Selecting the Right Alternative Donor: Comparison of Outcomes of HLA-Mismatched Alternative Donor Hematopoietic Cell Transplantations in Adult Patients with Acute Lymphoblastic Leukemia Regarding KIR-Ligand Mismatch

Background: Allogeneic hematopoietic cell transplantation from cord blood (CBT) or haploidentical donor (HIDT) has proved to be a viable option for patients unable to find a human leukocyte antigen (HLA)-matched conventional donor. However, few single-centered studies have compared outcomes of the two alternative donor transplantations in adult acute lymphoblastic leukemia (ALL).

Aim: Our aim is to compare outcomes of adult ALL patients treated with CBT with those treated with HIDT. Secondly, we tried to find out factor that could be considered in selecting donor for alternative donor transplantation to improve clinical outcomes.

Methods: We compare outcomes of adult ALL patients treated with CBT (n=134) between 2008 and 2022 with those treated with HIDT (n=47) between 2018 and 2022.Conditioning regimen for CBT consisted of total body irradiation (12Gy) plus cytarabine (9g/m²) plus fludarabine (150mg/m²). Graft-versus-host disease (GVHD) prophylaxis was attempted using tacrolimus and mycophenolate mofetil. For HIDT, we used reduced toxicity conditioning (RTC) regimen including fludarabine (150mg/m²) plus busulfan (6.4mg/kg). GVHD prophylaxis was done with ATG 1.5mg for 4 days plus short course methotrexate (MTX) (15mg/m²) plus tacrolimus. Killer cell immunoglobulin-like receptor (KIR) ligand mismatch was calculated in both GVH and HVG direction using HLA-KIR ligand database, and its effect on clinical outcomes of alternative donor transplantation was investigated.

Results: In HIDT, CD34+ cell count (´10⁶/kg) was significantly higher (7.4 vs. 0.1, P<0.001), and the patients were older (49 vs. 33.5, P<0.001) compared to CBT. Other characteristics, such as cytogenetic risk and leukocyte count at diagnosis, did not show significant difference between the two groups. Incidence of acute GVHD II-IV (51.1% vs. 50.0%), III-IV (25.5% vs. 17.9%), and moderate to severe chronic GVHD (10.8% vs. 11.0%) between HIDT vs. CBT were similar. After median follow-up of 39.4 months (range: 5.5-167.4), HIDT showed higher cumulative incidence of relapse (CIR) compared to CBT (47.9% vs. 18.9%, P<0.001), which remained true even after stratifying by age subgroup (Age<40: P=0.033, Age≥40: P=0.002). Conversely, CBT showed higher non-relapse mortality (NRM) than HIDT (22.8% vs. 9.0%, P=0.038), but the same was true only in the older age subgroup (Age<40: P=0.484, Age≥40: P=0.002). These effects summed up to produce comparable results in estimated 3-year disease-free survival (DFS) of CBT and HIDT (58.4% vs. 43.5%, P=0.103), especially in the older subgroup (38.7% vs. 40.7%, P=0.897). Multivariate analysis also revealed that HIDT was related with higher relapse (HR 3.47; 95%CI 1.92-6.27, P<0.001), and CBT with higher NRM (HR 4.18; 95%CI 1.42-12.3, P=0.009).

GVH direction KIR ligand mismatching worked for lower 3-year DFS in CBT (40.3% vs. 74.4%, P=0.007), although CIR and NRM did not show significant difference (27.8% vs. 15.6%, P=0.178; 35.0% vs 18.4%, P=0.057, respectively). On the other hand, the mismatch resulted in lower incidence of acute GVHD II-IV (32.0% vs. 54.2%, P=0.011). The trend remained when such effect was investigated separately in CBT and HIDT, but it was not statistically significant (33.3% vs. 53.1%, P=0.082; 28.6% vs. 57.6%, P=0.053, respectively).

Conclusions: Our data shows higher NRM in CBT, and unexpected higher relapse rate in HIDT. Age of the patient should be considered when choosing CBT, because in older patients CBT showed notably worse NRM compared to HIDT. Higher incidence of relapse in HIDT may be attributable to ALL disease characteristics itself or insufficient elimination of leukemic cells by RTC regimen with stronger GVHD prophylaxis.

Similar to what is known in acute myeloid leukemia, GVH direction KIR incompatibility showed protective effect against acute GVHD in ALL. However, the mismatch resulted in worse DFS in ALL patients treated with CBT, but not in those treated with HIDT. Possible explanation to this phenomenon is that the number of NK cells in cord blood was insufficient for mismatch-induced graft-versus-leukemia effect, while its immune response against infected cells was suppressed by the same mechanism that resulted in lower acute GVHD.

Keywords: Acute lymphoblastic leukemia, Cord blood transplantation, Haploidentical donor transplantation, HLA-mismatched transplantation, Killer-cell immunoglobulin like receptor