A Randomized Phase II Trial of 5-Day Versus 10-Day Schedules of Decitabine for Older Patients with Previously Untreated Acute Myeloid Leukemia

Background: Older patients (pts) with acute myeloid leukemia (AML) are often not candidates for intensive cytotoxic chemotherapy due to frailty or comorbidities, and decitabine has been shown to improve their survival compared to supportive care or low-dose cytarabine. However, the optimal dosing schedule of decitabine in this setting is not well-established. This phase II randomized study was conducted to evaluate relative safety and efficacy of two different schedules of decitabine in this population.

Methods: Adult pts ≥60 years of age with newly diagnosed AML were randomized using a Bayesian adaptive design to receive decitabine at a dose of 20 mg/m² IV daily for either 5 or 10 consecutive days. Pts <60 years of age were also eligible if deemed not to be suitable candidates for intensive cytotoxic chemotherapy. Pts were required to have a performance status of ≤3, total bilirubin <2 mg/dl and creatinine <2.5 mg/dl. Pts who had received prior hypomethylating agents for an antecedent hematologic malignancy were ineligible. Pts could receive up to 3 consecutive cycles of induction at the randomly assigned dose at approximately monthly intervals. Responding pts received additional consolidation cycles every 4-8 weeks for up to 24 total cycles. The primary endpoint was to compare the response rates of decitabine x 5 days versus decitabine x 10 days. Secondary endpoints included the response duration, survival and the safety of the 2 schedules.

Results: Between 2/2013 and 3/2017, 63 pts have been randomized to receive decitabine for either 5 days (n=24) or 10 days (n=39). Baseline characteristics of the 2 arms were well-balanced and are shown in Table 1. Among the total cohort (both arms), the median age was 77 years (range, 57-90 years); 49% had poor-risk cytogenetics, 43% had therapy-related or secondary AML and 35% had a TP53 mutation. The overall response rate (CR + CRi + PR) was 46% for the 5-day arm and 38% for the 10-day arm (P=0.56). The CR rates were 29% and 28%, respectively (P=0.94). No baseline characteristics were observed that predicted for superior response to either of the treatment arms. The median number of cycles to best response in both arms was 2 (range, 1 to 5 cycles for both arms). Early mortality was similar with both dosing schedules. The 30-day mortality rates for the 5-day and 10-day arms were 4% and 8%, respectively (P=0.58), and the 60-day mortality rates were 21% and 26%, respectively (P=0.66).

The median duration of follow-up was 34 months. Among the 11 responding pts in the 5-day arm and the 15 responding pts in the 10-day arm, the median remission durations were 9.4 and 6.4 months, and the 1-year continued remission rates were 25% and 27%, respectively (P=0.56). No pts in either arm underwent subsequent allogeneic stem cell transplantation. Survival outcomes did not differ between the 2 arms. The median EFS was 3.0 months for the 5-day regimen and 3.4 months for the 10-day regimen; 1-year EFS rates were 14% and 13%, respectively (P=0.83; Figure 1A). The median OS was 4.9 and 7.1 months, respectively; 1-year OS rates were 26% and 24%, respectively (P=0.62; Figure 1B). There was no difference in survival by treatment arms when pts were stratified by TP53 mutation status.

Conclusions: In this relatively poor-risk cohort of older adults with newly diagnosed AML, decitabine given for either 5 or 10 consecutive days resulted in similar response rates, early mortality and survival.

View largeDownload slide

View largeDownload slide

Close modal