Are the SORG and OPTImodel, Tokuhashi and Tomita algorithms still suitable as predictors of survival in patients with vertebral metastases in routine clinical practice?

Are the SORG and OPTImodel, Tokuhashi and Tomita algorithms still suitable as predictors of survival in patients with vertebral metastases in routine clinical practice? | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 6 February 2025 V1 Latest version Share on Are the SORG and OPTImodel, Tokuhashi and Tomita algorithms still suitable as predictors of survival in patients with vertebral metastases in routine clinical practice? Authors : Julián Cabria Fernández 0000-0002-9140-5710 [email protected] , Pablo González-Herráez Fernández , Javier Mateo Negreira , and Pedro Arcos González Authors Info & Affiliations https://doi.org/10.22541/au.173883524.47342656/v1 Published Cancer Medicine Version of record Peer review timeline 276 views 163 downloads Contents Abstract Supplementary Material Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Objectives: To evaluate the performance of the Tokuhashi, Tomita, SORG machine learning (SORG ML), and OPTImodel algorithms as survival predictors for vertebral metastases in clinical practice. Materials and Methods: A retrospective study (2013–2023) analyzed 573 patients from Cabueñes University Hospital (Asturias, Spain). Thirty-two demographic, epidemiological, clinical, and analytical variables were considered, including diagnosis chronology and survival. Results: Among the 573 patients studied, 272 (47.4%) presented visceral metastases at the time of diagnosis. A total of 362 patients (63.2%) had associated comorbidities. The most frequent primary histological diagnoses in these patients were lung 147, (25.7%), prostate 146 (25.5%), breast 118 (20.6%), kidney 30 (5.2%), and colorectal 29 (5.1%). The median survival of the cohort was 185 days. The accuracy rates for the Tokuhashi, SORG ML, OPTImodel, and Tomita algorithms were 0.5509, 0.4812, 0.3404, and 0.3858, respectively. The models with the highest accuracy rates in specific time segments were Tokuhashi (77.5% for less than 6 months) and OPTImodel (90.8% for more than 1 year). The areas under the curve (AUC) for survival intervals were as follows: Tokuhashi at 42 days (73.19%), 90 days (79.3%), and 365 days (82.73%); Tomita at 42 days (69.27%), 90 days (76.82%), and 365 days (78.79%); SORG ML at 42 days (52.77%), 90 days (51.69%), and 365 days (51.38%). Conclusions: All models showed relatively low accuracy. The newer models (OPTImodel, SORG ML) did not outperform the traditional Tomita and Tokuhashi in predicting survival for vertebral metastases patients. Are the SORG and OPTImodel, Tokuhashi and Tomita algorithms still suitable as predictors of survival in patients with vertebral metastases in routine clinical practice? Julián Cabria Fernández i , Pablo González-Herráez Fernández i ,Javier Mateo Negreira i Pedro Arcos González ii . i Department of Orthopedic Surgery and Traumatology, Cabueñes University Hospital, Gijón (Asturias, Spain). ii Departament of Medicine. University of Oviedo. Corresponding Author: Julián Cabria Fernández, [email protected] , Ctra Piles-Infanzón km 1146, Urbanización el Pisón nº 105, CP 33203, Gijón, Asturias, Spain. Telephone 66561842 jabbrv-ltwa-all.ldf jabbrv-ltwa-en.ldf FUNDING This research did not receive specific funding from public sector agencies, commercial entities, or non-profit organizations. CONFLICT OF INTEREST None declared. ETHICAL CONSIDERATIONS The present study has been exempted by the Research Ethics Committee of the Principality of Asturias (Spain), with registration ID CEImPA 2024.307. ACKNOWLEDGMENTS We extend our special gratitude to the Principado de Asturias Health Research Institute (ISPA), particularly to Paula Fernández Martínez, for their indispensable support in methodological and statistical advice and management. CONDENSED ABSTRACT The primary aim of our study is to evaluate the usefulness of four predictive models for survival in patients with vertebral metastases: Tokuhashi, Tomita, SORG machine learning (SORG ML), and OPTImodel. This single-center retrospective study included 573 patients diagnosed between 2013 and 2022. A total of 32 variables applicable to the four models were analyzed to compare estimated survival with observed outcomes. Our findings indicate that, while recent models such as SORG ML and OPTImodel offer new perspectives, they do not surpass the accuracy of the classic Tomita and Tokuhashi models in generalized populations of patients with vertebral metastases. ABSTRACT Objectives: To evaluate the performance of the Tokuhashi, Tomita, SORG machine learning (SORG ML), and OPTImodel algorithms as survival predictors for vertebral metastases in clinical practice. Materials and Methods: A retrospective study (2013–2023) analyzed 573 patients from Cabueñes University Hospital (Asturias, Spain). Thirty-two demographic, epidemiological, clinical, and analytical variables were considered, including diagnosis chronology and survival. Results: Among the 573 patients studied, 272 (47.4%) presented visceral metastases at the time of diagnosis. A total of 362 patients (63.2%) had associated comorbidities. The most frequent primary histological diagnoses in these patients were lung 147, (25.7%), prostate 146 (25.5%), breast 118 (20.6%), kidney 30 (5.2%), and colorectal 29 (5.1%). The median survival of the cohort was 185 days. The accuracy rates for the Tokuhashi, SORG ML, OPTImodel, and Tomita algorithms were 0.5509, 0.4812, 0.3404, and 0.3858, respectively. The models with the highest accuracy rates in specific time segments were Tokuhashi (77.5% for less than 6 months) and OPTImodel (90.8% for more than 1 year). The areas under the curve (AUC) for survival intervals were as follows: Tokuhashi at 42 days (73.19%), 90 days (79.3%), and 365 days (82.73%); Tomita at 42 days (69.27%), 90 days (76.82%), and 365 days (78.79%); SORG ML at 42 days (52.77%), 90 days (51.69%), and 365 days (51.38%). Conclusions: All models showed relatively low accuracy. The newer models (OPTImodel, SORG ML) did not outperform the traditional Tomita and Tokuhashi in predicting survival for vertebral metastases patients. KEYWORDS Spinal metastasis, Prognosis, Tokuhashi, Tomita, SORG machine learning, OPTImodel INTRODUCTION Bone metastases occur in the most advanced stage of tumor disease, and their most frequent location is the vertebrae, often being multiple at the time of diagnosis 1 . Metastases account for up to 90% of all malignant bone tumors in adulthood 2 3 , and their incidence is increasing due to several factors, including improved survival of patients receiving targeted oncological therapies and enhanced diagnostic accuracy with more precise imaging tools 4 . The disease is typically diagnosed in its advanced stages, and treatment is predominantly palliative. Surgical treatment with curative intent is only feasible in a limited number of cases. Therefore, there is consensus that a minimum survival of three months is necessary for the temporary loss of quality of life caused by extensive surgery to be clinically justified for the patient 5 . Proper selection of candidates is crucial, and various prognostic algorithms have been developed based on clinical, analytical, and histological variables. Unfortunately, there is currently no clear consensus on the superiority of one algorithm over others 6 . Among the most commonly used algorithms are those developed by Tokuhashi in 1990 7 and revised in 2005 8 , by Tomita in 2001 9 , by Bauer in 1995 10 and modified in 2008 11 , by Van Der Linden in 2005 12 , by Katagiri in 2014 13 , by Ghori in 2015 14 , the OPTImodel 15 and NESMS 16 models (both from 2016), and the SORG algorithm from 2016 17 , along with its updated version in nomogram form from 2017 18 and its 2019 machine learning-based version (SORG ML) 19 . The objective of this study is to evaluate the current performance of the Tokuhashi, Tomita, SORG ML, and OPTImodel algorithms as survival predictors in patients with vertebral metastases in routine clinical practice. MATERIALS AND METHODS A retrospective observational study was conducted on 622 patients diagnosed with vertebral metastases at Cabueñes University Hospital (Gijón, Spain) between 2013 and 2023. Inclusion criteria were as follows: patients had to be over 18 years old at the time of diagnosis, have an imaging test (CT, MRI, bone scintigraphy, etc.) confirming the location of the metastases, and have a complete record of all variables collected in their clinical history. The present study has been exempted by the Research Ethics Committee of the Principality of Asturias (Spain), with registration ID CEImPA 2024.307. The studied variables included: sex, age at diagnosis, number of bone metastases, number of vertebral metastases, presence or absence of visceral metastases, feasibility of treating these metastases, presence or absence of lymphatic metastases, presence or absence of brain metastases, presence or absence of pathological fractures, functional status using the ECOG and Karnofsky scales, degree of neurological impairment based on the ASIA scale, primary tumor diagnosis, presence or absence of comorbidities according to the Charlson scale, use of prior systemic therapy, and body mass index (BMI). Analytical variables recorded included: hemoglobin (g/dL), platelets (×10³/μL), leukocytes (×10³/μL), lymphocytes (×10³/μL), neutrophils (×10³/μL), serum creatinine (mg/dL), international normalized ratio (INR) of prothrombin time, alkaline phosphatase (IU/L), and serum albumin (g/L). Additionally, the vital status of each patient (alive or deceased) at the study’s end date (February 2024) was compared with the survival estimates provided by the Tokuhashi, Tomita, OPTImodel, and SORG ML predictive models. When applying the Tokuhashi, Tomita, SORG ML, and OPTImodel algorithms, estimated survival scores at 90, 180, and 365 days were obtained, respectively. For the SORG ML algorithm, the estimated survival percentages for these intervals were directly provided by its online tool 20 . For the distribution analysis of the variables, central tendency parameters such as mean, median, standard deviation (SD), and 95% confidence intervals (CI95%) were used. Sensitivity, specificity, and positive and negative predictive values of each algorithm were calculated based on confusion matrices. Receiver operating characteristic (ROC) curves were used to establish the corresponding area under the curve (AUC), and the Kolmogorov-Smirnov test was used to assess the normality of variables. To allow comparability with studies employing different methodologies, ROC curves for the algorithms studied were established at 42, 90, and 365 days of survival, matching the survival prediction thresholds of the SORG ML method. This could not be done for the OPTImodel method as it does not employ a quantitative scale. Following bibliographic recommendations, AUCs above 0.8 were deemed optimal, while those below 0.7 were considered insufficient. Statistical analysis was performed using SPSS version 22.0. RESULTS Of the 622 patients identified with vertebral metastases at the time of diagnosis, 49 were excluded for not meeting inclusion criteria. Specifically, 12 were excluded due to the unavailability of diagnostic imaging for analysis, and 37 due to missing data in the collected variables. The final number of patients included in the study was 573. The mean age at diagnosis was 69.1 years (SD = 11.4). Of the patients, 361 (63%) were male, with a mean BMI of 27.1 (SD = 5.33), and 362 (63.2%) had comorbidities as assessed by the Charlson Comorbidity Index. In terms of functional status, 287 patients (50%) had a Karnofsky score of 70, indicating good functional condition, while 143 (25%) had an initial score of 60 or less (Median = 70, Q1 = 60, Q3 = 80). A total of 516 patients (90.1%) had no spinal cord involvement according to the ASIA scale. Table 1 presents the analytical values of the patients at the time of diagnosis. The mean number of vertebral metastases at diagnosis was 2.89 (SD = 1.6), with a mean of 3.93 (SD = 1.88) total bone metastases per patient at diagnosis. A total of 272 patients (47.5%) had visceral metastases, of which 160 (27.9%) were potentially treatable. Additionally, 305 (53.3%) had lymphatic metastases, and 29 (5.1%) had brain metastases. Of the total sample, 342 patients (56.6%) had received prior systemic therapy, either for the primary tumor or its metastases. The survival percentage at the end of the 10-year study interval was 8.63%. The most frequent primary tumor was lung cancer, with 147 cases (25.6%), including both small-cell and non-small-cell subtypes, followed by prostate cancer with 146 cases (25.3%) and breast cancer with 118 cases (20.14%). Table 2 shows the frequencies of the other histological diagnoses. The highest survival rates were observed in prostate and breast cancers, while the lowest were found in bladder, pancreatic, and colorectal cancers. Regarding the models, the Tokuhashi algorithm divides survival into three categories: low (less than 6 months; 0–8 points), intermediate (6–12 months; 9–11 points), and high (greater than 1 year; 12–15 points) 9 . In our study, 252 patients (44.11%) were classified in the low survival group, of which 188 (74.9%) were correctly categorized as alive or deceased. In the intermediate survival group (9–11 points), which included 199 patients (34.8%), only 30 (15.2%) were correctly categorized. In the high survival group (12–15 points), which included 122 patients (21.09%), 93 (77.5%) were correctly categorized. Sensitivity was 0.67 for the low survival group, 0.42 for the intermediate group, and 0.43 for the high group, with specificities of 0.78, 0.66, and 0.92, respectively. Positive predictive values were 0.75 for the low survival group, 0.89 for the intermediate group, and 0.73 for the high survival group. Negative predictive values were 0.7, 0.89, and 0.73, respectively. The overall accuracy of the Tokuhashi algorithm was 0.5682 (CI 95%: 0.5063–0.5895). In our study, the Tokuhashi method demonstrated adequate predictive capacity for survival below 6 months and above 1 year, but its predictive ability declined significantly for intermediate survival values. As shown in Figure 1, the ROC curves provided AUCs of 73.19% (CI 95%: 68.44–77.95%) for survival at 42 days, 79.3% (CI 95%: 75.62–82.99%) for 90 days, and 82.73% (CI 95%: 79.29–86.17%) for 365 days. The Tomita algorithm considers four categories: 2–4 points (more than 2 years), 4–6 points (1–2 years), 6–8 points (6–12 months), and 8–10 points (less than 3 months) 11 . In our study, 201 patients (35.2%) were in the 2–4 points group, of which 87 (43.3%) were correctly categorized as alive or deceased. In the 4–6 points group, which included 143 patients (25%), only 22 (15.4%) were correctly categorized. In the 6–8 points group, with 99 patients (17.3%), 29 (29.3%) were correctly categorized, while in the 8–10 points group, with 128 patients (22.4%), 91 (71.1%) were correctly categorized. Sensitivity for the Tomita method was 0.68 for the 2–4 points group, 0.25 for the 4–6 points group, 0.19 for the 6–8 points group, and 0.45 for the 8–10 points group, with specificities of 0.74, 0.75, 0.83, and 0.9, respectively. Positive predictive values were 0.43, 0.15, 0.29, and 0.71, and negative predictive values were 0.9, 0.84, 0.74, and 0.75, respectively. The overall accuracy of the algorithm was 0.4011 (CI 95%: 0.3606–0.4426). Similar to Tokuhashi, the Tomita algorithm showed high predictive capacity for extreme survival values (<3 months) but lower accuracy for survival over two years, with prediction success rates below 30% for intermediate categories. For the Tomita method, the AUCs on the ROC curves were 69.27% (CI 95%: 64.08–74.47%) for 42 days, 76.82% (CI 95%: 72.84–80.79%) for 90 days, and 78.79% (CI 95%: 75.15–82.42%) for 365 days (Figure 2: ROC curves for Tomita at 42 (2A), 90 (2B), and 365 days (2C)). For OPTImodel, four categories are defined: D (less than 90 days), C (90 to 180 days), B (180 to 365 days), and A (more than 365 days) 17 . Figure 3 shows the Kaplan-Meier curve for each OPTImodel interval. In our study, 36 patients (6.28%) were in group A, of which 34 (94.4%) were correctly categorized as alive or deceased; 134 patients (23.4%) were in group B, with 16 (11.9%) correctly categorized; 213 patients (37.2%) were in group C, with 31 (14.6%) correctly categorized; and 185 patients (32.1%) were in group D, with 110 (59.5%) correctly categorized. The sensitivity values for the groups were 0.16 for group A, 0.22 for group B, 0.38 for group C, and 0.54 for group D. Specificity values were 0.99, 0.76, 0.63, and 0.8, respectively. Positive predictive values were 0.95 for group A, 0.12 for group B, 0.14 for group C, and 0.59 for group D, with corresponding negative predictive values of 0.66, 0.87, 0.86, and 0.76. The overall accuracy of the OPTImodel algorithm was 0.3368 (CI 95%: 0.2981–0.3773). As observed with other methods, OPTImodel showed high prediction success for extreme survival values (greater than 1 year or less than 90 days) but poor predictive success rates (below 15%) for intermediate survival values. Due to the study’s methodology, ROC curves could not be obtained for OPTImodel as it does not use a quantitative scale. The SORG ML method does not group patients by prognosis but directly calculates the probability of survival at 42, 90, and 365 days 21 . Due to this difference from other methods, calculations were made directly for each survival threshold. For the 42-day survival threshold, sensitivity was 0.21, specificity was 0.8, the positive predictive value was 0.8, and the negative predictive value was 0.21. The accuracy of the SORG ML algorithm for the 42-day threshold was 0.3333 (CI 95%: 0.2947–0.3737). For the 90-day threshold, sensitivity, specificity, positive predictive value, and negative predictive value were 0.75, 0.20, 0.63, and 0.31, respectively. The accuracy for this threshold was 0.5561 (CI 95%: 0.5143–0.5974). For the 365-day threshold, sensitivity, specificity, positive predictive value, and negative predictive value were 0.24, 0.8, 0.42, and 0.63, respectively. The accuracy of the SORG ML algorithm for the 365-day threshold was 0.3333 (CI 95%: 0.2947–0.3737). For the 42-day threshold, the AUC on the ROC curve was 52.77% (CI 95%: 46.65–58.89%); for the 90-day threshold, it was 51.69% (CI 95%: 46.75–56.62%); and for the 365-day threshold, it was 51.38% (CI 95%: 46.41–56.34%). Figure 4 shows the ROC curves for SORG ML at 42 (4A), 90 (4B), and 365 days (4C). The method with the highest accuracy was SORG ML at 365 days (0.5877), followed by SORG ML at 90 days (0.5561) and Tokuhashi (0.5482). Table 3 summarizes the accuracy and AUC of the prognostic algorithms, ranking them from highest to lowest accuracy and showing the AUC for each method. Only the Tokuhashi method achieved AUC values greater than 70% across all intervals, while Tomita yielded results close to this threshold, and SORG ML showed significantly lower values. DISCUSSION The predictive accuracy of the studied methods contrasts, particularly in the case of the SORG ML algorithm, with the numerous studies supporting its use. These studies mainly involve external validations 24 25 26 and comparative analyses 6 22 . However, our results find support in studies such as those by Truong et al 23 . or Hibberd et al 27 . Regarding the Tokuhashi method, external validation attempts, like the one conducted by Hernandez-Fernandez et al 27 ., have shown results consistent with ours. We conducted a literature review that identified 45 relevant articles from which we extracted the AUC values of the algorithms at 3 months, 6 months, and 1 year, as these are the most frequently reported intervals. We also recorded the overall conclusions and whether the sample included only surgically treated patients for vertebral metastases or, as in our study, made no distinction between treated and untreated patients. The average AUCs from the review were as follows: 0.66, 0.63, and 0.67 for Tokuhashi at 1, 3, and 12 months, respectively; 0.61, 0.63, and 0.69 for Tomita; 0.73, 0.67, and 0.75 for SORG ML; and 0.67 overall for OPTImodel. The data from our study align with published results for the Tomita and Tokuhashi methods but not for SORG ML, where our AUC values were barely above random chance (0.51–0.52), as shown in Table 4, which lists the average AUCs from the reviewed studies. Only two studies applied any variant of the SORG method to a sample that did not distinguish between operated and non-operated patients 23 29 . In the study by Tarabay et al 30 ., the AUC decreased significantly (0.59 at 1 year), while in Su et al 32 ., the AUC remained at 0.78. The other 17 studies involving SORG ML used samples exclusively comprising surgically treated patients 6 17 18 19 22 24 25 26 29 31 33 34 35 36 37 38 . The discrepancy between our results and those of other authors for the SORG ML method could be partially explained by differences in the patient populations studied, as our sample included both operated and non-operated patients without distinction. Moreover, biases may exist in validating prognostic algorithms in studies that include only surgically treated patients, as this subgroup is subjectively more favorable. Consequently, the majority of patients diagnosed with vertebral metastases who do not undergo surgery are excluded from such analyses. The main limitations of this study relate to the fact that our patients were older and had more comorbidities compared to those in other studies 25 . Small variations were also noted in the epidemiology of primary tumors, with a higher proportion of prostate and colorectal cancers in our sample, potentially attributable to the older age of the patients or geographic variations. Our sample was broad and heterogeneous, covering a 10-year study period, which may have introduced differences in the prognosis of older versus more recent patients. Additionally, quantitative comparisons with the OPTImodel method could not be performed due to the lack of AUC data. We did not record which patients underwent surgery, which could provide additional information on prognostic subcategories. This research group is currently conducting a study comparing outcomes in surgically and non-surgically treated patients with vertebral metastases. Multidisciplinary oncological committee decisions were also not recorded, as these were often absent from the retrospective review of clinical records for many patients in the study. We believe this could be a relevant factor for guiding prognosis. CONCLUSIONS Overall, the methods compared in our study demonstrated relatively low predictive accuracy. In our patient sample, the OPTImodel and SORG ML methods failed to outperform the classic Tomita and Tokuhashi methods in predicting survival in patients with vertebral metastases. Our findings align with those of other authors regarding the Tokuhashi and Tomita methods but not for SORG ML. It seems, therefore, that none of the four methods studied possesses sufficient prognostic quality for generalization in clinical practice, which may significantly limit their utility. Nonetheless, individual clinical parameters such as general health status, disease extent (e.g., visceral metastases, multiple bone metastases), and analytical values remain fundamental factors in evaluating these patients. REFERENCES 1. Al Farii H, Aoude A, Al Shammasi A, Reynolds J, Weber M. Surgical Management of Metastatic Spine Disease: A Review of the Literature and Proposed Algorithm. Global Spine J. March 2023;13(2):486-98. 2. Ziu E, Viswanathan VK, Mesfin FB. Spinal Metastasis. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2023 [cited September 3, 2023]. Available at: http://www.ncbi.nlm.nih.gov/books/NBK441950/ 3. Ramírez M, Codina Frutos G, Vergés R, Tortajada JC, Núñez S. Estrategias de tratamiento en la metástasis vertebral. Necesidad de comités multidisciplinarios desde la perspectiva del cirujano. Narración de la literatura. Revista Española de Cirugía Ortopédica y Traumatología. May 2023;S1888441523001285. 4. Hong SH, Chang BS, Kim H, Kang DH, Chang SY. An Updated Review on the Treatment Strategy for Spinal Metastases from the Spine Surgeon’s Perspective. Asian Spine J. October 2022;16(5):799-811. 5. Smeijers S, Depreitere B. Prognostic Scores for Survival as Decision Support for Surgery in Spinal Metastases: A Performance Assessment Systematic Review. Eur Spine J. October 2021;30(10):2800-24. 6. Yan Y, Zhong G, Lai H, Huang C, Yao M, Zhou M, et al. Comparing the Accuracy of Seven Scoring Systems in Predicting Survival of Lung Cancer Patients With Spinal Metastases. Spine (Phila Pa 1976). July 15, 2023;48(14):1009-16. 7. Tokuhashi Y, Matsuzaki H, Toriyama S, Kawano H, Ohsaka S. Scoring System for the Preoperative Evaluation of Metastatic Spine Tumor Prognosis. Spine (Phila Pa 1976). 1990;15(11):1110-3. 8. Tokuhashi Y, Matsuzaki H, Oda H, Oshima M, Ryu J. A Revised Scoring System for Preoperative Evaluation of Metastatic Spine Tumor Prognosis. Spine (Phila Pa 1976). 2005;30(19):2186-91. 9. Tomita K, Kawahara N, Kobayashi T, Yoshida A, Murakami H, Akamaru T. Surgical Strategy for Spinal Metastases. Spine (Phila Pa 1976). 2001;26(3):298-306. 10. Bauer HC, Wedin R. Survival After Surgery for Spinal and Extremity Metastases: Prognostication in 241 Patients. Acta Orthop Scand. 1995;66(2):143-6. 11. Leithner A, Radl R, Gruber G, et al. Predictive Value of Seven Preoperative Prognostic Scoring Systems for Spinal Metastases. Eur Spine J. 2008;17:1488-95. 12. van der Linden YM, Dijkstra SPDS, Vonk EJA, Marijnen CAM, Leer JWH, Group for TDBMS. Prediction of Survival in Patients With Metastases in the Spinal Column. Cancer. 2005;103(2):320-8. 13. Katagiri H, Okada R, Takagi T, Takahashi M, Murata H, Harada H, et al. New Prognostic Factors and Scoring System for Patients With Skeletal Metastases. Cancer Med. October 2014;3(5):1359-67. 14. Ghori AK, Leonard DA, Schoenfeld AJ. Modeling 1-Year Survival After Surgery on the Metastatic Spine. Spine J. 2015;15(11):2345-50. 15. Bollen L, van der Linden YM, Pondaag W, Fiocco M, Pattynama BPM, Marijnen CAM, et al. Prognostic Factors Associated With Survival in Patients With Symptomatic Spinal Bone Metastases: A Retrospective Cohort Study of 1,043 Patients. Neuro Oncol. July 2014;16(7):991-8. 16. Schoenfeld AJ, Le HV, Marjoua Y, et al. Assessing the Utility of a Clinical Prediction Score Regarding 30-Day Morbidity and Mortality Following Metastatic Spinal Surgery: The New England Spinal Metastasis Score (NESMS). Spine J. 2016;16:482-90. 17. Paulino Pereira NR, Janssen SJ, et al. Development of a Prognostic Survival Algorithm for Patients With Metastatic Spine Disease. J Bone Joint Surg Am. 2016;98:1767-76. 18. Paulino Pereira NR, McLaughlin L, Janssen SJ, van Dijk CN, Bramer JAM, Laufer I, et al. The SORG Nomogram Accurately Predicts 3- and 12-Month Survival for Operable Spine Metastatic Disease: External Validation. J Surg Oncol. 2017;115(8):1019-27. 19. Karhade AV, Thio QCBS, Ogink PT, Shah AA, Bono CM, Oh KS, et al. Development of Machine Learning Algorithms for Prediction of 30-Day Mortality After Surgery for Spinal Metastasis. Neurosurgery. July 1, 2019;85(1):E83-91. 20. The SORG Machine Learning Algorithm for Predicting 30-Day Mortality of Patients Undergoing Surgery for Spinal Metastatic Disease [Internet]. [cited September 3, 2023]. Available at: https://sorg-apps.shinyapps.io/spinemets/ 21. Dardic M, Wibmer C, Berghold A, Stadlmueller L, Froehlich EV, Leithner A. Evaluation of Prognostic Scoring Systems for Spinal Metastases in 196 Patients Treated During 2005-2010. Eur Spine J. October 2015;24(10):2133-41. 22. Ahmed AK, Goodwin CR, Heravi A, Kim R, Abu-Bonsrah N, Sankey E, et al. Predicting Survival for Metastatic Spine Disease: A Comparison of Nine Scoring Systems. The Spine Journal. October 1, 2018;18(10):1804-14. 23. Truong VT, Al-Shakfa F, Roberge D, Masucci GL, Tran TPY, Dib R, et al. Assessing the Performance of Prognostic Scores in Patients With Spinal Metastases from Lung Cancer Undergoing Non-surgical Treatment. Asian Spine J. August 2023;17(4):739-49. 24. Karhade AV, Ahmed AK, Pennington Z, Chara A, Schilling A, Thio QCBS, et al. External Validation of the SORG 90-Day and 1-Year Machine Learning Algorithms for Survival in Spinal Metastatic Disease. Spine J. January 2020;20(1):14-21. 25. Bongers MER, Karhade AV, Villavieja J, Groot OQ, Bilsky MH, Laufer I, et al. Does the SORG Algorithm Generalize to a Contemporary Cohort of Patients With Spinal Metastases on External Validation? Spine J. October 2020;20(10):1646-52. 26. Shah AA, Karhade AV, Park HY, Sheppard WL, Macyszyn LJ, Everson RG, et al. Updated External Validation of the SORG Machine Learning Algorithms for Prediction of Ninety-Day and One-Year Mortality After Surgery for Spinal Metastasis. Spine J. October 2021;21(10):1679-86. 27. Hibberd CS, Quan GMY. Accuracy of Preoperative Scoring Systems for the Prognostication and Treatment of Patients With Spinal Metastases. Int Sch Res Notices. August 15, 2017;2017:1320684. 28. Hernandez-Fernandez A, Vélez R, Lersundi-Artamendi A, Pellisé F. External Validity of the Tokuhashi Score in Patients With Vertebral Metastasis. J Cancer Res Clin Oncol. September 2012;138(9):1493-500. 29. Yang JJ, Chen CW, Fourman MS, Bongers MER, Karhade AV, Groot OQ, Lin WH, Yen HK, Huang PH, Yang SH, Schwab JH, Hu MH. International External Validation of the SORG Machine Learning Algorithms for Predicting 90-Day and One-Year Survival of Patients With Spine Metastases Using a Taiwanese Cohort. Spine J. October 2021;21(10):1670-78. 30. Tarabay B, Gennari A, Truong VT, Shen J, Dib R, Newmann N, Al-Shakfa F, Yuh SJ, Shedid D, Boubez G, Wang Z. Which Scoring System Is the Most Accurate for Assessing Survival Prognosis in Patients Undergoing Surgery for Spinal Metastases from Lung Cancer? A Single-Center Experience. World Neurosurg. December 2022;168:e408-e417. 31. Wick JB, Kalistratova VS, Jr DP, Fine JR, Boozé ZL, Holland J, Vander Voort W, Hisatomi LA, Villegas A, Conry K, Ortega B, Javidan Y, Roberto RF, Klineberg EO, Le HV. A Comparison of Prognostic Models to Facilitate Surgical Decision-Making for Patients With Spinal Metastatic Disease. Spine (Phila Pa 1976). April 15, 2023;48(8):567-76. 32. Su CC, Lin YP, Yen HK, Pan YT, Zijlstra H, Verlaan JJ, Schwab JH, Lai CY, Hu MH, Yang SH, Groot OQ. A Machine Learning Algorithm for Predicting 6-Week Survival in Spinal Metastasis: An External Validation Study Using 2,768 Taiwanese Patients. J Am Acad Orthop Surg. September 1, 2023;31(17):e645-e656. 33. Karhade AV, Fenn B, Groot OQ, Shah AA, Yen HK, Bilsky MH, Hu MH, Laufer I, Park DY, Sciubba DM, Steyerberg EW, Tobert DG, Bono CM, Harris MB, Schwab JH. Development and External Validation of Predictive Algorithms for Six-Week Mortality in Spinal Metastasis Using 4,304 Patients From Five Institutions. Spine J. December 2022;22(12):2033-41. 34. Li Z, Guo L, Guo B, Zhang P, Wang J, Wang X, Yao W. Evaluation of Different Scoring Systems for Spinal Metastases Based on a Chinese Cohort. Cancer Med. February 2023;12(4):4125-36. 35. Karhade AV, Thio QCBS, Ogink PT, Bono CM, Ferrone ML, Oh KS, Saylor PJ, Schoenfeld AJ, Shin JH, Harris MB, Schwab JH. Predicting 90-Day and 1-Year Mortality in Spinal Metastatic Disease: Development and Internal Validation. Neurosurgery. October 1, 2019;85(4):E671-E681. 36. Zegarek G, Tessitore E, Chaboudez E, Nouri A, Schaller K, Gondar R. SORG Algorithm to Predict 3- and 12-Month Survival in Metastatic Spinal Disease: A Cross-Sectional Population-Based Retrospective Study. Acta Neurochir (Wien). October 2022;164(10):2627-35. 37. Chanplakorn P, Budsayavilaimas C, Jaipanya P, Kraiwattanapong C, Keorochana G, Leelapattana P, Lertudomphonwanit T. Validation of Traditional Prognosis Scoring Systems and Skeletal Oncology Research Group Nomogram for Predicting Survival of Spinal Metastasis Patients Undergoing Surgery. Clin Orthop Surg. December 2022;14(4):548-56. 38. Denisov AA, Zaborovsky NS, Ptashnikov DA, Mikhailov DA, Masevnin SV, Smekalenkov OA. Comparison of Prognostic Scales for Patients With Metastatic Spine Disease. Orthop Rev (Pavia). January 28, 2021;12(4):8822. Supplementary Material File (metas figures.docx) Download 1.02 MB File (metas tables.docx) Download 26.89 KB Information & Authors Information Version history V1 Version 1 06 February 2025 Peer review timeline Published Cancer Medicine Version of Record 3 Feb 2026 Published Copyright This work is licensed under a Non Exclusive No Reuse License. Authors Affiliations Julián Cabria Fernández 0000-0002-9140-5710 [email protected] Hospital de Cabuenes View all articles by this author Pablo González-Herráez Fernández Hospital de Cabuenes View all articles by this author Javier Mateo Negreira Hospital de Cabuenes View all articles by this author Pedro Arcos González Universidad de Oviedo Facultad de Medicina y Ciencias de la Salud View all articles by this author Metrics & Citations Metrics Article Usage 276 views 163 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Julián Cabria Fernández, Pablo González-Herráez Fernández, Javier Mateo Negreira, et al. Are the SORG and OPTImodel, Tokuhashi and Tomita algorithms still suitable as predictors of survival in patients with vertebral metastases in routine clinical practice?. Authorea . 06 February 2025. DOI: https://doi.org/10.22541/au.173883524.47342656/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.173883524.47342656/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ff4b9530b4d41e2',t:'MTc3OTM3ODQ5MQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();