Eight RCTs (n=3,083 patients) were included in the review. Several trials did not describe or undertake randomisation, allocation concealment, or withdrawals; none of the trials were double-blind. Trial sample size ranged from 61 to 784 patients.
Irinotecan-based versus etoposide-based therapy (six trials): Compared with etoposide-based treatment, irinotecan-based treatment showed a statistically significant improvement in overall survival (HR 0.84, 95% CI 0.75 to 0.93; I2=45%; five trials) and progression-free survival (HR 0.83 95% CI 0.73 to 0.95; I2=0; two Western trials) in patients with extensive disease small-cell lung cancer. There was no significant difference in overall response rates. Compared with etoposide-based treatment, irinotecan-based treatment had a statistically significantly greater risk of grade 3/4 diarrhoea (OR 8.94, 95% CI 5.30 to 15.07; I2=40%), but a statistically significantly lower risk of grade 3/4 anaemia, leukopenia, neutropenia, neutropenic fever, and thrombocytopenia; the number needed to harm ranged from three to 26 patients. Subgroup analysis of the Western trials for overall survival provided similar results to the main analysis. Sensitivity analysis did not significantly alter the results for overall survival or progression-free survival.
Topotecan-based therapy versus etoposide-based therapy (two trials): There was no significant difference in overall survival between topotecan-based and etoposide-based therapy. Meta-analyses could not be performed for progression-free survival, overall response rates, or adverse events due to conflicting findings.
There was no evidence of publication bias.