Carvedilol in patients with compensated cirrhosis: The ongoing benefits of definitive randomised trials over meta-analysis in patients with small varices

Carvedilol reduces the risk of decompensation and mortality in patients with compensated cirrhosis in a competing-risk meta-analysisJournal of HepatologyVol. 77Issue 4PreviewWhether non-selective β-blockers can prevent decompensation of cirrhosis warrants clarification. Carvedilol might be particularly effective since its intrinsic vasodilatory activity may ameliorate hepatic vascular resistance, a major mechanism of portal hypertension in early cirrhosis. We assessed whether carvedilol may prevent decompensation and improve survival in patients with compensated cirrhosis and clinically significant portal hypertension (CSPH). Full-Text PDF Reply to: “Carvedilol in patients with compensated cirrhosis: The ongoing benefits of definitive randomised trials over meta-analysis in patients with small varices”Journal of HepatologyVol. 79Issue 1PreviewWe thank McPhail et al. for their interest in our recent individual patient data meta-analysis (IPD-MA) assessing the value of carvedilol in compensated cirrhosis.1 However, we would like to highlight relevant methodological issues with several points raised in their letter,2 just to enhance the picture by improving the focus. Full-Text PDF We read with interest the article by Villanueva et al.[1]Villanueva C. Torres F. Sarin S.K. Shah H.A. Tripathi D. Brujats A. et al.Carvedilol reduces the risk of decompensation and mortality in patients with compensated cirrhosis in a competing-risk meta-analysis.J Hepatol. 2022; 77: 1014-1025Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar on the use of carvedilol in patients with compensated cirrhosis. This is a timely article given the ongoing discussions on the appropriateness of carvedilol in patients with clinically significant portal hypertension in the wake of changes in endoscopic surveillance during the COVID-19 pandemic. Their use of individual patient data is to be commended. The authors present a meta-analysis using patient level data and suggest that the benefit of carvedilol over either placebo or endoscopic variceal ligation (EVL) warrants its more widespread use. While of interest, there are significant concerns regarding this statement based on a small meta-analysis and especially, as we contend, one where pooling of studies with different control groups is controversial. The authors suggest that there is clear evidence to suggest carvedilol is of benefit in patients to prevent decompensation, but a closer look at their data suggests otherwise and we would caution against an overly optimistic interpretation, particularly in light of ongoing clinical trials in overlapping patient groups. It is noteworthy that the control group is not homogeneous across the meta-analysis, with placebo used in two studies[2]Villanueva C. Albillos A. Genesca J. Garcia-Pagan J.C. Calleja J.L. Aracil C. et al.Beta blockers to prevent decompensation of cirrhosis in patients with clinically significant portal hypertension (PREDESCI): a randomised, double-blind, placebo-controlled, multicentre trial.Lancet. 2019; 393: 1597-1608Abstract Full Text Full Text PDF PubMed Scopus (299) Google Scholar,[3]Bhardwaj A. Kedarisetty C.K. Vashishtha C. Bhadoria A.S. Jindal A. Kumar G. et al.Carvedilol delays the progression of small oesophageal varices in patients with cirrhosis: a randomised placebo-controlled trial.Gut. 2017; 66: 1838-1843Crossref PubMed Scopus (61) Google Scholar and EVL used in two studies.[4]Tripathi D. Ferguson J.W. Kochar N. Leithead J.A. Therapondos G. McAvoy N.C. et al.Randomized controlled trial of carvedilol versus variceal band ligation for the prevention of the first variceal bleed.Hepatology. 2009; 50: 825-833Crossref PubMed Scopus (204) Google Scholar,[5]Shah H.A. Azam Z. Rauf J. Abid S. Hamid S. Jafri W. et al.Carvedilol vs. esophageal variceal band ligation in the primary prophylaxis of variceal hemorrhage: a multicentre randomized controlled trial.J Hepatol. 2014; 60: 757-764Abstract Full Text Full Text PDF PubMed Scopus (69) Google Scholar This likely reflects the presence of no or small varices in two studies (hence a placebo control is ethically warranted) and EVL as a control for medium/large varices in the other two studies. No assessment of publication bias is presented in the paper by Villanueva et al., which they report as a weakness of the individual patient analysis but do not explore further. In addition to this we note the authors use of a Log10 axis scale will over emphasise the effect toward the left-hand side, reducing the visual effect. Therefore in our repeat analysis (Fig. 1), while we also use the log scale, we use the generic inverse variance method as this is optimised against these multiple biases. In a further analysis (performed in MedCalc v20.118 [MedCalc Software, Ostend, Belgium] and RevMan v5.4 [The Cochrane Collaboration, 2020]), we perform an analysis where we limit the studies to those with a placebo control (the no or small varices group); in this analysis, the odds ratio does not demonstrate a significant effect for either decompensation (pooled odds ratio 0.40; 95% CI 0.12-1.36) or death (pooled odds ratio 0.54; 95% CI 0.21-1.38). Furthermore, Egger’s test for publication bias is also positive (p <0.001) in the placebo-controlled subgroup, indicating strong publication bias. This may be an effect of the authors’ strategy of seeking patient level data but casts further doubt on the use of non-selective beta-blockers in this group without further evidence or an appraisal of a wider set of studies. Therefore, assuming this meta-analysis to be inconclusive but informative for larger randomised trials, and assuming an annual decompensation/death rate of 0.25 and a hazard ratio of 0.5, then a minimum of 420 patients (not accounting for dropout) would be required to definitively answer this question in patients with no or small varices. This is higher than the pooled numbers in this analysis. In conclusion, while an important paper, this meta-analysis is unlikely to replace a larger prospective randomised-control trial which mitigates the bias we demonstrate in the reported meta-analysis. While it supports the use of carvedilol to prevent decompensation in patients with medium or large varices, the situation is not so clear in those with smaller varices. As the BOPPP (Beta-blockers or placebo for primary prophylaxis of oesophageal varices) study in patients with small varices is also collecting data on patients decompensating and has liver transplant as a competing event, there will be better quality data from this study in the future that can answer this question definitively. However, as our subgroup analysis demonstrates, there remains an unanswered clinical question for patients with compensated cirrhosis and small varices, and clinical practice should not change based on this meta-analysis. All authors are in part supported by the National Institute for Health Research, UK who funded the BOPPP study. The authors declare no conflicts of interest that pertain to this work. Please refer to the accompanying ICMJE disclosure forms for further details. All authors wrote and edited the work. MM conceived and performed the statistical analysis. The following are the supplementary data to this article: Download .pdf (.24 MB) Help with pdf files Multimedia component 1