Introduction Despite remarkable improvements in the survival of relapsed/refractory B-cell lymphoma (R/R BCL) patients following chimeric antigen receptor (CAR)-T cell therapy, there is room for improvement, as approximately 50% of the patients do not achieve long-term remission and 20% experience severe toxicity. We hypothesize that the limited efficacy is not only mediated by inefficient CAR-T cell expansion and longevity but also by a suppressive tumor microenvironment and adverse host response.

Materials & Methods We collected sequential blood samples from patients treated with CAR-T cells to monitor their immune response dynamics. Peripheral blood mononuclear cells (PBMCs) were isolated from blood samples (n = 167) collected from R/R BCL patients (n = 25; 11 DLBCL, 8 tBCL, 3 HGBCL, 2 PMBCL, 1 FL) treated with axicabtagene ciloleucel (axicel), tisagenlecleucel (tisacel). To date, 76 samples have been analyzed using flow cytometry (CD45, CD14, CD3, CD4, CD8, and G4S), and 30 samples have been subjected to single cell RNA-sequencing (scRNA-seq) and V(D)J sequencing of αβ T-cell receptors (TCR) using the 10x Chromium platform. The libraries were sequenced using the Illumina NovaSeq 6000 system, and the reads were aligned to a custom reference based on GRCh38, which included the addition of axicel and tisacel transcripts.

Results Flow cytometry analysis of the samples revealed distinct immune profiles and kinetics among patients at various time points. For example, at baseline (BL), there was a trend towards lower monocyte % in responders than non-responders (p = .12). Additionally, a low absolute lymphocyte-monocyte ratio calculated from differential blood counts was associated with progression (p = .13). As expected, the number of CAR-T cells in circulation was higher 1 week post-infusion (1w) and decreased by the 1-month (1m) timepoint (mean % 2.66 vs 0.59, p < .05).

scRNA-seq analysis enabled a more detailed characterization of a subset of samples. The assessment of CAR-T cell % with scRNA-seq correlated with flow cytometry data (R = 0.98, p < .0001). The distribution of different T cell subsets varied between patients, both overall and within the CAR-T cells. Notably, the proportion of CD4 T cells (p < .05) and specifically CD4 effector memory (Tem) cells (p < .05) among CAR-T cells was higher in patients who received tisacel (n = 3) compared to those who received axicel (n = 7). Moreover, for patients receiving tisacel, the overall proportion of CD4 Tem cells in circulation was nearly 15-fold higher at 1w than for axicel (mean % 36.28 vs 2.46, p < .05) and remained elevated at 1m (p < .05). The proportion of CD4 naïve T cells decreased during the treatment course (p < .01), and a similar trend was observed for CD8 naïve T cells. Furthermore, the proportion of CD8 naïve T cells at BL was associated with an improved response rate (p < .05), while no correlation was observed between any other CD8 or CD4 cell subsets and outcomes.

Investigation of the TCR sequences revealed that the vast majority (88.74%) of the persistent T cell clonotypes observed at multiple time points were CD8 Tem cells. In contrast, among the non-persistent clones, only 19.35% were CD8 Tem cells, with the predominant subtype being CD4 Tem cells (35.98%). For the T cells for which clonotype information was not available (30.92% of all T cells), the proportions of CD8 Tem cells and CD4 Tem cells were 60.20% and 14.60%, respectively. Only 0.38% of the T cells without detectable αβTCR were γδ T cells.

T cell clonotypic diversity was assessed by calculating the Shannon entropy index for each sample. There was a trend toward lower TCR diversity in non-responders compared to responders (p = .053). Reflecting on the T cell subsets, the Shannon entropy index correlated negatively with CD8 Tem cell % (R = -0.78, p < .0001) and positively with CD4 naïve T cell %, especially at BL (R = 0.79, p < .01).

Conclusions Our findings elucidate the complex immune landscape of patients with R/R BCL undergoing CAR-T therapy, revealing differences in immune profiles that clarify the host response to therapy. These insights not only deepen our understanding of the factors influencing CAR-T efficacy but also lay a foundation for future studies to identify patients who are more likely to benefit from CAR-T cell therapy.

This content is only available as a PDF.
2025
Sign in via your Institution