Melphalan versus cyclophosphamide combined with fludarabine and low dose total body irradiation (TBI) prior to ptcy-based allogeneic transplantation

Abstract A preparative regimen of fludarabine, cyclophosphamide (Cy) and low dose TBI 200cGy (FC-TBI) followed by post-transplant Cy (PTCy) is one of the most widely utilized conditioning platforms for nonmyeloablative peripheral blood stem cell transplantation (PBSCT). It provides reliable engraftment in patients receiving HLA-mismatched transplants, manageable regimen-related toxicity and low non-relapse mortality (NRM), even in older and less fit patients. However, relapse represents a major source of treatment failure, particularly in patients with higher risk myeloid malignancies and those with measurable residual disease (MRD) pre-transplant. To help mitigate the relapse risk, we substituted melphalan 100mg/m2 for pre-transplant Cy in the fludarabine/TBI 200cGy backbone (FM-TBI) followed by PTCy, tacrolimus and MMF. Twenty-eight consecutive patients with hematologic malignancy receiving PBSCT following FM-TBI from June '21 to Dec '24 was compared to 127 consecutive patients receiving FC-TBI from Feb '13 to Dec '24. The donor type was Haplo, MUD and MRD in 91%, 8% and 1% respectively. Of the 155 total patients, median age was 63 years (range, 19-77), comorbidity index was 3 (range, 0-10), DRI was high/very high in 25%, and most common transplanted diseases were AML [47], MDS [31] and ALL [24], and were similar between FC-TBI and FM-TBI patients. In contrast, FM-TBI patients were statistically more likely to receive MRD/MUD (29% vs. 5%, p<0.001) and have younger donors (median age 29 vs. 41 years, p<0.001). Median follow-up was 13.5 months and 67 months for FM-TBI and FC-TBI patients, respectively. Compared to FC-TBI patients, FM-TBI resulted in slower ANC and PLT engraftment (19 vs. 17 days, p=0.07 and 33 vs 27 days, p=0.032) with similar incidence of graft rejection (0 vs. 4%, p=0.59). No differences were seen in the incidence of acute or chronic GVHD or CMV reactivation. FM-TBI was associated with a significantly higher 1-yr NRM (36% vs. 7%, p<0.001) and lower 1-yr survival (67% vs. 82%, p=0.029), while 1-year relapse incidence was reduced (7% vs. 23%, p=0.031). Furthermore, the number of hospital days through day +100 was significantly longer following FM-TBI (20 vs. 11 days). In multivariable analysis, controlled for diagnosis and DRI, the use of FM-TBI was associated with higher NRM (HR 4.24, p=0.002) and a trend towards inferior survival (HR 1.92, p=0.08). In the context of reduced intensity PBSCT in older transplant recipients, substitution of melphalan for pre-transplant Cy resulted in higher regimen-related toxicity, resulting in increased NRM and requirement for inpatient hospitalization. Despite lower relapse incidence with FM-TBI, 1-yr survival favored FC-TBI. The use of FM-TBI RIC PBSCT should be reserved for younger, healthier transplant recipients. Alternative strategies are required to reduce relapse incidence in older, less fit PBSCT recipients with higher risk disease.