Back

TIM3 blockade with hypomethylating therapy restores NK and cytotoxic CD4+ T cell activity in patients with AML or MDS

Huuhtanen, J.; Forsten, S.; Ford, B.; Smolander, J.; Bruck, O.; Lundgren, S.; Kreutzman, A.; Dufva, O.; Kankainen, M.; Ilander, M.; Lahteenmaki, H.; Kasanen, T.; Klievink, J.; Leitner, J.; Steinberger, P.; Kontro, M.; Pelletier, M.; Lahdsemaki, H.; Sabatos-Peyton, C.; Rinne, M.; Porkka, K.; Peltonen, K.; Mustjoki, S.

2025-11-19 hematology
10.1101/2025.11.14.25339889 medRxiv
Show abstract

Occasional complete responses to immune checkpoint inhibitor therapies suggest that acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) are immune-sensitive when appropriately targeted. Here, we analyzed AML/MDS patients treated with anti-TIM3 sabatolimab and decitabine in a phase Ib clinical trial (NCT03066648) using single-cell RNA and T cell receptor (TCR) sequencing and functional co-culture assays. Unlike T cell restricted CTLA4 and PD1, TIM3 was broadly expressed across natural killer (NK), myeloid, and T cell populations. Therapy induced expansion of cytotoxic NK cell subsets and enhanced type I interferon signaling. Less than 1% of bone marrow CD8+ T cells showed canonical exhaustion phenotypes, and the treatment preferably expanded small CD8+ T cell clones in responders. Responders had more cytotoxic CD4+ T and B cells which was exemplified by a patient with an outstanding complete response of 23-months. Over >20% of this patients lymphocytes were CD4+ T-cell large granular lymphocyte leukemia (T-LGLL) cells that expressed a TCR capable of recognizing autologous blasts. SignificanceAnti-TIM3+decitabine offers a distinct mechanism compared to anti-PD1 and anti-CTLA4: it preferentially enhances innate and unconventional cytotoxic lymphocyte populations. Furthermore, our findings suggest that clonal expansion of self-antigen-specific T cells, such as T-LGLL, may augment anti-leukemic immunity in the context of immunotherapy.

Published in Cancer Immunology Research · training set

Matching journals

The top 6 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.