Recent published experiences comparing allogeneic cell transplantation (HCT) using matched sibling, unrelated or haploidentical donors, showed similar results. Here, we analyzed the experience of GATMO in these types of transplants.
Objective: to evaluate the impact on overall survival (OS), non relapse mortality (NRM), relapse (Rel) and disease free survival (DFS) in HCT depending on the type of donor. Secondary objective was to evaluate early outcomes as GVHD, engraftment, CMV reactivation and fungal infections.
Materials and Methods: a retrospective analysis of 733 medical records was conducted, including patients who received an HCT in seven Argentinian centers performing the three types of transplants, from January 2012 to December 2016. Mean Age was 34 years (range 0-71 y), mean donors age 33 years (range 0-69 y); sixty percent of patients and donors were males. Among the whole group, 301 patients (41%) received a matched sibling HCT (Sib), 257 (35%) an unrelated donor HCT (UD) and 175 (24%) an haploidentical donor HCT (Haplo). Prevalent diseases were acute myeloid leukemia (AML) 34%, acute lymphoid leukemia (ALL) 25%, myelodisplastic syndromes (MDS) 13% and lymproprolipherative diseases (LPD) 10%. Hematopoietic cell transplantation-comorbidity index (HCT-CI) was low risk in 58%, and 62% of the patients presented with late stage disease. A myeloablative-conditioning regime (MAC) was used in 71% of the patients.
OS and DFS were tested with Kaplan Meier; NRM and Rel with cumulative incidence. Dichotomic variables were tested with Chi2. We conducted a multivariate analysis with Cox Regression.
Results: Median Follow-up was 1.6 years. There were no differences in age, sex, CMV pre transplant status, HCT-CI score, type of conditioning regime between the three groups, while there were significant differences in pre transplant disease stage (more late stages) (p=0.001) and incidence of AML (p=0.001) in the Haplo group. Donors were younger in the UD (p<0.001). Engraftment was slower in the Haplo group than in Sib and UD (≥15 days) (80% vs. 50% vs 48% respectively; p<0.001) and graft failure was higher in Haplo (15% vs. 4% and 5% respectively; p<0.001). CMV reactivation was also more frequent in the Haplo and UD setting (51% y 57% vs.28%; p<0.001). Acute Graft versus Host Disease (aGVHD) grade II-IV was more frequent in the UD group (41% UD vs. 27% Rel and 36% Haplo; p=0.004). Chronic Graft versus Host Disease and fungal infections incidence was similar between three groups.
Early mortality expressed as NRM at 100 days was 7.5% in Sib, 17% in UD and 16% in Haplo; 1 and 3 years NRM (1-3 y) was 14-16%, 25-28% y 26-28% respectively (p<0.0001). Relapse rate (1-3 y) was 28-38% in Sib, 18-25% in UD, 40-49% in Haplo (p<0.0001). Subsequently, Haplo HCT showed lower DFS (1-3 y 58-46% Sib vs. 56-47% UD vs. 34-23% Haplo; p<0.0001) and OS (1-3 y 70-57% Sib vs. 63-50% UD vs. 48-31% Haplo; p<0.001).
Multivariate Analysis (Hazard Ratios with 95% confidence interval) showed that a Sib HCT was associated with a better OS and DFS vs. Haplo (OS HR: 1.79, 1.22-2.61) (DFS HR 1.48, 1.05-2.07), and also with lower NRM vs. UD/Haplo (HR: 2.59, 1.69-3.98). UD HCT remained as an independent variable associated with fewer Rel vs. other type of donors (HR: 1.67, 1.10-2.81). Advanced stage of disease at transplantation as well as HCT-CI were the other factors associated with survival outcomes
We performed subgroup analysis for the four main diseases. For AML patients Haplo donors resulted in lower OS (1-3 years 46%-34% vs. 61%-51% Sib and 65%-51% UD p=0.01). Similar differences were observed in patients with ALL (2 years OS 28% Haplo, 58% Sib and 45% UD; p=0.02) and MDS (2 years OS 20% Haplo, 66% Rel and 58% UD, p<0.001). No significant differences were observed for LPD (2 years OS 59% Haplo, 50% Rel and 30% UD; p=0.44)
Conclusion: Compared to UD, Haplo HCT resulted in similar NRM and higher relapse, with significant worse OS in almost all the soubgroups analysed. Prospective trials are needed to establish the best strategy for patients requiring a HCT.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.