No abstract

Follicular lymphoma (FL) is the most common indolent form of non-Hodgkin lymphoma. Histological transformation of FL to a more aggressive form of lymphoma occurs with a linear incidence of 2-3% per year and is associated with poor outcome. Divergent clonal evolution and an altered tumour microenvironment (TME) have both been implicated in the transformation process. However, the phenotypic consequences of this evolution and its implication in reshaping the TME remain unknown. To address this knowledge gap we performed single cell whole genome (scWGS) and single cell whole transcriptome sequencing (scWTS) of paired pre/post transformation samples of 11 FL patients. We further performed scWTS analysis of additional 11 FL samples from patients that had not undergone transformation within 7 years. Our comprehensive single cell analysis revealed the evolutionary dynamics of transformation at unprecedented resolution. Computational integration of scWGS and scWTS allowed us to identify gene programs upregulated and positively selected during evolution. Furthermore, our scWTS analysis revealed a shifting TME landscape, with an exhausted CD8 T cell signature emerging during transformation. Using multi-color immunofluorescence we transferred these findings to a novel TME based biomarker of transformation, subsequently validated in 2 independent cohorts of pretreatment FL samples. Taken together, our results provide a comprehensive view of the combined genomic and phenotypic evolution of malignant cells during transformation, and the shifting cross-talk between malignant cells and the TME.

Introduction: Histological transformation from follicular lymphoma (FL) to aggressive B-cell lymphoma (tFL) is a disease course altering event linked to poor prognosis for affected patients. From a biological point of view, it is paradigmatic of disease dynamics with distinct clinical stages that project onto genetically and phenotypically divergent states. Aim: By applying a series of high-dimensional single cell (sc)RNA and DNA profiling techniques, we aimed to characterize the clonal and phenotypic evolution of tumor B cells and to reveal dynamic interactions with components of the tumor microenvironment (TME) during transformation. Methods: We included 11 tFL patients with paired FL (tFL-FL) and DLBCL (tFL-DLBCL) timepoint biopsies, and 11 indolent FL controls (with > 6y of follow-up without evidence of progression or transformation). Single cell whole transcriptome (scWTS) and BCR sequencing was performed for all samples and single cell whole genome sequencing (scWGS) for transformation pairs. Results: In each transformation pair, BCR sequencing confirmed the clonal relationship between FL and DLBCL timepoints. Clustering of scRNA data from each pair showed an inverse correlation between transcriptional similarity and time between the two biopsies. Some tFL-FL cells could always be found within the tFL-DLBCL clusters. Therefore, we labeled these cells as presumed “early-DLBCL cells”. Phylogenetic analysis using scWGS data showed distinct FL and DLBCL clones, and “mixed-clones” composed of cells from both timepoints in most pairs. FL cells in DLBCL clones were favored to represent precursor cells of transformation. DLBCL cells in FL clones likely represent residual FL cells after transformation, and could be found in the majority of the pairs. Divergent evolution from FL to DLBCL with specific copy number abnormalities unique to each timepoint was the most common mode of evolution during transformation, and only one pair showed linear evolution. Integrative analysis of scRNA and scDNA data highlighted that samples with the fewest genomic changes showed the least transcriptomic changes and vice versa. Differential expression and gene set enrichment analysis of malignant cells identified “MYC targets V1“ as the main pathway enriched in tFL-DLBCL cells in comparison to tFL-FL cells. Cells from the indolent control FL cases had a significantly lower MYC score than cells from pre-transformed FL. In parallel to the insights into tumor cell evolution, scRNAseq analysis also revealed significant shifts in TME composition, from T cells with a TFH and central memory phenotype in tFL-FL samples, to cells with an exhausted cytotoxic phenotype in tFL-DLBCL samples. Conclusion: Applying high-dimensional scRNA and DNA profiling techniques we identified precursor cell populations of transformation at the genomic and phenotypic level and linked genomic and phenotypic evolution with shifting TME composition in a comprehensive disease evolution model of transformation. Citation Format: Clementine Sarkozy, Shaocheng Wu, Katsuyoshi Takata, Tomohiro Aoki, Susana B Neriah, Katy Milne, Brad Nelson, Andrew Weng, David Scott, Jeffrey W Craig, Christian Steidl, Andrew Roth. Integrated single cell analysis reveals co-evolution of malignant B cells and the tumor microenvironment in transformed follicular lymphoma [abstract]. In: Proceedings of the Third AACR International Meeting: Advances in Malignant Lymphoma: Maximizing the Basic-Translational Interface for Clinical Application; 2022 Jun 23-26; Boston, MA. Philadelphia (PA): AACR; Blood Cancer Discov 2022;3(5_Suppl):Abstract nr A19.