Back

Germline HLA-DQ genotype influences response to CAR T-cell therapy in patients with large B-cell lymphoma

Le Mene, M.; Allain, V.; Villemonteix, J.; Cuffel, A. A.; Taupin, J.-L.; di Blasi, R.; Thieblemont, C.; Caillat-Zucman, S.

2024-01-12 allergy and immunology
10.1101/2024.01.11.24301087 medRxiv
Show abstract

CD19-directed chimeric antigen receptor (CAR) T cells have greatly improved the prognosis of relapsed/refractory large B-cell lymphoma (rrLBCL), yet treatment failure occurs in more than half of patients, usually in the first 3 months after treatment. While they primarily act through CAR-dependent, HLA-independent recognition of tumor targets, CAR-T cells may also indirectly contribute to long-term tumor immunosurveillance by stimulating endogenous immunity. We hypothesized that HLA diversity, measured by the HLA evolutionary divergence (HED) metric which reflects the breadth of the immunopeptidome presented to host T cells, could influence antitumor response after CAR T-cell therapy, as seen after immune chekpoint inhibitor treatment. We studied 127 rrLBCL patients treated with commercial CAR-T cells in our center, of whom 50 % achieved durable response. We observed no impact of diversity at any HLA locus, except for HED-DQA1 that was surprisingly negatively associated with response. Analysis of the distribution of HLA-DQ alleles according to clustering of HED values pointed to the DQ1/DQ1 genotype as an independent predictor of durable response and lower incidence of relapse/progression. These findings highlight the unsuspected role of germline HLA-DQ molecules in the response to CAR-T cells and suggest an important contribution of cross-talk between CAR-T cells and endogenous immune cells. Key PointsO_LIGermline HLA-DQ genotype is an independent predictor of durable response and lower incidence of relapse/progression after CAR T-cell therapy in rrLBCL C_LIO_LIHLA-DQ1/DQ1 genotype could influence the host immune response after CAR T-cell therapy and increase the chances of a durable response C_LI

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.