Impact of Various Combinations of Potentially Immunogenic Therapeutics on the T Cell Response in Acute Myeloid Leukemia

Combinations of immunotherapeutic treatment strategies are gaining momentum in the treatment of AML. We need to better understand the mechanisms of immune responses and the relevant antigen structures that play a role in newer immunotherapies. Here we are investigating the influence of immunotherapies alone and in combination on the specific immune response.

We tested the influence of several immunogenic therapeutics on its own or in combination on the antigen-specific immune responses by specific T cells which were targeted against leukemic cells of AML patients. We used ELISpot, colony forming immunoassays (CFI) and flow cytometry to investigate the influence of these drugs on the antigen-specific immune responses by CTL stimulated with leukemia-associated antigens (LAA). The drugs were anti-PD-1 (Nivolumab), anti-CTLA4 (Ipilimumab), All-trans Retinoic Acid (ATRA), 5-Azacitidine (AZA), Lenalidomide (Len). Each drug was tested on its own as well as in combination with anti-PD-1.

In 20 PBMCs from AML patients that showed an immune response against at least one LAA in vitro, we detected a relevant frequency of immune responses against different LAA (WT-1/PRAME/NPM-1). PBMC were stimulated with the LAA that showed the strongest response to test the above immunotherapeutics alone or in combination with anti-PD-1. We tested immune responses against leukemic cells, focusing on leukemic stem cells.

The immunological effects were enhanced when anti-PD-1 was added to the stimulated specific T cells. 13 out of 20 patients showed a significant reduction in colonies, with a median reduction of 25 %. With AZA, we observed a significant reduction of 34% in 12 out of 20 patient samples, and with anti-PD-1+AZA the colony reduction was even higher, with a median colony reduction of 56% in 17 out of 20 patient samples. With Len, the measured reduction was 34% in 12 of 20 patient samples and with anti-PD-1+Len in 15 of 20 patient samples with a mean reduction of 44%. If only the responders are considered, some reduction rates were even higher: with anti-PD-1 39 %, with AZA 57 %, with anti-PD-1+AZA 66 % and with Len/anti-PD-1+Len 57/59 %.

Four patients were analyzed in IFNγ ELISpots. All data were normalized to the respective LAA. ELISpot results were compared to single CFI results. The median increase in stimulation with anti-PD-1 was 1.5-fold. The other immunotherapeutics alone or in combination with anti-PD-1 showed the following fold changes. ATRA 1.6-fold, anti-PD-1+ ATRA 1.6-fold, AZA 1.4-fold, anti-PD-1+AZA 2.0-fold, anti-CTLA4 2.2-fold, anti-PD-1+ anti-CTLA4 2.7-fold, Len 3.0-fold, and anti-PD-1+Len 3.8-fold. Especially AZA, CTLA4 and Len and its combinations resulted in a higher stimulation than anti-PD-1 alone. The ELISpot results cannot be compared with the CFI results in all cases, but this was feasible for some results.

Taken together, combinations of immunotherapeutic approaches increase antigen-specific immune responses against leukemic cells. Especially the combination of LAA peptides with the anti-PD-1 antibody and one further immunomodulating drug, like AZA, could be an interesting option for further studies.

Schrezenmeier:Amgen: Other: Fees; Omeros: Other: Fees; Sanofi: Other: Fees; Sobi: Other: Fees and travel support, Research Funding; Alexion, AstraZeneca Rare Disease: Other: Fees and travel support, Research Funding; F. Hoffmann-La Roche: Other: Fees; Apellis: Other: Fees; Novartis: Other: Fees and travel support, Research Funding.