Genome and prognosis models : towards the personalized therapy for patients with B-cell lymphoma

Abstract

B-cell lymphoma is a heterogeneous cancer derived from B lymphocytes, with over 60 subtypes classified by the World Health Organization (WHO). Common subtypes include diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL), and mantle cell lymphoma (MCL). Several factors have been suggested to be associated with lymphomagenesis and inferior outcomes, such as different types of genetic alternations, viral infection, and tumor microenvironment. Using high-throughput sequencing on the genome and transcriptome of tumors at the bulk tissue or single-cell levels, we investigated the factors contributing to lymphomagenesis and disease progression across various types of B-cell lymphomas. Furthermore, leveraging data from multiple cohorts, we developed a prognostic tool to predict outcomes for DLBCL patients.

In study I, we investigated the genetic and transcriptomic profiles of 2805 DLBCL patients, focusing on those with early disease progression within two years after first-line rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) treatment. We observed distinct mutational patterns in patients with poor outcomes, including enriched mutations in TP53, MYD88, SPEN, MYC, and unique expression profiles characterized by dysregulation of cell migration, T-cell activation, PI3K, and NF-kB signaling. Additionally, we developed a robust and independent prognostic tool based on the expression of 24 genes to identify DLBCL patients with early disease progression. This tool can also be combined with cell-of-origin subtypes and DNA molecular subtypes to further stratify patients.

In study II, we explored the association between chronic hepatitis B virus (HBV) infection and FL in the Chinese population using the data derived from genomic and transcriptomic sequencing. Compared to HBV surface antigen negative (HBsAg-) FL patients, HBsAg+ FL patients were younger, had higher histology grades, and poorer prognoses. Their tumors exhibited enhanced mutagenesis and distinct genetic profiles, characterized by a lack of IGH/BCL2 translocations and CREBBP mutations but more mutations in another set of genes (TNFAIP3, FAS, HIST1H1C, TP53, PIM1, etc.). Transcriptomic analyses showed that HBsAg+ FLs exhibited expression signatures similar to those of activated B-cell-like DLBCL (ABC-DLBCL). These data suggest that HBsAg+ FLs may represent a distinct FL subtype.

In study III, we studied tumor evolution and microenvironment of MCL from primary to relapse by integrating data from whole-genome sequencing (WGS), single-cell RNA sequencing (scRNA-seq), and single-cell B-cell receptor sequencing (scBCR-seq). We observed significant inter-patient, intra-patient, and intra-tumor heterogeneity in MCL. Additionally, MCL cells may receive survival signals from T cells via CD40-CD40L and CD70-CD27 interactions and from myeloid cells via BAFF-mediated signals. Furthermore, we demonstrated divergent clonal evolution patterns through disease progression in individual patients. Our results indicate that intra-tumor heterogeneity is likely a key factor contributing to MCL relapse, complicating therapeutic eradication. Subsequently, residual tumor clones may acquire new mutations and copy number changes, or migrate into favorable tissue microenvironments, promoting tumor cell proliferation and leading to disease relapse.

In study IV, we conducted a systematic characterization of mutational signatures using WGS data from 929 tumors across nine subtypes of B-cell malignancies. We identified 16 single-base substitution (SBS), four double-base substitution (DBS), and three indel signatures, including three lymphoid-specific signatures and several associated with DNA repair deficiencies and specific tumor types. Additionally, we extracted 6,040 kataegis events (highly clustered mutations) and confirmed three previously identified kataegis signatures associated with activation-induced cytidine deaminase (AID), polymerase h (POLH), and apolipoprotein B mRNA editing catalytic polypeptide-like (APOBEC) activities. Our extensive cohort further revealed that the AID-mediated kataegis signature was linked those subtypes with poorer prognostic outcomes.

In conclusion, our extensive genomic and transcriptomic studies provide valuable insights into the complex mechanisms driving lymphomagenesis and disease progression in different types of B-cell lymphoma. Our findings have the potential to advance the development of personalized medicine for lymphoma patients.