Cisplatin-based chemotherapy improves survival in appropriately selected patients with stage IV non-small cell lung cancer (NSCLC). However, cisplatin-based regimens have well-known dose-related toxicities, particularly renal insufficiency and neurotoxicity. On the basis of prior preclinical and phase I studies, we initiated a phase II study of SPI-77 (STEALTH) Liposomal Cisplatin) in patients with stage IIIB and IV NSCLC who failed previous treatment with platinum. Disease in all subjects had progressed during therapy, failed to respond, or progressed within 3 months after discontinuing the platinum-based chemotherapy. Between January and June 1999, 13 patients were enrolled at our institution. Patient characteristics included: seven women, six men; median age, 61 years; median Karnofsky performance status, 80%; median number of prior chemotherapy regimens, two (range, 1-3). All patients had adequate hepatic and renal function. SPI-77 was administered at a dose of 260 mg/m(2) IV every 3 weeks. A median of two cycles (range 1-6) were given; the total number of cycles was 35. Among the 12 patients evaluable for response, two had (17%) stable disease and ten (83%) had progressive disease. The median survival was 24.3 weeks, and the median follow-up was 43.9 weeks. Toxicity could be evaluated in all subjects. Moderate anemia (46% of cycles, <or=grade 2; 3% of cycles, >or=grade 3) with minimal granulocytopenia and thrombocytopenia (26% of cycles grade 1; 0% of cycles, >or=grade 2) were the most notable manifestations of myelosuppression. Grade 3 nonhematological toxicities included dyspnea (8%), fatigue (8%), and pain (8%). There were no grade 4 toxicities. These data suggest that this liposomal cisplatin formulation does not have appreciable activity in this population of patients with NSCLC who had received prior platinum-based chemotherapy. The lack of encouraging results from SPI-77 use in other phase I and II studies resulted in early closure of this trial by the manufacturer.