Is There a Place for Adjuvant Chemotherapy in the Treatment of Locally Advanced Cervical Cancer?

Abstract

Findings on the efficacy of adjuvant chemotherapy (ACT) of locally advanced cervical cancer (LACC) after the concurrent chemoradiation (CCRT) therapy were inconsistent, and the OUTBACK trial was expected to shed some light regarding the topic. Its results on ACT in LACC were negative, with the conclusion of not to use it. The objective of this review was to present the inconsistencies of previous studies, along with the OUTBACK trial in more detail, and to rethink whether its results provide an unambiguous and definite answer to the optimal position of ACT in the treatment of LACC. To critically appraise the OUTBACK trial and understand the consequences of its results, we used only randomized controlled studies (RCTs) on ACT in LACC that have been included in high-quality systematic reviews and meta-analyses. We calculated the pooled prediction intervals using a random effects meta-analysis of all published randomized studies including the OUTBACK trial. After combining the OUTBACK trial with the results of four previous randomized trials, the pooled hazard ratio for overall survival benefit of CCRT + ACT was 0.95 (95% CI 0.75; 1.20). The pooled hazard ratio of the four previous trials was 1.00 (95% CI 0.69; 1.44). The OUTBACK trial improved the precision of the pooled estimate, but the clinical heterogeneity and the consequent prediction intervals are still very wide, and with 95% reliability, we can expect that if the new study, using a similar approach to the ACT, on a randomly selected patient population from the presented five trials is conducted, its hazard ratio for overall survival after ACT would be between 0.47 and 1.93. In conclusion, there is an absolute need for further research in order to optimally define the position of ACT in the treatment of LACC.

Keywords: adjuvant chemotherapy; concurrent chemoradiation; locally advanced cervical cancer; uterine cervical neoplasms.

Figure 1

Forest plot of the effects of CCRT + ACT compared to CCRT alone on overall survival; gray squares represent each study hazard ratio (HR) for dying from any cause and whiskers represent 95% confidence intervals (CI); the size of the squares represent the weight of the study inverse to the study variance; black diamonds represent the pooled hazard ratio for older trials and for all trials (older and OUTBACK trial) calculated using a random effects model with a restricted maximum-likelihood method; whiskers from the lower “overall” diamond represent 95% prediction interval (PI); studies are sorted by the year of the end of enrollment.