Neutrophil-lymphocyte ratio (NLR) is a marker of systemic inflammatory response and its elevation has recently been shown to be a poor prognostic factor in many malignancies including colon, prostate and bladder cancer. The primary aim of this study was to assess the prognostic impact of NLR in a clinically annotated cohort of patients with glioblastoma multiforme (GBM). We hypothesised that elevated NLR would be associated with worse prognosis. Between 2004 and 2009, 137 patients had surgery for GBM and were assessed for consideration of adjuvant therapy at our institution. Of these, 84 patients with an evaluable pre-corticosteroid full blood count result were identified and included in the final analysis. Median overall survival was 9.3 months (range 0.7-82.1). On univariate analysis, age >65 years, gender, ECOG performance status ≥2, frontal tumour, extent of surgical resection, completion of adjuvant chemoradiation protocol and NLR > 4 were significantly correlated with overall survival. Patients with NLR > 4, had a worse median overall survival at 7.5 months versus 11.2 months in patients with NLR ≤ 4 (hazard ratio 1.6, 95 % CI 1.00-2.52, p = 0.048). On multivariate analysis NLR > 4 remained an independent prognostic indicator for poor outcome. These data are an important reminder of the potential relevance of host immunity in GBM. In our cohort, NLR > 4 conferred a worse prognosis independent of other well established prognostic factors. If validated in other cohorts NLR may prove to be a useful addition in predicting prognosis in GBM patients. The demonstration that host immunity plays a role in GBM biology suggests that investigation of emerging therapies which modulate host immune response are warranted in this disease.

2062 Background: GBM is the most common and aggressive primary brain tumor. The neutrophil to lymphocyte ratio (NLR) gives a measure of systemic inflammatory response and lymphopenia, both of which are poor prognostic factors in many malignancies. No published study has assessed the prognostic impact of NLR in GBM. Methods: Patients treated for GBM at our regional referral centre with assessable complete blood count at first presentation (prior to corticosteroid therapy or surgery) were identified. Medical notes were reviewed to extract demographic and treatment data. Survival curves were estimated via Kaplan-Meier method and compared via log-rank method. Multivariate analysis was performed via Cox proportional hazards regression modelling. Results: A total of 86 patients were identified, of which 65(76%) were male. Median age at diagnosis was 58 years (range: 18–76). At the time of analysis all patients still alive had ≥ 2 years follow-up. Median overall survival (OS) was 9.3 months (range: 1-82). 57% completed the standard adjuvant Stupp protocol and 43% discontinued early due to disease progression or treatment toxicity. Median OS was 11.2 months in patients with NLR<4 and 7.5 months in patients with NLR>4 (HR 0.59, p=.04). Other significant prognostic factors based on univariate analysis were consistent with published data (Table). After correcting for known prognostic factors NLR remained a significant predictor of survival (Table). Conclusions: Recent advances in immunotherapy have highlighted the importance of the immune system in the treatment and prognosis of cancer patients. Many GBM patients are on corticosteroids for a significant proportion of their disease course which may abrogate the effects of host immunity on outcome. Nevertheless, we have shown that NLR at diagnosis is an independent predictor of survival in GBM patients. Investigation of therapies which harness the immune response are warranted in this disease. [Table: see text]