In this issue of Haematologica, Salvatore and colleagues compare outcomes after transplantation of grafts from a haploidentical relative or a HLA-matched sibling for adults with acute myeloid leukemia in first complete remission.1 They conclude survival was inferior after transplantation of peripheral blood or bone marrow from a haploidentical relative compared to an HLA-matched sibling for patients with intermediate-risk cytogenetics. On the other hand, for patients with high-risk cytogenetics, survival did not differ by donor type despite a non-significant reduction in relapse risk after haploidentical transplant. Non-relapse mortality risks were higher after haploidentical transplantation negated any survival advantage to be expected with the modest reduction in relapse after haploidentical transplantation for AML with high-risk cytogenetics. The observation of a reduction in relapse risk, albeit non-significant, after haploidentical transplantation is intriguing. While it is tempting to attribute this to an enhanced graft-versus-host leukemia effect in the setting of an HLA-mismatched transplant, the reduction in relapse risk was only seen for patients with high-risk cytogenetics. Can this be explained by differences in transplant conditioning regimen intensity? The study population received both myeloablative and reduced intensity transplant conditioning regimens. Among patients with intermediate risk cytogenetics, reduced intensity conditioning was associated with higher relapse. Yet, in the group of patients with high-risk cytogenetics, relapse risks did not differ by transplant conditioning regimen intensity leading us to conclude this merits further investigation.
These data raise a fundamental question: when should we select an HLA-mismatched relative instead of an HLA-matched sibling? If an HLA-matched sibling is medically unfit or unwilling to donate, an HLA-mismatched relative could be the obvious choice for a number of reasons including, but not limited to, the ease of availability of the donor and timing of transplantation. Yet, when an HLA-matched sibling is medically fit and willing to donate are there circumstances that warrant selection of a haploidentical relative? A recent joint report from the European Society for Blood and Marrow Transplant and the Center for International Blood and Marrow Transplant explored whether post-transplant cyclophosphamide can nullify the detrimental effect of HLA mismatch for acute myeloid and lymphoblastic leukemia.2 The report showed haploidentical siblings donated to adult patients younger than 55 years and offspring donated to those 55 years and older. After adjusting for risk factors associated with survival the study concluded an HLA-matched sibling was a better choice than an offspring in patients 55 years and older. In the group with patients aged 18–54 years, a comparison of haploidentical to HLA-matched sibling transplant did not reveal differences in survival. The characteristics of the patients studied in two reports and their numbers differ21 and this is the most likely explanation for the differences between the two reports. As the report by Salvatore and colleagues did not consider donor-recipient relationship, we do not know whether the effect of cytogenetic risk on survival may be explained by donor-recipient relationship and patient age on survival. However, both these reports present more questions in regards to donor selection.
Donor age is associated with survival after unrelated donor transplantation.3 Survival is better after transplantation of grafts from younger donors after adjustment for donor-recipient HLA-match. Donor age is challenging to study in the setting of HLA-matched sibling transplants as generally the age of siblings falls within the same decade. Others have compared transplantation of grafts from a young unrelated donor and an HLA-matched sibling in older adults with hematologic malignancy and confirm there is no survival advantage when a young unrelated donor is chosen in favor of an older HLA-matched sibling.4 So, is there a potential advantage to selecting an offspring who is likely to be about 2–3 decades younger than the parent? The effects of donor age on adults with hematologic malignancy undergoing haploidentical transplantation has been studied by others.5 In that report, the age of the patient (≥55 years) rather than the age of the donor was associated with higher mortality.5 The study did not identify any donor factors that were associated with mortality.5 It is worth noting that the numbers of haploidentical transplants available for study are modest when compared to the numbers of HLA-matched sibling and unrelated donor transplants. Therefore, with the increasing numbers of haploidentical transplants performed, it is incumbent upon the community of transplant physicians to carefully evaluate the effects of characteristics of haploidentical donors on transplant outcomes.
Lastly, how can we best study donor selection for hematopoietic cell transplant? There is general agreement that when treatment options are being studied, a randomized trial is the gold standard. Planning and executing randomized trials is more easily said than done. In the context of related donor transplantation, subjects must have an HLA-matched sibling and a haploidentical relative for randomization. This in itself is limiting, as several more subjects will have a suitable haploidentical relative rather than an HLA-matched sibling. Secondly, we do not know whether there are differences amongst the haploidentical relatives and should randomization consider donor-recipient relationship. Thirdly, are physicians willing to randomize patients with an HLA-matched sibling to receive a haploidentical relative? While some may not, others may find this unacceptable. Regardless of the complexities of conducting randomized trials there is no denial in the lengthy duration of these trials and the expense incurred. Hence, there is reliance on data collected by large transplant registries to better understand the effect of donor types on transplant outcomes. In the meantime, the report by Salvatore and colleagues compels us to select an HLA-matched sibling when such a donor is available. A haploidentical relative is a suitable alternative when an HLA-matched sibling is not available.
References
- Salvatore D, Labopin M, Ruggeri A. Outcomes of hematopoietic stem cell transplantation from unmanipulated haploidentical versus matched sibling donor in patients with acute myeloid leukemia in first complete remission with intermediate or high-risk cytogenetics: a study from the Acute Leukemia Working Party of the European Society for Blood and Marrow Transplantation. Haematologica. 2018; 103(8):1317-1328. PubMedhttps://doi.org/10.3324/haematol.2018.189258Google Scholar
- Robinson TM, Fuchs EJ, Zhang MJ. Related donor transplants: has posttransplant cyclophosphamide nullified the detrimental effect of HLA mismatch?. Blood Adv. 2018; 2(11):1180-1186. PubMedhttps://doi.org/10.1182/bloodadvances.2018018291Google Scholar
- Kollman C, Spellman SR, Zhang MJ. The effect of donor characteristics on survival after unrelated donor transplantation for hematologic malignancy. Blood. 2016; 127(2):260-267. PubMedhttps://doi.org/10.1182/blood-2015-08-663823Google Scholar
- Alousi AM, Le-Rademacher J, Saliba RM. Who is the better donor older hematopoietic transplant recipients: an older-aged sibling or a young, matched unrelated volunteer?. Blood. 2013; 121(13):2567-2573. PubMedhttps://doi.org/10.1182/blood-2012-08-453860Google Scholar
- McCurdy SR, Zhang MJ, St. Martin A. Effect of donor characteristics on haploidentical transplantation with posttransplant cyclophosphamide. Blood Adv. 2018; 2(3):299-307. PubMedhttps://doi.org/10.1182/bloodadvances.2017014829Google Scholar