Impact of corticosteroid (CS) exposure on acquired resistance to immune checkpoint inhibitor (ICI) therapy in patients (pts) with metastatic renal cell carcinoma (mRCC): A single centre experience.

Abstract

404 Background: ICI therapy with Ipilimumab and Nivolumab (I/N) is a first line standard of care option in pts with intermediate – poor risk mRCC, with the longest durable response in this setting. Nonetheless, a high percentage of pts eventually progress on ICI treatment. Currently contributing factors for acquired ICI resistance remain unclear. The impact of CS use on acquired ICI resistance remains controversial and data in mRCC pts is particularly lacking. Methods: We conducted a retrospective analysis of consecutive mRCC pts with primary non-progression to first line I/N. We analysed the use of concurrent CS for all indications including immune related adverse events (IRAE) and compared baseline characteristics of pts who received CS (Cohort A) vs those who did not (Cohort B). Association between CS use, progression and mortality was analysed using multivariable Cox regression model. Adjustment for casemix included IMDC risk score, best response, presence of sarcomatoid/rhabdoid features and brain/liver metastases. Progression free (PFS) and overall survival (OS) were estimated using Kaplan-Meier analysis. Results: 64 pts treated with 1L I/N at the Royal Marsden Hospital with primary non-progression between 2016 - 2022 were included. The commonest histological subtype was clear cell (87.5%) with 30% having sarcomatoid/rhabdoid features. All pts had intermediate - poor IMDC risk, with 34% having poor IMDC risk. 47 pts received CS for any indication (Cohort A) and of these 44 pts for IRAE. 17 pts received no CS (Cohort B). Baseline characteristics including IMDC risk, grade and presence of liver/brain metastases were not statistically different between cohorts. 57 pts were eligible for survival analysis with 33 progression events. Median follow up was 22.5 months overall. At multivariable analysis no factors were significantly associated with PFS or OS. Median PFS was 16 months (95% CI 10, 33) in cohort A vs 25 months (95% CI 8, not reached [NR]) in cohort B. PFS was 39% (95% CI 23%, 55%) in Cohort A vs 50% (95% CI 19%, 75%) in Cohort B at 2 years, and 30% (95% CI 15%, 47%) vs 38% (95% CI 10%, 66%), at 3 years respectively. Median OS was 55 months (95% CI 21, NR) in cohort A and NR in cohort B. OS was 60% (95% CI 41%, 74%) in cohort A vs 65% (95% CI 23%, 88%) in cohort B at 2 years and 55% (95% CI 36%, 71%) vs 65% (95% CI 23%, 88%) at 3 years. Cumulative doses and time of CS initiation will be included in the final analysis to determine their impact on above outcomes. Conclusions: Although no statistically significant association was seen between CS use and PFS or OS, our results suggest a trend towards shorter time to progression and survival with CS use in pts on 1L I/N for mRCC. While further analyses regarding the dosage and timing of CS are underway, given this is a single centre retrospective analysis, these results would need to be confirmed in larger studies.