Abstract 5922: B cell subtype is a predictive and pharmacodynamic biomarkers for combination therapy in stage III-IV melanoma: a simon phase II trial

Background: Assessing biomarkers to predict immune-related adverse events (irAEs) is crucial for optimizing melanoma treatment with immune checkpoint inhibitors (ICIs). T and B lymphocytes are key mediators of autoimmunity and are implicated in the development of irAEs. Here we focus on analysis of specific B cell subsets, as predictive and pharmacodynamic biomarkers for irAEs in a phase II trial combining nivolumab (3 mg/kg), ipilimumab(1 mg/ml), and the IL-6 receptor inhibitor tocilizumab (4 mg/kg) in unresectable/advanced stage III/IV melanoma patients. Methods: We analyzed cryopreserved PBMC from melanoma patients collected at baseline and week 7, and categorized by irAE severity. Patients were categorized as either experiencing severe irAEs (grade 3-4) (TOX) or no toxicity or grade 1-2 irAEs (NT). We used multi-parameter flow cytometry panel to analyze B cells. PBMCs analyzed by a 35-color antibody B cell panel by spectral flow cytometry. The data underwent quality control, and normalization and initial clustering. CD19+ B cells were concatenated and clustered by FlowJo software. Fifteen B cell clusters was obtained and phenotyped using Cluster Explorer. Statistical analyses were performed using multiple logistic regression and Mann-Whitney U test, with a significance level set at 0.05. Results: Among the 70 patients administered treatment, we observed a best overall response rate of 57%. Grade 3-5 irAEs occurred in 22% of patients by week 24. We analyzed 58 quality-controlled baseline samples and 54 7-week samples. Fifteen distinct clusters were identified after clustering B cell populations using UMAP and re-clustering using PHATE. Two of these clusters showed significant differences in both multivariate logistic regression and Mann-Whitney U tests. TOX patients had significantly lower baseline naive B cell counts than NT patients (p=0.016), suggesting naïve B cells as a predictive marker for toxicity. A significant increase in late memory B cells at week 7 was a key pharmacodynamic marker (p=0.003). Conclusion: Our findings suggest that naive B cells could serve as a toxicity biomarker in combination therapy with ipilimumab, nivolumab, and tocilizumab. Late memory B cells emerged as a promising pharmacodynamic biomarker. Immunological monitoring may help elucidate the mechanisms of irAEs and predict irAE occurrence. Citation Format: Teruyuki Mizutani, Amrutesh Puranik, Tomoaki Muramatsu, Joanna Handzlik, Ankita Mitra, Diana Kimono, Stanzin Idga, Eddy Yang, Ramin Herati, Michelle Krogsgaard, Mehnert Janice. B cell subtype is a predictive and pharmacodynamic biomarkers for combination therapy in stage III-IV melanoma: a simon phase II trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5922.