• Lola B Chambless, Heather M Kistka, Scott L Parker, Laila Hassam-Malani, Matthew J McGirt, and Reid C Thompson.
• Vanderbilt University Medical Center, 1161 21st Avenue South, Rm T-4224 MCN, Nashville, TN, 37232-2380, USA.
• J. Neurooncol. 2015 Jan 1;121(2):359-64.

AbstractThe Karnofsky Performance Scale (KPS) score is a widespread metric to stratify patient prognosis and determine appropriate management in glioblastoma multiforme(GBM). Low preoperative KPS values have been associated with shorter overall survival (OS). However, surgical resection can have a dramatic effect on a patient's functional status which subsequently alters their KPS score. To determine the predictive value of preoperative verses postoperative KPS scores in terms of OS in patients with GBM. We conducted a retrospective review of 163 patients who underwent initial surgical intervention for pathologically proven GBM at our institution between 2003 and 2013. Pre and postoperative performance status, demographic, operative, and treatment variables were recorded for each patient. Multivariate regression analysis identified predictors of prolonged OS. The adequacy index was calculated to compare the predictive value of preoperative and postoperative KPS score. Median preoperative and postoperative KPS scores were 70 and 80, respectively. Overall, 92 (57 %) patients experienced an improvement in their KPS score, 40 (25 %) remained stable, and 29 (18 %) declined. Higher postoperative KPS (P = 0.0001), radiation therapy (P < 0.0001), younger age (P = 0.0443) and the absence of diabetes (P = 0.0006) were each independently associated with increased OS in a multivariate regression model. Postoperative KPS score has superior predictive value compared to pre-operative KPS score (A = 0.758 vs. 1.002). Postoperative KPS scores have superior predictive capabilities in terms of OS in GBM and should replace preoperative KPS scores when estimating prognosis in this population.

28 keywords...

hide…