Celiac Disease Is Not Associated with Chronic Hepatitis C
- First Online:
- Cite this article as:
- Thevenot, T., Boruchowicz, A., Henrion, J. et al. Dig Dis Sci (2007) 52: 1310. doi:10.1007/s10620-006-9360-5
- 35 Views
KeywordsCeliac diseaseHepatitis C
Celiac disease (CD) is characterized by malabsorption due to a local immune response to dietary gluten against the mucosa of the small intestine in genetically predisposed patients bearing the human leukocyte antigen (HLA) DR3-DQ2 or DR4-DQ8. CD has a wide spectrum of gastrointestinal and extraintestinal manifestations, including dermatitis herpetiformis, type 1 diabetes mellitus, and autoimmune thyroiditis . It has been hypothesized that hepatitis C virus (HCV) may trigger immunologic gluten intolerance in susceptible people . However, this assumption is still a matter of debate. Recently, we diagnosed four patients having both CD and chronic hepatitis C, three of them having a well-defined route of transmission, suggesting that there is no obvious link between the two diseases.
A 45-year-old woman had been suffering from CD for a long time and treated with a gluten-free diet. She had been transfused at the age of 20 due to ferriprive anemia. Chronic hepatitis C (genotype 1a) was discovered 15 years later because of fatigue. In 1997, the patient received a 6-month course of interferon α-2a therapy and experienced mild diarrhea during treatment despite the fact that she had continued her gluten-free diet.
A 63-year-old Italian woman was first referred in 1983 for hypertransaminasemia, proved 10 years later to be caused by HCV infection (genotype 1b). No obvious route of viral transmission was demonstrated except for complicated surgery for a broken kneecap following an automobile accident in 1971. The patient did not remember if she had been transfused. In 1996, she suffered from histologically proven dermatitis herpetiformis, successfully treated with dapsone and a gluten-free diet. Distal duodenal histology revealed flattened duodenal folds compatible with CD. At that time a liver biopsy examination showed active chronic hepatitis C (METAVIR score A2F3). The patient received only 3 months of interferon α-2b therapy because of nonresponse and poor tolerance. A second treatment was attempted in 2004 with peginterferon α-2a plus ribavirin but after 7 months of therapy she was still a nonresponder. During both antiviral therapies administered under a gluten-free diet, she denied any abdominal discomfort.
Patient 3 was a 49-year-old woman who was transfused after an abortion in 1987. A few weeks later, she developed jaundice, and non-A, non-B chronic hepatitis was diagnosed. Chronic iron-deficiency anemia led to a diagnosis of CD in 2000, with positive IgA antiendomysial antibodies and IgA antigliadin antibodies. CD was also confirmed on duodenal biopsies. In 2003, a combination of peginterferon α-2b plus ribavirin was administered under a gluten-free diet without any gastrointestinal symptoms.
This 49-year-old nurse was diagnosed with genotype 1a chronic active hepatitis (METAVIR score A1F2) in 2000. The route of transmission was probably occupational because the patient reported multiple needlesticks. In 2002, she experienced fatigue and diarrhea. A slight iron-deficiency anemia was discovered, requiring an upper gastrointestinal endoscopy, which revealed CD on distal duodenal biopsies. Antiendomysium IgA antibodies and antigliadin IgA antibodies were positive, at 40 and 94 UI/mL respectively. The patient received peginterferon α-2a, 180 μg/week, plus ribavirin, 800 mg/day, for 1 year while continuing her gluten-free diet. This treatment was well tolerated and permitted a sustained virological response.
Recently, it was hypothesized that nonintestinal inflammatory diseases may trigger immunologic gluten intolerance in susceptible people and HCV was thought to be a suitable candidate . It is well known that HCV causes secondary autoimmune disorders  such as mixed cryoglobulinemia, lichen planus, thyroiditis, and Sjogren’s syndrome. It has also been suggested that CD is another autoimmune syndrome associated with HCV. The T-cell response to HCV is restricted to HLA-DQ2 , the class II HLA allele linked to CD, and other autoimmune diseases . This immunopathogenic pathway could be the hypothetic link between CD and HCV. Few cases have reported an association between CD and chronic hepatitis C [5–15] and only two studies have investigated the prevalence of specific CD antibodies in HCV-positive patients [2, 16]. Using serologic screening, Fine et al. discovered that 1.2% (3/259) of HCV-positive patients were found to have significantly more CD than other forms of liver disease or than controls without liver disease . In another study , Durante-Mangoni et al. observed a higher prevalence of pretreatment antitransglutaminase IgA antibodies tested on stored serum samples (3.6%) in a cohort of 195 treated HCV-positive patients compared to a control group of 225 HCV-negative patients (0.4%). History of blood transfusion or drug addiction was investigated and only one patient had been a drug addict. This higher prevalence of CD in both studies supports a true association between CD and chronic hepatitis C. However, no mention was made of the route of transmission of HCV-infected patients in the first study  and caution must be advised in the second retrospective study  because possible route of transmission could not have been accurately reported. In the latter study, the prevalence of antitransglutaminase IgA antibodies in treated HCV-positive patients was 24% (6/25) and 0.6% (1/170) in patients with and without CD-like symptoms, respectively. In all patients with positive IgA antitransglutaminase antibodies, the duodenal histologic picture was compatible with CD. So the author concluded that interferon α may precipitate symptom onset in the great majority of silent CD cases, and that CD-like symptoms (diarrhea, anemia, marked weight loss) occurring on interferon therapy were significantly associated with the presence of antitransglutaminase antibodies before treatment.
This association between CD and HCV has not been found elsewhere, pointing out a possible bias in this relation. In a recent letter, Teml and Vogelsang drew our attention to the possible role of blood transmission in anemic patients with a silent form of CD . Among 488 celiac patients followed up at their outpatient unit, 3 women (0.6%) had chronic HCV infection. Two of them had received blood transfusions because of extensive blood loss during delivery, and the other had experienced drug addiction. In our four patients, two received blood transfusions, one had evidence of occupational exposure, and the other probably received blood following a road accident. Our patient 1 and perhaps also patient 2 were transfused because of hematological disorders related to CD, i.e., anemia in case 1 and perhaps vitamin K deficiency in case 2. A French prospective study, still under way and conducted by the Association Nationale des Hépato-gastroentérologues des Hôpitaux Généraux (ANGH), started to investigate in June 2003 the relation between chronic hepatitis C and CD by assessing the prevalence of antigliadin IgA and IgG and IgA antiendomysial antibodies in patients infected with HCV . Upper gastrointestinal endoscopy was systematically considered for patients having at least one positive CD antibody, and four biopsies were taken from the distal duodenum. No evidence of CD was found among the first 425 HCV-positive patients screened.
Possible activation of CD due to the treatment of hepatitis C infection is another controversial point. CD activation during interferon or interferon plus ribavirin therapy has recently been observed, suggesting that it is important to screen all HCV-positive patients for CD in order to prescribe a gluten-free diet before starting antiviral therapy [5–8, 12]. Ribavirin could also activate a latent CD by promoting a T-helper (Th) 1 cytokine-mediated immune response and by suppressing the Th2 response . Indeed, Adinolfi et al. reported a HCV-positive patient relapsing to interferon alone without any particular side effects during this first treatment, but manifesting a severe clinical form of CD during combined interferon/ribavirin therapy. In this report, ribavirin seemed to play a critical pathogenic role . None of our four patients experienced abdominal discomfort requiring antiviral therapy withdrawal. Only patient 1 had mild transient diarrhea while remaining on a full interferon dosage and under a gluten-free diet.
Finally, this association between CD and chronic hepatitis C seems to be a rare event, as confirmed by a French study conducted in a large population of HCV-positive patients. The link between these two diseases is probably biased by the route of transmission of HCV-infection. Indeed, iron-deficiency anemia or a reduction of coagulation factors is very common in silent CD, and blood transfusion could be the main factor contributing to the increase in cases reporting this association. In patients suffering from CD-like symptoms or anemia, screening for specific CD-antibodies should be considered before starting treatment for chronic hepatitis C. If CD is confirmed, a strict gluten-free diet before and during anti-HCV therapy should always be initiated since interferon or interferon/ribavirin may precipitate an unrecognized CD.