Haploidentical Transplant Using Selective Ex-vivo TCR-Αβ Depleted and Upfront Memory T-Cell Add-Back Is Associated with Lower Chronic Gvhd and More Favorable Gvhd-Free/Relapse-Free Survival As Compared to Matched Related and Unrelated Donor Transplant for Adults with Hematological Malignancies

Background

Haploidentical hematopoietic cell transplantation (HCT) with selective T-cell depletion (Haplo-TCD) and memory T-cell add-back has been adopted by our centers with good results. Using propensity score matching, we previously demonstrated that this approach produced better outcomes compared with post-transplant cyclophosphamide approach. [Victor LWT, et al. Blood (2022) 140 (Supplement 1):10500–10501]. Here, we compare the efficacy of Haplo-TCD with conventional donor sources, including matched-related donors (MRD), matched-unrelated donors (MUD), and mismatched unrelated donors (MMUD).

Methods

In this multicenter retrospective analysis, we evaluated the outcomes of 488 adult patients with hematological malignancies who underwent allogeneic HCT between January 2017 and December 2023. Our cohort included patients who received transplants from Haplo-TCD (n=160), MRD (n=160), MUD (n=143) and MMUD (n=25), following either myeloablative (n=182) or reduced-intensity conditioning (n=306) regimen.

Results

Median age of patients at the time of HCT was 52 years old (Range: 17-74). Patients in the MRD group (median age, 48) were younger than those in other groups. The most common diagnoses in all groups were acute myelogenous leukemia and myelodysplastic syndrome.

At a median follow-up of 34 months (Range: 5-90 months), the 3-year overall survival (OS) for Haplo-TCD, MRD, MUD, and MMUD HCT were 71.1%, 67.2%, 58.2%, and 60.6%, respectively (p=0.167). The corresponding 3-year event-free survival (EFS) were 61.9%, 58.1%, 46.5%, and 40.9%, respectively (p=0.019). In univariate analysis, the 3-year cumulative incidence (CI) relapse rate differed significantly between Haplo-TCD (20.2%) vs MUD (35.7%) (p=0.009) as well as vs MMUD (42.6%) (p=0.007) but was not statistically significant when compared to MRD (26.2%) (p=0.353). There was no statistically significant difference in CI of non-relapse mortality (NRM) at 3 years across all groups: Haplo-TCD 16.2%, MRD 15.7%, MUD 17.8%, and MMUD 16.5% (p=0.973).

The difference in CI of grade 2-4 acute GVHD (aGVHD) at day 180 was not statistically significant among the groups. However, CI of grade 3-4 aGVHD differed significantly (Haplo-TCD 6.9% vs MRD 14.4% vs MUD 4.9% vs MMUD 16.2%; p=0.0155). Remarkably, the 3-year chronic GVHD (cGVHD) incidence was the lowest in Haplo-TCD group (Haplo-TCD 8.1% vs MRD 56.3% vs MUD 27.6% vs MMUD 20.0%; p<0.001). Differences in relapse and cGVHD have resulted in the most favorable 3-year GVHD-free/Relapse-free survival (GRFS) in Haplo-TCD group (Haplo-TCD 54.4% vs MRD 10.9% vs MUD 26.5% vs MMUD 18.0%; p<0.001).

Patients with higher HCT comorbidity index (HCT-CI) ≥1 (p<0.001), high/very high disease risk index (p<0.001) and CMV seropositive status (p=0.003) were shown to have worse OS and EFS in multivariate analysis. Notably, Haplo-TCD was associated with significantly more favorable outcomes in EFS (p=0.03), cGVHD (p<0.001) and GRFS (p<0.001).

For patients with low/intermediate risk DRI, the 3-year OS were 80.3% for Haplo-TCD, 75.6% for MRD, 61.3% for MUD, and 79.0% for MMUD (p=0.026), with corresponding 3-year EFS of 72.8%, 66.6%, 51.0%, and 55.4% (p=0.069), and 3-year GRFS of 63.0%, 13.9%, 28.7%, and 30.0% (p<0.001). In patients with high/very high risk DRI, the 3-year OS were 53.3% for Haplo-TCD, 46.3% for MRD, 50.0% for MUD, and 33.3% for MMUD (p=0.66), with corresponding 3-year EFS of 45.5%, 36.8%, 23.1%, and 20.0% (p=0.078), and 3-year GRFS of 37.9%, 3.4%, 19.0%, and 0% (p<0.001).

Conclusions

Although non-randomized, our findings provide strong evidence that Haplo-TCD results in outcomes equivalent to those achieved with MRD and more favorable than those with MUD, particularly in terms of cGVHD and GRFS. This approach offers an additional alternative for patients lacking HLA-matched donors, while also obviating the cost and time associated with searching for unrelated donors.