A systematic review of chemotherapy trials in several tumour types was performed by The Swedish Council of Technology Assessment in Health Care (SBU). The procedures for the evaluation of the scientific literature are described separately (Acta Oncol 2001; 40: 155-65). This overview on chemotherapy for epithelial ovarian cancer is based on a total of 176 scientific reports. Five meta-analyses including 17,291 patients, 33 prospective randomised studies including 12,340 patients, 36 prospective studies including 3,593 patients and one retrospective study including 421 patients. The studies include approximately 33,642 patients. The conclusions reached can be summarized into the following points: Radically operated patients with low-risk early ovarian cancer (stage IA or IB non-clear-cell well-differentiated carcinomas or borderline tumours) have a very good prognosis and there is no indication for adjuvant therapy. Radically operated patients with high-risk early ovarian cancer (clear cell carcinomas or FIGO stage IA or IB moderately or poorly differentiated carcinomas or stage IC) have a substantial risk for micrometastatic disease. However, the role of adjuvant chemotherapy is unclear and such therapy should, thus, only be used within clinical trials. The median overall survival for patients with advanced (FIGO stages II-IV) ovarian cancer randomised to paclitaxel/platinum-containing chemotherapy in three large studies ranged between 36-39 months. Compared with historical data, this represents a six to seven times longer median survival time than after surgery only. The probability for long-term survival for patients treated with a paclitaxel/platinum combination is too early to define. In two prospective randomised trials in advanced ovarian cancer, paclitaxel in combination with cisplatin has provided a survival benefit over cyclophosphamide/cisplatin. Based on these trials, paclitaxel/cisplatin is considered to be the standard treatment. This choice of standard therapy might, however, be questioned based on the results of the hitherto largest randomised study in advanced ovarian cancer, ICON3, which is, as yet only available in abstract form. It compared paclitaxel/carboplatin with carboplatin only or a platinum combination (cyclophosphamide/doxorubicin/cisplatin). There were no statistically significant differences in progression-free or overall survival. The drug regimen in the control arms of the previous studies showing superiority of the paclitaxel-cisplatin combination may not have been the optimal non-paclitaxel platinum-containing regimen. Three randomised studies have compared carboplatin/paclitaxel with cisplatin/paclitaxel. All three are hitherto only published as abstracts with short follow-up precluding survival analyses. None of them shows any difference in response rates. All three show less toxicity and one also better quality of life with carboplatin. Thus, there are preliminary data supporting the substitution of cisplatin with carboplatin. Intraperitoneal therapy with cisplatin caused improved survival compared with intravenous therapy in one ramdomised study. Further studies have shown trends to better survival and longer progression-free interval with intraperitoneal therapy. The accrual to studies on intraperitoneal chemotherapy has been poor reflecting that it is a cumbersome and not easily accepted treatment. In advanced ovarian cancer, no convincing advantage has been shown from more dose-intensive chemotherapy, without cytokines or bone marrow stem cell support, compared with standard doses. High response rates are achieved with high-dose chemotherapy with stem cell support in the salvage situation but response duration is short. Phase III studies evaluating high-dose chemotherapy in the first-line situation are ongoing. Until supportive controlled clinical trials are presented, high-dose chemotherapy should be confined to clinical trials. Tumour response is frequently observed on re-treatment with the same drugs as given first-line in patients sensitive to first-line platinum-based chemotherapy with a long progression-free interval. Thus, in these patients treatment with a platinum/paclitaxel combination might be recommended. albeit based on limited data. In patients resistant to first-line therapy, a number of single agents induce tumour responses in the range of 10-30%. The literature does not permit general treatment recommendations in these patients, which are recommended to be included in controlled clinical trials.
