ORIGINAL ARTICLES

NEOMYCIN IN THE TREATMENT OF HEPATIC COMA

Les patients atteints d'affections hépatiques chroniques présentent fréquemment un état de dénutrition ; des apports nutritionnels adaptés sont alors indispensables pour permettre d'en réduire les effets négatifs sur l'évolution clinique. On peut remédier à certaines complications des hépatopathies chroniques telles que la rétention hydrique par le régime alimentaire ou la restriction des fluides mais cela ne doit pas se faire aux détriments de l'équilibre nutritionnel. Une évaluation nutritionnelle est donc impérative chez ces patients non seulement au début de leur affection mais également à intervalles réguliers sur le long terme. Une évaluation raisonnablement fiable peut être obtenue en associant les antécédents cliniques et diététiques, l'examen clinique et les mesures anthropométriques. Sur cette base, les besoins nutritionnels peuvent être définis et optimisés au cas par cas. Il est indispensable que la prise en charge nutritionnelle constitue un élément à part entière du traitement de ces patients.

Patients with chronic liver disease are often malnourished, and the provision of adequate nutritional support is essential to help minimize its detrimental effects on clinical outcome. Certain complications of chronic liver disease, such as fluid retention, are managed with dietary manipulation or restriction, but this should not be at the expense of the adequacy of the diet as a whole. It is, therefore, essential that nutritional status is assessed in this patient population not only at presentation but also at long-term intervals. A reasonably reliable assessment can be made by combining data from clinical and dietary histories, the physical examination and an anthropometric evaluation. With such information, feeding can be targeted and optimized on an individual basis. The inclusion of nutritional management in the mainstream rather than the periphery of patient care is essential.

Hepatic encephalopathy (HE) is a metabolic-neurophysiologic syndrome that occurs in patients with liver disease. One of the main pathogenic mechanisms is represented by circulating toxins produced from the intestinal metabolism of nitrogenous compounds. The therapeutic approach to HE includes drugs that eliminate ammoniaproducing bacteria. The aim of this study was to evaluate the efficacy and safety of rifaximin, a nonabsorbable antibacterial agent, in the treatment of chronic HE in a large population of cirrhotic patients. One hundred thirty-six cirrhotic patients (100 men, 36 women; mean age, 59.5 years) with clinical and biochemical signs of mild HE were studied under three different treatment protocols: one open study in which rifaximin (1200 mg/day) was administered for 21 days to 80 patients; and two randomized controlled studies, one in which rifaximin (1200 mg/day, 20 patients) and neomycin (3000 mg/day, 15 patients) were administered for 21 days and one in which rifaximin (1200 mg/day, 9 patients) and lactulose (40 gm/day, 12 patients) were administered for 21 days. In the open study, after 5 days of treatment, ammonia reached normal values in each patient; after 7 days, electroencephalographic abnormalities were present only in a few cases; and after 15 days, no patients showed neurologic signs of HE. In the rifaximin versus neomycin study, blood ammonia levels and neurologic signs significantly decreased during both treatments, but rifaximin induced an earlier disappearance of clinical and biochemical signs of HE. In the rifaximin versus lactulose study, both drugs were efficacious in reducing the neurologic signs of HE, but some patients treated with lactulose presented mild side effects at the beginning of the therapy, which disappeared in the course of treatment. This study confirms the usefulness of rifaximin in the management of cirrhotic patients with mild HE; the absence of significant side effects suggests its use even in more severe cases of HE, either alone or in association with other drugs.

To quantitate ammonia production by the intestinal flora, ammonia levels in arterial blood and the venous effluent of the small and large bowel of conventional, selectively decontaminated, germ-free and gnotobiotic rats were measured.

When the anaerobic flora was removed by decontamination a significant decrease in ammonia levels was observed in the effluent of both the small and large intestine. Decontamination of aerobic flora did not result in depression of ammonia production. Gnotobiotic rats colonised with an anaerobic flora or with a mixed aerobic and anaerobic flora, showed a slight increase in ammonia levels. No increase in ammonia production was observed when rats were colonised with aerobic flora. These results indicate that the Enterobacteriaceae were not responsible for ammonia generation.

The increase in ammonia levels after colonisation with anaerobic or mixed anaerobic/aerobic flora did not completely restore ammonia levels, despite reaching bacterial counts which were comparable to those in conventional rats. This may be explained by the limited number of species with which the rats were colonized. The finding that aerobic flora does not significantly contribute to ammonia production suggests that neomycin, known to be exclusively effective against aerobic flora, must have other effects to explain improvement of hepatic encephalopathy.

Lactulose and neomycin have, besides influencing ammonia production of the intestinal flora, been proposed to reduce glutamine-dependent ammonia formation. To test this hypothesis we determined the effects of lactulose and neomycin on the release or uptake of ammonia, urea, and amino acids across the intestine of freely moving healthy pigs. Blood was sampled from catheterized piglets (20 ± 0.8 kg; n = 6) before and 1, 2, 3 and 6 h after a standard pig meal (750 g, 12% protein). After a week of lactulose (Legendal; 2 × 60 g/day) or neomycin (8 g/day) treatment this procedure was repeated. Electromagnetic portal and small bowel flow measurements were carried out in separate groups of pigs. Flow measurements were independent of the kind of food ingested. No significant alterations in flow could be detected during the 6 h study period. Portal and porto-arterial ammonia differences were significantly decreased after lactulose (−20%) and neomycin (−35%) treatment. α-Amino-nitrogen absorption decreased in both groups as compared to controls, but this decrease did not reach significance. Systemic and portal glutamine levels as well as intestinal glutamine utilization were significantly lower in the treatment groups. Citrulline and glutamate levels and intestinal production decreased after treatment. In this in vivo model, ammonia liberation after protein meals decreased in animals pretreated with lactulose or neomycin. The decreased systemic and consequently intestinal glutamine utilization may contribute to a reduction of endogenous ammonia formation in the gut wall. Diminished absorption from the gut of α-amino-nitrogen may, however, also contribute to a decrease in ammonia production. These results are in accord with earlier observations in vitro and in vivo in Wistar rats. The mechanisms by which neomycin and lactulose lower systemic glutamine levels remain to be clarified.