EBMT 2020 | HSCT in AML: matched siblings versus haploidentical donors

The number of haploidentical transplant are actually booming or increasing rapidly, both in the Europe and in the States. The first comparison is that we believe, in the acute leukemia, work in part in also was done by CIBMTR. I mean, China compared unrelated transplant to haploidentical transplant. We show that the haploidentical transplant are not worse than nine out of 10, and mostly also 10 after 10 unrelated transplantation. With the invention of the post-transplant cyclophosphamide in the States, 70-80% of the haplo transplant in Europe are with post transplant cyclo or the Chinese approach with ATG. We dared to start comparing the haploidentical transplant to sibling transplantation because always sibling transplant, there was the gold standard for a transplantation, at least in patients with fair CR.

In a couple of papers, we compare the haploidentical transplant to sibling transplantation. The first one was published in Lancet Hematology this year in the more than a hundred thousand transplant in various hematological malignancies. Because the transplant related mortality was higher with the haplo mainly due to infection, at the end of the day is a result with the sibling transplant were better, but relapsed at least for intermediate risk and low risk hematological malignancies. According to the disease risk index, it was lower with the haplo transplant.

We then zoomed on AML. Again, it depends on the cytogenetic risk of AML. Results were somewhat different in high risk and intermediate risk male. For instance, in high risk male, there was lower relapsed risk with the haploidentical transplant. But overall, the sibling transplant were better.

We also hooked with the CIBMTR to compare elderly sibling to offspring haploidentical transplant, because there was some ideas that a younger offspring, a haploidentical transplant will be better than elderly sibling. There usually is a difference in ages, about 30 years. But we failed to show the advantage of the haploidentical transplant. In haploidentical transplant, both in the age of 18 to 54 and then 55 to 75, there was less chronic GVHD. But in the patient above the age of 54 with the offspring haplo, there were more mortality, non-engraftment, and transplant-related mortality. Therefore, in the end of the day, sibling transplantation were better.

We also compared in a relapsed refractory AML, a factory ML haplo versus sibling, and here again, the relapsed rate was lower with haplo, but the overall disease-free survival and the GRFS. GVHD free relapse, free survival were better with sibling. As for the anti-leukemia effect, there was some notion that because of the broad HLA disparities, a graft-versus-leukemia effect would be stronger with haplo, but we failed to show it in a leukemia paper published already in 2016.

Then, for second transplant in a recent paper, we compared haplo to the sibling or unrelated. The same donor, sibling or unrelated for the first transplant, or changing the donor to haplo in the second. Again, there was no stronger GVL effect. But when in May we published a comparison on behalf of the acute leukemia, between haplo in the sibling in unrelated with post-transplant cyclo. In here, the leukemia relapse was lower with haplo.

Overall, the result of haplo in haploidentical transplant are improving substantially. It’s still a second best in most of the cases to a sibling transplantation, but in some cases, the relapse rate is lower with the haploidentical transplant. The main reason for the superiority of the sibling is the high non-transplant related mortality with haplo, and mainly due to infection and maybe, or most probably, or it’s conceivable, that when we will know better how to reduce the infection rate in haploidentical transplant and improve the immune reconstitution, result will be the same with haploidentical transplant. But for a disease that need the strong GVL effect, maybe result will be better with haplo.

We’re still seeing that the graft versus leukemia effect is stronger with haploidentical transplant. The problem is that there is the issue of losing of haplotype lost that was discovered by the Sommerfeld group and especially professor Vargo. This is the mechanism of relapse in 24% of the haploidentical transplant while it’s only 11 or something percent in sibling transplantation. Because of this mechanism, we think that the end result is that not in all cases we see the stronger GVL or the less relapse of the haploidentical transplant.