Five trials (942 patients) were included in the review; one was a phase III trial, the remaining four were phase II trials. All scored trials 3 out of 5 for quality and reported randomisation, allocation concealment, and drop-outs.
There were no statistically significant differences between once weekly and once every three weeks paclitaxel-based chemotherapy for overall survival (five trials), progression-free survival (four trials), or overall response rate (five trials). No differences were observed for carboplatin/paclitaxel trials.
Median survival was 9.8 months for once-weekly chemotherapy and 10.7 months for chemotherapy once every three weeks. Median progression-free survival was 5.2 months for once-weekly chemotherapy and 4.7 months for chemotherapy once every three weeks.
The most commonly reported grade 3 and 4 adverse events were haematological toxicity (anaemia, leukocytopenia, thrombocytopenia or granulocytopenia), fever, and peripheral neuropathy. There was a lower incidence of neutropenia (OR 0.47, 95% CI 0.27 to 0.83), febrile neutropenia (OR 0.46, 95% CI 0.21 to 0.98) and peripheral neuropathy (OR 0.50, 95% CI 0.33 to 0.76) and a higher incidence of anaemia (OR 2.08, 95% CI 1.20 to 3.58) for once-weekly chemotherapy compared with once every three weeks chemotherapy. There was no difference in treatment-related deaths.
No evidence of publication bias was found (funnel plots and test results not shown), although this was a cautious interpretation as there were only five trials in the analysis.