A comparative assessment of molecular-based prognostic models in CMML

Several prognostic systems integrating clinical, cytogenetic, and molecular parameters have been developed to estimate risk and inform treatment in chronic myelomonocytic leukemia (CMML). Recently, the molecular International Prognostic Scoring System (IPSS-M) was introduced for risk stratification in myelodysplastic syndromes (MDS), demonstrating improved prognostic accuracy over the mutation-agnostic Revised International Prognostic Scoring System (IPSS-R) and potentially offering a novel tool for risk assessment in this population. We aimed to assess whether the applicability of the IPSS-M extends to CMML while providing a comprehensive comparison of all major molecular-based integrated models. Baseline clinical and molecular data were collected from 340 patients with CMML. The most frequent mutations were TET2, SRSF2, ASXL1, RUNX1, and NRAS. The IPSS-M stratified patients into 6 risk categories, with median overall survival (OS) of 18.5, 5.1, 3.9, 2.65, 1.7, and 1.1 years, corresponding to very low to very high risk disease (P < .001). Additionally, the 4-year cumulative incidence of acute myeloid leukemia evolution was 4.2%, 12.1%, 19.4%, 25.9%, 32.8%, and 26.7%, respectively (P = .008). Both CMML-specific prognostic scoring system (CPSS)-Mol and IPSS-M improved OS discrimination compared to the Mayo molecular and Groupe Francophone des Myélodysplasies models. CPSS-Mol outperformed CPSS, and IPSS-M was superior to IPSS-R. CPSS-Mol demonstrated the highest prognostic accuracy for predicting leukemic evolution, establishing it as the superior overall model. Importantly, IPSS-M was reliably applicable in CMML and displayed prognostic accuracy comparable to CPSS-Mol. Furthermore, all models retained predictive validity in patients receiving frontline hypomethylating agent therapy, suggesting that using the IPSS-M is unlikely to adversely affect outcomes when guiding treatment decisions, particularly in community settings in which CMML is often grouped with MDS.