Portal de Periódicos da CAPES

Markers of Cystic Fibrosis—Associated Liver Disease

2001; Lippincott Williams & Wilkins; Volume: 32; Issue: 4 Linguagem: Inglês

10.1097/00005176-200104000-00005

1536-4801

Michael R. Narkewicz,

Pediatric Hepatobiliary Diseases and Treatments

As the pulmonary and nutritional care of children and adults with cystic fibrosis (CF) improves, CF–associated liver disease is emerging as a significant medical issue in need of clear definition, a diagnostic algorithm, and effective therapy. To date, a major drawback in the study of CF–associated liver disease has been the lack of sensitive and specific biomarkers for identifying the disorder and its prevalence. Current prevalence data suggest that symptomatic elevations of transaminases or alkaline phosphatase occur in 20 to 56% of patients with CF, hepatomegaly in about 30%, and cirrhosis in 5 to 15%(1). However, as a recent report from the Cystic Fibrosis Foundation Hepatobiliary Disease Consensus Group (1) acknowledges, “Because there are no sensitive diagnostic markers of liver involvement in CF, current prevalence rates should be considered estimates that most likely underestimate the true risk.” In seeking the means to prevent the progression of CF–associated liver disease, it seems that therapeutic intervention is more likely to be successful early in the course, not after the development of clinically apparent cirrhosis. Herein lies the problem with CF–associated liver disease. We know that cirrhosis can develop in children without any abnormalities in standard liver biochemical measurements. However, there is a myriad of possible causes for biochemical abnormalities of liver function in patients with CF other than CF–associated liver disease, including such insults as medications, infection, and malnutrition. Thus, there is a need for sensitive, specific, and minimally invasive markers of early CF–associated liver disease. Various tests have been proposed to develop such markers, including markers of hepatic inflammation and injury such as standard liver biochemistries, high–molecular-weight alkaline phosphatase (2), and glutathione-S-transferase β1 (3). The two latter tests have, in small studies, been found to be early markers of CF–associated liver disease, but these observations have yet to be confirmed. In two other small studies, TGF-β(4) and collagen type VI (5) correlated with hepatic fibrosis in CF. Measures of hepatic synthetic function, such as prothrombin time and fasting serum bile acids (6), have not yet proven to be sensitive indicators of CF–associated liver disease. Liver biopsy can give direct evidence of fibrosis (7), but the risks of the procedure may outweigh the potential benefits of early identification. Thus, investigators are still looking for indirect tests of hepatic function that would preclude the need for liver biopsy as a reliable early marker of CF–associated liver disease. Galactose clearance, an indirect measure of hepatic blood flow, has been investigated in a small number of CF subjects and results suggest that abnormalities may correlate with the presence of hepatic fibrosis (8). Studies using caffeine clearance as a measure of hepatic function have suggested that CF subjects without obvious liver disease have normal caffeine metabolism (9). A small study of subjects with CF liver disease also did not find any consistent abnormalities in caffeine metabolism (8). This suggests that caffeine clearance will not prove to be a satisfactory early screen for CF–associated liver disease. A study using doppler measurement of hepatic venous flows as an early indicator of fibrosis found that this measurement too was unreliable (10). There has been more experience with hepatic scintigraphy (8,11,12), which apparently can be used as a predictor of improvement in transaminases and in biliary excretion after ursodeoxycholic-acid treatment in patients with CF–associated liver disease (11). However, no data exists on the sensitivity of scintigraphy for the early diagnosis of CF–associated liver disease. Unfortunately, all of these studies suffer either from inadequate subject numbers, inconsistent entry criteria, and/or the lack of a correlative study. For this issue of the Journal of Pediatric Gastroenterology and Nutrition, Gremse et al. investigated monoethylglycinexylidide (MEGX) formation to determine whether this test might be used to identify patients with early CF–associated liver disease. They evaluated a small number of subjects with CF with normal liver function by conventional tests and compared them to non–CF control participants also without obvious liver disease (13). MEGX is a metabolite of lidocaine that is produced by the hepatic mixed function oxidases. Low production of MEGX after a standard dose of intravenous lidocaine indicates either low activity of the hepatic mixed function oxidases or of a portosystemic shunt, such as might occur with portal hypertension. Although the CF patients in this study did not have any clinically apparent liver disease, they demonstrated a decreased formation of MEGX compared with non–CF control participants. This decrease was not the result of differences in hepatic blood flow as determined by simultaneous indocyanine-green clearance. These data suggest that there may be an inherent difference in certain aspects of hepatic metabolism in CF patients that must be considered when interpreting these studies. An alternative explanation is that the CF subjects in this study actually had subclinical liver disease that directly affected their MEGX production. Unfortunately this study was not designed to compare MEGX production in CF subjects with and without liver disease. Interestingly, previous studies of other hepatic metabolic pathways using caffeine metabolism have not demonstrated differences between patients with CF and control subjects (9,14), which suggests that there may be abnormalities of the hepatic mixed function oxidase system unique to CF patients. This MEGX study suggests that CF–specific normative data for indirect tests of hepatic function are required before their use in the study of liver disease in patients with CF. The unique aspect of the study by Gremse et al. is the finding of altered MEGX production in subjects with CF who have no standard markers for liver disease (13). Although the findings may indicate that MEGX production cannot be used as an early marker for CF–associated liver disease, they should serve as a stimulus for studies of the mechanism underlying the altered lidocaine metabolism in CF. This same group has previously published a study on the use of MEGX production in the prediction of liver disease in children other than those with CF (15). The levels of MEGX formation in their liver-disease group were similar to levels found in the CF subjects in this study. However, the number of patients studied (28 with chronic liver disease (15) and 19 with CF (13)) is inadequate to make any statistical comparisons. While attractive for the evaluation of CF liver disease, further studies of indirect tests of hepatic function must assess CF subjects both with and without CF–associated liver disease and provide a comparison to non–CF control participants and age-appropriate normative data. This can only be accomplished by large studies designed with significant power and strict entry criteria. By their very nature, cooperative multicenter studies will require the pediatric gastroenterology community to lead the way in the study of this important issue.