Challenges for Conducting Clinical Trials Evaluating Maintenance Chemotherapy In Acute Myeloid Leukemia (AML): a Cancer and Leukemia Group B (CALGB) Study

Recent advances in frontline therapy for newly diagnosed AML include increased dose of anthracycline during induction, multi-agent regimens, high dose cytarabine (HiDAC) as consolidation for core binding factor (CBF) patients (pts), and allogeneic transplantation (alloHCT) during 1^st remission for poor risk pts. Prolonged low-dose cytotoxic maintenance therapy does not appear to improve clinical outcomes, but investigation of novel maintenance strategies remains appealing, affording pts an opportunity to receive new therapies without increasing the considerable toxicities of induction/ consolidation. The CALGB performed a phase II study of induction and risk-adapted consolidation, followed by one year of maintenance therapy with decitabine, for newly diagnosed AML pts <60 years of age (CALGB 10503). Induction was daunorubicin 90 mg/m² for 3 days, etoposide 100 mg/m² for 3 days, and cytarabine 100 mg/m² for 7 days (3+3+7). Reinduction (2+2+5) was given for residual disease on day 14. CBF pts in CR received 3 cycles of HiDAC; all other pts received autologous transplantation (autoHCT) if eligible, or HiDAC if not. Poor risk pts in 1^st CR received alloHCT off study, when appropriate. Maintenance decitabine (20mg/m² intravenously for 4–5 days, every 6 weeks for 8 cycles) began within 60–90 days after consolidation for pts with neutrophil ≥1000/uL and platelets ≥75,000/uL. Clinical results for the efficacy of decitabine maintenance for one year are not yet mature; the study closed to accrual July 30, 2010. In a previous study for a slightly more favorable patient cohort that was treated with identical induction and consolidation (CALGB 19808), 29% of all enrolled pts (214/732) and 39% of CR pts were registered to a randomized investigational maintenance therapy. To examine the feasibility of conducting a subsequent randomized trial of maintenance therapy, we analyzed reasons for study discontinuation for all pts enrolled on CALGB 10503; 473 pts had “response” or “off study” data forms submitted, to date. Complete remission* (CR) was achieved in 349 (74%). Reasons for induction failure were primary refractory AML (15%), induction death (5%), and other (6%, including those withdrawn from protocol for alternative therapy presumably due to persistent disease). The most common reason for not receiving maintenance therapy was the presence of AML, either primary refractory as noted or early relapse (7% of CR pts). In pts who achieved CR, the most common reason for not receiving maintenance was removal from the protocol for alloHCT (71/349, 20%). 9% refused further protocol therapy after consolidation, perhaps from “treatment fatigue.”

Correlative investigations for CALGB 10503 are ongoing, including identification of novel prognostic markers, targets for treatment, and markers of minimal residual disease. Clearly, clinical investigation of maintenance therapy as part of a comprehensive treatment approach for frontline therapy of AML (inclusive of homogeneous induction and consolidation therapy) requires considerable up-front enrollment in order to reach maintenance accrual goals. Randomized maintenance studies will likely need intergroup participation for timely completion. However, we conclude that the benefits of up-front enrollment (relative to a post-consolidation enrollment strategy) outweigh the accrual burdens. Benefits include uniform pre-maintenance therapy and the potential for novel discovery from correlative studies. Study designs for future maintenance trials for AML in 1^st CR must balance expediency with the more comprehensive approach employed in CALGB 10503.

Blum: Celgene: Research Funding. Off Label Use: decitabine in AML. DeAngelo: Deminimus: Consultancy.