Impact of Cytoreductive Therapy and Depth of Response before Allogeneic Transplantation for Advanced Myelodysplastic Syndrome

Introduction

Cytoreduction either with hypomethylating agents (HMA) or induction chemotherapy (IC) before hematopoietic cell transplantation (HCT) for patients with advanced myelodysplastic syndromes (MDS) has been used in an attempt to decrease post-HCT relapse, but the benefit for post-HCT long-term survival remains a debatable issue. We investigated the impact of previous therapy, and depth of response, as well as donor source on the outcome of allo-HCT for patients with advanced MDS.

Method

After excluding patients who had chronic myelomonocytic leukemia, we analyzed 303 advanced MDS patients (96 (32%) with RAEB-I, 148 (49%) with RAEB-2, and 59 (19%) with tAML) who underwent transplantation from a matched related/unrelated (MSD/URD, n=75/12) or haploidentical (n= 216) donor (HID) after preparation with myeloablative conditioning regimens (identical with BUCY for MSD and BUCY+ATG for HID) between year 2015 and 2018.

Results

Baseline characteristics

Median age was 42 years old. Of the 303, 142 (47%) received only best supportive care (BSC) before HCT; 18 (6%) received IC, 54 (18%) received HMA, and 89 (29%) received both (IC+HMA). Before therapy, 45 (15%), 141 (46%) and 117 (39%) patients had intermediate 1,2 or high-risk scores per the International Prognostic Scoring System (IPSS). Thirty-six (12%) achieved complete remission (CR), 81 (27%) had marrow CR, 160 (53%) had stable disease, and 26 (9%) had progressive disease before HCT defined by International Working Group (IWG). Patients in CR or marrow CR were grouped together as responders (OR) while patients not in CR were grouped together as non-responders (NR) at HCT. Patients had more high-risk disease by WHO and IPSS in the IC and IC+HMA groups. Other patient and donor characteristics known to affect outcomes such as age, time from diagnosis to HCT, IPSS-R, donor source are comparable between groups.

Outcome for the entire cohort

CR rate at HCT was 6%, 33%, and 30% for HMA, IC, and IC+HMA groups, respectively (P =.001) while OR rate was 37%, 56%, and 70%, respectively (P =.001). With a median follow-up of 727 days, 2-year disease-free survival (DFS) was 77%, 69%,60%, and 62% for BSC, HMA, IC, and IC+HMA groups, respectively (P = .075); 2-y DFS was 77%, 72%,and 56% for untreated, OR, and NR groups, respectively (P = .026); 2-y DFS was 77%, 78%,and 60% for untreated, CR, and non-CR groups, respectively (P = .008); 2-y DFS was 80% and 57% for MSD/URD and HID groups, respectively (P = .051). Multivariate analyses revealed that older age (hazard ratio [HR], 1.03; P< .0001); higher-risk histologic subtypes (HR, 2.3; P =.003), time from diagnosis to HCT (HR,1.005; P =.027), haploidentical donor (HR, 1.9; P =.012) and NR at HCT (HR, 1.6; P =.05) were poor prognostic factors for DFS.

Outcome for patients with EB2 and t-AML

Focusing on the 207 patients who had BM blasts of 10% or higher before therapy, CR rate at HCT was 6%, 35%, and 33% for HMA, IC, and IC+HMA groups, respectively (P =.007) while OR rate was 44%, 59%, and 74%, respectively (P =.001); 2-year DFS was 77%, 68%,58%, and 62% for BSC, HMA, IC, and IC+HMA groups, respectively (P = .20); 2-y DFS was 77%, 71%,and 50% for untreated, OR, and NR groups, respectively (P = .025); 2-y DFS was 77%, 77%,and 58% for untreated, CR, and non-CR groups, respectively (P = .015); 2-y DFS was 75% and 64% for MSD/URD and HID groups, respectively (P = .43).

Propensity score analysis

To allow for potential confounding factors between treatments that could influence outcome, a 1:1 ratio propensity score matching was also performed. Included in the propensity score model were: WHO stage, IPSS score, age, time from diagnosis to HCT, and donor source. We were able to pair-match 104 untreated with 104 cytoreductive and 2-y DFS were 80% vs 68% (p= .14); 2-y DFS was 80%, 76%,and 63% for untreated, OR, and NR groups, respectively (P = .19); 2-y DFS was 80%, 94%,and 64% for untreated, CR, and non-CR groups, respectively (P = .015); 2-y DFS was 85% and 68% for MSD/URD and HID groups (P = .044) and the rate was 82% vs 70% after excluding EB-1 (p= .30). After excluding IC group (n=8), 2-year DFS was 80%, 67%, and 75% for BSC, HMA, and IC+HMA groups, respectively (P = .28).

Conclusion

In this analysis, various therapy approaches before HCT did not lead to different transplantation outcomes. There was no evidence of a benefit in post-HCT outcome associated with prior cytoreductive therapy for advanced MDS. Further randomized studies need to delineate the role of debulking strategy.