Non-Engrafting CD8-Depleted Human Leukocyte Antigen (HLA) Mismatched Unrelated Donor Lymphocyte Infusion (DLI) Achieves Remissions in Transplant-Naive Myelodysplastic Syndrome (MDS) and Secondary Acute Myeloid Leukemia (sAML)

Introduction: Hypomethylating agents (HMAs) are standard of care in higher risk MDS but are effective in only ~50% of patients. In cases that are refractory to HMA or progress to sAML, median survival is only 4-6 months. We designed a first-in-human cellular therapy of CD8-depleted, CD4+ mismatched unrelated donor lymphocyte infusion (CD4+ MMUDLI) administered in patients who had not received a prior allogeneic hematopoietic cell transplant (HCT). The mismatched donor reverse immune tolerance to elicit an antitumor response and induce remission, while depletion of CD8+ T cells prevents the risk of graft-versus-host disease (GVHD) and permanent engraftment (Symons. BBMT. 2008).

Methods: This completed phase I trial with expansion cohort enrolled patients with MDS refractory to at least 4 cycles of HMA or sAML from 2020-2023. Patients who received prior allogeneic HCT were excluded. DLI donors were selected from a pool of 30 unrelated donors at the New York Blood Center (NYBC) based on class II HLA mismatch, CMV serology, and ABO compatibility. Patients received a standard of care cytarabine-based induction chemotherapy regimen chosen by their primary physician. In parallel, donors underwent leukapheresis at the NYBC and the cell product was shipped to the Moffitt Cancer Center and depleted of CD8⁺ T cells using the CliniMACS® system with CD8 reagent. The CD4+ MMUDLI was infused fresh within 24-36 hours after completing induction chemotherapy. The primary safety endpoint was the maximal tolerated dose (MTD). The trial employed a 3+3 dose escalation schema where DL1 was 1x10⁶ CD4 cells/kg, DL2 was 1x10⁷ CD4 cells/kg, and DL3 was 5x10⁷ CD4 cells/kg, and then enrolled an additional 10 subjects at the MTD. Dose limiting toxicities (DLTs) were defined as grade 2-4 acute GVHD, grade 3-4 cytokine release syndrome (CRS), grade 2-4 infusion reactions, aplasia beyond 56 days in the presence of detectable donor chimerism, or grade 4 organ toxicity by CTCAE 5.0. The primary efficacy endpoint was complete remission (CR) rate by day 56.

Results: Nineteen patients were treated (16 sAML, 3 MDS). Of these, 18 had relapsed/refractory disease. All AML patients had adverse risk disease by European LeukemiaNet (ELN) criteria, and all MDS patients had very high risk disease by revised international prognostic scoring system (IPSS-R) and molecular IPSS (IPSS-M). The most common somatic mutation was TP53 (n=11), followed by ASXL1 (n=4) and RUNX1 (n=3), and 13 patients had complex cytogenetics. Median age was 67 (range: 51-76) years. The median follow-up time was 12.8 months (range 4.9-21 months).

A suitable donor was identified for all patients despite a donor pool of only 30 donors. All cell collections, shipments, and processing were successful. All products met the desired CD4+ T cell dose, <3% CD8+ T cells, and >70% cell viability.

No DLTs were observed at any dose level, confirming DL3 as the MTD. Two patients at DL3 experienced grade 2 CRS that resolved, one requiring tocilizumab. There were no cases of grade 3-4 CRS and no cases of GVHD.

By dose level, CR was achieved in 0/3 (0%) patients at DL1, 1/3 (33%) patients at DL2, and 7/13 (54%) patients at DL3. Among TP53 mutated cases, 3/8 (38%) treated at DL3 achieved CR, including 2 with no minimal residual disease after therapy. All 8 patients who reached CR proceeded to allogeneic HCT, and 6 remain in remission at last follow-up. For responders the median PFS was not reached, and 12-month PFS was 73%, compared to 1.4 months and 0% in non-responders.

By day 56 post-treatment, donor DNA was not detectable in the blood or marrow of any of the patients.

Cytokine profiles were performed for subjects in the dose escalation portion of the trial (n=9). Patients who achieved CR had significantly higher levels of IL6 on day +1 (p=0.01) and a trend towards higher levels of IL2, IL15, IFN-gamma, and TNF-alpha in the first 7 days after DLI compared to patients who did not achieve CR.

Conclusions: This completed trial of induction chemotherapy followed by CD4+ MMUDLI confirmed a MTD of 5x10⁷ CD4 cells/kg. At this dose, CR rates were over 50%, comparing favorably to historical 20-30% CR rates with chemotherapy alone in comparable patients including relapsed MDS or sAML, and those with TP53 mutations. We observed dose dependent improvements in response rates, which correlated with post-treatment cytokine levels. These encouraging results suggest further development of CD4+ MMUDLI for treatment of MDS and sAML.