475 Exposure to antibiotic interventions as a potential risk factor for developing immune-related cutaneous adverse events in patients treated with immune checkpoint blockade for cancer

<h3>Background</h3> Immune checkpoint blockade (ICB) treatment for cancer is complicated due to the occurrence of immune-related cutaneous adverse events (ircAE) in up to 50% of cases. However, the underlying mechanisms for ircAE remain unclear. Here, we investigate the role of skin microbiome perturbation through antimicrobial interventions as a potential risk factor for ircAE occurrence. <h3>Methods</h3> This is a retrospective analysis of 11,499 patients with solid tumors who were treated with ICB at a National Cancer Institute designated cancer center (Memorial Sloan Kettering Cancer Center) at any time during a 5-year period . Only the first ICB regimen was considered per patient. Regimens with a duration of one day were excluded from analysis (N=1562; 13%). Any antimicrobial intervention was recorded; the primary endpoint of ircAE was assessed using international classification of diseases (ICD) codes and confirmed by an oncodermatologist based on chart review. The cause-specific hazard ratio (HR) for developing an ircAE while on immunotherapy following exposure to any antibiotic intervention was evaluated using a Cox proportional hazards model adjusted for sex, race, ethnicity, and age at start of ICB. Antibiotic intervention was considered for the duration of the intervention and the following two months, and was assessed as a time-varying covariate. A secondary analysis modeled antibiotic interventions classified by microbiome perturbation potential (low, medium/high). <h3>Results</h3> Overall, the incidence of documented and confirmed ircAE was 7.5% (N=741 ircAE to 9,937 patients), with 150 ircAE occurring within 2 months of antibiotic exposure and 591 without antibiotic exposure. The cause-specific HR for developing an ircAE was 1.64 (95% CI 1.36 – 1.96, p &lt;0.001) for patients who received antibiotics compared to those who did not. The increased HR was maintained for antibiotics classified as both low (1.49, 95% CI 1.13 – 1.96) and medium/high (1.74, 95% CI 1.40 – 2.18) for their microbiome perturbation potential. <h3>Conclusions</h3> Exposure to antibiotics in patients on ICB was associated with a significantly increased risk of developing an ircAE, regardless of whether antibiotics had low or medium/high microbiome perturbation potential. This finding suggests a role of the skin microbiome in ircAE and provides additional evidence that antibiotics should be prescribed cautiously in cancer patients on ICB. Of note, this study does not disentangle the effect of the antibiotic from the reason it was prescribed. Thus, further studies are warranted to delineate the causal mechanisms underlying antibiotic exposure, microbiome perturbation, and ircAE development. <h3>Ethics Approval</h3> This study was approved by MSK IRB 16-458.