Re-Treatment with Low-Dose 5-Aza-2′-Deoxycytidine (Decitabine) Results in Second Remissions of Previously Responsive MDS Patients.

At low doses, azanucleosides with DNA methyltransferase inhibitory activity like 5-azacytidine and 5-aza-2′-deoxycytidine (DAC) have significant single-agent activity in MDS and AML. However, the optimal treatment duration is still a matter of debate: while at least two consolidating courses after best response are usually performed, the response to treatment with these drugs after a relapse of MDS has not been systematically studied. We report on 22/91 patients (pts) with MDS (24%) having received DAC both initially and, at relapse, as re-treatment. Median age was 71 years (range, 51–81). Characteristics of MDS at time of initial treatment according to FAB: 3 RA, 10 pts with RAEB, 8 pts with RAEB-t (i.e. AML according to WHO) and 1 pt with CMML. According to the IPSS at time of initial treatment, 5/22 (23%) pts were intermediate-1 (int-1), 4/22 (18%) pts int-2 and 13/22 (59%) pts were high-risk. Patients had initially received a median of 6 DAC courses (range, 2–6), in 12/22 (55%) pts resulting in a complete remission (CR), in 6/22 (27%) pts in a partial remission (PR) and in 4/12 (18%) pts in hematologic improvement (HI). 6/12 pts. with hematologic CR had also had a complete cytogenetic response. Re-treatment of relapse with DAC started a median 11 months (range, 3–27) after the last course of initial treatment. Characteristics of MDS at time of re-treatment: 1 pt RA, 13 pts RAEB, 6 pts RAEB-t and 2 pts sAML. During re-treatment, pts received a median of 3 courses (range, 1–6), with 10/22 (45%) responses: 1 CR, 2 PR (all 3 had CR at initial treatment) and 7 HI (2 CR at initial treatment, 4 PR, 1 HI). 12/22 (55%) pts did not show objective responses to re-treatment, including 3 with primary resistance to the first re-treatment course. Median survival of all pts (n=22) from start of first DAC course was 28 months (range, 15–50+) 13/22 (59%) pts transformed to AML. Median time from second response to relapse was 4 months (range, 1–16), thus three times shorter than first response. The pt with CR both at initial treatment (9 months duration) and at re-treatment (5 months duration) received a second re-treatment with 3 courses of DAC (total of 15 courses), without achieving another response. Compared to pts not achieving a second response, responders to DAC re-treatment were less frequently in the IPSS high-risk group (40% versus 83%), and had lower median sLDH at re-treatment start (197 U/L versus 282 U/l). Age, FAB subtype at retreatment, previous response to DAC (including response duration and achievement of a cytogenetic remission) did not differ between both groups.

In conclusion, re-treatment resulted in objective responses in 45% of previously DAC-responsive patients. However, quality and duration of second remissions were inferior, implying that DAC-responsive patients might benefit more from continuation of the initial treatment than from re-treatment. Ongoing multi-institutional trials allow continuation of initial DAC treatment beyond 6 courses.