New Predictive Index for Survival in Symptomatic Spinal Metastases

Objective: Patients with spinal metastases have broad variability in morbidity, mortality, and survival. Existing prognostic scoring systems have limited predictive value. Our aim is, given recent advances in surgical and medical care for patients with cancer and spinal metastases, to develop a new survival index with superior prognostic value. Methods: We completed a retrospective analysis on 77 patients who received surgery for metastatic tumors to the spine, of patient factors like pathologic subtype, age, neurologic examination, type of surgical procedure, Hauser Ambulation Index, and a novel scoring system for degree of tumor burden in several organ systems, among others. A survival index will be derived from the patient factors that, when measured preintervention, best predicted survival post intervention. Results: Although primary organ or pathologic type was not predictive of survival for patients with metastatic disease in this population, the degree of lung tumor burden (LTB) and preoperative Hauser Ambulation Index were predictive of survival. After a multivariable analysis of >20 different patient factors, the Jenkins Survival Index (JSI, a 0–21 scale) was constructed using a machine-learning system as the sum of the HAI (0–9 scale) and LTB score (0–3 scale) multiplied by 4 (JSI = HAI + 4 · LTB, Rho = −0.588, P < 0.0001). The JSI had a positive predictive value of 92% compared with 54.1% and 56.9% for Tokuhashi and Tomita scales, respectively. Conclusions: The JSI predicts in a meaningful way survival outcomes for patients symptomatic from spinal metastases, which will be of value to oncologists and other clinicians treating patients with metastatic disease.