Moderate precision of prognostic scoring systems in a consecutive, prospective cohort of 544 patients with metastatic spinal cord compression.

Abstract

Improved survival among cancer patients and diverse conclusions from recent studies make it relevant to reassess the performance of the Tokuhashi Revised score and the Tomita score. The aim of this study was to validate and compare these two scoring systems in a recent and unselected cohort of patients with metastatic spinal cord compression (MSCC). In 2011, we conducted a prospective cohort study of 544 patients who were consecutively admitted with MSCC to one treatment facility. Patients estimated survival were assessed with the Tokuhashi Revised score and the Tomita score and compared to the observed survival. We assessed how precise the scoring systems predicted survival with McNemar's test. The prognostic value was illustrated with Kaplan-Meier curves, and the individual prognostic components were analyzed with Cox regression analysis. The mean age was 65years (range 20-95), and 57% of the patients were men. The majority of tumors were lung (23%), prostate (21%), and breast tumors (18%). The overall precision of predicted survival was 58.7% for the Tokuhashi Revised score and 52.9% for the Tomita score. The observed survival in each of the scoring groups categorized by the scoring systems was statistically significantly different (p<0.0001). The Tokuhashi Revised score and the Tomita score are useful in categorizing patients into prognostic groups, and the individual components have important prognostic values. The Tokuhashi Revised score was most precise in predicting survival. However, due to the relatively low precision, we suggest that a modification of both scoring systems is necessary.