The Rise of the HLA-Mismatched Unrelated Donor Transplant in the U.S.

I met Diane^* in Colorado soon after beginning my first hematology job in 2014. She was a 58-year-old who was referred for allogeneic hematopoietic cell transplantation (HCT) for high-risk myelodysplastic syndromes (MDS). She had no children or living siblings. Population data predicted that due to her Native American ancestry, Diane’s chance of finding a human leukocyte antigen (HLA)-matched unrelated transplant donor using the worldwide registries was relatively low. Sure enough, her donor search yielded only 7/8 HLA-mismatched donors, and I chose the best of the available options: bone marrow from a young ABO-matched male who harbored a single antigen mismatch in HLA-A.

Diane received myeloablative conditioning and graft-versus-host disease (GVHD) prophylaxis with tacrolimus and methotrexate. Shortly after engraftment, she developed acute GVHD of the skin, liver, and gastrointestinal (GI) tract. She was hospitalized for months, ultimately succumbing to steroid-refractory GI GVHD. Hers was my first encounter with non-relapse mortality as an attending physician. I regretted my decision to use an HLA-mismatched donor, aware of the literature associating HLA class I mismatches with higher rates of GVHD and death.¹  Amid Diane’s suffering, I vowed that I would never again willingly use an HLA-mismatched unrelated donor in my clinical practice.

As I pondered the advances in the field of allogeneic HCT for the “Year’s Best” lineup of articles in The Hematologist, I repeatedly returned to this story, recalling the many patients who lack HLA-matched donors due to their ancestral heritage and the negative consequences that often result from selecting an incompatible transplant donor. Therefore, as an homage to the past and a guiding light to the future, I chose an article that exemplifies one of the greatest developments of 2024: the rise of the “safe and effective” HLA-mismatched unrelated donor transplant in the United States.

In their article, Brian C. Shaffer, MD, and colleagues present observational research using data collected by the Center for International Blood and Marrow Transplant Research.²  The analysis cohort included more than 10,000 adults with acute leukemia or MDS transplanted between 2017 and 2021 with either HLA-matched (8/8) or HLA-mismatched (7/8) unrelated donors. Patients were additionally categorized as having received calcineurin inhibitor (CNI)-based GVHD prophylaxis (with other agents, including antithymocyte globulin [ATG]) or post-transplant cyclophosphamide (PTCy)-based GVHD prophylaxis. Multivariable models for transplant-related endpoints used HLA-matched unrelated donors with CNI-based GVHD prophylaxis as the reference group. Relative to this group, the use of HLA-mismatched unrelated donors with PTCy-based GVHD prophylaxis was associated with similar overall survival (OS) (hazard ratio [HR] = 0.9, p=0.2) and relapse (HR=1.1, p=0.2); superior GVHD-relapse-free survival (GRFS) (HR=0.7, p<.0001); and lower risk of non-relapse mortality (HR=0.8, p=0.03), grade II-IV GVHD (HR=0.7, p<0.0001), grade III-IV acute GVHD (HR=0.6, p=0.01), and moderate-severe chronic GVHD (HR=0.5, p<0.0001). Furthermore, use of PTCy-based GvHD prophylaxis was associated with similar three-year adjusted OS and GRFS between HLA-matched and HLA-mismatched donors. Taken together, real-world contemporary data from this large, primarily U.S. cohort shows the tremendous impact of PTCy on closing the gap between HLA-matched and single-loci HLA-mismatched unrelated donor transplantation.

However, in the same issue of the Journal of Clinical Oncology, Esteban Arrieta-Bolaños, DrMBCC, MSc, PhD, and colleagues published an analysis of HLA-mismatched unrelated donor transplantation on behalf of the European Society for Blood and Marrow Transplantation (EBMT), showing somewhat discordant results.³  Their cohort included 17,292 adults with hematologic malignancies who underwent HLA-matched or HLA-mismatched (one or two loci) unrelated donor transplants between 2005 and 2020 and were reported to the EBMT registry. Multivariable models demonstrated that the use of HLA-mismatched unrelated donors was associated with significantly inferior OS (HLA 9/10: HR=1.2, p<0.001; HLA 8/10: HR 1.3, p<0.001) and lower GRFS (HLA 9/10: HR=1.2, p<0.001; HLA 8/10: HR=1.2, p=0.007) relative to HLA-matched unrelated donors. Further, any positive impact of PTCy-based GVHD prophylaxis on outcomes was limited to the HLA-matched unrelated donor setting; PTCy did not appear to mitigate the inferior outcomes associated with HLA-mismatched donors.

An editorial accompanying both articles nicely summarizes possible explanations for the discrepancy between the U.S. and European data.⁴  Importantly, regional differences in leukemia and transplant practices, such as the widespread use of ATG in Europe, appear to limit extrapolation of findings across regions, which has also been observed in other recent transplant studies.⁵  Nevertheless, if the U.S. data hold up, and ongoing and upcoming clinical trials can further improve outcomes using unrelated donors with HLA mismatches, we can rest assured that almost every patient will have a suitable unrelated donor for allogeneic transplantation.⁶ 

In the decade since Diane’s death, both the field of transplantation and my career within it have evolved. I now lead a transplant and cell therapy clinic in Northern California focused on acute lymphoblastic leukemia. Most of my patients identify as Hispanic or Latino, and a sizable proportion are of Asian ancestry. I receive a “poor” donor search result (indicating a lack of HLA-matched donors) from our HLA team almost weekly. My reaction to these reports is now muted, nonchalant, or even enthusiastic. We proceed with transplant planning using the optimal haploidentical or, increasingly, HLA-mismatched unrelated donor. The moral dilemma I held onto for years — whether to offer a patient without other options an HLA-mismatched unrelated donor transplant — has essentially dissolved.