Prognostic models for spinal metastatic disease: evolution of methodologies, limitations, and future opportunities

Predictive Modeling for Spinal Metastatic Disease

Spinal metastasis is exceedingly common in patients with cancer and its prevalence is expected to increase. Surgical management of symptomatic spinal metastasis is indicated for pain relief, preservation or restoration of neurologic function, and mechanical stability. The overall prognosis is a major driver of treatment decisions; however, clinicians' ability to accurately predict survival is limited. In this narrative review, we first discuss the NOMS decision framework used to guide decision making in the treatment of patients with spinal metastasis. Given that decision making hinges on prognosis, multiple scoring systems have been developed over the last three decades to predict survival in patients with spinal metastasis; these systems have largely been developed using expert opinions or regression modeling. Although these tools have provided significant advances in our ability to predict prognosis, their utility is limited by the relative lack of patient-specific survival probability. Machine learning models have been developed in recent years to close this gap. Employing a greater number of features compared to models developed with conventional statistics, machine learning algorithms have been reported to predict 30-day, 6-week, 90-day, and 1-year mortality in spinal metastatic disease with excellent discrimination. These models are well calibrated and have been externally validated with domestic and international independent cohorts. Despite hypothesized and realized limitations, the role of machine learning methodology in predicting outcomes in spinal metastatic disease is likely to grow.

Prediction of 5-year survival in soft tissue leiomyosarcoma using a machine learning model algorithm

BACKGROUND AND OBJECTIVES

Leiomyosarcoma (LMS) is associated with one of the poorest overall survivals among soft tissue sarcomas. We sought to develop and externally validate a model for 5-year survival prediction in patients with appendicular or truncal LMS using machine learning algorithms.

METHODS

The Surveillance, Epidemiology, and End Results (SEER) database was used for development and internal validation of the models; external validation was assessed using our institutional database. Five machine learning algorithms were developed and then tested on our institutional database. Area under the receiver operating characteristic curve (AUC) and Brier score were used to assess model performance.

RESULTS

A total of 2209 patients from the SEER database and 81 patients from our tertiary institution were included. All models had excellent calibration with AUC 0.84-0.85 and Brier score 0.15-0.16. After assessing the performance indicators according to the TRIPOD model, we found that the Elastic-Net Penalized Logistic Regression outperformed other models. The AUCs of the institutional data were 0.83 (imputed) and 0.85 (complete-case analysis) with a Brier score of 0.16.

CONCLUSION

Our study successfully developed five machine learning algorithms to assess 5-year survival in patients with LMS. The Elastic-Net Penalized Logistic Regression retained performance upon external validation with an AUC of 0.85 and Brier score of 0.15.

[Translated article] Treatment strategies in vertebral metastasis. Need for multidisciplinary committees from the perspective of the surgeon. Narration of literature

Improvements in cancer diagnosis and treatment have improved survival. Secondarily, the number of patients who present a vertebral metastasis and the number with some morbidity in relation to these metastases also increase. Vertebral fracture, root compression or spinal cord injury cause a deterioration of their quality of life. The objective in the treatment of the vertebral metastasis must be the control of pain, maintenance of neurological function and vertebral stability, bearing in mind that in most cases it will be a palliative treatment. The treatment of these complications needs a multidisciplinary approach, radiologists, interventional radiologists, oncologists and radiation therapists, spine surgeons, but also rehabilitation or pain units. Recent studies show that a multidisciplinary approach of these patients can improve quality of life and even prognosis. In the present article, a review and reading of the literature on the multidisciplinary management of these patients is carried out.

External Validation of the Radiographic Investigation of the Distal Extension of Fractures Into the Articular Surface of the Tibia (RIDEFAST Study)

OBJECTIVE

To test the external validity of the fracture to plafond (FTP-length of fracture/distance to plafond) ratio to rule out distal intra-articular fractures (DIA) in distal tibial shaft fractures at an independent tertiary trauma center.

DESIGN

Retrospective cohort study.

SETTING

Two Level 1 trauma centers.

PATIENTS

Two hundred seventeen patients with a distal tibial shaft fracture in the model cohort and 146 patients in the validation cohort.

INTERVENTION

Radiographic measurements to calculate FTP ratio.

MAIN OUTCOME MEASUREMENTS

Calibration plots, area under receiver operating characteristic curve (AUC), and decision curve analyses to evaluate the external validity of FTP ratio to determine DIA.

RESULTS

The AUC for the anteroposterior (AP) FTP ratio was 0.83 [95% confidence interval (CI) 0.78-0.88] in the model data set and 0.86 (95% CI 0.80-0.91) in the validation data set. The AUC for the lateral FTP ratio was 0.82 (95% CI 0.77-0.87) in the model data set and 0.82 (95% CI 0.75-0.88) in the validation data set. The previously established AP FTP cutoff ratio of 0.61 had a 94% negative predictive value in the model cohort and a 100% negative predictive value in the validation cohort.

CONCLUSION

The FTP ratio is an effective and externally validated screening tool to rule out DIA in distal tibia shaft fractures.

LEVEL OF EVIDENCE

Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.

Prognostic scores for survival as decisional support for surgery in spinal metastases: a performance assessment systematic review

PURPOSE

To review the evidence on the relative prognostic performance of the available prognostic scores for survival in spinal metastatic surgery in order to provide a recommendation for use in clinical practice.

METHODS

A systematic review of comparative external validation studies assessing the performance of prognostic scores for survival in independent cohorts was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines. Eligible studies were identified through Medline and Embase until May 2021. Studies were included when they compared at least four survival scoring systems in surgical or mixed cohorts across all primary tumor types. Predictive performance was assessed based on discrimination and calibration for 3-month, 1-year and overall survival, and generalizability was assessed based on the characteristics of the development cohort and external validation cohorts. Risk of bias and concern regarding applicability were assessed based on the 'Prediction model study Risk Of Bias Assessment Tool' (PROBAST).

RESULTS

Twelve studies fulfilled the inclusion criteria and covered 17 scoring systems across 5.130 patients. Several scores suffer from suboptimal development and validation. The SORG Nomogram, developed in a large surgical cohort, showed good discrimination on 3-month and 1-year survival, good calibration and was superior in direct comparison with low risk of bias and low concern regarding applicability. Machine learning algorithms are promising as they perform equally well in direct comparison. Tokuhashi, Tomita and other traditional risk scores showed suboptimal performance.

CONCLUSION

The SORG Nomogram and machine learning algorithms outline superior performance in survival prediction for surgery in spinal metastases. Further improvement by comparative validation in large multicenter, prospective cohorts can still be obtained. Given the heterogeneity of spinal metastases, superior methodology of development and validation is key in improving future machine learning systems.

Interhospital transfer status for spinal metastasis patients in the United States is associated with more severe clinical presentations and higher rates of inpatient complications

OBJECTIVE

In patients with metastatic spinal disease (MSD), interhospital transfer can potentially impact clinical outcomes as the possible benefits of transferring a patient to a higher level of care must be weighed against the negative effects associated with potential delays in treatment. While the association of clinical outcomes and transfer status has been examined in other specialties, the relationship between transfer status, complications, and risk of mortality in patients with MSD has yet to be explored. The purpose of this study was to examine the impact of transfer status on in-hospital mortality and clinical outcomes in patients diagnosed with MSD.

METHODS

The National (Nationwide) Inpatient Sample (NIS) database was retrospectively queried for adult patients diagnosed with vertebral pathological fracture and/or spinal cord compression in the setting of metastatic disease between 2012 and 2014. Demographics, baseline characteristics (e.g., metastatic spinal cord compression [MSCC] and paralysis), comorbidities, type of intervention, and relevant patient outcomes were controlled in a multivariable logistic regression model to analyze the association of transfer status with patient outcomes.

RESULTS

Within the 10,360 patients meeting the inclusion and exclusion criteria, higher rates of MSCC (50.2% vs 35.9%, p < 0.001) and paralysis (17.3% vs 8.4%, p < 0.001) were observed in patients transferred between hospitals compared to those directly admitted. In univariable analysis, a higher percentage of transferred patients underwent surgical intervention (p < 0.001) when compared with directly admitted patients. After controlling for significant covariates and surgical intervention, transferred patients were more likely to develop in-hospital complications (OR 1.34, 95% CI 1.18-1.52, p < 0.001), experience prolonged length of stay (OR 1.33, 95% CI 1.16-1.52, p < 0.001), and have a discharge disposition other than home (OR 1.70, 95% CI 1.46-1.98, p < 0.001), with no significant difference in inpatient mortality rates.

CONCLUSIONS

Patients with MSD who were transferred between hospitals demonstrated more severe clinical presentations and higher rates of inpatient complications compared to directly admitted patients, despite demonstrating no difference in in-hospital mortality rates.

Clinician Experiences in Treatment Decision-Making for Patients with Spinal Metastases: A Qualitative Study

BACKGROUND

Effective management of metastatic disease requires multidisciplinary input and entails high risk of disease-related and treatment-related morbidity and mortality. The factors that influence clinician decision-making around spinal metastases are not well understood. We conducted a qualitative study that included a multidisciplinary cohort of physicians to evaluate the decision-making process for treatment of spinal metastases from the clinician's perspective.

METHODS

We recruited operative and nonoperative clinicians, including orthopaedic spine surgeons, neurosurgeons, radiation oncologists, and physiatrists, from across North America to participate in either a focus group or a semistructured interview. All interviews were audiorecorded and transcribed verbatim. We then performed a thematic analysis using all of the available transcript data. Investigators sequentially coded transcripts and identified recurring themes that encompass overarching patterns in the data and directly bear on the guiding study question. This was followed by the development of a thematic map that visually portrays the themes, the subthemes, and their interrelatedness, as well as their influence on treatment decision-making.

RESULTS

The thematic analysis revealed that numerous factors influence provider-based decision-making for patients with spinal metastases, including clinical elements of the disease process, treatment guidelines, patient preferences, and the dynamics of the multidisciplinary care team. The most prominent feature that resonated across all of the interviews was the importance of multidisciplinary care and the necessity of cohesion among a team of diverse health-care providers. Respondents emphasized aspects of care-team dynamics, including effective communication and intimate knowledge of team-member preferences, as necessary for the development of appropriate treatment strategies. Participants maintained that the primary role in decision-making should remain with the patient.

CONCLUSIONS

Numerous factors influence provider-based decision-making for patients with spinal metastases, including multidisciplinary team dynamics, business pressure, and clinician experience. Participants maintained a focus on shared decision-making with patients, which contrasts with patient preferences to defer decisions to the physician, as described in prior work.

CLINICAL RELEVANCE

The results of this thematic analysis document the numerous factors that influence provider-based decision-making for patients with spinal metastases. Our results indicate that provider decisions regarding treatment are influenced by a combination of clinical characteristics, perceptions of patient quality of life, and the patient's preferences for care.

Predicting tumor-specific survival in patients with spinal metastatic renal cell carcinoma: which scoring system is most accurate?

OBJECTIVE

Although several prognostic scores for spinal metastatic disease have been developed in the past 2 decades, the applicability and validity of these models to specific cancer types are not yet clear. Most of the data used for model formation are from small population sets and have not been updated or externally validated to assess their performance. Developing predictive models is clinically relevant as prognostic assessment is crucial to optimal decision-making, particularly the decision for or against spine surgery. In this study, the authors investigated the performance of various spinal metastatic disease risk models in predicting prognosis for spine surgery to treat metastatic renal cell carcinoma (RCC).

METHODS

Data of patients who underwent surgery for RCC metastatic to the spine at 2 tertiary centers between 2010 and 2019 were retrospectively retrieved. The authors determined the prognostic value associated with the following scoring systems: the Tomita score, original and revised Tokuhashi scores, original and modified Bauer scores, Katagiri score, the Skeletal Oncology Research Group (SORG) classic algorithm and nomogram, and the New England Spinal Metastasis Score (NESMS). Regression analysis of patient variables in association with 1-year survival after surgery was assessed using Cox proportional hazard models. Calibration and time-dependent discrimination analysis were tested to quantify the accuracy of each scoring system at 3 months, 6 months, and 1 year.

RESULTS

A total of 86 metastatic RCC patients were included (median age 64 years [range 29-84 years]; 63 males [73.26%]). The 1-year survival rate was 72%. The 1-year survival group had a good performance status (Karnofsky Performance Scale [KPS] score 80%-100%) and an albumin level > 3.5 g/dL (p < 0.05). Multivariable-adjusted Cox regression analysis showed that poor performance status (KPS score < 70%), neurological deficit (Frankel grade A-D), and hypoalbuminemia (< 3.5 g/dL) were associated with a higher risk of death before 1 year (p < 0.05). The SORG nomogram, SORG classic, original Tokuhashi, and original Bauer demonstrated fair performance (0.7 < area under the curve < 0.8). The NESMS differentiates survival among the prognostic categories with the highest accuracy (area under the curve > 0.8).

CONCLUSIONS

The present study shows that the most cited and commonly used scoring systems have a fair performance predicting survival for patients undergoing spine surgery for metastatic RCC. The NESMS had the best performance at predicting 1-year survival after surgery.

Treatment Strategy for Metastatic Spinal Tumors: A Narrative Review

Metastatic spinal tumors are common, and their rising incidence can be attributed to the expanding aging population and increased survival rates among cancer patients. The decision-making process in the treatment of spinal metastasis requires a multidisciplinary approach that includes medical and radiation oncology, surgery, and rehabilitation. Various decision-making systems have been proposed in the literature in order to estimate survival and suggest appropriate treatment options for patients experiencing spinal metastasis. However, recent advances in treatment modalities for spinal metastasis, such as stereotactic radiosurgery and minimally invasive surgical techniques, have reshaped clinical practices concerning patients with spinal metastasis, making a demand for further improvements on current decision-making systems. In this review, recent improvements in treatment modalities and the evolution of decision-making systems for metastatic spinal tumors are discussed.

Impact of body fat distribution and sarcopenia on the overall survival in patients with spinal metastases receiving radiotherapy treatment: a prospective cohort study

Introduction: An increasing number of patients is diagnosed with spinal metastases due to elevated cancer incidence and improved overall survival. Patients with symptomatic spinal bone metastases often receive radiotherapy with or without surgical stabilisation. Patients with a life expectancy of less than 3 months are generally deemed unfit for surgery, therefore adequate pre-treatment assessment of life expectancy is necessary. The aim of this study was to assess new factors associated with overall survival for this category of patients.Patients and methods: Patients who received radiotherapy for thoracic or lumbar spinal metastases from June 2013 to December 2016 were included in this study. The pre-treatment planning CT for radiotherapy treatment was used to assess the patient's visceral fat area, subcutaneous fat area, total muscle area and skeletal muscle density on a single transverse slice at the L3 level. The total muscle area was used to assess sarcopenia. Furthermore, data were collected on age, sex, primary tumour, Karnofsky performance score, medical history, number of bone metastases, non-bone metastases and neurological symptoms. Univariable and multivariable cox regressions were performed to determine the association between our variables of interest and the survival at 90 and 365 days.Results: A total of 310 patients was included. The median age was 67 years. Overall survival rates for 90 and 365 days were 71% and 36% respectively. For 90- and 365-day survival, the Karnofsky performance score, muscle density and primary tumour were independently significantly associated. The visceral or subcutaneous fat area and their ratio and sarcopenia were not independently associated with overall survival.Conclusions: Of the body morphology, only muscle density was statistically significant associated with overall survival after 90 and 365 days in patients with spinal bone metastases. Body fat distribution was not significantly associated with overall survival.

Predictive Analytics in Spine Oncology Research: First Steps, Limitations, and Future Directions

The potential of big data analytics to improve the quality of care for patients with spine tumors is significant. At this moment, the application of big data analytics to oncology and spine surgery is at a nascent stage. As such, efforts are underway to advance data-driven oncologic care, improve patient outcomes, and guide clinical decision making. This is both relevant and critical in the practice of spine oncology as clinical decision making is often made in isolation looking at select variables deemed relevant by the physician. With rapidly evolving therapeutics in surgery, radiation, interventional radiology, and oncology, there is a need to better develop decision-making algorithms utilizing the vast data available for each patient. The challenges and limitations inherent to big data analyses are presented with an eye towards future directions.