Metaanálisiscombinación de tratamiento endoscópico y farmacológico para prevenir el resangrado por varices esofágicas en la cirrosis hepática

Abstract

Background: The best option to prevent variceal rebleeding in cirrhosis, whether endoscopic or drug (non-selective beta-blockers) therapy remains controversial. Recent clinical guidelines recomend the combination of endoscopic (band ligation) and drug (beta-blockers) therapy as the first-line therapeutic option to prevent variceal rebleeding in cirrhosis (Hepatology. 2007; 46: 922. Am J Gastroenterol. 2007; 102: 2086). This recomendation is not supported by the last consensus conference of Baveno IV, when the evidence to support this afirmation was wear (J Hepatol. 2005; 43: 167). Purpose: To determine the efficacy of endoscopic (sclerotherapy or banding) plus drug therapy compared to either therapy alone to prevent variceal rebleeding and to improve survival. Data sources: English and non-English-language randomized trials appearing in MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials and Cochrane Database of Systematic Reviews, and conference proceedings until 30 December 2007. Study selection: Randomized-controlled trials comparing endotherapy plus betablockers with endoscopic therapy or beta-blockers to prevent rebleeding in cirrhosis. Data extraction: Two coauthors independently extracted data regarding interventions, rebleeding and mortality. We pooled data according to the Mantel–Haenszel fixedeffects model and if the heterogeneity was significant, according to the random-effects model. Primary outcomes measures were pooled relative risk for rebleeding and Peto odds ratio for mortality. Secondary outcomes were variceal rebleeding, variceal recurrence, and adverse events. Meta-regression and stratified analysis were used to explore heterogeneity. Data synthesis: 23 trials (1860 patients) contributed to the meta-analysis, which revealed a statistically significant reduction in upper digestive rebleeding in favour of combination therapy compared with endoscopic therapy (0.68, 95%CI 0.52-0.89; moderate heterogeneity I2 =61%) or beta-blockers (0.71, CI 0.59-0.86; no heterogeneity). Combination therapy also reduced variceal rebleeding and recurrence. Reduction in mortality by combination therapy was not statistically significant compared with endoscopic (0.78; CI 0.58-1.07) or drug therapy (0.70; CI 0.46-1.06). Effect sizes were independent of the endoscopic procedure. Meta-regression or stratified analysis did not identify any trial-level covariate associated with intevention effect. Limitations: Variability among trials regarding quality, populations under study, endoscopic procedures and length of follow-up. Conclusions: Endoscopic plus drug therapy reduces the risk of upper digestive and variceal rebleeding in cirrhosis compared with either therapy alone, and constitutes the best option for secondary prophylaxis of variceal bleeding, yet the benefit on survival is not statistically significant.