O10.09 * REDUCTION IN NEUTROPHIL-LYMPHOCYTE RATIO DURING INITIAL CONCURRENT CHEMORADIOTHERAPY IS PROGNOSTIC FOR SURVIVAL OF GLIOBLASTOMA PATIENTS

Abstract

INTRODUCTION: Elevated baseline neutrophil-lymphocyte ratio (NLR) has been associated with poorer outcome in multiple tumour sites. In the context of glioblastoma (GBM), elevated NLR prior to any initial therapy, including surgery or corticosteroids, or prior to second surgery for recurrent disease predicted worse outcome. The aim of this retrospective cohort study was to assess if change in NLR during initial focal radiotherapy (RT) and concurrent temozolomide (TMZ) predicted outcome in newly diagnosed GBM patients. METHODS: A retrospective chart review was performed for patients treated concurrently with TMZ and RT, at our tertiary referral centre between Jan 2004 and Sept 2010, for histologically confirmed GBM. Age, baseline ECOG performance status (ECOG-PS) and extent of surgery were recorded. Baseline NLR, change in NLR, and time-weighted mean dexamethasone dose (TWMdex) were collected between 2 weeks prior to commencing RT and 10 weeks following the start of RT. Overall survival (OS) was calculated using Kaplan Meier method. Univariate (UVA) and multivariable (MVA) analyses using cox proportional hazard models evaluated variables associated with OS, accounting for colinearity of variables. RESULTS: There were 394 patients who met eligibility and were included in the final analysis. Median age was 54 (range 18-73). With median follow up of 15.2 months (range 1.6-134.6), OS was 66.4% (CI 61.4-70.7) and 31.1% (CI 26.5-35.8) at 1 and 2 years respectively. Median baseline dexamethasone dose was 8 mg (0-30 mg) and median TWMdex dose was 4.94 mg (0.0 -27.4 mg). On UVA, older age (HR 1.034, p 2 (HR 1.076, p < 0.001) were associated with shorter survival whereas baseline NLR below the median of 7.5 (HR 0.628, p < 0.0001) and decrease in NLR during treatment (HR 0.644, p < 0.0001) were associated with better survival. On MVA, decrease in NLR remained significantly prognostic for OS (HR 0.763, p = 0.02; CI 0.603 - 0.964) in addition to age (HR 1.032, p < 0.001; CI 1.018-1.045), TWMdex (HR 1.039, p = 0.002; CI 1.014-1.065) and ECOG-PS (HR 0.593, p = 0.008; CI 0.430-0.873). Although, baseline NLR was no longer statistically significant (HR 0.814, p = 0.09; CI 0.641 - 1.032), patients with low baseline NLR and further reduction in NLR during treatment had significantly longer survival (p < 0.0001). The association between change in NLR and change in dexamethasone dose over time was weak (r = 0.22). CONCLUSION: In patients with newly diagnosed glioblastoma, any reduction in neutrophil-lymphocyte ratio during initial concurrent chemoradiotherapy was an independent prognostic factor for longer survival, accounting for other prognostic factors including dexamethasone dose. Patients with low baseline NLR and further reduction in NLR during treatment had the best prognosis.