Impact of Pre-Transplant Molecular and Cytogenetic Remission on Outcomes of Allogeneic Stem Cell Transplant in Patients with Myelodysplastic Syndrome

Background: Allogeneic stem cell transplant (allo-SCT) carries a high relapse rate (~35% within 2 years) in patients (pts) with myelodysplastic syndrome (MDS). Pre-transplant measurable residual disease (MRD) increases the risk of relapse and death in acute myeloid leukemia (Thol et al, Blood 2018). However, the impact of pre-transplant MRD on post-transplant outcomes is less clear in MDS. Prior studies have compared outcomes between myeloablative (MAC) and reduced intensity conditioning (RIC) in pts with pre-transplant MRD persistence based on either cytogenetic remission alone or ultra-deep genomic sequencing with a limited 10-gene panel, suggesting reduced risk of relapse in pts receiving MAC (Festuccia et al, Biol Blood Marrow Transplant 2016; Dillon et al, JCO PO 2021). However, pre-transplant MRD assessment using a commercial next-generation sequencing (NGS) panel in addition to cytogenetic testing to assess post-transplant outcomes has not been evaluated in MDS.

Methods: We conducted a multicenter retrospective review of MDS pts who underwent allo-SCT from 2015-2022. Pts with FISH/cytogenetic (FC) and/or molecular abnormalities at diagnosis and MRD assessment by commercial myeloid-panel NGS (with a variant-allele frequency detection limit up to 5%) and FC performed on peripheral blood (PB) or bone marrow (BM) aspirate within 3 months prior to transplant were included for analysis. The Kaplan-Meier method and log-rank tests were used to estimate overall survival (OS). Gray's test/cumulative incidence functions were used to estimate relapse-free survival (RFS) with non-relapse mortality (NRM) as competing risk.

Results: Table 1 includes baseline pt characteristics. Of 81 pts, 69 pts (85.2%) had residual FC and/or molecular abnormalities (MRD+) at pre-transplant PB/BM assessment. Twelve (14.8%) pts were in both FC and molecular remission (MRD-) prior to transplant. With a median follow-up of 13.9 months [IQR 6.8 - 38.9], 43 (53%) died and 24 (30%) pts experienced a relapse. Twenty-three of 69 (33%) MRD+ pts had relapsed disease with a 38% cumulative incidence of relapse (CIR) at 2 years. In contrast, only 1/12 (8.3%) MRD- pts experienced a relapse after 7 years, with chromosomal loss of TP53 not present at initial diagnosis. CIR at 2 years was 33% and 41% (p=0.207) for MRD+ pts who received MAC vs RIC, respectively with NRM of 34.5% associated with MAC vs 28.2% for RIC (p=0.61).

TP53 mutations were most frequently associated with relapse in 8/24 (33%) pts with a median OS of 9.8 months [8.2 - NR]. Besides the previously defined 10-gene panel (Dillon et al, JCO PO 2021), relapses were also driven by SRSF2, U2AF1, KRAS, SETBP1 and NF1 mutations present in the original clone. Six pts had DNMT3A, TET2 and/or ASXL1 (DTA) mutations only on pre-transplant NGS, with 2/6 pts relapsing without the re-emergence of DTA mutations.

Median OS was not reached for MRD- pts and was 14.1 months [11.35-NR] for MRD+ pts, p = 0.061 [Fig.1]. OS probability at 2 years was 35% for MRD+ pts vs 75% for MRD- pts. Type of conditioning regimen, RIC vs MAC, did not impact OS (p= 0.6). An equal number of deaths, 35% each, were attributable to GvHD and relapse, respectively. Thirteen (30%) pts died from other causes, primarily infections and bleeding complications. NRM was 25% and 36% for MRD- and MRD+ cohorts, respectively.

Pts received a median of 6 cycles [range, 2-36] of hypomethylating agent (HMA) +/- venetoclax prior to transplant with no association between the number of cycles and MRD status (p=0.80). The incidence of grade 3-4 acute GvHD and chronic GvHD requiring systemic therapy did not significantly differ, at 25% and 18.8% (p= 0.696) and 33.3% and 31.8% (p= 1.00) in MRD+ and MRD- pts, respectively.

Conclusions: MRD assessment using commercially available NGS panels and cytogenetic testing that are routinely performed in MDS pts in real-world practice can predict risk of relapse. Pre-transplant MRD is associated with worse OS irrespective of the intensity of conditioning regimen. Although limited in statistical power due to a small sample size of MRD- pts in our cohort, further studies with a larger cohort are underway to better clarify the role of MRD on transplant outcomes in MDS. Our data also highlights that only a small minority of MDS pts (<15%) achieve complete molecular and cytogenetic remission at the time of transplant; strategies to eliminate the pre-transplant MDS clone are urgently needed.