Graft-versus-host disease (GVHD) remains a major cause of morbidity and mortality after allogeneic hematopoietic cell transplantation (allo-HCT). Optimizing GVHD prevention strategies remains a key concern to improve patient survival and quality of life.

In Europe, anti-thymocyte globulin (ATG) is widely used as GVHD prophylaxis. However, post-transplant cyclophosphamide (PTCy), initially introduced in the haploidentical transplant setting, has more recently been extended to mismatched unrelated donors (MMUD). In matched related (MRD) and matched unrelated (MUD) allo-HCT, studies comparing PTCy to the traditional ATG strategy have shown controversial results which do not allow a specific strategy to be recommended. Here, we report a representative analysis based on large national data to complement existing evidence.

We retrospectively analyzed patients from the SFGM-TC registry who underwent a first allo-HCT from MRD or MUD for acute myeloid leukemia (AML) or myelodysplastic syndromes (MDS) between 2014 and 2022. Patients received GVHD prophylaxis with: immunosuppressive agents (IS) alone (group NONE), PTCy with or without (+/-) IS (group PTCy), ATG +/- IS (group ATG), PTCy+ATG +/- IS (group PTCy+ATG). We restricted our analysis specifically to allo-HCT with reduced intensity conditioning (RIC) and peripheral blood stem cells (PBSC). IS regimens included mainly calcineurin inhibitor (CNI) + antimetabolite (methotrexate or mycophenolate mofetil), or CNI alone. RIC regimens consisted in: Fludarabine+Busulfan, Thiotepa-Busulfan-Fludarabine (TBF), total body irradiation (TBI) based regimen, Baltimore like regimens, and other regimens.

A total of 3702 patients were analyzed: group PTCy n=103, group ATG n=3340, group PTCy+ATG n=46 and group NONE n=213. The groups were comparable in age at allo-HCT, gender, disease type, remission status at transplant and HCT-CI score but differed significantly in transplant year, donor type, blood group compatibility, RIC regimen and IS regimen. Of note, in the PTCy and PTCy+ATG groups, 14% of allo-HCT were performed without additional IS, compared to 0% in the other groups. In the NONE group, 65% of allo-HCT used MRD donors versus 45% in the remaining groups. These variables (except transplant year) were included as adjustment factors in multivariable analysis.

At day 100, the cumulative incidence of grade ≥ II acute GVHD (aGVHD) was 26%, 29%, 22% and 24% (p=0.169) and grade III-IV aGVHD was 8%, 10%, 7% and 6% (p= 0.333) in PTCy, ATG, NONE and PTCy+ATG groups respectively. At 2 years, the cumulative incidence of moderate-severe/extensive chronic GVHD (mod-ext cGVHD) was 14%, 19%, 28% and 26% in PTCy, ATG, NONE and PTCy+ATG groups respectively (p=0,012). In adjusted multivariable analysis, PTCy was associated with a significant reduction in grade III-IV aGVHD compared to ATG (HR 0.46, p=0.039). Regarding mod-ext cGVHD, PTCy and ATG were both significantly superior to NONE but PTCy seemed to be superior to ATG in lowering mod-ext cGVHD risk (HR 0.61, p=0.088). In addition, PTCy+ATG was not significantly different from other groups but the statistical power was limited by the small number of patients.

The cumulative incidence of relapse (RI) at 2 year was 35%, 36%, 28% and 33% was in PTCy, ATG, NONE and PTCy+ATG groups respectively and not significantly different between groups. In multivariable analysis, PTCy group experienced a significantly lower non-relapse mortality (NRM) compared to NONE (HR 0.45, p=0.036) but did not differ with ATG group (HR 0.60, p=0.150). An analysis of toxicity patterns and causes of death will be presented.

With a median follow-up of 17.6 months for the overall population, the 2-year OS, PFS and GRFS was 59.8% [58.1 - 61.5], 53.8% [52.1 - 55.6] and 42.3% [40.7 - 44.1] respectively. These outcomes were not significantly influenced by donor type (MRD and MUD) or by GVHD prophylaxis group in multivariable analysis.

In conclusion, in this large retrospective cohort of MRD/MUD and RIC allo-HCT in AML/MDS, PTCy significantly reduced severe aGVHD compared to ATG, that might contribute to lower the NRM. However, no significant differences were observed between PTCy and ATG regarding cGVHD, RI, OS, PFS or GRFS. These results are consistent with previous studies explaining the overall trend toward a democratization of PTCy, although a robust international prospective randomized study would be necessary before changing current practices.

This content is only available as a PDF.
2025
Sign in via your Institution