493P - Improving patient selection for immuno-oncology phase I trials: An external validation of five prognostic scores at Claudius Regaud Institute of Toulouse, Oncopôle (IUCT-O)

Abstract

BackgroundWe aimed to compare the performance of 5 prognostic scores (RMH: Royal Marsden Hospital, MDACC: MD Anderson Clinical Center, MDA-ICI: MD Anderson Immune Checkpoint Inhibitors, GRIm: Gustave Roussy Immune Score and LIPI: Lung Immune Prognostic Index) in predicting overall survival (OS) in phase 1 patients treated with immune checkpoint inhibitors (ICI). MethodsWe reviewed records of patients with advanced solid tumors enrolled in phase 1 ICI trials between 2015 and 2018 at IUCT-O. We compared the performance of prognostic scores using Akaike criterion, discriminatory ability (Harrell’s C, the Royston’s D) and proportion of explained variation (R²) statistics. Primary endpoint was OS. ANC: Absolute Neutrophil Count ALC: Absolute Lymphocyte count (d)NLR: (Derived) Neutrophil / Lymphocyte ratio PS: Performance statusTable493PTableRMHMDACCMDA-ICIGRImLIPISites of metastases > 2✓✓LDH > ULN✓✓✓✓LDH > 466✓Albumin < 35G/L✓✓✓Gastrointestinal tumor✓PS≥1✓PS>1✓Age > 52 years✓Platelet count > 300✓ANC > 4.9✓ALC < 1.8✓liver metastases✓NLR > 6✓dNLR > 3✓AIC1310.71290.01296.41293.51296.9CH0.600.670.640.660.65Dadj0.670.940.810.980.84R² adj0.0960.1760.1360.1860.145 ResultsA total of 259 patients were included. Median age was 63 years (range 18-83). Main primary cancers were melanoma (18.5%), head and neck (16.2%), lung (12.7%) and bladder (9.7%). With a median follow up of 15 months (95% CI: [11.6;17.5]), median OS was 12.5 months (95%CI=[10.3;16.0]). All scores were associated with OS: Hazard Ratio (HR)=1.98 [1.41;2.78] for RMH score 2-3 vs 0-1, HR=1.68 [1.09;2.60] for MDA score 2 and 3.65 [2.42;5.51] for score 3-5 vs 0-1, HR=1.18 [0.77;1.81] for MDA-ICI score 3; HR=2.70 [1.74;4.17] for score 4 and HR=4.85 [2.62;8.98] for 5-6 vs 0-2, HR=2.70 [1.92;3.79] for GRIm score 2-3 vs 0-1 and finally 1.86 [1.25;2.78] for LIPI score 1 and HR=3.86[2.43;6.13] for score 2 vs 0. MDA and GRIm scores obtained more significant results for discrimination than RMH, MDA-ICI and LIPI (Table). ConclusionsThe utilization of theses scores could allow a better patients selection in early trials, especially during the critical periods of dose escalation and proof-of-concept expansion cohorts. Legal entity responsible for the studyThe authors. FundingHas not received any funding. DisclosureAll authors have declared no conflicts of interest.