Pixantrone is a potentially more effective, less cardiotoxic alternative to doxorubicin for patients with aggressive non-Hodgkin lymphoma (aNHL). This phase I/II non-comparative study evaluated pixantrone in place of doxorubicin in the standard CHOP regimen (cyclophosphamide, doxorubicin, vincristine, and prednisone), i.e. CPOP (cyclophosphamide, pixantrone, vincristine, and prednisone), in patients with relapsed aNHL who had previously received CHOP ± rituximab. Patients were administered pixantrone on day 1 of each 21-day cycle. Phase I (n = 35) dose escalation from 80 mg/m(2) to 180 mg/m(2) established the phase II (n = 30) dose as 150 mg/m(2). In phase II, 20 patients (67%) received all six planned cycles. The objective response rate was 73%, complete response/complete response unconfirmed (CR/CRu) rate was 47%, and median overall survival was 17.9 months. Myelosuppression was nearly universal. Six patients (20%) developed febrile neutropenia. Overall, left ventricular ejection fraction (LVEF) declines ≥10% occurred in 14 patients; declines seemed transient and unrelated to dose. Symptomatic cardiac failure occurred in four patients; however, pre-existing conditions confounded causality.