1508P - Retrospective analysis of immunotherapy prognostic scores in advanced NSCLC at Nottingham University Hospitals (UK)

Abstract

BackgroundThe Royal Marsden Hospital (RMH) and Gustave Roussy Institute (GRIm) have prospectively validated prognostic scores to aide patient selection for phase I immunotherapy trials in many tumour sites, including NSCLC. These include number of metastatic sites, LDH, Albumin and Neutrophil-to-Lymphocyte Ratio (NLR). GRIm score stratified 155 patients to high-risk and low-risk categories, according to LDH, Albumin and NLR, with median OS of 4 months in high-risk cohort and 17 months in low-risk group. NSCLC was the predominant tumour site in this analysis including 30 stage IV patients. This score had similar design to RMH score, except for number of metastatic sites instead of NLR. Methods165 consecutive Nottinghamshire patients with stage IIIB/IV NSCLC on PD1/PDL1 immune-checkpoint inhibitors from December 2015 to December 2018 have been retrospectively analysed. At data cut-off in April 2019, 148 - 151 patients were censored for median TTF, PFS and OS, according to GRIm - RMH scores respectively: 17 patients were unclassified into GRIm (10.3%) and 14, into RMH (8.5%) for LDH unavailability; 70 patients remain alive (42.4%), 31 of whom still receiving immunotherapy (18.8%). ResultsTable1508PTablemTTF (cycles)mPFS (months)mOS (months)GRIm score0-1 80 patients (95% CI) Pembrolizumab-61 p Atezolizumab-13 p Nivolumab-3 p9 (9-13) 9 (7.5-11) 7 (4.6-10.5) 67 (6.9-10) 7 (7.5-11) 6 (3.6-7.3) 310 (10-13.5) 11 (11-15) 7.5 (5.2-9.1) 62-3 68 patients (95% CI) Pembrolizumab-59 p Atezolizumab-6 p Nivolumab-3 p3 (3.8-7) 3 (4-7.5) 3.5 (0.3-8.1) 12 (3-5.7) 2 (3-6.1) 3 (0.04 - 5.4) 14 (5.1-8.6) 5 (5-9.1) 6 (1.3-10) 1RMH score0-1 93 patients (95% CI) Pembrolizumab-78 p Atezolizumab-12 p Nivolumab-3 p9 (8.6-12.4) 8 (8.8-13.2) 11 (4.7-12.1) 67 (6.9-9.9) 6.5 (7.2-10.7) 7 (3.3-8) 310 (10-13.5) 11 (10.6-14.6) 8 (5.5-9.5) 62-3 58 patients (95% CI) Pembrolizumab-46 p Atezolizumab-9 p Nivolumab-3 p2.5 (3.4-7) 3 (3.7-8.1) 2 (1.35-4.2) 12 (2.6-5.4) 2.5 (2.8-6.2) 3 (0.9-3.4) 13 (4.5-7.9) 4 (4.8-9) 3 (1.7-7.3) 1 ConclusionsGRIm / RMH scores on our population have consistently shown predictive and prognostic relevance, with more favourable TTF, PFS and OS in low-risk groups across any immunotherapy. On treatment stratification, Atezolizumab outcomes show better correlation with RMH score and Nivolumab remains underrepresented at our centre. Our results are consistent with international multi-institutional studies at Gustave Roussy (France) on Lung Immune Prognostic Index (LIPI) and Princess Alexandra Hospital (Australia) on modified LIPI. Disease control has been observed when immune-related adverse events have occurred, suggesting baseline autoimmune antibodies may be predictive of response and immune-related toxicity, a currently open line of observational prospective research. Legal entity responsible for the studyNottingham University Hospitals NHS Trust at UK. FundingHas not received any funding. DisclosureAll authors have declared no conflicts of interest.