Metastatic melanoma patient outcomes since introduction of immune checkpoint inhibitors in England between 2014 and 2018.

Abstract

e19310 Background: Population-based research provides real world information about prescribing trends, as well as benefits and harms associated with new interventions first evaluated in randomized clinical trials (RCTs). The national health service in England collects data on systemic anti-cancer therapy (SACT) prescribed to patients. We retrospectively examined outcomes including overall survival (OS) and treatment toxicity for patients (pts) prescribed checkpoint inhibitors (CPIs) approved for 1st line use in metastatic melanoma. Methods: National Cancer Registration and Analysis Service (NCRAS) datasets were used to identify metastatic melanoma pts prescribed 1st line CPIs and calculate survival outcomes. The Hospital Episode Statistics database was used to identify emergency department attendances and emergency hospital admissions. Results: Between April 2014 and March 2018, 5,465 melanoma pts were prescribed SACT. Of these, 2,322 pts received 1st line CPIs for metastatic disease at 56 different hospitals in England: 1,174 were prescribed pembrolizumab (pembro), 724 ipilimumab (ipi), 52 nivolumab (nivo), 372 ipinivo. Prescribing reflected national access, starting with ipi in July 2014, replaced by pembro in Nov 2015 and ipinivo prescriptions started in July 2016. By 2018, 1 in 3 pts were receiving ipinivo as treatment of choice. There was high concordance for 3yr OS with RCT outcomes for ipi (32%) and ipinivo (58%), but pembro outcomes were poorer (40%). Pts prescribed pembro were significantly older (51% vs. 12% >70 years) and frailer (38% vs. 61% ECOG PS 0) than pts prescribed ipinivo (p<0.0001). 30-day rates of emergency hospital attendances/admissions from the earliest SACT and last SACT treatment dates were significantly higher for ipinivo (16%/37% and 19%/55%) compared with pembro (8%/17% and 14%/29%). Colitis was the most common IrAE associated with hospitalization: 25% with ipinivo and 4% with pembro. 30-day mortality from last SACT was 3.8% with ipinivo and 9.1% with pembro (p<0.0001). Conclusions: Linked real world datasets demonstrated OS outcomes remarkably similar to RCTs associated with ipi and ipinivo, while poorer outcomes associated with pembro may reflect patient selection bias. Ipinivo generated the highest rates of emergency hospital visits and admissions, but 30-day mortality was not compromised, potentially reflecting efficient and effective management of complications. Hospital-level outcomes and 30-day mortality causes are now being assessed to explore variations in care.