Does Total Neoadjuvant Treatment Improve Overall Survival in Rectal Cancer? No, It Does Not

Abstract

Recently Kong et al. presented a meta-analysis comparing total neoadjuvant treatment (TNT: chemotherapy plus [chemo]radiation) with neoadjuvant chemoradiation (control) for rectal cancer. Overall survival (OS) drew the authors’ attention as the most important outcome used in decision-making. A higher pooled 3-year OS rate was reported in the TNT group than in the control group (84.9% vs 82.3%; p = 0.006). However, this evaluation might be at risk for bias due to methodologic limitations. To reduce this risk, the authors repeated the meta-analysis using a different methodology. They included only randomized studies comparing TNT with neoadjuvant (chemo)radiation and not the combination of both randomized and nonrandomized studies. They quantified the difference in OS, taking into account all deaths and not only those occurring in the first 3 years. They included updated instead of early results. Moreover, they incorporated two recent publications. The study enrolled 2176 patients from five randomized studies (Table 1; eFig. 1). The supplementary material details the methods and results. The difference in OS was measured using the hazard ratio (HR). A subtotal metaanalysis of the trials with neoadjuvant chemotherapy in the TNT arm of the study delivered for 3 months or longer was performed because, to our knowledge, no evidence exists to show that shorter neoadjuvant or adjuvant chemoradiotherapy can reduce the risk of distant metastasis in colorectal cancer (the primary goal of TNT). The meta-analysis showed no difference in OS, neither in all the trials nor in the subset of the four studies using neoadjuvant chemotherapy for 3 months or longer. The HR of death in the TNT group compared with the control group was 0.89, 95% confidence interval (CI) 0.75–1.06, p = 0.18, heterogeneity p = 0.42, I = 0% and 0.88, 95% CI 0.69–1.12, p = 0.29, heterogeneity p = 0.27, I = 23%, respectively (Fig. 1). The results are consistent with those obtained in a meta-analysis by Liu et al., wherein the respective HR was 0.88 (95% CI 0.74–1.05; p = 0.15). A limitation of the meta-analysis was the use of literature-based data rather than individual patient data. Moreover, the follow-up period was rather short, ranging from 3.9 to 7.0 years. Notably, the trial designs were somewhat heterogeneous. That is, the studies did not address the same question exactly (see remarks in Table 1). Only the GCR-3 trial had a pure design. Two large studies included in this meta-analysis showed an absolute 7% reduction in the rate of distant metastasis, resulting in an improvement of 3-year disease-free survival (DFS) after TNT. This benefit likely will translate into improved OS after extended follow-up evaluation. Such a scenario is possible because findings show 3-year DFS to be a surrogate end point for 6to 7-year OS in stage 3 colon cancer. However, this surrogacy has not been shown in stage 2 colon cancer, nor in rectal cancer. One reason for this is the shorter survival after relapse in the adjuvant Supplementary Information The online version contains supplementary material available at https://doi.org/10.1245/s10434021-10432-0.