Abstract
Introduction: Allogeneic stem cell transplantation (SCT) is the only known curative treatment in higher-risk (HR) myelodysplastic syndromes (MDS). We sought to evaluate the impact of pre-SCT factors on post SCT outcomes in HR-MDS. This knowledge has the potential to optimize the use of SCT in HR-MDS.
Methods: This was a single center retrospective study. We identified patients (pts) presenting with newly diagnosed HR (IPSS-R > 3.5 points) MDS (WHO 2016) between 2000-2023. Pts who underwent SCT were identified. Responses were evaluated using the IWG 2023 criteria. Myeloablative conditioning (MAC) was defined as busulfan AUC ≥20,000 μmol/min or ≥16,000 μmol/min with thiotepa. Time-to-event endpoints were counted from SCT D0. Cumulative incidence (CI) of post-SCT relapse and treatment-related mortality (TRM) were evaluated as competing risks (Fine-Gray). Multivariate analysis (MVA) was performed using a Cox proportional hazards regression.
Results: 2045 pts with HR-MDS were identified, of which 427 (21%) underwent SCT. In transplanted pts, the median age was 59 years (range 18-76) and 269 (63%) were male. IPSS-R risk was intermediate in 108 (25%), high in 139 (33%), and very high in 180 (42%). Complex cytogenetics were present in 171 (40%) and TP53 mutations in 82/240 (34%) (20/240 [8%] monoallelic and 62/240 [26%] biallelic). MAC conditioning was performed in 255 (60%) and RIC in 172 (40%). Post-transplant cyclophosphamide (PTCy) was administered in 179 (42%).
At the time of SCT, 65/427 (15%) had transformed to acute myeloid leukemia (AML) and were excluded from further analysis. In 362 pts who underwent SCT as MDS, the median follow-up time was 60.6 months (m) and the median overall survival (OS) from SCT day 0 was 23.9 m (5-year 40%). The median progression-free survival (PFS) was 14.4 m (5-year 38%), the 5-year CI of relapse was 37%, and the 5-year CI of TRM was 25%. No significant differences in OS, PFS, CI of relapse, or CI of TRM were observed between different age categories. No significant differences in CI of relapse or CI of TRM were observed between MAC and RIC.
OS significantly improved over time (median 11.8 m in 2000-2010, 28.2 m in 2011-2016, 40.2 m in 2017-2023; p=0.01). OS was comparable between IPSS-R intermediate (median 140.7 m, 5-year 61%) and high risk (103.1 m, 5-year 59%) but significantly worse in very high risk (8.2 m, 5-year 15%; p<0.0001). Pts with TP53 wild type (n=137) had excellent OS (median not reached [NR], 5-year 69%) while those with TP53 monoallelic (n=19, 9.1 m, 5-year 20%) and TP53 biallelic (n=62, 6.8 m, 5-year 5%; p<0.0001) had poor OS. Pts with complex cytogenetics without TP53 mutations (n=24) had OS comparable to TP53-wild type, non-complex pts (n=113) (median NR vs NR, 5-year 61% vs 70%; p=0.41). In the pts who underwent SCT following no therapy or a single line of therapy, OS was 22.8 m in untreated (n=29), 15.3 m with hypomethylating agent (HMA; n=217), NR with intensive chemotherapy (IC; n=30), and 34.8 m with venetoclax-based regimens (n=38). Untreated pts had no significant difference in OS compared to those who received HMA (p=0.80), whereas pts who received IC had significantly better OS vs HMA (p=0.02). Pts who received venetoclax-based regimens had a trend towards improved OS vs HMA (p=0.06) and comparable to IC (p=0.83). Immediate pre-SCT IWG 2023 response had no significant impact on OS (median 27.3 m with CR, 19.7 m with CRL/CRh/PR, 50.3 m with HI, and 51.5 m with no response; p=0.78). However, immediate pre-SCT blasts of 10-19% (n=29) was associated with significantly worse OS compared to 0-9% blasts (n=327) (median 7.1 vs 28.1 m; p=0.001). Pts with 0-4% (n=267) and 5-9% (n=60) blasts had comparable OS (median 27.2 vs 31.8 m; p=0.58).
By MVA considering age, IPSS-R, TP53 status, HCT-CI, PTCy, year of SCT, and therapy prior to SCT, factors significantly associated with worse OS were IPSS-R very high risk (HR 1.77, p=0.02), TP53 (HR 3.45, p<0.001 for monoallelic; HR 4.21, p<0.001 for biallelic), and HCT-CI (HR 1.11, p=0.03). Receipt of a venetoclax-based regimen prior to SCT was associated with a HR for death of 0.61 but not meeting statistical significance (p=0.2).
Conclusions: SCT is a curative strategy in 69% of patients with TP53 wild-type MDS. Use of venetoclax as a pre-SCT cytoreductive regimen should be explored in larger studies. Outcomes remain poor in TP53-mutated disease and alternative approaches are needed in this population.
