Abstract A134: Single-cell roadmap of the evolution of T-cell response during anti-LAG3 and anti-PD1 combination treatment in metastatic melanoma patients

Abstract

Abstract Despite the impressive impact of CTLA4 and PD1-targeted immuno-oncologic (IO) therapies, a large proportion of patients fail to respond. The observed variance in treatment efficacy has been linked to heterogeneity in the immune cell distribution of individual patients. Lymphocyte activation gene 3 (LAG3, CD223) is one of the newest IO targets entering the clinic. Triggering of LAG3 on T-cells by HLA-DR -ligands has a well-established role in the negative regulation of T-cell function. Although for example on activated regulatory T-cells (Treg) LAG3 is widely expressed, the detailed effects of LAG3 on various cell types are still unknown. We profiled over 30,000 CD45+ lymphocytes cells using a novel paired single-cell RNA and T-cell receptor (TCR) αβ chain (10x Genomics) sequencing method for peripheral blood samples from two patients receiving anti-LAG3 and anti-PD1 combination treatment in multicentre phase I trial. The sequenced cells were from IO-treatment naïve metastatic melanoma patients from before, after 4 weeks and after 12 weeks of the start of therapy. For validation, we performed TCRβ-sequencing and flow cytometry analysis of a larger cohort of melanoma patients (n = 12) enrolled in the same study. To gain in-depth understanding of the immune cell subsets, we used a recently described method to seek matching mutual neighbors between patients to normalize the interpatient variation to enable a systematic comparison across patients. To identify phenotypic clusters, we used a graph theory-based clustering method SNN-Cliq and built predictive machine learning classifiers to assess the reproducibility of learnt clusters. After optimizing our choice of input genes and parameters, we identified 16 distinct clusters that define the T-cell roadmap of anti-LAG3 and anti-PD1 treatment that includes 6 CD8+, 8 CD4+ (including Treg cluster), and 2 other clusters. We identified 4 CD8+ T-cell clusters of increasing cytotoxicity profile using a cytotoxicity score and the pseudotime algorithm Monocle3. The most highly cytotoxic clusters increased and a cluster of lower cytotoxicity score decreased during the treatment. In addition, we defined 2 CD8+ exhaustion clusters. During treatment, we observed a decrease in the exhausted T-cell cluster defined by LAG3 and PDCD1 expression, but an increase in TIGIT+ exhaustion cluster. Furthermore, ordering the cluster of FOXP3+ Treg cells along pseudotime trajectory revealed two different fates for Treg cells, where the other fate showed significantly decreased expression of LAG3 and PDCD1, adding evidence of the effect of the treatment on Treg cells. TCRαβ analysis revealed 19 individual expanded clonotypes of size of 100 sequenced individual cells. The true transcriptomic heterogeneity of identical clonotypes was revealed as most clonotypes spanned several of the 16 different clusters, challenging our view of clonal expansion. Furthermore, we were able to assess the temporal phenotypic changes in individual clones during the course of immunotherapy. Most clonotypes, including cytotoxic and exhausted clones, had more homogenous transcriptomes before the start of the treatment, but diversified during the therapy, suggesting a release of immunologic break in these clones. In summary, we defined the evolution of immunological response to anti-LAG3 and anti-PD1 therapy cell by cell, and described an increase in the heterogeneity of cytotoxicity in CD8+ cells, distinct fates of Treg cells and the release of transcriptomic profiles of individual T-cell clonotypes during IO treatment. Citation Format: Jani Huuhtanen, Henna H.E. Hakanen, Tapio Lönnberg, Olli Dufva, Katriina Peltola, Siru Mäkelä, Micaela Hernberg, Petri Bono, Kreutzman Anna, Satu Mustjoki. Single-cell roadmap of the evolution of T-cell response during anti-LAG3 and anti-PD1 combination treatment in metastatic melanoma patients [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr A134.