1007 Background: We had previously shown that sequential A→T is a reference treatment as first-line CT for MBC patients (pt) (JCO 2004, (22), 2587–2593). However, responses are short-lived and time to progression (TTP) is also short. Maintenance chemotherapy (MCT) with LPD could improve TTP without relevant additional toxicity. Methods: All pt received an induction CT with 6 cycles (cy) of A→T. Pt with a complete or partial response (CR, PR), or stable disease (SD) were randomly assigned to either LPD (40 mg/m2 q4wk × 6 cy) or observation (O) from May 2002 to Dec 2006. Eligible pt had adequate bone marrow, renal, hepatic and cardiac functions (by left ventricular ejection fraction, LVEF). The study hypothesis assumed a median TTP since start of induction CT in O arm of 10.5 months (m), and a prolongation of TTP of 66% with MCT. Therefore, 154 pt (77 per arm) were necessary (one-sided a and 1-β errors of 0.01 and 0.8). Results: Pt accrual was completed in December 06. 122 and 136 pt (60/66 in LPD vs. 62/70 in O) are currently evaluable for efficacy and safety. Median age was 57 yr (30–74). Status disease in pt assigned to LPD vs. O: 2 (3%) vs. 6 (9%) of pt had CR, 31 (47%) vs. 43 (61%) had PR; 33 (50%) vs. 21 (30%) had SD. 291 cy of LPD were administered (median 6, range 0–6). Neither relevant LVEF decrease nor clinical congestive failure were seen. Nausea/vomiting and alopecia incidence was negligible. G3 hand-foot syndrome was present in 3 pt/3 cy (5/1%). Median TTP in LPD arm was 16.04 m (14.06–18.02), vs. 9.96 m (8.87–11.05) in O arm, p=0.0001. Conclusion: Maintenance therapy with LPD significantly prolongs TTP in MBC pt after a first-line CT without significant clinical toxicity. No significant financial relationships to disclose.