Valproic acid shows in-vitro activity against AML blasts and synergizes with ATRA in differentiation induction. We reported in-vivo activity of VPA and VPA+ATRA in patients with MDS and AML. Here we present follow-up data on 119 patients treated between February, 2002, and July, 2005.

87 patients were started on VPA monotherapy. Addition of ATRA was planned for patients who did not respond or relapsed after an initial response. 32 patients were treated with VPA +ATRA from the start. VPA was targeted to reach serum concentrations of 50–110 μg/ml. Median treatment duration was 4 months (<1–42) for VPA and 3 months (<1–35) for ATRA.

25 patients (21%) achieved responses according to International Working Group (IWG) criteria for MDS. The response rate was 28% in MDS (n=58) and 15% in AML (n=61). One patient achieved CR, and another patient achieved PR. 23 patients showed hematologic improvement. The median response duration was 4 months (2–33). 45 patients (38%) had stable disease, and 49 (41%) showed progressive disease.

Among the 25 responders, 17 relapsed. 10 of these subsequently received ATRA, which resulted in a second remission in 4. Median response duration of the second responses (20 months, n=4) was significantly longer than the first response (4 months, n= 25, p=0,008). Of 94 nonresponders, 25 received additional ATRA without response. The response rate among the 87 patients on VPA monotherapy was higher than among the 32 patients receiving VPA+ATRA from the start (24% vs. 13%).

We found a response rate of 42% in MDS with a normal blast count (PSA, RSA, RCMD, RSCMD; n=36), 6% in RAEB (RAEB I+II; n=18), 15% in AML (n=61), and 0% in CMML (n=4). The difference in response rate was significant when RSA, RCMD, and RSCMD were taken together and compared with all other disease types (p = 0.001). Treatment responses were correlated with risk assessment according to the International Prognostic Scoring System. Patients with low-risk MDS (IPSS low + int-1) had a significantly higher response rate than all other risk groups (p = 0.049). The response rate was 44% in low-risk (n=16), 29% in intermediate-1 (n=28), 10% in intermediate-2 (n=10), and 0% in high-risk MDS (n=4). We compared response duration in patients with PSA, RSA, RCMD, and RSCMD, and patients with RAEB I+II and AML, using the Kaplan-Meyer method. We found a significantly longer response duration in good-risk patients (3 months (2–26) vs. median not reached (2–33); p=0,043). Bone marrow blast count was predictive of response (p = 0.004). For cytogenetic risk groups, there was a trend of lower responses in patients with a high-risk karyotype (11%), compared to an intermediate (17%) or low-risk karyotype (28%). 72 patients had a bone marrow blast count >5%. In 3 of these patients, the blast count normalized, 4 achieved a reduction by more than 50%, and 3 had a complete peripheral blast clearance.

We conclude that VPA has a high response rate in low-risk MDS. In patients responding to VPA monotherapy, subsequent combination with ATRA may lead to prolonged responses. For patients with high-risk MDS, combination regimens including inhibitors of signal transduction, or demethylating agents should be investigated.

Author notes

Corresponding author