New Multivariate Model Enhances Prognosis for HMA-Treated MDS

A recent study examined the effectiveness of the Molecular International Prognostic Scoring System (IPSS-M) in predicting outcomes for patients with myelodysplastic syndromes (MDS) treated with hypomethylating agents (HMAs). The IPSS-M, which integrates genomic data with clinical and cytogenetic information, has been useful in risk stratifying MDS patients at diagnosis. However, its utility in patients receiving HMA therapy has not been fully established. The study also introduced a multivariate model for better risk stratification in this specific patient group.

Researchers presented their study at the Twelfth Annual Meeting of the Society of Hematologic Oncology in Houston, Texas. They conducted a retrospective analysis of 455 patients with MDS who received frontline HMA therapy at a tertiary cancer center. They assessed patient characteristics and bone marrow data at diagnosis, applying both univariate and multivariate analyses using Cox proportional hazards models to determine factors affecting overall survival.

Improved Prognosis for MDS After HMA Treatment

Risk stratification with the IPSS-R (revised International Prognostic Scoring System) showed a median overall survival (OS) of 60 months for very low-risk patients, 87.4 months for low-risk patients, 36.2 months for intermediate-risk patients, 23.7 months for high-risk patients, and 12.8 months for very high-risk patients, with a concordance index of 0.698 (P<.0001). The IPSS-M produced comparable results, with a median OS of 60 months in low-risk patients, 40.8 months in moderate-low-risk patients, 44.2 months in moderate-high-risk patients, 33.4 months in high-risk patients, and 16.6 months in very high-risk patients. The concordance index for the IPSS-M in HMA-treated patients was 0.694, indicating similar predictive power as the IPSS-R.

To enhance risk prediction, the study developed a multivariate model using factors identified in the univariate analysis, such as age at diagnosis; reductions in hemoglobin and platelet levels; bone marrow blast percentage; cytogenetic risk; and mutations in TP53, PTPN11, and U2AF1. This model enabled more precise stratification into five or six risk categories, offering improved prognostic accuracy compared with both IPSS-R and IPSS-M.

Further validation of this scoring system is recommended to confirm its utility in clinical practice.

Reference

Chien KS, Li Z, Urrutia S, et al. Risk stratification in hypomethylating agent-treated patients with myelodysplastic syndromes: a multivariate model. Abstract #MDS-531. Presented at the Twelfth Annual Meeting of the Society of Hematologic Oncology. September 4-7, 2024; Houston, Texas.