Response to Faitot et al “Two-stage Hepatectomy Versus 1-Stage Resection Combined With Radiofrequency for Bilobar Colorectal Metastases: A Case-matched Analysis of Surgical and Oncological Outcomes”
The article by Faitot et al1 compares the retrospective experiences of 2 expert liver centers comparing 1-stage (S1) with 2-stage (S2) strategies for extensive colorectal liver metastases (CRLM). Essentially, S2 utilizes portal vein ligation/embolization and major hepatectomies and has long been considered as the standard of care for the surgical management of this situation whereas the newer S1 approach uses minor hepatectomies often combined with intraoperative ablation in a sole procedure. The aim of the article was to compare the outcomes of both strategies, and the authors conclude that the strategies are equivalent, a point against which we would like to argue.
The choice of methodology, a retrospective case-matched paired series, is unfortunately frequently used in the surgical literature as an ersatz randomized trial. Nevertheless, such methodology is biased and underpowered (only 78 paired patients). Moreover, the fact that the proponents of the 2 strategies were opposed for a long time and have now come together might be perceived as an attempt for peace at the price of scientific perception.
S2 hepatectomies are concluded to offer better oncological control, first, by removing more micrometastases and, second, by lower stimulation of the remaining micrometastatic disease. For the first point, Cauchy et al2 have shown that extensive hepatectomy is not the best strategy for advanced CRLM, especially if more than 12 cycles of chemotherapy are required preoperatively. From a theoretical point of view, if a patient has bilateral CRLM, he or she also has bilateral dormant micrometastases, and we cannot be convinced that performing a right extended hepatectomy could impact positively on the micrometastases remaining in the left liver. The authors state that intraliver recurrence after ablation would be higher than after resection and cite a retrospective comparison by Tanis et al as proof of the local recurrence rates of ablation in the CLOCC trial versus the EPOC trial. How can we accept this assertion when patients in the CLOCC study were technically unresectable at entry and had a median of 4 CRLMs compared with those of the EPOC trial who were easily resectable at entry and had a median of 1 CRLM? In the same manner, the authors cite a retrospective appraisal of early recurrences after CRLM treatment by Vigano et al, who reported that intraoperative radio-frequency ablation would be a strong prognostic factor for recurrence, forgetting to add that 0-mm margin resections were also such a prognostic factor of recurrence, and finally, omitting to observe that both of these factors are only markers of disease severity and are normally present when patients have extensive surgery for extended disease. For this reason, we disagree with the conclusions reached by Faitout et al. Many liver surgeons now believe that a parenchyma-saving approach is strategically superior to a more radical approach3 and that nonanatomical resections are superior to anatomical resections.
With regard to the second point, there are at least as many experimental articles reporting a higher rate of biological stimulation after ablation as that seen after liver resection4 than those reporting the opposite effect.5 Moreover, Elias et al6 have reported clear stimulation of CRLM by portal vein mobilization, a major component of the S2 approach. This high level of cytokine stimulation might explain the high level of dropout between the first and second stages. The S2 strategy is like a biological Russian roulette. When successfully completed, high survival rates are observed, but when patients do not proceed to the second stage, survival is significantly poorer. For this latter group, 5-year survival was inferior to that seen after chemotherapy alone.