What Is Known
- Liver involvement in children with celiac disease ranges from asymptomatic hypertransaminasemia to more severe conditions, including autoimmune liver disease and cirrhosis.
- Available studies report a high burden of liver involvement in celiac children at the diagnosis (asymptomatic hypertransaminasemia 15%–57%, autoimmune liver disease 1%–2%).
- Data on risk factors for liver involvement are discordant.
What Is New
- Prevalence of liver involvement (alanine transaminase >40 U/L) in children with celiac disease at the diagnosis is low (27/700 patients, 3.9%).
- Younger age is the only risk factor.
- Celiac disease–associated autoimmune liver disease is rare (1/700 patients, 0.14%).
Celiac disease (CD) is a systemic autoimmune disorder triggered by gluten ingestion in genetically predisposed subjects (1,2). Beyond enteropathy, CD may be associated with multiple extraintestinal manifestations (3–5). Among these, liver involvement is one of the most common (6–8). Hepatic manifestations of CD range from asymptomatic elevation of liver enzymes (also known as “celiac hepatitis”) to more severe hepatic disorders, including autoimmune hepatitis, primary biliary cholangitis, and sclerosing cholangitis (9–13). Data on the prevalence of liver involvement in CD are highly heterogeneous. In adults, hypertransaminasemia (HTS) has been reported in 11% to 42% of patients (14–18), with a 2011 meta-analysis estimating its prevalence at 27% (19). In children, there are few prevalence studies. Available data indicate that an elevation of serum transaminases is present in 15% to 57% of celiac children at the diagnosis (36% according to a 2013 meta-analysis) (20–24), and reported prevalence of autoimmune liver disease in celiac children is 1% to 2% (23,25). Data on risk factors for liver involvement are also discordant. We sought to evaluate prevalence, clinical course, and risk factors for liver involvement in a large cohort of children with CD.
MATERIALS AND METHODS
All children (age 0–18 years) who received a diagnosis of CD from March 2010 to April 2016 at Referral Center for Celiac Disease of the Institute for Maternal and Child Health “Burlo Garofolo,” in Trieste, Italy, were prospectively enrolled to be included in our institutional pediatric CD database. Informed consent was obtained from all patients’ parents or caregivers. Institutional approval from our Institute's independent bioethics committee was obtained. Baseline data of all patients were collected at the time of diagnosis: sex, age, presenting symptoms, associated autoimmune diseases, familial history of CD, and other autoimmune diseases. Serum anti-tissue transglutaminase 2 (anti-TG2) IgA was measured using an enzyme-linked immune sorbent assay (Eu-tTG, Eurospital, Italy). A cut-off value of 7 U/mL was used. In case of IgA deficiency, anti-TG2 immunoglobulin G (IgG) was measured. Serum endomysial antibodies (EMAs) positivity was evaluated by immunofluorescence on human umbilical cord sections by an experienced analyst (T.N.). Human leukocyte antigen (HLA) genotyping was performed when necessary to reach a diagnosis. Susceptibility alleles for CD were determined by polymerase chain reaction assay with allele-specific primers (Eu-Gene-Risk kit, Eurospital, Italy). Other blood and serum tests were performed according to our laboratory guidelines. We reviewed the following parameters, when available: hemoglobin, aspartate transaminase (AST), alanine transaminase (ALT), and gamma-glutamyltransferase (GGT). Ferritin levels were measured when clinically indicated. HTS was considered to be present if ALT levels were >40 U/L. Isolated elevation of AST was not considered, because this finding may be nonspecific or due to hemolysis in the blood sample (26). In patients with HTS, serum transaminases were retested after at least 12 months of a gluten-free diet (GFD). Further tests for liver disorders were performed as clinically indicated. Duodenal biopsies were obtained during diagnostic esophagogastroduodenoscopy. For each subject, 4 specimens were taken: 2 from the duodenal bulb and 2 from the distal duodenum. The histological description was based on the Marsh classification (27). Diagnosis of CD was made according to available clinical guidelines (1,28,29). After 2012, the new European Society for Paediatric Gastroenterology Hepatology and Nutrition (ESPGHAN) diagnostic guidelines for CD were used; therefore, in patients with clinical symptoms highly suggestive for CD (diarrhea with weight loss, failure to thrive, and/or iron-deficiency anemia), plus a permissive HLA haplotype, positivity of EMA, and an anti-TG2 IgA antibody serum concentration over 10 times the upper limit of normal, CD was diagnosed without intestinal biopsy. A diagnosis of potential CD was made in patients with clinical signs and symptoms suggestive of CD plus a positive CD-specific serology and a CD-compatible HLA haplotype who lacked diagnostic mucosal findings (ie, Marsh 0 or 1) if anti-TG2 IgA antibody deposits were present in the duodenal mucosa, since these have been shown to represent an early marker of CD-specific immune response that can be detected before evidence of intestinal damage on histology (1,30–32). Mucosal anti-TG2 IgA deposits were investigated by direct double immunofluorescence according to a method described previously (30). Briefly, 6 frozen duodenal sections per patient were incubated with a mouse antihuman TG2 antibody (1:100 dilution; CUB7402, NeoMarkers, Fremont, CA, USA) followed by an Alexa Fluor 594-conjugated donkey antimouse IgG secondary antibody (1:100 dilution; Invitrogen, Waltham, MA, USA) and then by a fluorescein rabbit antihuman IgA antibody (1:80 dilution; Dako, Denmark). The multicolor analysis was performed using an Axioplan2 fluorescence microscope (Carl Zeiss, Germany) to localize the anti-TG2 IgA antibody deposits.
Statistical analysis: HTS prevalence was calculated on the population with transaminase levels available. For the risk analysis, subjects with HTS were compared with those with normal ALT levels (control group). Chi-square or Fisher test and Wilcoxon-Mann-Whitney test were applied, according to the variable distribution, to establish the crude association between HTS and population characteristics. Continuous variables were reported as median and interquartile range (IQR); categorical data as number and percentage. Multivariate logistic regression models were constructed with HTS as outcome and population characteristics as explanatory variables. The variables included in the final model were those with a P value of 0.10 in the univariate logistic model; to dichotomize continuous variables a cut-off value was fixed using the first quartile. Data on risk analysis were reported as odds ratio (OR) and 95% confidence interval (95% CI).
Between March 2010 and April 2016, 806 patients received a diagnosis of CD (median age 7.39 years, range 0.8–17.9 years). Of these, transaminases levels were available for 700 patients (86.8%), who were therefore included in the prevalence analysis. Fifty-six of 700 patients (8.0%) were diagnosed with potential CD, whereas 24 patients (3.4%) received a diagnosis without biopsy according to the 2012 ESPGHAN guidelines. Twenty-seven patients (3.9%) had serum levels of ALT >40 U/L (HTS group). Table 1 compares subjects with HTS and control group for the main characteristics collected. Median ALT and AST levels in the HTS group were 57 U/L (IQR: 49–80 U/L) and 67 U/L (IQR: 53–85 U/L), respectively. GGT levels were normal (<20 U/L) in all but 5 patients (median 8.0 U/L, IQR 6.0–12.5). Of these 5, 2 also had elevated ALT levels.
At univariate analysis, median age was significantly lower in the HTS group as compared to patients with normal ALT levels (2.58 vs 7.10 years, P < 0.0001). Prevalence was highest in patients younger than 2 years of age (7/40, 17.5%) (Fig. 1). No difference in the male/female ratio was noted between the 2 groups. Presence of symptoms at presentation was similar between HTS and control group (88.5% vs 73.6%, P = 0.26), but when only signs and symptoms of malabsorption were considered (diarrhea, failure to thrive, weight loss, anemia, or low ferritin level) a statistically significant difference was observed (15.8% in the control group vs 40.7% in the HTS group, P = 0.002). Hemoglobin and ferritin levels median values were also significantly lower in the HTS group (P = 0.01 and P = 0.02, respectively). A trend toward higher anti-TG2 values (median 128 vs 94 U/L, P = 0.06) and toward less frequent EMA positivity (19/27, 70.4%, vs 574/673, 85.3%, P = 0.051) was observed in the HTS group but did not reach statistical significance. Among patients with HTS in whom intestinal biopsy was performed, all patients had Marsh 3 changes, except 2 who had Marsh 2 changes. No patient in the potential CD group had elevated ALT levels. No statistically significant difference was noted between control and HTS groups in terms of associated autoimmune diseases, and familial history of CD or other autoimmune diseases.
At multivariate logistic analysis (which included age, malabsorption symptoms, anti-TG2 levels, hemoglobin levels, and ferritin levels as variables), only age ≤4.27 years still correlated with an increased risk of HTS (OR 3.73; 95% CI: 1.61–8.66). Statistical significance was not retained for signs or symptoms of malabsorption (OR 2.18; 95% CI: 0.91–5.22, P = 0.08).
Of the 27 patients with HTS, 9 were lost to follow-up because they lived far from our institute and declined to return for further evaluation. The remaining 18 patients were retested after at least 12 months from the start of GFD (median 31.5 months, range 12–82 months). All but 3 patients had normal ALT levels on a GFD (median value in patients who normalized ALT levels: 10 U/L, IQR 7–14). Three patients still had elevated ALT levels: in the first patient, ALT was 83 U/L at the last available follow-up visit (after 22 months of GFD), and had significantly reduced from the diagnosis (175 U/L). Good adherence to GFD was reported, but EMA was still positive; therefore, significant gluten intake was considered to be present. The second patient had only slightly elevated ALT levels after 4 years of GFD (47 U/L), which had progressively decreased from the diagnosis (104 U/L at diagnosis, 68 U/L after 18 months of GFD). Anti-TG2 IgA and EMA were negative; a slowly resolving form of celiac hepatitis was hypothesized. Both patients were completely asymptomatic, and other liver disorders (autoimmune liver disease, viral hepatitis, fatty liver disease, Wilson disease, alpha-1 antitrypsin deficiency) were excluded. The third patient was a 12-year-old girl who had elevated liver enzymes after 6 months of GFD (ALT 133 U/L, AST 83 U/L) and elevated GGT (138 U/L). Notably, she was one of the few patients with elevated GGT at CD diagnosis. Abdominal ultrasonography showed inhomogeneous liver structure and splenomegaly, whereas magnetic resonance cholangiography showed diffuse irregular stenosis of biliary ducts. Autoantibodies for autoimmune liver disease were negative, whereas a liver biopsy was consistent with sclerosing cholangitis. Concurrent inflammatory bowel disease was excluded.
The prevalence of HTS and autoimmune liver disease in our cohort were 3.9% and 0.14%, respectively. These figures are significantly lower than those previously reported in medical literature for liver involvement in pediatric CD. Previous studies found HTS in 15% to 57% of celiac children at the diagnosis (20–24), with a 2013 meta-analysis estimating its prevalence 36% (24), whereas prevalence of autoimmune liver disease was estimated at 1% to 2% (23,25). Several factors may concur to explain this difference. First, to increase specificity of our findings, we choose to exclude patients with an isolated elevation of AST, because this finding may be nonspecific. Previous studies on prevalence of HTS in children with CD often included also patients with isolated AST elevation, and when only patients with elevated ALT when considered, the prevalence of abnormal findings was actually lower (14.9%–27.7%) (20,21). Transaminases reference intervals can vary among clinical laboratories, and slightly different upper limits of normal have been used in different studies. We chose 40 U/L as the upper limit of normal for serum ALT levels since this is the upper limit of normal adopted by our laboratory, and it is consistent with the results of a recent meta-analysis on transaminases levels in healthy children (33). The median value of ALT in the HTS group was only 1.8 times the upper limit of normal and the highest value of ALT was 175 U/L. Notably, GGT were normal in all but 5 patients. Of these 5, 2 also had elevated ALT levels, and 1 of these 2 was eventually diagnosed with sclerosing cholangitis, whereas in the other both values normalized after GFD. This observation underscores that abnormal GGT values are not common in patients with CD at the diagnosis, and should raise suspicion of a cholestatic condition, especially when associated with abnormal ALT levels.
Data on risk factors for liver involvement are also discordant, with younger age, poor growth, severe villous atrophy, low ferritin level, and elevated autoantibodies being associated to HTS in several studies. In our study malabsorption signs/symptoms were associated with liver involvement at univariate analysis, but after multivariate analysis only age remained as a significant risk factor. It is possible that different age distribution among studies influences the observed prevalence of HTS. In our cohort, median age was 6.95 years, with 41 of 700 patients younger than 2 years of age. This is in accordance with international data showing that age at CD diagnosis is now higher than that in the past (34,35). We could provide data for 700 (86.8%) patients from the whole cohort of 806 patients, whereas for 106 patients ALT levels were not available. Clinical characteristics among these 2 groups were similar except for age, which was lower in the second group (4.5 years, IQR 2.4–8 vs 6.9 years, IQR 4.1–10.4, P < 0.0001). Although this may represent a bias, the overall prevalence of HTS is unlikely to be significantly affected. In fact, estimating prevalence of HTS in the untested group of patients using data from the 700 patients for whom ALT levels were available and applying an age-adjusted 95% CI would result in 4 to 11 (mean 7) extra cases of HTS. This would lead to an overall estimated prevalence of 31 to 38 (mean 34) cases on the whole population of 806 patients (4.2%, 95% CI 3.1%–5.8%). When compared to the observed prevalence (3.9%, 95% CI 2.6%–5.5%), the difference was not statistically significant (P = 0.79).
The majority of patients (96.6%) in our study had intestinal biopsy performed, and all but 2 patients with HTS had villous atrophy (Marsh 3 changes), whereas no cases of HTS were found in the potential CD group. This observation is also consistent with previous reports on the association between elevated transaminases and intestinal atrophy, both in children and in adults (17,36). In agreement with previous studies, most of our patients with HTS normalized transaminases levels on a GFD, with only 1 patient presenting with a slow fall of ALT levels despite a good compliance to GFD.
Overall, our results indicate a low prevalence of liver involvement in children with CD at the diagnosis. This seems to be in accordance with the results of the most recent studies. In 2014 data from the Spanish National Registry of Celiac Disease found “abnormal liver enzymes” in 13.6% of patients, even though it is not clear which enzymes were evaluated and which cut-off were chosen (37). In 2016, Äärelä et al (36) reported ALT >30 U/L in 14.7% of patients, whereas Lee et al (38) reported elevated transaminases (both ALT and AST) in 15.1%. Remarkably, prevalence of autoimmune liver disease was also very low (0.14%) compared to historical data (1%–2%) (23,25).
The reasons for these changes in the epidemiology of hepatic manifestations of CD are not known. We speculate that it may be related to an increased awareness of CD among the medical community leading to earlier diagnoses, and to the changing epidemiology of CD with increasing age at the diagnosis (39,40), because HTS is more common in younger patients, also in our cohort.
In conclusion, our data show that the prevalence of liver involvement in children with untreated CD at the time of diagnosis, both as asymptomatic HTS and autoimmune liver disease, is now lower than previously reported, possibly due to changing epidemiology of CD. Younger age is confirmed to be the single most important risk factor for HTS. Transaminases levels among patients with HTS are usually mildly elevated, and spontaneous resolution following GFD is the norm. CD-associated autoimmune liver disease is rare.
1. Husby S, Koletzko S, Korponay-Szabó IR, et al. European Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines for the diagnosis of coeliac disease. J Pediatr Gastroenterol Nutr
2. Lebwohl B, Sanders DS, Green PHR. Coeliac disease. Lancet
3. Jericho H, Sansotta N, Guandalini S. Extraintestinal manifestations of celiac disease: effectiveness of the gluten-free diet
. J Pediatr Gastroenterol Nutr
4. Pinto-Sánchez MI, Bercik P, Verdu EF, et al. Extraintestinal manifestations of celiac disease. Dig Dis
5. Ventura A, Martelossi S. Dental enamel defects and coeliac disease. Arch Dis Child
6. Rubio-Tapia A, Murray JA. The liver in celiac disease. Hepatology
7. Marciano F, Savoia M, Vajro P. Celiac disease-related hepatic injury: Insights into associated conditions and underlying pathomechanisms. Dig Liver Dis
8. Anania C, De Luca E, De Castro G, et al. Liver involvement in pediatric celiac disease. World J Gastroenterol
9. Volta U, Rodrigo L, Granito A, et al. Celiac disease in autoimmune cholestatic liver disorders. Am J Gastroenterol
10. Gillett HR, Cauch-Dudek K, Jenny E, et al. Prevalence of IgA antibodies to endomysium and tissue transglutaminase in primary biliary cirrhosis. Can J Gastroenterol
11. Lawson A, West J, Aithal GP, et al. Autoimmune cholestatic liver disease in people with coeliac disease: a population-based study of their association. Aliment Pharmacol Ther
12. Caprai S, Vajro P, Ventura A, et al. Autoimmune liver disease associated with celiac disease in childhood: a multicenter study. Clin Gastroenterol Hepatol
13. Villalta D, Girolami D, Bidoli E, et al. High prevalence of celiac disease in autoimmune hepatitis detected by antitissue transglutaminase autoantibodies. J Clin Lab Anal
14. Hagander B, Berg N, Brandt L, et al. Hepatic injury in adult coeliac disease. Lancet
15. Bardella MT, Fraquelli M, Quatrini M, et al. Prevalence of hypertransaminasemia
in adult celiac patients and effect of gluten-free diet
16. Korpimäki S, Kaukinen K, Collin P, et al. Gluten-sensitive hypertransaminasemia
in celiac disease: an infrequent and often subclinical finding. Am J Gastroenterol
17. Zanini B, Baschè R, Ferraresi A, et al. Factors that contribute to hypertransaminasemia
in patients with celiac disease or functional gastrointestinal syndromes. Clin Gastroenterol Hepatol
18. Castillo NE, Vanga RR, Theethira TG, et al. Prevalence of abnormal liver function tests in celiac disease and the effect of a gluten-free diet
in the US population. Am J Gastroenterol
19. Sainsbury A, Sanders DS, Ford AC. Meta-analysis: coeliac disease and hypertransaminasaemia. Aliment Pharmacol Ther
20. Bonamico M, Pitzalis G, Culasso F, et al. Hepatic damage in celiac disease in children. Minerva Pediatr
21. Farre C, Esteve M, Curcoy A, et al. Hypertransaminasemia
in pediatric celiac disease patients and its prevalence as a diagnostic clue. Am J Gastroenterol
22. Arslan N, Büyükgebiz B, Oztürk Y, et al. The prevalence of liver function abnormalities in pediatric celiac disease patients and its relation with intestinal biopsy findings. Acta Gastroenterol Belg
23. Di Biase AR, Colecchia A, Scaioli E, et al. Autoimmune liver diseases in a paediatric population with coeliac disease—a 10-year single-centre experience. Aliment Pharmacol Ther
24. Vajro P, Paolella G. Pediatric celiac disease, cryptogenic hypertransaminasemia
, and autoimmune hepatitis. J Pediatr Gastroenterol Nutr
25. Ventura A, Magazzù G, Greco L. Duration of exposure to gluten and risk for autoimmune disorders in patients with celiac disease. SIGEP Study Group for Autoimmune Disorders in Celiac Disease. Gastroenterology
26. Lippi G, Salvagno GL, Montagnana M, et al. Influence of hemolysis on routine clinical chemistry testing. Clin Chem Lab Med
27. Marsh M. Gluten, major histocompatibility complex, and the small intestine. A molecular and immunobiologic approach to the spectrum of gluten sensitivity (’celiac sprue’). Gastroenterology
28. Revised criteria for diagnosis of coeliac disease. Report of Working Group of European Society of Paediatric Gastroenterology and Nutrition. Arch Dis Child
29. National Institute for Health and Clinical Excellence (2009). Coeliac disease: recognition and assessment of coeliac disease. London: National Institute for Health and Clinical Excellence. Available from: www.nice.org.uk/CG86
. Accessed September 10, 2018.
30. Salmi TT, Collin P, Korponay-Szabó IR, et al. Endomysial antibody-negative coeliac disease: clinical characteristics and intestinal autoantibody deposits. Gut
31. Salmi TT, Collin P, Jarvinen O, et al. Immunoglobulin A autoantibodies against transglutaminase 2 in the small intestinal mucosa predict forthcoming coeliac disease. Aliment Pharmacol Ther
32. Not T, Ziberna F, Vatta S, et al. Cryptic genetic gluten intolerance revealed by intestinal antitransglutaminase antibodies and response to gluten-free diet
33. Stirnadel-Farrant HA, Galwey N, Bains C, et al. Children's liver chemistries vary with age and gender and require customized pediatric reference ranges. Regul Toxicol Pharmacol
34. Kivelä L, Kaukinen K, Lähdeaho M-L, et al. Presentation of celiac disease in Finnish children is no longer changing: a 50-year perspective. J Pediatr
35. Tapsas D, Hollén E, Stenhammar L, et al. The clinical presentation of coeliac disease in 1030 Swedish children: changing features over the past four decades. Dig Liver Dis
36. Äärelä L, Nurminen S, Kivelä L, et al. Prevalence and associated factors of abnormal liver values in children with celiac disease. Dig Liver Dis
37. Cilleruelo ML, Roman-Riechmann E, Sanchez-Valverde F, et al. Spanish National Registry of Celiac Disease: incidence and clinical presentation. J Pediatr Gastroenterol Nutr
38. Lee GJ, Boyle B, Ediger T, et al. Hypertransaminasemia
in newly diagnosed pediatric patients with celiac disease. J Pediatr Gastroenterol Nutr
39. Volta U, Caio G, Stanghellini V, et al. The changing clinical profile of celiac disease: a 15-year experience (1998-2012) in an Italian referral center. BMC Gastroenterol
40. Almallouhi E, King KS, Patel B, et al. Increasing incidence and altered presentation in a population-based study of pediatric celiac disease in North America. J Pediatr Gastroenterol Nutr