Which one is a valuable surrogate for predicting survival between Tomita and Tokuhashi scores in patients with spinal metastases? A meta-analysis for diagnostic test accuracy and individual participant data analysis

Abstract

This study is to estimate the diagnostic accuracy of Tokuhashi and Tomita scores that assures 6-month predicting survival regarded as a standard of surgical treatment. We searched PubMed, EMBASE, European PubMed central, and the Cochrane library for papers about the sensitivities and specificities of the Tokuhashi and/or Tomita scores to estimate predicting survival. Studies with cut-off values of ≥9 for Tokuhashi and ≤7 for Tomita scores based on prior studies were enrolled. Sensitivity, specificity, diagnostic odds ratio (DOR), area under the curve (AUC), and the best cut-off value were calculated via meta-analysis and individual participant data analysis. Finally, 22 studies were enrolled in the meta-analysis, and 1095 patients from 8 studies were included in the individual data analysis. In the meta-analysis, the pooled sensitivity/specificity/DOR for 6-month survival were 57.7 %/76.6 %/4.70 for the Tokuhashi score and 81.8 %/47.8 %/4.93 for Tomita score. The AUC of summary receiver operating characteristic plots was 0.748 for the Tokuhashi score and 0.714 for the Tomita score. Although Tokuhashi score was more accurate than Tomita score slightly, both showed low accuracy to predict 6 months residual survival. Moreover, the best cut-off values of Tokuhashi and Tomita scores were 8 and 6, not 9 and 7, for predicting 6-month survival, respectively. Estimation of 6-month predicting survival to decide surgery in patients with spinal metastasis is quite limited by using Tokuhashi and Tomita scores alone. Tokuhashi and Tomita scores could be incorporated as part of a multidisciplinary approach or perhaps interpreted in the context of a multidisciplinary approach.