Myelodysplastic/myeloproliferative overlap neoplasms: Clinicopathological features, treatment strategies, and outcomes — a UK multicentre cohort Study

Abstract Background: Myelodysplastic/Myeloproliferative overlap neoplasms (MDS/MPN) are rare, clonal malignancies displaying clinicopathological features from both MDS and MPN. The World Health Organisation 2022 classification (WHO2022) define MDS/MPN as; chronic myelomonocytic leukaemia (CMML), MDS/MPN with neutrophilia (MDS/MPN-N), MDS/MPN with SF3B1 mutation and thrombocytosis (MDS/MPN-SF3B1-T), and MDS/MPN, not otherwise specified (NOS). There is a need for real-world evidence regarding phenotype, therapeutic approaches and survival. Aims: Evaluation of clinicopathological features, treatment strategies and outcomes of patients diagnosed with ‘non-CMML’ MDS/MPN from 17 UK centres. Methods: Chart reviews were conducted using electronic records. MDS/MPN were defined according to WHO2022 classification. Baseline demographics, molecular profiles, treatment strategies and clinical outcomes were presented using descriptive statistics. Progression was defined as progression to AML or increasing myelofibrosis (MF; ≥ grade 2 reticulin), and ≥1 grade higher than baseline. Results: 89 MDS/MPN patients were identified, diagnosed between November 2010 and April 2025. 63 (71%) had MDS/MPN-NOS, 18 (20%) MDS/MPN-N, and 8 (9%) MDS/MPN-SF3B1-T. Median age at diagnosis was 74 (range (r),42-96), and 65% were male. Median follow up was 18 months (r,0-111). 29% had an antecedent diagnosis of MPN (n=18) or MDS (n=7), and 11 treated with either hydroxyurea (HU; n=9), anagrelide (ANA; n=3), hypomethylating agent (HMA; n=1) or underwent allogeneic transplantation (allo-HCT; n=1). At diagnosis, 26% (n=23) had splenomegaly and 16% (n=14) antecedent thrombosis. Median presenting blood counts; hemoglobin 10.6 g/dL (59-185), WBC 11.96 x109/L (1.2-189.6) and platelets 222 x109/L (2-1447). Bone marrow analysis showed MF grade 0 (n=22), 1 (n=21), 2 (n=19) and 3 (n=6). Karyotypic abnormalities were detected in 29% (n=26). Mutational analysis revealed aberrations in epigenetic regulators (n=63; 71%), splicing factors (n=61; 69%), signalling pathways (n=57; 64%), transcription factors (n=10; 10%) and others (n=32; 36%). Of note, only 1 had pathogenetic TP53. Characterizing the MDS/MPN-NOS cohort as per ‘genomic signature’ (Paloma et al2020) split the cohort into: ‘CMML-like’ (n=16; 25%) biallelic TET2 (n=2, 13%), TET2-SRSF2 (n=5; 31%) and RUNX1-SRSF2 (n=2; 13%); ‘aCML-like’ (n=11; 18%) ASXL1-SRSF2 (n=3; 27%) and ASXL1-SETBP1 (n=2; 18%) and ‘MDS/MPN-RS-T-like’ (n=14; 22%) SF3B1 (n=5; 35.7%) and SF3B1-JAK2 (n=3; 21%); and STAG2 (n=2; 3%). Treatment was initiated in 57%; 29 (33%) received HU, 7 (8%) ANA, 13 (15%) HMA, 6 (7%) ruxolitinib, and 5 underwent allo-HCT. 47% were red cell transfusion requiring and 38% dependent. In total, 29 (33%) received best supportive care (BSC); 26 (29%) anti-platelets, 7 (8%) anti-coagulation, and 11 (12%) ESA. Median overall survival (OS) for entire cohort was 28.9 months (95% CI 21.9, 49.5) and median progression free survival (PFS) not reached. For the entire cohort, anaemia (<10 g/dL) at diagnosis was associated with worse OS (22.8 months) versus not (31.2 months; p=0.05). The MDS/MPN-N cohort had the shortest OS at 11 months, compared to 26.6 and 61.8 months for MDS/MPN-NOS, and MDS/MPN-SF3B1-T (p=0.02). Median OS for each MDS/MPN-NOS ‘genomic’ sub-group was 15.3 (aCML-like), 20.8 (CMML-like), 58.9 (MDS/MPN-RS-T-like), and 59.8 (Others) months (p=0.002). In MDS/MPN-NOS, ASXL1 and RUNX1 mutations (n=18) were associated with a shorter PFS of 20.5 months (p<0.0001). MDS/MPN-NOS with > 2 mutations (n=31) had a shorter OS of 22.8 compared to 59.8 months (n=28; p=0.03) for ≤2 mutations, and shorter PFS of 49.5 months (p=0.004). Regarding non-allo-HCT therapies, no clear survival benefit was observed compared to BSC/monitoring (p=0.98). Leukemic transformation occurred in 14% (n=12; 11/12 MDS/MPN-NOS) at a median of 20 months (r, 5-57). For upfront allo-HCT, median OS post-HCT was relatively poor at 20.7 months. Conclusion: MDS/MPN neoplasms present therapeutic challenges due to heterogenic phenotypes, lack of clear survival benefit with current non-transplant therapies, poor OS post allo-HCT and a paucity of new agents. Genomic characterisation aids prognostic estimation across all subtypes. Moreover, we confirm the prognostic relevance of ‘genomically stratifying’ the MDS/MPN-NOS group as per the seminal work of Paloma et al, aiding therapeutic decisions. Better prognostic models and novel therapeutic options are urgently required.