Prognostic impact of pre-transplant IPSS-m risk downstaging in Myelodysplastic Syndromes

Abstract Introduction Mutations incorporated risk stratification tool in MDS (IPSS-M) have offered greater prognostic precision and outperformed the traditional IPSS-R in all risk categories. While typically used at diagnosis, the role of IPSS-M as a dynamic tool to re-stratify prior to a definite treatment for MDS, like allo-HCT remains unexplored. Whether such a strategy, before transplant, would inform more prognostic clues and impact outcomes in MDS is less well studied. Methods: We identified 130 myeloid neoplasm patients from 1300 records at Karmanos Cancer Institute (2016–2025). Mutations were detected via a 54-gene TruSight Myeloid Panel (Illumina MiSeqDx). IPSS-M scores were manually calculated. Chi-square tests were used for categorical variables and Kaplan-Meier curves for survival outcomes. Results: We analyzed 130 MDS patients with paired samples at diagnosis and pre-transplant. Median age was 66 years (range 23–95), with 61% male. Patients were 75% White, 10% Black, and 15% other, with a significant difference between White and Black patients (p = 0.00001). Karyotype was similar at both timepoints: normal (33% vs 31%, p = 0.79), abnormal (52% vs 48%, p = 0.54), and complex (15% vs 21%, p = 0.33). Median blasts declined from 2% to 0.4% (p = 0.49). For diagnosis and pre transplant time points, median ANC, hemoglobin, and platelets (diagnosis vs pre-transplant) were 1.1 vs 1.2 (p = 0.30), 9.9 vs 9.0 g/dL (p = 0.61), and 68 vs 70 × 10³/μL (p = 0.99). WHO subtypes (diagnosis vs pre-transplant) were LB (43% vs 42%, p = 1), IB1 (12% vs 10%, p = 0.60), IB2 (51% vs 43%, p = 0.20), SF3B1 (5% vs 4%, p = 1), and 5q (3% vs 1%, p = 0.30). IPSS-M risk categories at diagnosis versus pre-transplant were as follows: Very Low (3% vs 7%, p = 0.25), Low (18% vs 15%, p = 0.50), Moderate Low (15% vs 19%, p = 0.41), Moderate High (17% vs 19%, p = 0.75), High (17% vs 16%, p = 1.00), and Very High (30% vs 24%, p = 0.33). From diagnosis to pre-transplant, IPSS-M for 59 patients (45%) remained unchanged, 33 (25%) were upstaged, and 38 (30%) down-staged. Normal karyotype was more frequent in the downstaged group (53%) vs unchanged (25%) and upstaged (30%) (p = 0.019), while complex karyotype was highest in unchanged (26%) and lowest in downstaged (5%) (p = 0.040). Bone marrow blasts were similar across groups (median 3% for unchanged and downstaged and 4% for upstaged, P=0.40). WHO 2022 subtypes varied across IPSS-M trajectories: MDS-LB (49% unchanged, 33% upstaged, 69% downstaged; p = 0.002), MDS-IB2 (34%, 48%, 8%; p = 0.001), with no significant differences in MDS-IB1, SF3B1, or 5q. HMA monotherapy was used in 39% of unchanged, 40% of upstaged, and 34% of downstaged patients (p = 0.80). HMA with venetoclax was used in 8%, 3%, and 10%, respectively (p = 0.47). Intensive induction (7+3) was given in 10%, 6%, and 10% (p = 0.77). Luspatercept was used exclusively in the downstaged group (5%) (p = 0.09). TP53MTwas most frequent in the unchanged group (48%) compared to upstaged (17%) and downstaged (15%) (p = 0.001). TET2MT was enriched in the upstaged group (33%) vs unchanged (20%) and downstaged (10%) (p = 0.04). No significant differences were observed in DNMT3AMT, SF3B1MT, SRSF2MT, or STAG2MT across groups. In the unchanged group, the median IPSS-M score was 0.2 at diagnosis and – 0.2 pre-transplant (range –1.4 to 2.3 and –3.1 to 1.8). Among upstaged patients, the score increased from –0.6 to 0.9 (range –1.4 to 1.5 and –1.3 to 3.1). In the downstaged group, it declined from 1.9 to –0.3 (range –0.3 to 3.9 and –2.0 to 2.0), reflecting shifts in risk burden. Median OS was 64 months in the downstaged group vs 22 months in the unchanged/upstaged group (p=0.0001). Median RFS was 60 vs 6 months, respectively (p=0.0002). Rates of all-grade GVHD were comparable across the groups. Acute GVHD was seen in 35% of unchanged, 30% of upstaged, and 26% of downstaged patients (p = 0.62), while chronic GVHD was seen in 35%, 27%, and 33%, respectively (p = 0.71). Conclusion: Pre-transplant downstaging of IPSS-M was associated with significantly improved post-transplant survival and relapse outcomes, supporting its role as a dynamic prognostic tool. These results support the integration of serial IPSS-M assessments into restratifying risks before transplant to better reflect treatment response and evolving disease biology. Ongoing work focuses on large-scale validation at multiple time points to study the impact of treatment on risk strata.