Use of Post-Treatment Clinical Data To Predict Response to Decitabine.

Best response of AML or MDS to decitabine (DAC) may require > 3 courses. Identification of pts whose response to initial courses suggests they are unlikely to respond further to subsequent courses would allow such pts to receive other therapies more expeditiously. 104 pts, median age 69, received DAC (20 mg/m2 IV daily × 5 days every 5 weeks) for newly- diagnosed MDS (80, 76 IPSS int-2 or high) or AML (24) from 9/00-5/07. After 3 courses, 33 of the 104 were in CR, 6 were dead, and 32 had been removed from study, with the remaining 32 receiving more DAC. The probability of CR decreased to < 10% on all courses after the 5th:

Results were qualitatively similar for AML and MDS. We divided the 32 pts who received ≥ 4 courses according to whether after course 3 their:

1. marrow showed a response, defined as a p < 0.05 difference between the pre-Rx and post course 3 blast count,

2. platelet count increased > 30, 000 if < 100,000 pre-Rx, or

3. neutrophil count increased > 500 if < 500 pre Rx; criteria 2 and 3 were adapted from those of the IWG. 8/14 pts with vs. 1/11 without a marrow response after course 3 entered CR on subsequent courses (p = .02, 6/10 vs. 1/10 considering only MDS).

The corresponding figure for platelet response/no response was 5/8 vs. 2/11 (3/5 vs. 2/9 for MDS). The median number of courses after the 3rd given pts with marrow response was 2 (range 1–10) vs. 3 (range 1–12) for pts without marrow response. Our data suggest that if CR is the goal, pts should not receive more than 5 courses of DAC and no more than 3 if marrow response has not occurred after the 3rd.