Better discriminative capacity of blast and blast-mol scores in chronic myelomonocytic leukemia (CMML)

Abstract

Abstract Introduction: Prognostic assessment in chronic myelomonocytic leukemia (CMML) remains a challenge due to the variable clinical course and the biological complexity of the disease.Risk stratification is key for guiding treatment decisions and estimate prognosis. Two of the most used scoring systems are CMML-specific prognostic scoring system, CPSS and CPSS-mol, however they give limited weight to clinical variables that might influence disease prognosis. Recently Tefferi A et al. (Blood 2025) have described a specific prognostic scoring system acronymized “BLAST” which defines 3 categories [low-risk (0 points); intermediate-risk (1 point), and high-risk (2-4 points)] based on circulating blasts ≥2% (1 point), leukocytes ≥13 x 109/L (1 point) and presence of severe (2 points) or moderate (1 point) anemia. Molecular mutations and karyotype were added as a second step thus allowing the authors to develop a combined clinical-molecular risk model (BLAST-mol) that included the BLAST clinical risk variables and a 3-tiered molecular risk score. Aims: To validate in our series the prognostic value of the new “BLAST “and “BLAST-mol” scores in a CMML patient cohort and compare its performance with other CMML-specific scoring systems (CPPS and CPSS-mol) across the whole cohort. Methods: We retrospectively analyzed a cohort of 247 patients diagnosed with CMML between 2005 and 2024 from 5 institutions of the Spanish Group of Myelodysplastic Syndromes (GESMD). The median follow-up for patients alive was 2.22 years (range 0.09-15.34). CPSS, CPSS-mol, BLAST, and BLAST-mol scores could be assessed in 220 patients. Overall survival (OS) and leukemia free survival (LFS) were calculated for each prognostic index, using the Kaplan-Meier method. The discriminative ability of the CPSS, CPSS-Mol, BLAST and BLAST-mol scores was evaluated using the concordance index (Harrell's c-index). Results: The studied group included 151 (68.6%) males and 69 (31.4%) females. Median age at diagnosis was 71 years (range 27-95). According to 2022 WHO CMML classification, 185 (84.1%) were CMML-1 and 35 (15.9%) were CMML-2. In accordance with FAB classification: 174 (79.1%) were MD-CMML and 46 (20.9%) MP-CMML. At diagnosis, median hemoglobin, leucocytes, monocytes and bone marrow blast were 11.6 g/dL, 7.48x10e9/L, 1.510e9/L and 3.5 %, respectively. Forty-three (19.5%) of the total 220 patients evolved to AML. Thirty-two (14.5%) were transfusion dependent. The median OS of the whole cohort was 3.9 years (95%CI 3.2–4.6). According to the CPSS risk score: 127 (57.7%) were low risk, 56 (25.4) intermediate-1 risk, 32 (14.6%) intermediate-2 and 5 (2.3%) high risk. Regarding CPSS-mol: 69 (31.3%) were low, 55 (25%) intermediate-1, 57 (26%) intermediate-2 and 39 (17.7%) were high. These groups showed significantly different OS and EFS (p<0.001). Stratification according to BLAST score categorized patients into three risk groups: 114 (51.8%) low, 66 (30%) intermediate and 40 (18.2%) high. Following this new stratification, the median OS and EFS were as follows: low risk OS 6.8y (5.45-8.54); EFS 6.8y (5.45-8.32), intermediate OS 2.45y (1.67-3.99); EFS 1.94y (1.55-3.90), and high risk OS 1.79y (1.57-2.34) and EFS 1.79y (1.36-2.33) (p< 0.001). Later on, patients were stratified applying BLAST-mol risk score, defining 3 different risk groups low, intermediate and high risk being 125 (56.8%), 78 (35.5%) and 17 (7.7%) respectively. OS according to BLAST-mol score was: low risk 6.38y (5.28-8.10); intermediate 2.33y (1.79-3.78) and high risk 1.59y (0.41-2.06) (p< 0.001). Concerning EFS, the score showed high discriminative capacity with EFS of 6.14y (5.2-8.1) in low-risk patients, 1.94y (1.74-2.98) in intermediate and 1.49y (0.4-3.1) in high risk CMML patients. Overall, BLAST and BLAST-mol scores performed better in the OS and EFS discrimination than CPSS and CPSS-mol with concordance index (c-index) of 0.690 and 0.672, compared to 0.653 and 0.656 for OS and 0.688 and 0.672, compared to 0.663 and 0.665 for EFS, respectively. Summary/Conclusions:BLAST and BLAST-mol retained its prognostic performance in CMML population with higher discrimination for OS and better predictive capacity for AML progression when compared with CPSS and CPSS-Mol. These findings support the applicability of BLAST and BLAST-mol to CMML highlighting its potential utility in guiding clinical decision-making and therapeutic strategies. Further prospective validation in larger cohorts is warranted.