144 Identifying a composite signature for predicting immune-related adverse events in advanced melanoma patients treated with immune checkpoint blockade

• DOI: 10.1136/jitc-2024-sitc2024.0144
• Publication Date: 2024-11-05T14:58:54Z
• Authors: Evgenii S Bolshakov; Tatiana Vasileva; Sofya A Kust; Alena Frank; Maria Savchenko; Iris Wang; Tori Conroy; Nicholas R Merriam; Kseniia L Markova; Aleksandra Brunov...; Ekaterina Ushakova; Siune Ambarian; Alina Mulyukina; Evgeniy Shilov; Nazar Arutyunyan; Anna Shchetsova; Polina Shulga; Danil Spirin; Anastasiya Terenteva; Michael F Goldberg; Aleigha Lawless; Genevieve M Boland; Ryan J Sullivan; Alexander Zaytcev
• Keywords: Immune checkpoint; Signature (topology)

<h3>Background</h3> While immune checkpoint blockade (ICB) produces long-lasting remission in some patients, they may cause serious related adverse events (irAEs) that can result morbidity and mortality. irAEs have been increasingly recognized to stem from autoimmune-like activation of the adaptive system. Here, we analyzed pre-treatment peripheral blood mononuclear cells (PBMCs) patients with advanced melanoma treated ICB uncover processes underlying irAE occurrence if onset be predicted. <h3>Methods</h3> Pre-treatment PBMCs were profiled using flow cytometry bulk RNA sequencing. We developed a cellular signature 17 cell populations selected open-source data. Principal component analysis was then used evaluate these samples cohort. Additionally, gene-based 55 reported irAE-associated genes ssGSEA for gene calculation. <h3>Results</h3> cohort (n=47) either anti-PD-1 (pembrolizumab; n=23) or plus anti-CTLA-4 (nivolumab + ipilimumab; n=24) analyzed. The severe incidence 20% (10/47). Patients had significantly different (PC2) at baseline those without (ROC-AUC=0.78, p=0.01; figure 1). In cohort, showed differentially expressed (ROC-AUC=0.76, 2). Finally, an independent differential population identified clusters distinguished between (figure 3). Interestingly, who post-ICB treatment presented higher frequencies memory effector CD4+ CD8+ T cells, which overlap our signature. <h3>Conclusions</h3> constructed complementary expression-based signatures via supervised approach datasets. These signatures, coupled analysis, distinct expression patterns among later treatment. As such, findings indicate potential clinical utility identifying probability developing irAEs, allowing physicians mitigate side effects timely manner improve patient outcomes. Future studies will investigate why how occur, based on receiving <h3>Ethics Approval</h3> All signed informed consent Dana-Farber/Harvard Cancer Center IRB approved protocol 11-181.