Validation of a post-hypomethylating agent failure prognostic model in myelodysplastic syndromes patients treated in a randomized controlled phase III trial of rigosertib vs. best supportive care

This is the authors' abstract. We don't add key points for this paper.

The hypomethylating agents (HMAs) azacitidine and decitabine are standard therapies for patients with higher-risk myelodysplastic syndromes (MDS) 1 , 2 . Although treatment with these agents may improve overall survival (OS), eventually every patient becomes refractory to or relapses after treatment with HMAs, and has a dismal outcome 3 . Clinical trials enrolling patients at the time of HMA failure (HMAF) have often used the International Prognostic Scoring System (IPSS) or the Revised IPSS (IPSS-R) as eligibility criterion for trial entry 4 , 5 . We have previously shown that the IPSS, IPSS-R and other commonly used MDS prognostic models (e.g., the World Health Organization classification-based Prognostic Scoring System, and the MD Anderson Prognostic Scoring System) have limited predictive power in the HMAF setting 6 . We therefore developed and validated a new model to predict outcome of patients after HMAF 6 that includes six clinical variables: age; performance status; complex cytogenetics (>4 abnormalities); marrow blast percentage >20%; platelet count; and red cell transfusion dependency 6 . This model separates patients into two risk categories: lower-risk, with a median OS of 11.0 months (95% confidence interval (CI) 8.8–13.6); and higher-risk disease, with median OS of 4.5 months (95% CI 3.9–5.3) 6 . The model was subsequently validated in an independent cohort of 223 MDS patients derived from the Groupe Francophone des Myélodysplasies database 7 . In the validation cohort, 82 patients were classified in the lower-risk category and they had a median OS of 13 months compared to 141 higher-risk category patients with a median OS of 5 months ( p < 0.001) 7 .