1356P - Increase in S100B and LDH as early outcome predictors for non-responsiveness to anti-PD-1 monotherapy in advanced melanoma

Abstract

BackgroundImmune-checkpoint inhibitors have revolutionized advanced melanoma care. Despite major improvements in survival, many patients do not derive long-term benefit and treatment could cause severe side-effects. Since several new (combination) therapies are or will become available, early prediction of non-responsiveness (NR) to anti-PD-1 monotherapy becomes relevant in order to enable early switch to next line (combination) treatment and avoid nonsensical care. MethodsAdvanced melanoma patients treated with pembrolizumab (PEM) who started PEM between June 2014 and August 2016 were included in this retrospective analysis. S100 and lactate dehydrogenase (LDH) levels were routinely determined prior to initiation of PEM and every 3-weeks during treatment. NR to treatment was defined as progressive disease or death at 6 months after start of PEM. Biomarker response characteristic (BReC) plots were obtained and LDH and S100 response cut-offs were determined based on two criteria: specificity for NR>95% and feasibility to use in clinical practice. Next, sensitivity, specificity and predictive values were generated per follow-up time point (week 3, 6, 9, 12 and 15). Results166 advanced melanoma patients were included. The BReC analyses showed clear relations between an early (week 3 to 15) increase in tumor biomarker and NR. An increase of>50% in LDH or a>100% increase in S100 above the upper limit of normal compared to baseline at any follow-up interval was determined as criterion to positively test for NR. Obtained specificity ranged from 0.97-0.98 and the positive predictive value ranged from 0.92-0.96 for the studied follow-up intervals. Obtained sensitivity for detecting non-responsiveness ranged from 0.25 (95% CI; 0.16-0.35) at 3 weeks of follow-up to 0.35 (95% CI; 0.24-0.47) at 12 weeks of follow-up. ConclusionsAn early increase in S100 and/or LDH are strong parameters for non-responsiveness to PD-1 blockade in advanced melanoma. Prospective confirmation is needed before clinical implementation. Legal entity responsible for the studyNetherlands Cancer Institute. FundingNetherlands Cancer Institute. DisclosureE.A. Rozeman: Travel / Accommodation / Expenses: NanoString; Travel / Accommodation / Expenses: MSD. J.V. van Thienen: Advisory / Consultancy: Pfizer; Advisory / Consultancy: Novartis. J.B.A.G. Haanen: Advisory / Consultancy, Research grant / Funding (institution): BMS; Advisory / Consultancy, Research grant / Funding (institution): MSD; Advisory / Consultancy: Pfizer; Advisory / Consultancy: AZ/MedImmune; Advisory / Consultancy: Roche/Genentech; Advisory / Consultancy: Ipsen; Advisory / Consultancy: Bayer; Advisory / Consultancy: Immunocore; Advisory / Consultancy, Research grant / Funding (institution): Novartis; Advisory / Consultancy, Research grant / Funding (institution): Neon Therapeutics; Advisory / Consultancy: Celsius; Advisory / Consultancy: Seattle Genetics; Advisory / Consultancy: Gadet. C.U. Blank: Advisory / Consultancy, Research grant / Funding (institution): BMS; Advisory / Consultancy: MSD; Advisory / Consultancy, Research grant / Funding (institution): Novartis; Advisory / Consultancy: Pfizer; Advisory / Consultancy: Lilly ; Advisory / Consultancy: GSK; Advisory / Consultancy: GenMab; Research grant / Funding (institution): NanoString; Advisory / Consultancy: Roche; Advisory / Consultancy: Pierre Fabre. All other authors have declared no conflicts of interest.