LAUSR

9532 Background: Both pembrolizumab (PEMBRO) and combination ipilimumab + nivolumab (IPI+NIVO) are FDA-approved immunotherapies for advanced melanoma (AM). These two treatment regimens have different toxicity profiles which may impact health care resource utilization (HCRU). Our aim was to compare real-world risk of hospitalization and emergency department (ED) visits within 12 months of starting the two treatment regimens. Methods: A retrospective cohort study was conducted in patients ≥18 years old with AM initiating PEMBRO or IPI+NIVO between Jan 1, 2016 – Dec 30, 2017. Patients were identified from 12 US academic medical centers and affiliated satellite clinics. Data were abstracted through chart review. All-cause hospitalizations or ED visits and the rates per patient per month (PPPM) through 12 months of follow-up were calculated. Utilization was compared between PEMRBO and IPI+NIVO using multivariate logistic regression analysis. Results: 400 patients were included, 200 each PEMBRO and IPI+NIVO with mean (SD) follow-up time of 10 (3) and 10 (4) months, respectively. The PEMBRO cohort had poorer Eastern Cooperative Group (ECOG) performance status at treatment start, 71% ECOG 0 or 1 vs 88% (p < .001); more diabetes, 21% vs 13% (p = .045); a trend towards more heart disease, 18% vs 12% (p = .067); were more likely to be PD-L1 expression positive, 77% vs 63% (p = .011); and less likely to harbor a BRAF mutation, 35% vs 50% (p = .003). The proportion with at least one hospitalization through 12 months was 17% PEMBRO vs 24% IPI+NIVO. Less than 2% of patients had more than one admission and none had more than two, regardless of cohort. Unadjusted mean (SD) PPPM hospitalizations were .016 (.037) for PEMBRO and .020 (.038) for IPI+NIVO. The adjusted odds ratio for any hospitalization with PEMBRO was 0.55 (95% CI .31, .97; p = .039) vs. IPI+NIVO. ED visits occurred in 18% vs 21% in PEMBRO and IPI+NIVO respectively, with no difference in covariate-adjusted analysis (p = .147). Conclusions: Patients receiving PEMBRO had a significantly lower probability of hospitalization and similar probability of ER visits compared with IPI+NIVO in the real world through 12 months.