Thirteen RCTs (n= 4,144) were included in meta-analyses, nine of which reported time-to-event statistics. Seven RCTs were published as full papers and six in abstract form only. Methodological quality varied between trials. Almost all trials reported intention-to-treat analyses. However, only seven trials reported the concealment of randomisation. Masking of intervention and outcome assessment was inadequate in most trials.
When the studies were pooled, for RCTs evaluating induction thalidomide, thalidomide was significantly associated with an improvement of overall survival when added to standard non-transplantation therapy (HR 0.67, 95% confidence interval, CI: 0.56, 0.81, p<0.0001; four RCTs).
For RCTs evaluating maintenance thalidomide following ASCT, thalidomide was not significantly associated with an improvement in overall survival (HR 0.61, 95% CI: 0.37, 1.01, p=0.05; four RCTs). When excluding one trial that evaluated induction and maintenance therapy, thalidomide was significantly associated with a survival advantage (HR 0.49, 95% CI: 0.32, 0.74, p=0.0007; three RCTs).
The addition of thalidomide to standard induction therapy significantly improved myeloma response rate (RR 1.50, 95% CI: 1.33, 1.68, p<0.00001; nine RCTs), and complete response to induction thalidomide (RR 2.82, 95% CI: 1.80, 4.41, p<0.00001; nine RCTs). Induction thalidomide was significantly associated with an increased risk of venous thromboembolism (VTE) (RR 2.56, 95% CI: 1.89, 3.49, p<0.00001; 10 RCTs), as was maintenance thalidomide (RR 1.95, 95% CI: 1.15, 3.30, p<0.01; three RCTs).
Statistically significant heterogeneity was observed in the outcome of overall response rate (p=0.0009, I2 = 69.8%) and complete response rate (p=0.06, I2 = 47.1%). Sensitivity analyses of response rates between published and unpublished studies did not materially affect the results.