Small intestinal bacterial overgrowth and Helicobacter pylori: can they be cause of thrombocytopenia in patients with chronic liver disease?

To The Editor: We read the recent investigation by Kalambokis et al. (1) with interest. In this randomized placebo-controlled study the authors treated 10 cirrhotic patients with rifaximin, administered at a dose of 1,200 mg daily for 4 weeks, and 13 patients were in the placebo group. They observed a significant increase in platelet count in patients with thrombocytopenia who received rifaximin, while this parameter remained unaltered in all thrombocytopenic patients in the placebo group. The Authors attributed this finding to the potential eradication of small intestinal bacterial overgrowth obtained administering high-dose rifaximin, since it has been previously demonstrated that in cirrhotic patients the increased intestinal permeability may lead some bacteria to migrate into the systemic circulation. This can theoretically decrease platelet count through several mechanisms: i) enhancing nitric oxide production and splanchnic vasodilation, thus leading to increased portal hypertension; ii) increasing circulating levels of tumor necrosis factor-α and interferon-γ that may act both centrally - suppressing the growth and differentiation of bone marrow megakaryocytes - and peripherically decreasing thrombopoietin liver synthesis; iii) stimulating hypergammaglobulinemia with the formation of antibodies reacting with platelet surface antigens (2,3). However, another possible explanation of their findings has to be taken into consideration. In fact, it is well known that H. pylori (Hp) infection has a high prevalence in cirrhotic patients (4). Moreover, Hp has been associated to systemic effects similar to those previously described for intestinal bacterial overgrowth, and, in particular, favors the development of anti-platelet antibodies, thereby causing secondary immune thrombocytopenia (2,3,5). Previous studies demonstrated that rifaximin, administered alone or in combination with other antibiotics for a short period (7-10 days), is able to eradicate H. pylori infection in patients with Hp-related chronic active gastritis (6). Given these hypotheses and considering that cirrhotic patients are often treated with chronic/cyclic antibiotic therapy for selective intestinal decontamination as prophylaxis for hepatic encephalopathy or spontaneous bacterial peritonitis, the significant increase in platelet count in the cirrhotic patients with thrombocytopenia observed by Kalambokis et al. could be due to the eradication of Hp infection likely present in them, with the consequent decrease of circulating anti-platelet antibodies and peripheral platelet destruction. Our observations suggest that further studies aimed at investigating the effects of long-term rifaximin administration on cirrhosis-associated thrombocytopenia should consider not only the presence of intestinal bacterial overgrowth but also the potential presence of chronic Hp infection.