Application of morphometric analysis to patients with lung cancer metastasis to the spine: a clinical study

Abstract

OBJECTIVE Predicting the survival rate for patients with cancer is currently performed using the TNM Classification of Malignant Tumors (TNM). Identifying accurate prognostic markers of survival would allow better treatment stratification between more aggressive treatment strategies or palliation. This is especially relevant for patients with spinal metastases, who all have identical TNM staging and whose surgical decision-making is potentially complex. Analytical morphometrics quantifies patient frailty by measuring lean muscle mass and can predict risk for postoperative morbidity after lumbar spine surgery. This study evaluates whether morphometrics can be predictive of survival in patients with spinal metastases. METHODS Utilizing a retrospective registry of patients with spinal metastases who had undergone stereotactic body radiation therapy, the authors identified patients with primary lung cancer. Morphometric measurements were taken of the psoas muscle using CT of the lumbar spine. Additional morphometrics were taken of the L-4 vertebral body. Patients were stratified into tertiles based on psoas muscle area. The primary outcome measure was overall survival, which was measured from the date of the patient's CT scan to date of death. RESULTS A total of 168 patients were identified, with 54% male and 54% having multiple-level metastases. The median survival for all patients was 185.5 days (95% confidence interval [CI] 146–228 days). Survival was not associated with age, sex, or the number of levels of metastasis. Patients in the smallest tertile for the left psoas area had significantly shorter survival compared with a combination of the other two tertiles: 139 days versus 222 days, respectively, hazard ratio (HR) 1.47, 95% CI 1.06–2.04, p = 0.007. Total psoas tertiles were not predictive of mortality, but patients whose total psoas size was below the median size had significantly shorter survival compared with those greater than the median size: 146 days versus 253.5 days, respectively, HR 1.43, 95% CI 1.05–1.94, p = 0.025. To try to differentiate small body habitus from frailty, the ratio of psoas muscle area to vertebral body area was calculated. Total psoas size became predictive of mortality when normalized to vertebral body ratio, with patients in the lowest tertile having significantly shorter survival (p = 0.017). Left psoas to vertebral body ratio was also predictive of mortality in patients within the lowest tertile (p = 0.021). Right psoas size was not predictive of mortality in any calculations. CONCLUSIONS In patients with lung cancer metastases to the spine, morphometric analysis of psoas muscle and vertebral body size can be used to identify patients who are at risk for shorter survival. This information should be used to select patients who are appropriate candidates for surgery and for the tailoring of oncological treatment regimens.