For many patients with MDS and AML, the disease is incurable, and therapy is palliative in nature. Allogeneic stem cell transplantation remains the only potentially curative therapy, but transplantation is limited to only a subset of patients. Therefore, there has been a great deal of interest in finding ways to reproduce the immune benefits of transplantation, or the graft-vs-leukemia response, but in other contexts and without the associated toxicities. 

The rationale for combining HMAs such as azacitidine or decitabine with immune checkpoint inhibitors (ICIs) was based on several preclinical models that have resulted in clinical trials examining a number of ICI combinations including anti–PD-1, PD-L1, and CTLA-4 agents. In a randomized study of azacitidine with or without the anti–PD-L1 agent atezolizumab, there was no significant benefit in the response rate or duration of responses in the combination arm, and the study was terminated early due to an unexpectedly high early death rate. This underscores how some toxicities associated with canonical ICIs (ie, anti–PD-1, PD-L1, and CTLA-4 agents) can make their administration challenging and has led us to explore differing immunological targets in MDS and AML, such as TIM-3 or CD47.

TIM-3 is a cell surface marker that does have immune checkpoint function, but it is also expressed on leukemic progenitors and it appears to be distinct from canonical ICIs. It is expressed on a number of immune effector cells, including T cells and natural killer cells, and is associated with a T-cell exhaustion phenotype. Given its expression on the leukemic progenitors, targeting TIM-3 may provide both immune-related and direct leukemic progenitor–related therapeutic activity. 

Sabatolimab is a TIM-3–directed antibody that is probably the furthest along in its development. The STIMULUS clinical trial program includes multiple studies evaluating sabatolimab as part of different combination therapies. In our phase 1b study of patients with higher-risk MDS, we administered azacitidine or decitabine followed by sabatolimab in escalating doses, and we evaluated the response rates and the duration of those responses, as well as the safety of this combination. Updated data from this study were reported at the 63rd American Society of Hematology Annual Meeting and Exposition in 2021. 

Given the experience with other immuno-oncology agents, we wanted to make sure that this combination could be safely administered to patients. In our analysis, we were particularly attuned to potential immune-related toxicities, and we did identify several, including arthralgias and rashes that were grade 3 or lower. In general, such toxicities could be managed with supportive care or steroids, and only 1 of 101 patients actually stopped therapy due to an adverse event. In contrast to anti–PD-1 or CTLA-4 agents, sabatolimab seems to have a different immunological phenotype, with fewer instances of pneumonitis and colitis. It can be administered to an older patient population, and it does not appear to be as challenging to coadminister with azacitidine. We also observed durable responses in patients across a number of high-risk mutation profiles.

Our experience suggested that the combination of TIM-3 inhibition with HMA therapy could be administered without adding excess toxicity. Hopefully, this means that patients can stay on the combination for longer periods and without dose interruptions. We await data from the currently enrolling phase 3 trial to see whether this approach will expand our treatment repertoire.