Based on murine data, we initiated a phase I/II clinical study to reduce PTCy from 50 mg/kg/day on days +3/+4 (Set 1) to 25 mg/kg/day on days +3/+4 (Set 2) or 25 mg/kg on day +4 only (Set 3) along with sirolimus and MMF (both starting day +5) for myeloablative HLA-haploidentical bone marrow HCT. Phase II used the Set 2 dosing and found no grade II-IV acute GVHD events and 13% chronic GVHD requiring systemic therapy, but with faster T-cell recovery and less frequent CMV reactivation and less severe BK virus cystitis compared with Set 1. Single-cell RNA sequencing was performed fresh in donor PBMCs and marrow on day 0 and in recipient PBMCs on days 0, +3, +5, +7, and +21 for the 19 phase I patients (n=5-7 per Set). Donor- and recipient-derived cells were separated using unique SNPs identified from day 0 samples. Data were batch corrected (Harmony), annotated (Seurat Azimuth), and analyzed in a pseudobulk manner. Alloreactive T cells were defined based either on CITEseq expression of CD69 or CD137 at day +3, +5, and/or +7 or based on matching with donor T-cell clones proliferating against recipient pre-HCT cells in mixed lymphocyte cultures in vitro . 594,070 cells were captured across 132 patient samples. Alloreactive CD4 + and CD8 + T cells were trackable and persisted despite PTCy; interestingly, given the HLA-haploidentical setting, alloreactive T cells commonly derived from memory cells in the donor. T-cell receptor diversity was more clonal at day +21 after Set 1 vs. Sets 2/3. Pathway analysis showed donor CD4 + and CD8 + T cells, natural killer cells, and monocytes had less DNA damage at days +5/+7 in Sets 2/3 vs. Set 1, an effect also seen within alloreactive T cells. At day +21, CD4 + T cells in Sets 2 and 3 had increased Th17 differentiation compared with Set 1, and Set 2 also had increased type I interferon signaling vs. Set 1. Donor CD8 + T cells and alloreactive CD8 + T cells within Sets 2/3 at days +5/+7/+21 had increases in pathways related both to GVHD but also responses to pathogens when compared to Set 1. Donor monocytes had increased interleukin-10 signaling at day +21 in Set 2 compared with Set 1. Cell chat analyses showed different patterns of cell-cell communications between Sets 2/3 and Set 1, particularly between donor and recipient T cells at day +5/+7 and between donor monocytes and T cells at day +21. Gene correlation network analysis across Sets and timepoints showed consistent associations at days +3, +5, +7, and +21 with non-relapse mortality. Reducing PTCy dosing results in less DNA damage, increased CD8 + T-cell pathways of GVHD and pathogen response, and modulations in cell-cell interactions. Even so, these effects did not increase clinical GVHD, even as they decreased infection. Gene modules were predictive of non-relapse mortality at day +3, suggesting that, irrespective of the PTCy dose, some patients already may be immunologically set up for failure very early post-HCT.
Schneider et al. (Sun,) studied this question.