Historical expectations with DNA methyltransferase inhibitor monotherapy in MDS: when is combination therapy truly “promising”?

Abstract

AbstractDNA methyltransferase inhibitors (DNMTIs) for patients with higher risk myelodysplastic syndromes (HR-MDS) have low complete remission rates and are not curative. Early DNMTI combination clinical trials in HR-MDS are often termed “promising,” but many randomized trials subsequently failed to show benefit. Clearer understanding of when a combination is likely to improve upon DNMTI monotherapy would inform randomized studies. We reviewed MDS azacitidine or decitabine monotherapy studies. We collected baseline demographics including International Prognostic Scoring System (IPSS) risk, DNMTI, disease characteristics; and response variables including survival and marrow and hematologic responses. Aggregate estimates across studies were calculated using meta-analyses techniques. Using a binomial design, we estimated the necessary operating characteristics to design a phase 2 study showing improved efficacy of a combination over monotherapy. Among 1908 patients, the overall response rate (ORR) was 24% (n = 464; 95% confidence interval [CI], 0.22-0.26): 267 complete response (CR, 14%), 68 partial response (4%), and 129 marrow complete remission (7%). Among 1604 patients for whom a hematologic response was reported, 476 (30%; 95% CI, 0.27-0.32) reported hematologic improvement (HI). More patients treated with azacitidine achieved HI (38%; 95% CI, 0.35-0.41) compared with decitabine (15%; 95% CI, 0.13-0.19), whereas the marrow ORR rate was higher with decitabine (29%; 95% CI, 0.26-0.33) compared with azacitidine (21%; 95% CI, 0.19-0.23). CR rates were similar between DNMTIs: 13% with azacitidine and 16% with decitabine. Variables that influence MDS response include the specific DNMTI backbone and the distribution of IPSS risk of patients enrolled on a trial. Considering these factors can help identify which early combination approaches are worth assessing in larger randomized trials.