Outcomes of allogeneic transplant in patients with DDX41 mutated myelodysplastic syndrome and acute myeloid leukemia

The DEAD-box helicase 41 ( DDX41 ) gene, located at 5q35.3 locus [1], is involved in interactions with spliceosome proteins and development of innate immune response [2, 3]. DDX41 -mutated acute myeloid leukemia (AML) and higher-grade myelodysplastic syndrome (MDS) are reported to have favorable outcomes [4 – 6]. In this study, we describe outcomes of patients with DDX41 -mutated MDS/AML undergoing allogeneic stem cell transplant (alloSCT). We retrospectively reviewed patients with DDX41 -mutated, WHO de ﬁ ned, MDS or AML [7]. The study was approved by Mayo Clinic Institutional Review Board. Patients with MDS/AML, found to have a DDX41 mutation on Next Generation Sequencing performed on peripheral blood or bone marrow aspirate, per institutional policy, were included. Germline testing was not performed. We reviewed the Mayo Clinic electronic medical records to determine patient demographics, transplant characteristics, and post-transplant outcomes. Disease risk was determined using the revised International Prognostic Scoring System for MDS and the European Leukemia Network 2017 risk strati ﬁ cation for AML [8, 9].

We retrospectively reviewed patients with DDX41-mutated, WHO defined, MDS or AML [7]. The study was approved by Mayo Clinic Institutional Review Board. Patients with MDS/AML, found to have a DDX41 mutation on Next Generation Sequencing performed on peripheral blood or bone marrow aspirate, per institutional policy, were included. Germline testing was not performed. We reviewed the Mayo Clinic electronic medical records to determine patient demographics, transplant characteristics, and post-transplant outcomes. Disease risk was determined using the revised International Prognostic Scoring System for MDS and the European Leukemia Network 2017 risk stratification for AML [8,9].
Primary objective was to assess the effect of alloSCT on overall survival (OS). Secondary objectives included assessment of cumulative incidence of relapse and non-relapse mortality (NRM) post-transplant.
Patient and transplant characteristics were summarized using descriptive statistics. Kaplan-Meier and log-rank tests were used to estimate OS and compare time to disease progression/relapse. We used Cox proportional hazards for time dependent variable to determine the effect of alloSCT on OS [10]. Post-transplant NRM was calculated using competing risk analysis. R 4. Among patients undergoing alloSCT, the median time to disease progression before alloSCT was similar to those who did not undergo alloSCT (3.1 vs. 2.3 years, P = 0.96), suggesting that patients in both alloSCT and non-alloSCT groups had comparable rates of disease progression.
Among patients with MDS, one (16.7%) had high-risk disease, four had intermediate-risk (66.7%) and one patient (16.7%) had low-risk disease. One patient with intermediate-risk MDS had transformed to AML before alloSCT. Four (30.8%) patients had persistent disease at the time of transplant. Nine (69.2%) patients were in complete morphological remission; two of whom (15.4%) had positive minimal residual disease testing. Of the 13 patients, six (46.15%) had received a second line of therapy before proceeding to alloSCT. Median time from diagnosis to alloSCT was 8 months (range 2-54 months), with a median follow-up after alloSCT of 10 months (range 2.7-28 months).
Median OS since diagnosis for patients undergoing alloSCT was 3.2 years. There was a trend towards inferior survival among patients undergoing alloSCT compared to those who did not have alloSCT (median OS 3.2 vs. 11.4 years, P = 0.06) (Fig. 1b). Overall survival did not differ among patients with a DDX41 VAF < 40% compared with VAF ≥ 40% (median OS 11.4 years vs. NA, P = 0.66).
Our cohort had a predominantly male predisposition and a later age of onset in the late sixties. These findings are consistent with those reported in previous studies [6]. Most of the patients with DDX41-mutated MDS were classified as MDS-EB. Studies have shown DDX41-mutated MDS-EB to have favorable outcomes [4,5].
Our study shows that patients with DDX41-mutated MDS/AML undergoing alloSCT had a trend towards lower survival. Although this finding did not reach statistical significance due to small sample size, patients undergoing alloSCT had a high rate of nonrelapse mortality (1-year NRM 44.8%) despite low HCT-CI of 0-2 in most patients.
The post-transplant mortality was most commonly due to infections, highlighting the need to use lower intensity conditioning regimens and optimize infectious disease prophylaxis in these patients.
Due to the retrospective nature of our study, the comparison between alloSCT and non-alloSCT cohort is not without bias. However, patients undergoing alloSCT usually have better performance status and lesser comorbidities compared to patients who do not undergo alloSCT. Therefore, the inferior survival of the alloSCT cohort in our study is likely related to the alloSCT itself. It was recently shown that alloSCT did not improve survival in DDX41-mutated AML [12]. All five patients in our cohort who died after alloSCT were in complete remission before transplant and died most commonly of infections (Fig. 1a). While the majority of patients in our study underwent alloSCT due to high-risk disease, a subset of patients was deemed to have intermediate-risk disease, 40% of whom had non-relapse mortality, suggesting that alloSCT can potentially be deferred in these patients until disease progression or relapse.
In conclusion, this is the first study assessing post-transplant outcomes in patients with DDX41-mutated MDS/AML. Our study highlights a high post-transplant NRM in these patients, suggesting that alloSCT might need to be reserved for disease progression or relapse.