Tumor Genomics and Microenvironment Characterization of Diffuse Large B-Cell Lymphoma in Asian and Pacific Islander Patients

Introduction

Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma worldwide. Multiple epidemiological studies have evaluated how differences in environmental factors, race and ethnicity impact clinical presentation and outcome in DLBCL, but limited data exist to address differences in disease biology across multiethnic cohorts. Here, we analyzed whole-exome (WES) and RNA sequencing (RNA-seq) in DLBCL tumors from the Hawaii Surveillance, Epidemiology, and End Results (SEER) residual tissue repository (RTR), using machine learning (ML) approaches to characterize molecular and transcriptomic features in a sample cohort enriched for Asian and Pacific Islander (PI) patients.

Methods

Formalin-fixed paraffin-embedded tissue samples for patients with DLBCL from the Hawaii SEER registry were collected, underwent quality control (QC), and were analyzed. For some cases, multiple samples from different biopsy sites were available. The ancestry of each patient was determined with PLINK, which utilizes single-nucleotide variant data to assess population stratification. PLINK presented membership in 26 populations belonging to 5 superpopulations; the superpopulation with the highest weight in each patient was chosen as the ancestral group. Somatic mutations were calculated using a supervised ML approach. A custom GRIDSS-based pipeline that identifies single breakends in genomic DNA (Cameron et al., Genome Biol, 2021) was used to detect genomic breaks or translocations. DLBCL subtypes were delineated using LymphGen (Wright et al.,Cancer Cell, 2020). Lymphoma microenvironment (LME) types were estimated as described in Kotlov et al. (Cancer Discov, 2021). Cell-of-origin (COO) and double-hit signature (DHITsig) (Ennishi et al., J Clin Oncol, 2019) status was estimated as described in Kotlov et al. (Cancer Discov, 2021) to determine DLBCL subgroups. Deconvolution by Kassandra was used to predict cell percentages from bulk RNA-seq data.

Results

Of 121 DLBCL samples received for 84 patients, 86 sample passed QC for WES, and 57 for RNA-seq. This cohort included 37 males and 28 females, with race self-reported as Asian in 42 cases, white in 10, PI in 6, “mixed” in 6, and “other” in 1 case. Age ranged 10-85+ years, with 22 patients alive and 43 deceased (25 due to lymphoma; 18 due to other reasons) after median follow-up of 87 months. In 59 patients with available WES data, estimation of global genetic ancestry identified 6 (10%) as admixed American (AMR), 42 (71%) as East Asian (EAS) and 11 (19%) as European (EUR). Genes most frequently mutated were KMT2D, SOCS1, and TP53. In 16 samples, translocations involving BCL2 (n=8), BCL6 (n=3), and MYC (n=5) were identified. One sample with BCL2-IGH structural variant was also classified as DHITsig-positive. COO analysis classified 36 samples as germinal center B cell-derived and 21 samples as activated B cell-derived subtypes.

LymphGen did not detect the prevalence of a particular DLBCL subtype among the samples analyzed, with 62% (53/86 WES samples) being unclassified/“Other” genetic subtype. In 7 cases, different LymphGen subtypes were observed in samples from the same patient: for instance, for one patient, samples were classified as either “other”, ST2 or N1 subtypes. Due to limited clinical data, it is unclear whether this represented heterogeneity of subtype at a single point in time, or, more likely, evolution of subtype at relapse.

Across 43 patients with RNA-seq data, 42% exhibited immune-depleted LME. Among the EAS group with available RNAseq data, 13(50%) had immune-depleted LME. Both of these prevalences are higher than the 30% described in initial datasets that included predominantly white patients with DLBCL.

Conclusion

Our analyses provide an initial characterization of DLBCL LME in Asian and PI patients, an under-represented population in prior studies. Our cohort exhibited higher incidence of immune-depleted LME than expected for DLBCL, and most cases did not map to a LymphGen classification. These findings suggest that currently used genetic classification systems may suboptimally represent DLBCL heterogeneity in non-EUR populations. Studies with larger and more diverse patient populations are warranted to better understand how ancestry, socioeconomic elements, and environmental factors contribute to DLBCL pathobiology and outcomes.

Lee:NCI: Other: This work was supported by the Hawaii Tumor Registry of the Univ of Hawaii Cancer Center through NCI SEER Contract Award HHSN261201300009I. The content is solely the responsibility of the authors. Does not necessarily represent the official views of NCI.. Zemskiy:BostonGene: Current Employment. Kriukov:BostonGene: Current Employment. Tkachuk:BostonGene: Current Employment. Zornikova:BostonGene: Current Employment. Nomie:BostonGene: Current Employment, Current equity holder in private company, Current equity holder in publicly-traded company. Bagaev:BostonGene: Current Employment, Current equity holder in private company, Current holder of stock options in a privately-held company, Patents & Royalties: BostonGene, Corp.. Koff:AbbVie: Consultancy; BeiGene: Consultancy; Viracta Therapeutics: Research Funding. Flowers:BeiGene: Consultancy; Eastern Cooperative Oncology Group: Research Funding; Denovo Biopharma: Consultancy; Morphosys: Research Funding; TG Therapeutics: Research Funding; Foresight Diagnostics: Consultancy, Current holder of stock options in a privately-held company; Acerta: Research Funding; Novartis: Research Funding; Xencor: Research Funding; Seagen: Consultancy; Sanofi: Research Funding; Bayer: Consultancy, Research Funding; Ziopharm National Cancer Institute: Research Funding; Allogene: Research Funding; Burroughs Wellcome Fund: Research Funding; BostonGene: Research Funding; EMD Serono: Research Funding; Takeda: Research Funding; Pharmacyclics: Research Funding; Genmab: Consultancy; Spectrum: Consultancy; 4D: Research Funding; Kite: Research Funding; Janssen Pharmaceuticals: Research Funding; Adaptimmune: Research Funding; Pfizer: Research Funding; Iovance: Research Funding; Cellectis: Research Funding; Celgene: Consultancy, Research Funding; Pharmacyclics / Janssen: Consultancy; Guardant: Research Funding; Genentech/Roche: Consultancy, Research Funding; N-Power Medicine: Consultancy, Current holder of stock options in a privately-held company; Nektar: Research Funding; Karyopharm: Consultancy; Gilead: Consultancy, Research Funding; Amgen: Research Funding; AstraZeneca: Consultancy; Cancer Prevention and Research Institute of Texas: CPRIT Scholar in Cancer Research: Research Funding; Bio Ascend: Consultancy; Bristol Myers Squibb: Consultancy; AbbVie: Consultancy, Research Funding.