Association of antibiotic exposure with overall survival and colitis in patients with stage III and IV melanoma receiving immune checkpoint inhibitors.

Abstract

56 Background: Recent studies suggest that changes in the gut microbiome modulate response to cancer treatment, including immune checkpoint inhibitors (ICI). Broad-spectrum antibiotics (Abx) are known to cause significant dysbiosis. We hypothesize that recent Abx exposure worsens outcomes in patients (pts) with stage III/IV melanoma (MM) receiving ICI. Methods: We identified MM pts treated with ICI from our institutional database. All received their first ICI between 2004-2019. Antibiotic exposure was defined as receipt of Abx within 3 months prior to the first infusion of ICI. The primary outcome was overall survival (OS) and the secondary outcome was immune-mediated colitis requiring intravenous (IV) steroids. Stage III and IV pts were analyzed separately for the primary analysis. Results: Of 568 pts in our database, 20% received Abx within the 3 months prior to ICI. 36% of pts had stage III disease and 26% of pts were treated with either adjuvant or neoadjuvant ICI. 1.6% of pts died of causes other than MM. The Abx+ and Abx- groups were balanced in terms of stage, race, age, sex, BRAF status, LDH, prior systemic therapies, and class of ICI received. Only 4 pts were hospitalized due to the infection prompting the Abx, and no pts died due to the infection. In the Stage IV group, Abx+ pts had worse OS on MV analysis (HR 1.6, 95% CI 1.1-2.2). Stage III Abx+ also had worse OS (HR 2.8, 95% CI 1.3-5.9). In a sensitivity analysis excluding pts who received IV Abx or were admitted due to infection, survival was still worse for Abx+ pts (HR 1.7, 95% CI 1.2-2.4). In a Fine-Grey competing risk MV model, Abx+ had a higher rate of immune-mediated colitis requiring IV steroids (HR 2.1, 95% CI 1.02-4.5). Conclusions: In MM pts treated with ICI, receipt of Abx within 3 months prior to ICI initiation was associated with decreased OS and increased colitis. Future research should include prospective studies to better define the risk/benefit profile of antibiotics in close proximity to ICI. [Table: see text]