/arm dexterity, severe impairment in leg use/walking (OR 1.9), and severe impairment of urinary function (OR 1.8) were associated with impaired driving ability at baseline. Of the 291 patients with baseline impairment and available 24-month follow-up data, 209 (71.8%) reported postoperative improvement in their driving ability. This improvement seemed to be mediated particularly through the achievement of the minimal clinically important difference (MCID) in neck pain and improvement in leg function/walking. Patients with improved driving at 24 months noted higher postoperative satisfaction (88.5% vs 62.2%, p < 0.01) and were more likely to achieve a clinically significant improvement in their quality of life (50.7% vs 37.8%, p < 0.01). CONCLUSIONS Nearly one-third of patients with CSM report impaired driving ability at presentation. Seventy-two percent of these patients reported improvements in their driving ability within 24 months of surgery. Surgical management of CSM can significantly improve patients’ driving abilities at 24 months and hence patients’ quality of life.

OBJECTIVE Chiari malformation type I (CM-I) is a congenital disorder occurring in 0.1% of the population. In symptomatic cases, surgery with posterior fossa decompression (PFD) is the treatment of choice. Surgery is, however, associated with peri- and postoperative complications that may require readmission or renewed surgical intervention. Given the associated financial costs and the impact on patients’ well-being, there is a need for predictive tools that can assess the likelihood of such adverse events. The aim of this study was therefore to leverage machine learning algorithms to develop a predictive model for 30-day readmissions and reoperations after PFD in pediatric patients with CM-I. METHODS This was a retrospective study based on data from the National Surgical Quality Improvement Program–Pediatric database. Eligible patients were those undergoing PFD (Current Procedural Terminology code 61343) for CM-I between 2012 and 2021. Patients undergoing surgery for tumors or vascular lesions were excluded. Unplanned 30-day readmission and unplanned 30-day reoperation were the main study outcomes. Additional outcome data considered included the length of hospital stay, 30-day complications, discharge disposition, and 30-day mortality. Training and testing samples were randomly generated (80:20) to study the 30-day readmission and reoperation using logistic regression, decision tree, random forest (RF), K-nearest neighbors, and Gaussian naive Bayes algorithms. RESULTS A total of 7106 pediatric patients undergoing PFD were included. The median age was 9.2 years (IQR 4.7, 14.2 years). Most of the patients were female (56%). The 30-day readmission and reoperation rates were 7.5% and 3.4%, respectively. Headaches (32%) and wound-related complications (30%) were the most common reasons for 30-day readmission, while wound revisions and evacuation of fluid or blood (62%), followed by CSF diversion-related procedures (28%), were the most common reasons for 30-day reoperation. RF classifiers had the highest predictive accuracy for both 30-day readmissions (area under the curve [AUC] 0.960) and reoperations (AUC 0.990) compared with the other models. On feature importance analysis, sex, developmental delay, ethnicity, respiratory disease, premature birth, hydrocephalus, and congenital/genetic anomaly were some of the variables contributing the most to both RF models. CONCLUSIONS Using a large-scale nationwide dataset, machine learning models for the prediction of both 30-day readmissions and reoperations were developed and achieved high accuracy. This highlights the utility of machine learning in risk stratification and surgical decision-making for pediatric CM-I.

OBJECTIVE Unplanned returns to the operating room (RORs) constitute an important quality metric in surgical practice. In this study, the authors present a methodology to compare a department’s unplanned ROR rates with national benchmarks in the context of large-scale quality of care surveillance. METHODS The authors identified unplanned RORs within 30 days from the initial surgery at their institution during the period 2014–2018 using an institutional documentation platform that facilitates the collection of reoperation information by providers in the clinical setting. They divided the procedures into 28 groups by Current Procedural Terminology and International Classification of Diseases, 9th and 10th Revision codes. They estimated national benchmarks of unplanned RORs for these procedure groups via querying the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) registry during the period 2014–2018. Finally, they numerically assessed the unplanned ROR rates at their institution compared with those calculated from the ACS NSQIP registry. RESULTS Using the above methodology, the authors were able to classify 12,575 of the cases performed in their department during the period of interest, including 6037 (48%) cranial cases and 6538 (52%) spinal or peripheral nerve cases. Among those, 161 (1.3%) presented with complications that required an unplanned ROR within 30 days from the initial surgery. The respective cumulative unplanned ROR rate in the ACS NSQIP registry during the same timeframe was 3.6%. Among 15 categories of cranial procedures, the cumulative unplanned ROR rate was 1.3% in the authors’ department and 5.6% in the ACS NSQIP registry. Among 13 categories of spinal and peripheral nerve procedures, the cumulative unplanned ROR rate was 1.3% in the authors’ department and 2.8% in the ACS NSQIP registry. Unplanned ROR rates at the authors’ institution were lower than the national average for each of the 28 procedure groups of interest. Yearly analysis of institutional ROR rates for the five most commonly performed procedures showed lower reoperation rates compared with the national benchmarks. CONCLUSIONS Using an institutional documentation tool and a widely available national database, the authors developed a reproducible and standardized method of comparing their department’s outcomes with national benchmarks per procedure subgroup. This methodology accommodates longitudinal quality surveillance across the different subspecialties in a neurosurgical department and may illuminate potential shortcomings of care delivery in the future.

BACKGROUND AND OBJECTIVES: Cervical disk arthroplasty (CDA) offers the advantage of motion preservation in the treatment of focal cervical pathology. At present, implant sizing is performed using subjective tactile feedback and imaging of trial cages. This study aims to construct interpretable machine learning (IML) models to accurately predict postoperative range of motion (ROM) and identify the optimal implant sizes that maximize ROM in patients undergoing CDA. METHODS: Adult patients who underwent CDA for single-level disease from 2012 to 2020 were identified. Patient demographics, comorbidities, and outcomes were collected, including symptoms, examination findings, subsidence, and reoperation. Affected disk height, healthy rostral disk height, and implant height were collected at sequential time points. Linear regression and IML models, including bagged regression tree, bagged multivariate adaptive regression spline, and k-nearest neighbors, were used to predict ROM change. Model performance was assessed by calculating the root mean square error (RMSE) between predicted and actual changes in ROM in the validation cohort. Variable importance was assessed using RMSE loss. Area under the curve analyses were performed to identify the ideal implant size cutoffs in predicting improved ROM. RESULTS: Forty-seven patients were included. The average RMSE between predicted and actual ROM was 7.6° (range: 5.8-10.1) in the k-nearest neighbors model, 7.8° (range: 6.5-10.0) in the bagged regression tree model, 7.8° (range: 6.2-10.0) in the bagged multivariate adaptive regression spline model, and 15.8° (range: 14.3-17.5°) in a linear regression model. In the highest-performing IML model, graft size was the most important predictor with RMSE loss of 6.2, followed by age (RMSE loss = 5.9) and preoperative caudal disk height (RMSE loss = 5.8). Implant size at 110% of the normal adjacent disk height was the optimal cutoff associated with improved ROM. CONCLUSION: IML models can reliably predict change in ROM after CDA within an average of 7.6 degrees of error. Implants sized comparably with the healthy adjacent disk may maximize ROM.

OBJECTIVE Dural arteriovenous fistulas are rare vascular malformations that affect the brain and spinal cord. Spinal dural arteriovenous fistulas (sdAVFs) are the most frequently encountered vascular malformation affecting the spinal cord. The object of this study was to evaluate the impact of treatment delays on the long-term neurological outcomes of either open surgical or interventional treatment of sdAVFs. METHODS In this retrospective, population-based cohort study, the authors examined consecutive patients with diagnosed sdAVFs at a tertiary care center between 2005 and 2020. Patients were assessed using the Aminoff-Logue disability scale (ALS) at various time points including symptom onset, primary care visit, first specialist outpatient visit, as well as both short and long-term follow-ups. The postoperative long-term ALS gait and bladder grades constituted the primary outcomes of the study. RESULTS Among the 34 patients included in the study, the median age was 65 years, and there was a male predominance (71%). Most lesions were in the lumbar region (47%). Significant worsening in ALS gait and bladder grades was observed preoperatively, followed by postoperative improvements (p < 0.05). There was no difference in outcomes between surgical and endovascular treatments. Older age (OR 1.10, 95% CI 1.03–1.17, p = 0.007), worse preoperative ALS gait grades (OR 5.12, 95% CI 2.18–12.4, p < 0.001), and longer time from first specialist outpatient visit to first treatment (OR 1.00, 95% CI 1.00–1.01, p = 0.040) were independently associated with worse long-term gait outcomes. Only the preoperative ALS bladder score was a predictor of worse long-term bladder function (OR 92.7, 95% CI 28.0–306.7, p < 0.001). CONCLUSIONS Both surgical and endovascular treatments for sdAVFs led to significant neurological improvements. However, treatment delays were associated with less favorable long-term outcomes. Prompt diagnosis and early intervention prior to symptom progression may enhance recovery and help to preserve neurological function.