Life Expectancy and Metastatic Spine Scoring Systems: An Academic Institutional Experience
A retrospective data collection study with application of metastatic spine scoring systems.Objectives:
To apply the Tomita and revised Tokuhashi scoring systems to a surgical cohort at a single academic institution and analyze spine-related surgical morbidity and mortality rates.Summary of Background Data:
Surgical management of metastatic spine patients requires tools that can accurately predict patient survival, as well as knowledge of morbidity and mortality rates.Methods:
An Oregon Health & Science University (OHSU) Spine Center surgical database was queried (years 2002–2010) to identify patients with an ICD-9 code indicative of metastatic spine disease. Patients whose only surgical treatment was vertebral augmentation were not included. Scatter plots of survival versus the Tomita and revised Tokuhashi metastatic spine scoring systems were statistically analyzed. Spine-related morbidity and mortality rates were calculated.Results:
Sixty-eight patients were identified: 45 patients’ (30 male patients, mean age 45 y) medical records included operative, morbidity, and mortality statistic data and 38 (26 male patients, mean age 54 y) contained complete metastatic spine scoring system data. Of the 38 deceased spine metastatic patients, 8 had renal cell, 7 lung, 4 breast, 2 chondrosarcoma, 2 prostate, 11 other, and 4 unknown primary cancers. Linear regression analysis revealed R2 values of 0.2570 and 0.2009 for the revised Tokuhashi and Tomita scoring systems, respectively. Overall transfusion, infection, morbidity, and mortality rates were 33% and 9%, and 42% and 9%, respectively.Conclusions:
Application of metastatic prognostic scoring systems to a retrospective surgical cohort revealed an overall poor correlation with the Tomita and revised Tokuhashi predictive survival models. Morbidity and mortality rates concur with those in the medical literature. This study underscores the difficulty in utilizing metastatic spine scoring systems to predict patient survival. We believe a scoring system based on cancer type is needed to account for changes in treatment paradigms with improved patient survival over time.