IntroductionAllogeneic hematopoietic cell transplantation (HCT) is a potentially curative therapy for several malignant hematologic diseases. As only 30% of patients will have a matched related donor, alternative donors play a significant role in HCT. Luznik et al pioneered the use of posttransplant cyclophosphamide (PTCy) based GVHD prophylaxis in haploidentical transplantation (haplo-HCT), which greatly expanded the pool of donors and nearly revolutionized the field. Despite the increasing use of haplo-HCT with PTCy, some questions remain, namely the selection of the best donor, graft source and conditioning regimen, and risk factors for adverse outcomes.MethodsRetrospective study conducted at two Brazilian centers. All patients with hematologic malignancies who underwent first HCT between 2010 and 2021, haploidentical with posttransplant cyclophosphamide, were included. The objective was to identify risk factors for adverse outcomes following haploidentical HCT with PTCy. Independent variables are detailed in table 1. Acute GVHD was graded according to the MAGIC criteria and chronic GVHD was diagnosed according to the NIH consensus criteria. Survival and cumulative incidence curves were built with the Kaplan-Meier and Gray methods, and compared with the logrank and Gray tests, respectively. Multivariate analyses were carried out with Cox models. GVHD, as an independent variable, was included as a time-dependent covariate.ResultsA total of 103 patients were included (table 1). In brief, median age was 39 y/o and most patients were male (58%); 71 patients had low or intermediate disease risk index, while 29 had high or very high-risk disease. Bone marrow was the preferred stem-cell (72%, followed by peripheral blood stem-cells - 28%) and reduced-intensity, the most frequent conditioning regimen (43%). Median follow-up for the whole cohort was 2.6 years.Overall survival at 2 years was 51.7% (95CI 42.4-63.0%, figure 1). Multivariable analyses are in table 2. Risk factors for death were age at transplant (HR = 1.03 for each year; 95CI 1.01-1.05; p = 0.002), high or very high disease risk index (HR = 2.73; 95CI 1.52-4.90; p = 0.0008), and compared with low or intermediate, and mother as the donor (HR = 3.50; 95CI 1.44-8.47; p = 0.006).Progression-free survival at 2 years was 45.8% (95CI 36.6-57.2%). In multivariate analysis (table 2), progression-free survival was significantly poorer for older patients (HR = 1.02; 95CI 1.01-1.04; p = 0.009), high or very high disease risk index (HR = 2.29; 95CI 1.30-4.04; p = 0.004), compared with low or intermediate, and mother as a donor (HR = 3.15; 95CI 1.37-7.22; p = 0.007; figure 2). Two-year relapse rate was 22.2% (95CI 15.2-32.5%). Risk factors for relapse (table 2) were high or very high disease risk index (HR = 3.55; 95CI 1.44-8.74; p = 0.006), compared with low or intermediate, and mother as the donor (HR = 2.85; 95CI 1.12-7.25; p = 0.007). Two-year non-relapse mortality was 32.0% (23.9-42.9%). The only risk factor we found was age at transplantation (HR = 1.03 for each year; 95CI 1.01-1.05; p = 0.02; table 2). We found no effect of GVHD, acute or chronic, on relapse, nor on non-relapse mortality.Rates of grades II-IV and III-IV acute GVHD and 2y-chronic GVHD were 30.4% (95CI 22.7-40.8%), 6.9% (95CI 3.4-14.1%) and 19.8% (95CI 13.2-29.7%), respectively. The only risk factor (table 2) identified for grades II-IV acute GVHD was tacrolimus (HR = 0.36; 95CI 0.14-0.95; p = 0.04) compared with cyclosporine. We have not carried out multivariate analysis for grades III-IV acute GVHD due to the low number of events (only 7). In multivariable analysis (table 2), peripheral blood graft was a risk factor for chronic GVHD (HR = 3.72; 95CI 1.53-9.01; p = 0.004; figure 3), compared with bone marrow, while tacrolimus was protective (HR = 0.27; 95CI 0.11-0.68; p = 0.005), compared with cyclosporine.ConclusionOur results show that mother as the donor was an important risk factor for poorer OS, PFS and relapse, and mothers might not the best choice of donor. Tacrolimus was protective for both grades II-IV aGVHD and for cGVHD, suggesting that tacrolimus may be more effective than cyclosporine in preventing GVHD. PBSC was a risk factor for cGVHD without any impact on relapse, and this result does not support the systematic use of PBSC in detriment of BM. As expected, age at transplant negatively impacted OS, PFS and NRM as well as high/very high DRI led to poorer OS, PFS and relapse. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.