Epigenetic Therapy with 5-Azacitidine, Valproic Acid, and ATRA in Patients with High-Risk AML or MDS: Results of the French VIVEDEP Phase II Study

Introduction: Promising results have been reported last year with a combination of 5-azacitidine (AZA), valproic acid (VPA), and all-trans retinoic acid (ATRA) in patients with AML/MDS treated at MDACC in a Phase I/II study (Soriano et al. Blood 2007). We report here on a similar study conducted in 9 centers between 7/2006 and 8/2007.

Methods: Patients with high-risk AML (AML in patients aged 70y+ unsuitable for intensive chemotherapy or early relapsing/refractory AML) or MDS (int-2/high IPSS without possibility of allogeneic SCT) were eligible. Treatment consisted of 6 cycles with AZA 75 mg/m²/d SC (d1-7), VPA 35 to 50 mg/kg/d PO (d1-7), and ATRA 45 mg/m²/d PO (d8-28). Cycle 1 was initiated at the hospital but cycles 2–6 were planned monthly in out-patients. Response was assessed after cycle 1, 3, and 6 (IWG AML criteria). VPA was started at 35 mg/kg/d and then increased at 50 mg/kg/d if well tolerated. Sixty-three patients were enrolled and 51 are evaluable.

Results: Patients characteristics were: M/F, 27/24; median age, 73y (50–87); median WBC, 2.3 × 10⁹/L; PS 0-1/2, 45/6 patients; median follow-up, 13 months. Forty-two patients had AML (31 de novo, 9 therapy-related, 2 post-MDS/MPD) and 9 had MDS. Only 6 patients had received prior intensive therapy. Cytogenetics was available in 46 patients with high-risk features in 26 of them (complex, -7). Twenty-nine patients stopped the treatment after 1–5 cycles: 13 due to disease progression, 11 due to toxic events (mostly infection), and 5 due to physician decision despite stable disease. VPA was associated with notable CNS toxicity at 50 mg/kg, but not at the 35 mg/kg dose level. ATRA-related symptoms (headaches, mucosal dryness) were noted. In the 22 patients who received the 6 cycles, re-hospitalization rate was 27, 41, 23, 18, 20, and 14% after cycle 1 to 6, respectively. Among these patients, 11 reached CR (6 after cycle 3) and 5 reached PR (4 after cycle 3). The CR/PR rate was thus 31%, reaching 35% in the 46 patients who did not interrupt the treatment in the absence of progression or toxic event. Table 1 gives CR/PR rate according to various patient subsets. In patients with de novo AML, CR/PR rate was 45%. Advanced age and high-risk cytogenetics did not influence the response rate. In multivariate analysis, cytogenetics but not age remained, however, a poor risk factor for OS (PS, WBC, and MDS/secondary AML being other significant factors). Ten of the 16 responders relapsed after a median response duration of 10.6 months. Median OS was 12 months (not reached in the responders). Results of sequential DNA methylation and gene expression profiles monitoring (#17 patients) will be presented.

Conclusion: This study confirms that epigenetic therapy with AZA, VPA, and ATRA yields a 35% response rate in patients with high-risk AML/MDS. Although randomized studies are needed (AZA ± HDAC inhibitors), this combined approach appears to be a good option to treat older patients with low WBC and favorable PS, whatever their cytogenetics. Maintenance options should be investigated in responding patients.

Table 1