Editorial: bariatric surgery to reduce the risk of liver cancer

PIROLA, CARLOS J.; SOOKOIAN, SILVIA

ALIMENTARY PHARMACOLOGY & THERAPEUTICS.

WILEY-BLACKWELL PUBLISHING, INC

Año: 2021 vol. 53 p. 1155 - 1156

0269-2813

Obesity and its systemic consequences are responsible for a continuing increase in morbidity and mortality rates,1 the latter in part explained by the alarming rise of cancer-associated deaths.The symbiotic relationship between obesity, type 2 diabetes, and nonalcoholic fatty liver disease may lead to hepatocellular carcinoma (HCC), particularly in patients who develop steatohepatitis and cirrhosis. In this issue of Alimentary Pharmacology & Therapeutics, Ramai and colleagues evaluated the impact of bariatric surgery on the incidence of HCC by a meta-analysis.The authors pooled adjusted and unadjusted odds ratios (ORs) extracted from abstracts or full texts on retrospective studies of patients undergoing bariatric surgery. A total of 18 423 546 controls and 1 091 204 bariatric patients was included. A quantitative estimate of HCC incidence showed an unadjusted OR of 0.40 that favoured bariatric surgery. However, high and significant heterogeneity (I2: 79%) was observed. This interesting report addressed an important topic. The study has the strength of a large sample size and the chance of comparing the incidence of HCC among subjects that undergo bariatric surgery versus presumably matched controls. Unfortunately, there are many unanswered questions. Most importantly, in some studies, the follow-up period is missing from the original reports, and most of the studies are based on population-basedcancer and/or clinical registries. Hence, an inquisitive reader can wisely raise questions and concerns regarding the robustness of the conclusion. The findings are certainly plausible. Prior studies showed that bariatric surgery is associated with a significant reduction in overall cancer incidence with the effect most marked in ?obesity-related? cancers such as breast and colon HCC would clearly fall into this category. Nevertheless, there remain significant challenges over both the magnitude and mechanism/s of risk reduction. Some of these are methodological with the inevitable heterogeneity in large cohorts lessening the generalizability of the estimates. There are many potential confounders that may affect the magnitude of the effect, matching of controlgroups consistently with the patients who underwent surgery in terms of age, sex, type 2 diabetes, and any other intervention, including medications, for instance statins or metformin, is unlikely tohave been uniform in the studies included. There is also a disbalance between the number of cases and population-based controls (~1:17), which not only might amplify the putative presence of uncontrolled confounding variables but also challenges the matching strategy that is used for stratification analysis. A question that cannot be answered by this meta-analysis is whether a change in BMI after surgery explains the reduction in HCC incidence. Likewise, it remainsunclear whether the effect was explained by the improvement of type 2 diabetes. Finally, Ramai et al?s study4 did not include information about underlying clinical conditions, particularly the presenceof liver cirrhosis pre-surgery. While HCC does occur in non-cirrhotic patients the majority of the risk would still be expected to arise in those with significant liver disease and quantification of this riskwould have important practical implications.It may be possible with future studies to improve the precision of effects estimates and answer some of the open questions. It is biologically plausible that bariatric surgery through obesity controldecreases HCC incidence and it is encouraging to see evidence to support this. The magnitude of risk reduction and key patient related factors which may increase or decrease ongoing risk requires further elucidation.