The aim of this study was to evaluate the toxicity and the activity of two non-cross-resistant platinum compounds: oxaliplatin (L-OHP) and cisplatin (CDDP) in platinum pretreated ovarian cancer patients. Chemotherapy consisted of L-OHP and CDDP given sequentially as 2 h infusions on day 1 at their standard recommended dose (130 mg/m2 for oxaliplatin, 100 mg/m2 for cisplatin) every 3 weeks. Dose reductions (20-35%) were planned according to baseline haematological and renal status, but the dose ratio between L-OHP and CDDP was always maintained at 1.3. Cycles were repeated until progression or treatment limiting toxicities. From September 1992 to November 1994, 25 patients with pretreated ovarian cancer entered this salvage programme. They had received a median number of three previous chemotherapy lines (1-7), one at least platinum based. Previously cisplatin had been given to 22 patients at a median total dose of 600 mg/m2 (170-1175), while 18 had received carboplatin to a median total dose of 1135 mg/m2 (200-2450). 9 patients had also received and were resistant to taxanes (paclitaxel, 6 patients, docetaxel, 3 patients), while the rest were considered ineligible for simultaneously ongoing single-agent taxane phase II trials. 13 and 12 patients, respectively, were considered to have platinum refractory and potentially sensitive disease, according to Markman's criteria. 77 cycles of L-OHP/CDDP were given, with a median of three cycles/patient (range 1-6) and were evaluable for toxicity. The limiting toxicity of the L-OHP/CDDP combination was a cumulative, sensory peripheral neuropathy, severe (> or = grade 3 CTC) after more than three cycles, but reversible within a few months of its discontinuation. Grade 3-4 (WHO scale) neutropenia and thrombopenia were seen in 35-40% of cycles, with one neutropenic treatment-related death (septic shock). 22 patients with measurable/evaluable disease were assessable for antitumoral activity. Two complete responses (CR) (8%) (one proven histologically at laparotomy (pCR)) and 8 partial responses (PR) (32%) for an overall objective response rate (ORR) of 40% (95% CI, 21-61%) (intent to treat). The median duration of response was 4 months. Seven responses were seen among 12 potentially platinum-sensitive tumours (58%, CI 95% 28-85%), while 3/13 platinum refractory patients (23%, CI 95% 5-54%) had an objective response. These encouraging results are the basis for new first- and second-line combination treatment programmes in ovarian carcinoma.