Performance assessment of the metastatic spinal tumor frailty index using machine learning algorithms: limitations and future directions

Advances in Surgery for Metastatic Disease of the Spine: An Update for Oncologists

STUDY DESIGN

Narrative review.

OBJECTIVE

Metastatic spine disease is an increasingly common clinical challenge that requires individualised multidisciplinary care from spine surgeons and oncologists. In this article, the authors describe the recent surgical advances in patients presenting with spinal metastases.

METHODS

We present an overview of the presentation, assessment, and management of spinal metastases from the perspective of the spine surgeon, highlighting advances in surgical technology and techniques, to facilitate multidisciplinary care for this complex patient group. Neither institutional review board approval nor patient consent was needed for this review.

RESULTS

Advances in radiotherapy delivery and systemic therapy (including immunotherapy and targeted therapy) have refined operative indications for decompression of neural structures and spinal stabilisation, while advances in surgical technology and technique enable these goals to be achieved with reduced morbidity. Formulating individualised management strategies that optimise outcome, while meeting patient goals and expectations, requires a comprehensive understanding of the factors important to patient management.

CONCLUSION

Spinal metastases require prompt diagnosis and expert management by a multidisciplinary team. Improvements in systemic, radiation, and surgical therapies have broadened operative indications and increased operative candidacy, and future advances are likely to continue this trend.

Machine Learning in Spine Oncology: A Narrative Review

STUDY DESIGN

Narrative Review.

OBJECTIVE

Machine learning (ML) is one of the latest advancements in artificial intelligence used in medicine and surgery with the potential to significantly impact the way physicians diagnose, prognose, and treat spine tumors. In the realm of spine oncology, ML is utilized to analyze and interpret medical imaging and classify tumors with incredible accuracy. The authors present a narrative review that specifically addresses the use of machine learning in spine oncology.

METHODS

This study was conducted in accordance with the Preferred Reporting Items of Systematic Reviews and Meta-Analysis (PRISMA) methodology. A systematic review of the literature in the PubMed, EMBASE, Web of Science, Scopus, and Cochrane Library databases since inception was performed to present all clinical studies with the search terms '[[Machine Learning] OR [Artificial Intelligence]] AND [[Spine Oncology] OR [Spine Cancer]]'. Data included studies that were extracted and included algorithms, training and test size, outcomes reported. Studies were separated based on the type of tumor investigated using the machine learning algorithms into primary, metastatic, both, and intradural. A minimum of 2 independent reviewers conducted the study appraisal, data abstraction, and quality assessments of the studies.

RESULTS

Forty-five studies met inclusion criteria out of 480 references screened from the initial search results. Studies were grouped by metastatic, primary, and intradural tumors. The majority of ML studies relevant to spine oncology focused on utilizing a mixture of clinical and imaging features to risk stratify mortality and frailty. Overall, these studies showed that ML is a helpful tool in tumor detection, differentiation, segmentation, predicting survival, predicting readmission rates of patients with either primary, metastatic, or intradural spine tumors.

CONCLUSION

Specialized neural networks and deep learning algorithms have shown to be highly effective at predicting malignant probability and aid in diagnosis. ML algorithms can predict the risk of tumor recurrence or progression based on imaging and clinical features. Additionally, ML can optimize treatment planning, such as predicting radiotherapy dose distribution to the tumor and surrounding normal tissue or in surgical resection planning. It has the potential to significantly enhance the accuracy and efficiency of health care delivery, leading to improved patient outcomes.

Baseline Frailty Measured by the Risk Analysis Index and 30-Day Mortality After Surgery for Spinal Malignancy: Analysis of a Prospective Registry (2011-2020)

OBJECTIVE

To evaluate the prognostic utility of baseline frailty, measured by the Risk Analysis Index (RAI), for prediction of postoperative mortality among patients with spinal malignancy (SM) undergoing resection.

METHODS

SM surgery cases were queried from the American College of Surgeons - National Surgical Quality Improvement Program database (2011-2020). The relationship between preoperative RAI frailty score and increasing rate of primary endpoint (mortality or discharge to hospice within 30 days, "mortality/hospice") were assessed. Discriminatory accuracy was assessed by computation of C-statistics (with 95% confidence interval [CI]) in receiver operating characteristic (ROC) curve analysis.

RESULTS

A total of 2,235 cases were stratified by RAI score: 0-20, 22.7%; 21-30, 11.9%; 31-40, 54.7%; and ≥ 41, 10.7%. The rate of mortality/hospice was 6.5%, which increased linearly with increasing RAI score (p < 0.001). RAI was also associated with increasing rates of major complication, extended length of stay, and nonhome discharge (all p < 0.05). The RAI demonstrated acceptable discriminatory accuracy for prediction of primary endpoint (C-statistic, 0.717; 95% CI, 0.697-0.735). In pairwise ROC comparison, RAI demonstrated superiority versus modified frailty index-5 and chronological age (p < 0.001).

CONCLUSION

Preoperative frailty, as measured by RAI, is a robust predictor of mortality/ hospice after SM surgery. The frailty score may be applied in clinical settings using a user-friendly calculator, deployed here: https://nsgyfrailtyoutcomeslab.shinyapps.io/spinalMalignancyRAI/.

Application of the NSE score (Neurology-Stability-Epidural compression assessment) to establish the need for surgery in spinal metastases of elderly patients: a multicenter investigation

PURPOSE

This retropective multicentric study aims to investigate the clinical applicability of the NSE score in the elderly, to verify the role of this tool as an easy help for decision making also for this class of patients.

METHODS

All elderly patients (> 65 years) suffering from spinal metastases undergoing surgical or non-surgical treatment at the authors' Institutions between 2015 and 2022 were recruited. An agreement group (AG) and non-agreement group (NAG) were identified accordingly to the agreement between the NSE score indication and the performed treatment. Neurological status and axial pain were evaluated for both groups at follow-up (3 and 6 months). The same analysis was conducted specifically grouping patients older than 75 years.

RESULTS

A strong association with improvement or preservation of clinical status (p < 0.001) at follow-up was obtained in AG. The association was not statistically significant in NAG at the 3-month follow-up (p 1.00 and 0.07 respectively) and at 6 months (p 0.293 and 0.09 respectively). The group of patients over 75 years old showed similar results in terms of statistical association between the agreement group and better outcomes.

CONCLUSION

Far from the need or the aim to build dogmatic algorithms, the goal of preserving a proper performance status plays a key role in a modern oncological management: functional outcomes of the multicentric study group showed that the NSE score represents a reliable tool to establish the need for surgery also for elderly patients.

Utilization of Machine Learning to Model Important Features of 30-day Readmissions following Surgery for Metastatic Spinal Column Tumors: The Influence of Frailty

STUDY DESIGN

Retrospective cohort study.

OBJECTIVE

The aim of this study was to determine the relative importance and predicative power of the Hospital Frailty Risk Score (HFRS) on unplanned 30-day readmission after surgical intervention for metastatic spinal column tumors.

METHODS

All adult patients undergoing surgery for metastatic spinal column tumor were identified in the Nationwide Readmission Database from the years 2016 to 2018. Patients were categorized into 3 cohorts based on the criteria of the HFRS: Low(<5), Intermediate(5-14.9), and High(≥ 15). Random Forest (RF) classification was used to construct predictive models for 30-day patient readmission. Model performance was examined using the area under the receiver operating curve (AUC), and the Mean Decrease Gini (MDG) metric was used to quantify and rank features by relative importance.

RESULTS

There were 4346 patients included. The proportion of patients who required any readmission were higher among the Intermediate and High frailty cohorts when compared to the Low frailty cohort (Low:33.9% vs. Intermediate:39.3% vs. High:39.2%, P < .001). An RF classifier was trained to predict 30-day readmission on all features (AUC = .60) and architecturally equivalent model trained using only ten features with highest MDG (AUC = .59). Both models found frailty to have the highest importance in predicting risk of readmission. On multivariate regression analysis, Intermediate frailty [OR:1.32, CI(1.06,1.64), P = .012] was found to be an independent predictor of unplanned 30-day readmission.

CONCLUSION

Our study utilizes machine learning approaches and predictive modeling to identify frailty as a significant risk-factor that contributes to unplanned 30-day readmission after spine surgery for metastatic spinal column metastases.

Frailty Indexes in Metastatic Spine Tumor Surgery: A Narrative Review

Quantification of preoperative frailty is an important prognostic tool in neurosurgical decision making. Metastatic spine tumor patients undergoing surgery are frail and have unfavorable outcomes that include an increased length of stay, unfavorable discharge disposition, and increased readmission rates. These undesirable outcomes result in higher treatment costs. A heterogeneous mixture of various frailty indexes is available with marked variance in their validation, leading to disparate clinical utility. The lack of a universally accepted definition for frailty, let alone in the method of creation or elements required in the formation of a frailty index, has resulted in a body of frailty literature lacking precision for predicting neurosurgical outcomes. In this review, we examine the role of reported frailty indexes in predicting postoperative outcomes after resection of metastatic spine tumors and aim to assist as a frailty guide for helping clinical decision making.

Can We Use Artificial Intelligence Cluster Analysis to Identify Patients with Metastatic Breast Cancer to the Spine at Highest Risk of Postoperative Adverse Events?

OBJECTIVE

Group patients who required open surgery for metastatic breast cancer to the spine by functional level and metastatic disease characteristics to identify factors that predispose to poor outcomes.

METHODS

A retrospective analysis included patients managed at 2 tertiary referral centers from 2008 to 2020. The primary outcome was a 90-day adverse event. A 2-step unsupervised cluster analysis stratified patients into cohorts using function at presentation, preoperative spine radiation, structural instability, epidural spinal cord compression (ESCC), neural deficits, and tumor location/hormone status. Comparisons were performed using χ2 test and one-way analysis of variance.

RESULTS

Five patient "clusters" were identified. High function (HIGH) had thoracic metastases and an Eastern Cooperative Oncology Group (ECOG) score of 1.0 ± 0.8. Low function/irradiated (LOW + RADS) had preoperative radiation and the lowest Karnofsky scores (56.0 ± 10.6). Estrogen receptor or progesterone receptor (ER/PR) positive patients had >90% estrogen/progesterone positivity and moderate Karnofsky scores (74.0 ± 11.5). Lumbar/noncompressive (NON-COMP) had the fewest patients with ESCC grade 2 or 3 epidural disease (42.1%, P < 0.001). Low function/neurologic deficits (LOW + NEURO) had ESCC grade 2 or 3 disease and neurologic deficits. Adverse event rates were 25.0% in the HIGH group, 73.3% in LOW + RADS, 24.0% in ER/PR, 31.6% in NON-COMP, and 60.0% in LOW + NEURO (P = 0.003).

CONCLUSIONS

Function at presentation, tumor hormone signature, radiation history, and epidural compression delineated postoperative trajectory. We believe our results can aid in expectation management and the identification of at-risk patients who may merit closer surveillance following surgical intervention.

Prevalence and Influence of Frailty on Hospital Outcomes After Surgical Resection of Spinal Meningiomas

OBJECTIVE

Frailty has been shown to affect patient outcomes after medical and surgical interventions. The Hospital Frailty Risk Score (HFRS) is a growing metric used to assess patient frailty using International Classification of Diseases, Tenth Revision codes. The goal of this study was to investigate the impact of frailty, assessed by HFRS, on health care resource utilization and outcomes in patients undergoing surgery for spinal meningiomas.

METHODS

A retrospective cohort study was performed using the 2016-2019 National Inpatient Sample database. Adult patients with benign or malignant spinal meningiomas, identified using International Classification of Diseases, Tenth Revision, Clinical Modification codes, were stratified by HFRS: low frailty (HFRS <5) and intermediate-high frailty (HFRS ≥5). Patient demographics, hospital characteristics, comorbidities, procedural variables, adverse events, length of stay (LOS), discharge disposition, and cost of admission were assessed. Multivariate regression analysis was used to identify predictors of increased LOS, discharge disposition, and cost.

RESULTS

Of the 3345 patients, 530 (15.8%) had intermediate-high frailty. The intermediate-high cohort was significantly older (P < 0.001). More patients in the intermediate-high cohort had ≥3 comorbidities (P < 0.001). In addition, a greater proportion of patients in the intermediate-high cohort experienced ≥1 perioperative adverse events (P < 0.001). Intermediate-high patients experienced greater mean LOS (P < 0.001) and accrued greater costs (P < 0.001). A greater proportion of intermediate-high patients had nonroutine discharges (P < 0.001). On multivariate analysis, increased HFRS (≥5) was independently associated with extended LOS (adjusted odds ratio [aOR], 3.04; P < 0.001), nonroutine discharge (aOR, 1.98; P = 0.006), and increased costs (aOR, 2.39; P = 0.004).

CONCLUSIONS

Frailty may be associated with increased health care resource utilization in patients undergoing surgery for spinal meningiomas.

The Role of Frailty and Sarcopenia in Predicting Major Adverse Events, Length of Stay and Reoperation Following En Bloc Resection of Primary Tumours of the Spine

STUDY DESIGN

Retrospective observational cohort study.

OBJECTIVE

En bloc resection for primary tumours of the spine is associated with a high rate of adverse events (AEs). The objective was to explore the relationship between frailty/sarcopenia and major perioperative AEs, length of stay (LOS), and unplanned reoperation following en bloc resection of primary spinal tumours.

METHODS

This is a unicentre study consisting of adult patients undergoing en bloc resection for a primary spine tumor. Frailty was calculated with the modified frailty index (mFI) and spine tumour frailty index (STFI). Sarcopenia was quantified with the total psoas area/vertebral body area ratio (TPA/VB) at L3 and L4. Univariable regression analysis was used to quantify the association between frailty/sarcopenia and major perioperative AEs, LOS and unplanned reoperation.

RESULTS

95 patients met the inclusion criteria. The mFI and STFI identified a frailty prevalence of 3% and 18%. Mean CT TPA/VB ratios were 1.47 (SD ± .05) and 1.83 (SD ± .06) at L3 and L4. Inter-observer reliability was .93 and .99 for CT and MRI L3 and L4 TPA/VB ratios. Unadjusted analysis demonstrated sarcopenia and mFI did not predict perioperative AEs, LOS or unplanned reoperation. Frailty defined by an STFI score ≥2 predicted unplanned reoperation for surgical site infection (SSI) (P < .05).

CONCLUSIONS

The STFI was only associated with unplanned reoperation for SSI on unadjusted analysis, while the mFI and sarcopenia were not predictive of any outcome. Further studies are needed to investigate the relationship between frailty, sarcopenia and perioperative outcomes following en bloc resection of primary spinal tumors.

A Machine Learning-Based Online Prediction Tool for Predicting Short-Term Postoperative Outcomes Following Spinal Tumor Resections

Background: Preoperative prediction of short-term postoperative outcomes in spinal tumor patients can lead to more precise patient care plans that reduce the likelihood of negative outcomes. With this study, we aimed to develop machine learning algorithms for predicting short-term postoperative outcomes and implement these models in an open-source web application. Methods: Patients who underwent surgical resection of spinal tumors were identified using the American College of Surgeons, National Surgical Quality Improvement Program. Three outcomes were predicted: prolonged length of stay (LOS), nonhome discharges, and major complications. Four machine learning algorithms were developed and integrated into an open access web application to predict these outcomes. Results: A total of 3073 patients that underwent spinal tumor resection were included in the analysis. The most accurately predicted outcomes in terms of the area under the receiver operating characteristic curve (AUROC) was the prolonged LOS with a mean AUROC of 0.745 The most accurately predicting algorithm in terms of AUROC was random forest, with a mean AUROC of 0.743. An open access web application was developed for getting predictions for individual patients based on their characteristics and this web application can be accessed here: huggingface.co/spaces/MSHS-Neurosurgery-Research/NSQIP-ST. Conclusion: Machine learning approaches carry significant potential for the purpose of predicting postoperative outcomes following spinal tumor resections. Development of predictive models as clinically useful decision-making tools may considerably enhance risk assessment and prognosis as the amount of data in spinal tumor surgery continues to rise.

Assessment of Spinal Metastases Surgery Risk Stratification Tools in Breast Cancer by Molecular Subtype

BACKGROUND

Breast cancer molecular features and modern therapies are not included in spine metastasis prediction algorithms.

OBJECTIVE

To examine molecular differences and the impact of postoperative systemic therapy to improve prognosis prediction for spinal metastases surgery and aid surgical decision making.

METHODS

This is a retrospective multi-institutional study of patients who underwent spine surgery for symptomatic breast cancer spine metastases from 2008 to 2021 at the Massachusetts General Hospital and Brigham and Women's Hospital. We studied overall survival, stratified by breast cancer molecular subtype, and calculated hazard ratios (HRs) adjusting for demographics, tumor characteristics, treatments, and laboratory values. We tested the performance of established models (Tokuhashi, Bauer, Skeletal Oncology Research Group, New England Spinal Metastases Score) to predict and compare all-cause.

RESULTS

A total of 98 patients surgically treated for breast cancer spine metastases were identified (100% female sex; median age, 56 years [IQR, 36-84 years]). The 1-year probabilities of survival for hormone receptor positive, hormone receptor positive/human epidermal growth factor receptor 2+, human epidermal growth factor receptor 2+, and triple-negative breast cancer were 63% (45 of 71), 83% (10 of 12), 0% (0 of 3), and 12% (1 of 8), respectively ( P < .001). Patients with triple-negative breast cancer had a higher proportion of visceral metastases, brain metastases, and poor physical activity at baseline. Postoperative chemotherapy and endocrine therapy were associated with prolonged survival. The Skeletal Oncology Research Group prognostic model had the highest discrimination (area under the receiver operating characteristic, 0.77 [95% CI, 0.73-0.81]). The performance of all prognostic scores improved when preoperative molecular data and postoperative systemic treatment plans was considered.

CONCLUSION

Spine metastases risk tools were able to predict prognosis at a significantly higher degree after accounting for molecular features which guide treatment response.

The predictive ability of routinely collected laboratory markers for surgically treated spinal metastases: a retrospective single institution study

PURPOSE

We aimed to identify effective routinely collected laboratory biomarkers for predicting postoperative outcomes in surgically treated spinal metastases and attempted to establish an effective prediction model.

METHODS

This study included 268 patients with spinal metastases surgically treated at a single institution. We evaluated patient laboratory biomarkers to determine trends to predict survival. The markers included white blood cell (WBC) count, platelet count, neutrophil count, lymphocyte count, hemoglobin, albumin, alkaline phosphatase, creatinine, total bilirubin, calcium, international normalized ratio (INR), platelet-to-lymphocyte ratio (PLR), and neutrophil-to-lymphocyte ratio (NLR). A nomogram based on laboratory markers was established to predict postoperative 90-day and 1-year survival. The discrimination and calibration were validated using concordance index (C-index), area under curves (AUC) from receiver operating characteristic curves, and calibration curves. Another 47 patients were used as a validation group to test the accuracy of the nomogram. The prediction accuracy of the nomogram was compared to Tomita, revised Tokuhashi, modified Bauer, and Skeletal Oncology Research Group machine-learning (SORG ML).

RESULTS

WBC, lymphocyte count, albumin, and creatinine were shown to be the independent prognostic factors. The four predictive laboratory markers and primary tumor, were incorporated into the nomogram to predict the 90-day and 1-year survival probability. The nomogram performed good with a C-index of 0.706 (0.702-0.710). For predicting 90-day survival, the AUC in the training group and the validation group was 0.740 (0.660-0.819) and 0.795 (0.568-1.000), respectively. For predicting 1-year survival, the AUC in the training group and the validation group was 0.765 (0.709-0.822) and 0.712 (0.547-0.877), respectively. Our nomogram seems to have better predictive accuracy than Tomita, revised Tokuhashi, and modified Bauer, alongside comparable prediction ability to SORG ML.

CONCLUSIONS

Our study confirmed that routinely collected laboratory markers are closely associated with the prognosis of spinal metastases. A nomogram based on primary tumor, WBC, lymphocyte count, albumin, and creatinine, could accurately predict postoperative survival for patients with spinal metastases.

The measurements of frailty and their possible application to spinal conditions: a systematic review

BACKGROUND CONTEXT

Frailty is associated with an increased risk of postoperative adverse events (AEs) within the surgical spine population. Multiple frailty tools have been reported in the surgical spine literature. However, the applicability of these tools remains unclear.

PURPOSE

Primary objective is to appraise the construct, feasibility, objectivity, and clinimetric properties of frailty tools reported in the surgical spine literature. Secondary objectives included determining the applicability and the most sensitive surgical spine population for each tool.

STUDY DESIGN

Systematic Review.

PATIENT SAMPLE

Studies reporting the use of a clinical frailty tool with a defined methodology in the adult surgical population (age ≥18 years).

OUTCOME MEASURES

Postoperative adverse events (AEs) including mortality, major and minor morbidity, length of stay (LOS), unplanned readmission and reoperation, admission to the Intensive Care Unit (ICU), and adverse discharge disposition; postoperative patient-reported outcomes (health-related quality of life (HRQoL), functional, cognitive, and symptomatic); radiographic outcomes; and postoperative frailty trajectory.

METHODS

This systematic review was registered with PROSPERO: CRD42019109045. Publications from January 1950 to December 2020 were identified by a comprehensive search of PubMed, Ovid, and Embase, supplemented by manual screening. Studies reporting and validating a frailty tool in the surgical spine population with a measurable outcome were included. Each tool and its clinimetric properties were evaluated using validated criteria and definitions. The applicability of each tool and its most sensitive surgical spine population was determined by panel consensus. Bias was assessed using the Newcastle-Ottawa Scale.

RESULTS

47 studies were included in the final qualitative analysis. A total of 14 separate frailty tools were identified, in which 9 tools assessed frailty according to the cumulative deficit definition, while 4 instruments utilized phenotypic or weighted frailty models. One instrument assessed frailty according to the comprehensive geriatric assessment (CGA) model. Twelve measures were validated as risk stratification tools for predicting postoperative AEs, while 1 tool investigated the effect of spine surgery on postoperative frailty trajectory. The modified frailty index (mFI), 5-item mFI, adult spinal deformity frailty index (ASD-FI), FRAIL Scale, and CGA had the most positive ratings for clinimetric properties assessed.

CONCLUSIONS

The assessment of frailty is important in the surgical decision-making process. Cumulative deficit and weighted frailty instruments are appropriate risk stratification tools. Phenotypic tools are sensitive for capturing the relationship between spinal pathology, spine surgery, and prehabilitation on frailty trajectory. CGA instruments are appropriate screening tools for identifying health deficits susceptible to improvement and guiding optimization strategies. Studies are needed to determine whether spine surgery and prehabilitation are effective interventions to reverse frailty.

Evaluating frailty, mortality, and complications associated with metastatic spine tumor surgery using machine learning-derived body composition analysis

OBJECTIVE

Cancer patients with spinal metastases may undergo surgery without clear assessments of prognosis, thereby impacting the optimal palliative strategy. Because the morbidity of surgery may adversely impact recovery and initiation of adjuvant therapies, evaluation of risk factors associated with mortality risk and complications is critical. Evaluation of body composition of cancer patients as a surrogate for frailty is an emerging area of study for improving preoperative risk stratification.

METHODS

To examine the associations of muscle characteristics and adiposity with postoperative complications, length of stay, and mortality in patients with spinal metastases, the authors designed an observational study of 484 cancer patients who received surgical treatment for spinal metastases between 2010 and 2019. Sarcopenia, muscle radiodensity, visceral adiposity, and subcutaneous adiposity were assessed on routinely available 3-month preoperative CT images by using a validated deep learning methodology. The authors used k-means clustering analysis to identify patients with similar body composition characteristics. Regression models were used to examine the associations of sarcopenia, frailty, and clusters with the outcomes of interest.

RESULTS

Of 484 patients enrolled, 303 had evaluable CT data on muscle and adiposity (mean age 62.00 ± 11.91 years; 57.8% male). The authors identified 2 clusters with significantly different body composition characteristics and mortality risks after spine metastases surgery. Patients in cluster 2 (high-risk cluster) had lower muscle mass index (mean ± SD 41.16 ± 7.99 vs 50.13 ± 10.45 cm2/m2), lower subcutaneous fat area (147.62 ± 57.80 vs 289.83 ± 109.31 cm2), lower visceral fat area (82.28 ± 48.96 vs 239.26 ± 98.40 cm2), higher muscle radiodensity (35.67 ± 9.94 vs 31.13 ± 9.07 Hounsfield units [HU]), and significantly higher risk of 1-year mortality (adjusted HR 1.45, 95% CI 1.05-2.01, p = 0.02) than individuals in cluster 1 (low-risk cluster). Decreased muscle mass, muscle radiodensity, and adiposity were not associated with a higher rate of complications after surgery. Prolonged length of stay (> 7 days) was associated with low muscle radiodensity (mean 30.87 vs 35.23 HU, 95% CI 1.98-6.73, p < 0.001).

CONCLUSIONS

Body composition analysis shows promise for better risk stratification of patients with spinal metastases under consideration for surgery. Those with lower muscle mass and subcutaneous and visceral adiposity are at greater risk for inferior outcomes.

Hospital Frailty Risk Score and healthcare resource utilization after surgery for metastatic spinal column tumors

OBJECTIVE

The Hospital Frailty Risk Score (HFRS) was developed utilizing ICD-10 diagnostic codes to identify frailty and predict adverse outcomes in large national databases. While other studies have examined frailty in spine oncology, the HFRS has not been assessed in this patient population. The aim of this study was to examine the association of HFRS-defined frailty with complication rates, length of stay (LOS), total cost of hospital admission, and discharge disposition in patients undergoing spine surgery for metastatic spinal column tumors.

METHODS

A retrospective cohort study was performed using the years 2016 to 2019 of the National Inpatient Sample (NIS) database. All adult patients (≥ 18 years old) undergoing surgical intervention for metastatic spinal column tumors were identified using the ICD-10-CM diagnostic codes and Procedural Coding System. Patients were categorized into the following three cohorts based on their HFRS: low frailty (HFRS < 5), intermediate frailty (HFRS 5-15), and high frailty (HFRS > 15). Patient demographics, comorbidities, treatment modality, perioperative complications, LOS, discharge disposition, and total cost of hospital admission were assessed. A multivariate logistic regression analysis was used to identify independent predictors of prolonged LOS, nonroutine discharge, and increased cost.

RESULTS

Of the 11,480 patients identified, 7085 (61.7%) were found to have low frailty, 4160 (36.2%) had intermediate frailty, and 235 (2.0%) had high frailty according to HFRS criteria. On average, age increased along with progressively worsening frailty scores (p ≤ 0.001). The proportion of patients in each cohort who experienced ≥ 1 postoperative complication significantly increased along with increasing frailty (low frailty: 29.2%; intermediate frailty: 53.8%; high frailty: 76.6%; p < 0.001). In addition, the mean LOS (low frailty: 7.9 ± 5.0 days; intermediate frailty: 14.4 ± 13.4 days; high frailty: 24.1 ± 18.6 days; p < 0.001), rate of nonroutine discharge (low frailty: 40.4%; intermediate frailty: 60.6%; high frailty: 70.2%; p < 0.001), and mean total cost of hospital admission (low frailty: $48,603 ± $29,979; intermediate frailty: $65,271 ± $43,110; high frailty: $96,116 ± $60,815; p < 0.001) each increased along with progressing frailty. On multivariate regression analysis, intermediate and high frailty were each found to be significant predictors of both prolonged LOS (intermediate: OR 3.75 [95% CI 2.96-4.75], p < 0.001; high: OR 7.33 [95% CI 3.47-15.51]; p < 0.001) and nonroutine discharge (intermediate: OR 2.05 [95% CI 1.68-2.51], p < 0.001; high: OR 5.06 [95% CI 1.93-13.30], p = 0.001).

CONCLUSIONS

This study is the first to use the HFRS to assess the impact of frailty on perioperative outcomes in patients with metastatic bony spinal tumors. Among patients with metastatic bony spinal tumors, frailty assessed using the HFRS was associated with longer hospitalizations, more nonroutine discharges, and higher total hospital costs.

Relevance of presenting risks of frailty, sarcopaenia and osteopaenia to outcomes from aneurysmal subarachnoid haemorrhage

INTRODUCTION

Aneurysmal subarachnoid haemorrhage (aSAH) is a condition with significant morbidity and mortality. Traditional markers of aSAH have established their utility in the prediction of aSAH outcomes while frailty markers have been validated in other surgical specialties. We aimed to compare the predictive value of frailty indices and markers of sarcopaenia and osteopaenia, against the traditional markers for aSAH outcomes.

METHODS

An observational study in a tertiary neurosurgical unit on 51 consecutive patients with ruptured aSAH was performed. The best performing marker in predicting the modified Rankin scale (mRS) on discharge was selected and an appropriate threshold for the definition of frail and non-frail was derived. We compared various frailty indices (modified frailty index 11, and 5, and the National Surgical Quality Improvement Program score [NSQIP]) and markers of sarcopaenia and osteopaenia (temporalis [TMT] and zygoma thickness), against traditional markers (age, World Federation of Neurological Surgery and modified Fisher scale [MFS]) for aSAH outcomes. Univariable and multivariable analysis was then performed for various inpatient and long-term outcomes.

RESULTS

TMT was the best performing marker in our cohort with an AUC of 0.82, Somers' D statistic of 0.63 and Tau statistic 0.25. Of the frailty scores, the NSQIP performed the best (AUC 0.69), at levels comparable to traditional markers of aSAH, such as MFS (AUC 0.68). The threshold of 5.5 mm in TMT thickness was found to have a specificity of 0.93, sensitivity of 0.51, positive predictive value of 0.95 and negative predictive value of 0.42. After multivariate analysis, patients with TMT ≥ 5.5 mm (defined as non-frail), were less likely to experience delayed cerebral ischaemia (OR 0.11 [0.01 - 0.93], p = 0.042), any complications (OR 0.20 [0.06 - 0.069], p = 0.011), and had a larger proportion of favourable mRS on discharge (95.0% vs. 58.1%, p = 0.024) and at 3-months (95.0% vs. 64.5%, p = 0.048). However, the gap between unfavourable and favourable mRS was insignificant at the comparison of 1-year outcomes.

CONCLUSION

TMT, as a marker of sarcopaenia, correlated well with the presenting status, and outcomes of aSAH. Frailty, as defined by NSQIP, performed at levels equivalent to aSAH scores of clinical relevance, suggesting that, in patients presenting with acute brain injury, both non-neurological and neurological factors were complementary in the determination of eventual clinical outcomes. Further validation of these markers, in addition to exploration of other relevant frailty indices, may help to better prognosticate aSAH outcomes and allow for a precision medicine approach to decision making and optimization of best outcomes.

The performance of frailty in predictive modeling of short-term outcomes in the surgical management of metastatic tumors to the spine

BACKGROUND CONTEXT

The concept of frailty has become increasingly recognized, and while patients with cancer are at increased risk for frailty, its influence on perioperative outcomes in metastatic spine tumors is uncertain. Furthermore, the impact of frailty can be confounded by comorbidities or metastatic disease burden.

PURPOSE

The purpose of this study was to evaluate the influence of frailty and comorbidities on adverse outcomes in the surgical management of metastatic spine disease.

STUDY DESIGN/SETTING

Retrospective analysis of a nationwide database to include patients undergoing spinal fusion for metastatic spine disease.

PATIENT SAMPLE

A total of 1,974 frail patients who received spinal fusion with spinal metastasis, and 1,975 propensity score matched non-frail patients.

OUTCOME MEASURES

Outcomes analyzed included mortality, complications, length of stay (LOS), nonroutine discharges and costs.

METHODS

A validated binary frailty index (Johns Hopkins Adjusted Clinical Groups) was used to identify frail and non-frail groups, and propensity score-matched analysis (including demographics, comorbidities, surgical and tumor characteristics) was performed. Sub-group analysis of levels involved was performed for cervical, thoracic, lumbar and junctional spine. Multivariable-regression techniques were used to develop predictive models for outcomes using frailty and the Elixhauser Comorbidity Index (ECI).

RESULTS

7,772 patients underwent spinal fusion with spinal metastasis, of which 1,974 (25.4%) patients were identified as frail. Following propensity score matching for frail (n=1,974) and not-frail (n=1,975) groups, frailty demonstrated significantly greater medical complications (OR=1.58; 95% CI 1.33-1.86), surgical complications (OR=1.46; 95% CI 1.15-1.85), LOS (OR=2.65; 95% CI 2.09-3.37), nonroutine discharges (OR=1.79; 95% CI 1.46-2.20) and costs (OR=1.68; 95% CI 1.32-2.14). Differences in mortality were only observed in subgroup analysis and were greater in frail junctional and lumbar spine subgroups. Models using ECI alone (AUC=0.636-0.788) demonstrated greater predictive ability compared to those using frailty alone (AUC=0.633-0.752). However, frailty combined with ECI improved the prediction of increased LOS (AUC=0.811), cost (AUC=0.768), medical complications (AUC=0.723) and nonroutine discharges (AUC=0.718). Predictive modeling of frailty in subgroups demonstrated the greatest performance for mortality (AUC=0.750) in the lumbar spine, otherwise performed similarly for LOS, costs, complications, and discharge across subgroups.

CONCLUSIONS

A high prevalence of frailty existed in the current patient cohort. Frailty contributed to worse short-term adverse outcomes and could be more influential in the lumbar and junctional spine due to higher risk of deconditioning in the postoperative period. Predictions for short term outcomes can be improved by adding frailty to comorbidity indices, suggesting a more comprehensive preoperative risk stratification should include frailty.

Artificial Intelligence-Driven Prediction Modeling and Decision Making in Spine Surgery Using Hybrid Machine Learning Models

Healthcare systems worldwide generate vast amounts of data from many different sources. Although of high complexity for a human being, it is essential to determine the patterns and minor variations in the genomic, radiological, laboratory, or clinical data that reliably differentiate phenotypes or allow high predictive accuracy in health-related tasks. Convolutional neural networks (CNN) are increasingly applied to image data for various tasks. Its use for non-imaging data becomes feasible through different modern machine learning techniques, converting non-imaging data into images before inputting them into the CNN model. Considering also that healthcare providers do not solely use one data modality for their decisions, this approach opens the door for multi-input/mixed data models which use a combination of patient information, such as genomic, radiological, and clinical data, to train a hybrid deep learning model. Thus, this reflects the main characteristic of artificial intelligence: simulating natural human behavior. The present review focuses on key advances in machine and deep learning, allowing for multi-perspective pattern recognition across the entire information set of patients in spine surgery. This is the first review of artificial intelligence focusing on hybrid models for deep learning applications in spine surgery, to the best of our knowledge. This is especially interesting as future tools are unlikely to use solely one data modality. The techniques discussed could become important in establishing a new approach to decision-making in spine surgery based on three fundamental pillars: (1) patient-specific, (2) artificial intelligence-driven, (3) integrating multimodal data. The findings reveal promising research that already took place to develop multi-input mixed-data hybrid decision-supporting models. Their implementation in spine surgery may hence be only a matter of time.

Evaluation of Deep Learning-Based Automated Detection of Primary Spine Tumors on MRI Using the Turing Test

Background

Recently, the Turing test has been used to investigate whether machines have intelligence similar to humans. Our study aimed to assess the ability of an artificial intelligence (AI) system for spine tumor detection using the Turing test.

Methods

Our retrospective study data included 12179 images from 321 patients for developing AI detection systems and 6635 images from 187 patients for the Turing test. We utilized a deep learning-based tumor detection system with Faster R-CNN architecture, which generates region proposals by Region Proposal Network in the first stage and corrects the position and the size of the bounding box of the lesion area in the second stage. Each choice question featured four bounding boxes enclosing an identical tumor. Three were detected by the proposed deep learning model, whereas the other was annotated by a doctor; the results were shown to six doctors as respondents. If the respondent did not correctly identify the image annotated by a human, his answer was considered a misclassification. If all misclassification rates were &gt;30%, the respondents were considered unable to distinguish the AI-detected tumor from the human-annotated one, which indicated that the AI system passed the Turing test.

Results

The average misclassification rates in the Turing test were 51.2% (95% CI: 45.7%–57.5%) in the axial view (maximum of 62%, minimum of 44%) and 44.5% (95% CI: 38.2%–51.8%) in the sagittal view (maximum of 59%, minimum of 36%). The misclassification rates of all six respondents were &gt;30%; therefore, our AI system passed the Turing test.

Conclusion

Our proposed intelligent spine tumor detection system has a similar detection ability to annotation doctors and may be an efficient tool to assist radiologists or orthopedists in primary spine tumor detection.

Comparison of three predictive scoring systems for morbidity in oncological spine surgery

Estimating complications in oncological spine surgery is challenging. The objective of this study was to compare the accuracy of three scoring systems for predicting perioperative morbidity after surgery for spinal metastases. One-hundred and five patients who underwent surgery between 2013 and 2019 were included in this study. All patients had scores retrospectively calculated using the New England Spinal Metastasis Score (NESMS), Metastatic Spinal Tumor Frailty Index (MSTFI), and Anzuategui scoring systems. The main outcome measure was development of a medical complication (minor or major) within 30 days of surgery. The predictive ability for each system was assessed using receiver operating characteristic analysis and calculations of the area under the curve (AUC). The average age for all patients was 61 years and 61/105 patients (58.1%) were male. The most common primary tumor origins were hematologic (23.8%), prostate (16.2%), breast (14.3%), and lung (13.3%). The overall 30-day complication rate was 36.2% and the rate of major complications was 21.9%. Among all patients who underwent oncological spine surgery, the NESMS score had the highest AUC for 30-day overall (AUC 0.64; 95% CI, 0.53 - 0.75) and major morbidity (AUC 0.68; 95% CI, 0.54- 0.81) in our population. However, the accuracy did not meet the threshold for clinical utility. Future prospective validation of these systems in other populations is encouraged.