Believe in haploidentical HSCT: less is better

In this issue of Blood, a prospective, multicenter study by Pulsipher et al demonstrates further validation of TCRαβ/CD19-depleted haploidentical hematopoietic stem cell transplantation (HSCT) showing ∼80% disease-free survival (DFS) at 2 years using killer immunoglobulin-like receptor (KIR) favorable donors for children with acute leukemia.¹ 

Back in 2010 in Rome, an exciting hypothesis was developed: improved patient outcomes with haploidentical transplantation could be obtained if we moved from a strategy of CD34⁺ selection to a negative depletion of αβT cells, thus removing the only cells capable of mediating graft-versus-host disease (GVHD). The Italian group pioneered the use of αβ T-cell/CD19 B-cell depletion (αβhaplo) for haploidentical HSCT and reported 90% DFS in children with life-threatening nonmalignant disorders.^2,Haplo is the new black, was the title Pulsipher used for his commentary³ to that paper.² The next step was to evaluate the efficacy of αβhaplo-HSCT in acute leukemia. In 80 patients with leukemia, with a median follow-up of 3 years, low rates of transplant-related mortality (TRM; 5%) and of relapse (26%)⁴ were documented. A year later, a retrospective multicenter analysis among 13 Italian centers compared 245 matched unrelated HSCT and 98 αβhaplo-HSCT and found the following: αβhaplo-HSCT and unrelated donors had comparable cumulative incidence of TRM and disease recurrence; αβhaplo-HSCT abrogated severe acute GVHD and led to faster neutrophil and platelet recovery; and αβhaplo-HSCT was superior to unmanipulated mismatched unrelated HSCT with a significantly lower TRM and improved chronic GVHD-free/relapse-free survival.⁵ 

However, further validation of the αβhaplo-HSCT approach was missing in a prospective, multicenter trial conducted on the other side of the ocean. Fast forward to 2022, and Pulsipher et al have elegantly and decisively addressed the issue of a prospective trial by investigating outcomes after αβhaplo-HSCT of 51 pediatric and young adult patients and comparing their outcome with a Center for International Blood and Marrow Transplant Research control group using other donor cell sources (see figure).¹ Although the present study confirms the low rates of acute and chronic GVHD and TRM reported by the European groups,^2-5 Pulsipher et al report, for the first time in the αβhaplo-HSCT setting, the superiority of a reduced toxicity (RTC) preparative regimen. In addition, they question the advantage of using a total body irradiation-based preparative regimen for patients with acute lymphoblastic leukemia. Finally, the KIR screening performed on donor/recipients’ pairs showed that 80% of patients have either a potential donor with a favorable ligand mismatch⁶ or a high B-cell content.⁷ However, in line with previous reports, using a haploidentical KIR favorable donor did not reduce the risk of relapse.

Engraftment is always the priority after HSCT, and graft rejection historically has represented a struggle in the haploidentical T-cell depleted setting. The results of the present study demonstrate that, using KIR-favorable TCRαβ/CD19-depleted haploidentical donors, it is possible to simultaneously diminish the risk of rejection and TRM. In the present study, all 4 rejection episodes (7.8%) occurred in the rATG/Flu/Thio/Mel RTC group. However, MRD⁻ patients, who received the same regimen and engrafted, had a 2-year DFS of 90%, which was significantly better than the 60% 2-year DFS observed in patients treated with a traditionally myeloablative approach. When analyzing the rate of graft rejection, it is important to note that the present study differs from the prior experiences by using ATG Thymoglobulin rather than ATLG Grafalon.^2-5 Thymoglobulin is derived from rabbit vaccination with human thymocytes. Grafalon is manufactured by immunizing rabbits with the human Jurkat leukemic T-cell line. Although closely related, there are significant differences in both the ligands and the pharmacokinetic profiles of ATG/ATLG, which could impact the effective depletion of residual host T cells, as well as the post-HSCT immune reconstitution. Thus, the use of ATG/ATLG may affect the likelihood of graft rejection, the risk of GVHD, and the risk of developing potentially fatal viral infections. Identifying the optimal timing and dose of ATG Thymoglobulin before αβhaplo-HSCT remains a major goal.

Pulsipher et al have confirmed that αβhaplo-HSCT is an ideal platform for cancer immunotherapy. However, a deeper understanding of the impact in variations in the inherent phenotypic characteristics of the donor, the graft composition, and the post-HSCT immune modulation are paramount for further improvements. The TCRαβ/CD19 depletion graft manipulation strategy transfers donor hematopoietic stem cells, committed hematopoietic progenitors (HSPC), and mature natural killer (NK) and γδ T cells, which overcome the limitations associated with the transplantation of CD34⁺ selected cells. Although the biology underlying NK alloreactivity and its impact in preventing leukemia relapse has been widely investigated, little is known about the role of donor γδ T cells. High-throughput sequencing of donor γδ T cells infused with the graft and longitudinally tracked post-HSCT may determine whether the γδ clonality, that occurs after HSCT, mirrors the γδ clonality of the original donor or whether unique clones develop after HSCT. In addition, it has been recently shown that the constitutive expression of NKG2A identifies a subset of Vδ2⁺ γδ T cells with an intrinsic hyperresponsiveness against cancer,⁸ whereas γδ T cells expressing FOXP3 might confer regulatory properties. Analyses of the HSPC composition of αβhaplo-HSCT and its relationship to clinical outcomes are also lacking. A study performed by Mantri et al has demonstrated that CD34⁺ absolute counts do not predict the HSPC content of cord blood.⁹ By correlating the HSPC graft composition data with patients' outcome, including leukemic relapse, we could identify the optimal αβhaplo-HSCT graft composition. Clarifying all these complex features is an ongoing area of research.

Following αβhaplo-HSCT, patients suffer suboptimal immune reconstitution, because of the ex vivo αβ T-cell depletion, and the pre-HSCT serotherapy. Thus, there is a need to expedite post-transplant immune reconstitution and ultimately improve the antileukemic efficacy. Specifically, identifying a combinatorial approach whereby various adaptive post-HSCT cell therapies with potential graft-versus-leukemia (GVL) effects, such as donor-derived CAR T cells,¹⁰ T-allo10 cells, or ex vivo expanded γδ T cells, are delivered after αβhaplo-HSCT, would boost post-HSCT immune reconstitution and increase GVL without increasing GVHD.

The intersection between HSCT and immunotherapy is probably the most fascinating area in the field. I predict that the combination of αβhaplo-HSCT with new cell therapies will dominate the next generation of pediatric clinical trials for cancer immunotherapy.

Conflict-of-interest disclosure: A.B. declares no competing financial interests.