Need Help?
Copied to clipboard!

Transcriptomic Profiling of CD8+ T Cells during Mosunetuzumab Therapy

This study characterizes longitudinal transcriptional and T-cell receptor (TCR) repertoire changes in circulating CD8+ T cells during treatment with mosunetuzumab, a CD20×CD3 bispecific antibody, in patients with previously untreated follicular lymphoma or marginal zone lymphoma enrolled on the investigator-initiated phase 2 BrUOG-401 trial (NCT04792502). Eligible participants had histologically confirmed, high-burden lymphoma requiring systemic therapy and had not received prior systemic lymphoma treatment. Mosunetuzumab was administered subcutaneously for eight 21-day cycles using step-up dosing during cycle 1. Clinical response was assessed by PET/CT after four cycles and at the end of treatment. The RNA-seq analysis focused on specimens obtained before treatment and during mosunetuzumab monotherapy, before exposure to response-adapted lenalidomide.

Peripheral blood mononuclear cells were collected prospectively at protocol-defined timepoints, cryopreserved, and subsequently used to isolate CD8+ T cells for bulk RNA sequencing. RNA-seq was performed on samples obtained at baseline (cycle 1), early during treatment (cycle 2), and at approximately 3 months of treatment (cycle 5). Analyses included differential gene-expression testing, gene-set enrichment analysis, transcription-factor and gene-regulatory network analysis, and digital cytometry using reference profiles derived from single-cell RNA-seq datasets. TCR repertoires were additionally inferred directly from the bulk RNA-seq data using TRUST4, with primary repertoire analyses based on TCRβ sequences.

The principal transcriptional response to mosunetuzumab occurred early during treatment. Cycle 2 samples demonstrated broad induction of CD8+ T-cell activation, proliferation, TCR-signaling, and immune-regulatory programs, accompanied by increased expression of genes including CD27, CD38, TNFRSF9, LAG3, HAVCR2, and CTLA4. Metabolic analyses identified prominent activation of cholesterol and fatty-acid metabolic pathways, including increased activity of the SREBF1 and SREBF2 regulatory programs. By cycle 5, much of the early activation and proliferation response had contracted toward the pretreatment state, while selected features persisted, including increased LAG3, HAVCR2, and LAYN expression and sustained SREBF2-associated cholesterol-metabolic activity. Digital cytometry supported a transient early increase in cycling and exhaustion-associated CD8+ T-cell states followed by partial normalization later in treatment.

TCR repertoire analysis demonstrated increased CD8+ T-cell clonality during therapy, including greater representation of hyperexpanded and dominant clonotypes at cycle 5. Repertoire overlap with baseline remained high, indicating that treatment-associated repertoire focusing predominantly reflected expansion of pre-existing clonotypes rather than replacement by newly detected clones. Higher cycle 5 clonality and greater abundance of the dominant clonotype were associated with complete response at the early response assessment.

Data available through dbGaP include de-identified raw RNA-seq FASTQ files from isolated peripheral-blood CD8+ T cells together with subject- and sample-level metadata necessary for interpretation of the sequencing data, including age, lymphoma histology, early clinical complete-response status, biological sex, and RNA-seq sample collection timepoint.