Seven RCTs with a total of 1,524 participants were included in the review.
There was no indication of publication bias (P=0.66).
Two-year overall survival was significantly greater with ERT than with LRT (risk ratio, RR=1.17, 95% confidence interval, CI: 1.02, 1.35, P=0.03). The result for 3-year overall survival still favoured ERT over LRT, but was no longer statistically significant (RR 1.13, 95% CI: 0.92, 1.39, P=0.23). No statistically significant heterogeneity was detected. The NNT for one additional survivor was 19 at 2 years and 42 at 3 years.
In sensitivity analyses, the removal of two of the trials using sequential or alternating strategies for chemotherapy had little effect on 2- or 3-year overall survival. However, after the removal of one large study which used concurrent chemotherapy in the ERT arm, but a sequential strategy for the LRT arm, there was no longer a significant benefit of ERT over LRT at 2 years' follow-up (RR, 1.09, 95% CI: 0.93, 1.28, P=0.28).
Studies using a twice daily regimen showed a statistically significant benefit of ERT over LRT for both 2-year overall survival (RR 1.44, 95% CI: 1.17, 1.83, P=0.001; NNT 6) and 3-year overall survival (RR 1.39, 95% CI: 1.02, 1.90, P=0.04; NNT 11). No significant benefit of ERT over LRT was seen for patients receiving a once daily regimen. Studies using platinum-based chemotherapy demonstrated a statistically significant benefit of ERT over LRT at both 2-year overall survival (RR 1.30, 95% CI: 1.10, 1.53, P=0.002; NNT 10) and 3-year overall survival (RR 1.35 95% CI: 1.07, 1.70, P=0.01; NNT 14). No significant benefit of ERT over LRT was seen for studies using non-platinum-based chemotherapy. Neither the studies using concurrent chemotherapy nor those not using concurrent chemotherapy showed a statistically significant difference between ERT and LRT for 2- or 3-year overall survival.
Results of the meta-regression found an absolute benefit in 2-year overall survival of 5% for ERT versus LRT when once daily radiation and platinum-based chemotherapy were used, compared with the referent category (once daily radiation and non-platinum based-chemotherapy). However, this did not reach statistical significance. When twice daily radiation and platinum-based chemotherapy were used, there was an 18% absolute benefit associated with ERT compared with LRT (P=0.002). There was a 9% absolute benefit in 3-year overall survival for ERT versus LRT when once daily radiation and platinum-based chemotherapy were compared with the referent category (P=0.03). When twice daily radiation and platinum-based chemotherapy were used, there was a 14% absolute benefit associated with ERT, compared with the referent category (P=0.004).