Navigating the controversy regarding antibiotic prophylaxis in acute variceal bleeding

Abstract

Antibiotic prophylaxis in patients with cirrhosis and acute variceal bleeding is part of the standard of care according to most clinical guidelines. However, with recent evidence arguing against antibiotic prophylaxis, the role of this intervention has become less clear.

Key Words: Cirrhosis, Acute variceal bleeding, Antibiotic prophylaxis, Endoscopic band ligation, Spontaneous bacterial peritonitis

Core Tip: Although antibiotic prophylaxis is currently a part of guideline-directed therapy for patients with cirrhosis and acute variceal bleeding, new and conflicting evidence has challenged this premise. While previous meta-analyses supported prophylaxis, a recent retrospective cohort study by Ichita et al regarding acute esophageal variceal bleeding found no significant differences in mortality, rebleeding, or bacterial peritonitis between prophylaxis and non-prophylaxis groups. A tailored approach that considers underlying liver function, the method of endoscopic treatment, and local antibiotic resistance patterns might be more appropriate.

TO THE EDITOR

We read with interest Ichita et al[1] published work regarding the effectiveness of antibiotic prophylaxis for patients with acute esophageal variceal bleeding (AVB) treated by band ligation. In this multicentric and retrospective cohort comparing mortality, rebleeding and spontaneous bacterial peritonitis among patients who did or did not receive prophylaxis, no significant differences were found, and this lack of statistical significance persisted after inverse probability of treatment weighting. Furthermore, subgroup and sensitivity analyses revealed no differences across Child-Pugh classes, specific antibiotics used or duration of therapy[1]. This finding challenges the current treatment paradigms and suggests that there should be a personalized approach to antibiotic use in these patients.

The subject at hand has been a matter of ongoing debate[2]. A systematic review and meta-analysis published by Chavez-Tapia et al[3,4] revealed that antibiotic prophylaxis decreased rebleeding, infection-related mortality, and all-cause mortality[3,4]. Since then, Western guidelines have chosen to endorse antibiotic prophylaxis, making it the standard of care[5-7]. Consistently, a more recent meta-analysis by Wong et al[8] reiterated the beneficial effects of prophylactic antibiotics on rebleeding, six-week mortality and infection-related mortality[8]. However, it is worth mentioning that endoscopic outcomes have improved since the shift from sclerotherapy to ligation[9], and several studies included in both meta-analyses might not accurately reflect current standards of care[3,4,8]. As a result, the role of indiscriminate prophylaxis in variceal bleeding has become less clear.

Like the authors, we also are concerned about antibiotic resistance and find their stance on the search for an individualized approach to be reasonable. Other studies, consistent with this one, suggest that the benefits of prophylaxis may not be as important as previously considered[10]. An observational study by Martínez et al[11] revealed that bacterial infections developed in up to one-fifth of patients with cirrhosis and AVB despite antibiotic prophylaxis. However, most infections were respiratory and were associated with advanced liver disease[11]. Similarly, Tandon et al[12] published a retrospective cohort in which patients with cirrhosis and AVB who were classified as Child-Pugh class A had lower rates of infection and lower mortality without antibiotic prophylaxis than those classified as class B or C[12]. However, how can this be tested prospectively in a randomized clinical trial?

Because antibiotic prophylaxis is a guideline-directed treatment according to European and American guidelines, prospectively comparing it to placebo in a randomized clinical trial could raise ethical concerns. A potential solution might lie in countries such as Japan, where a well-executed prospective study could be conducted, given that Japanese (1^st Edition) guidelines do not mandate clinicians to administer prophylaxis and it is not currently part of the standard of care[13].

Conversely, a prospective study comparing third-generation cephalosporins (i.e., ceftriaxone) vs an appropriate carbapenem (i.e., ertapenem) in the treatment of AVB at centers with high rates of extended-spectrum beta-lactamase-producing microorganisms could be undertaken. If improved outcomes were to be identified in the carbapenem group, it could be inferred that prophylaxis (with sufficient coverage) is beneficial in these patients. Regardless, in a retrospective study by Mücke et al[14], patients with cirrhosis and AVB were screened for multidrug-resistant organisms (MDROs) at the time of bleeding, and the efficacy of antibiotic prophylaxis was determined according to microbiological susceptibility testing. This study revealed that MDRO colonization was not associated with a greater risk of rebleeding, infection or death, even when antibiotic prophylaxis had insufficient coverage, suggesting that there is no benefit of providing adequate coverage[14].

To conclude, we commend the authors on their work and consider it relevant. Studies such as this one help us reflect on our practices and reconsider our recommendations as medicine continues to evolve. It is paramount to determine which patients, if any, benefit from antibiotic prophylaxis in the setting of cirrhosis and AVB. Furthermore, the beneficial effect of rifaximin on infections in patients with AVB might also be worth exploring. Although rifaximin use was found to be associated with a reduced risk of variceal bleeding in a retrospective cohort[15], a randomized clinical trial by Sharma et al[16] testing lactulose plus rifaximin or placebo for the reversal of hepatic encephalopathy showed reduced mortality with no differences in bleeding-related deaths[16]. It is possible that an individualized approach might be more appropriate, considering the degree of compensation of any given patient, the type of endoscopic hemostasis used and, possibly, the local patterns of antibiotic resistance.

All told, we maintain that prospective studies using current guideline-directed therapy (i.e., those with endoscopic band ligation) are warranted before we can safely disregard antibiotic prophylaxis in these patients.