Atypical Celiac Disease Presenting as Obesity-Related Liver Dysfunction

Asymptomatic liver involvement is more and more recognized in otherwise healthy, obese children ^(1–3). Loss of weight in these patients is an important target to achieve because it is most often followed by normalization of liver function tests, imaging, and histologic examination, thus reasonably confirming the causal role of weight excess with no need of further particular investigations. In the case of persisting liver disease despite a successful diet program, other causes of hypertransaminasemia and fatty liver should be ruled out. Hitherto, celiac disease has not been described as a condition possibly contributing to liver disease in obese patients. This may depend either on a true rarity of the association of these two conditions or on an insufficient correct investigation resulting from the bias of the seemingly improbable coexistence of celiac disease and obesity. However, there is some evidence that celiac disease is both an established cause of liver disease ⁽⁴⁾ and a condition possibly associated with obesity as well ^(5–7). Herein, we report for the first time a case of an obese child with weight-loss–resistant, longstanding liver disease. Liver abnormalities completely resolved only after gluten was withdrawn from diet after an unexpected diagnosis of celiac disease.

CASE REPORT

The patient, an Italian 9-year-old girl with a noncontributory past history, was admitted to our institution for evaluation of asymptomatic and persistent (>1 year) hypertransaminasemia of unknown origin. She had been breast fed for the first 5 months of life, and thereafter was weaned with no gastrointestinal problems. Obesity had been recorded during the second year of life, but had never been treated. She had not been treated previously with hepatotoxic drugs, had not undergone surgery, nor had she received blood or blood products.

At admission, physical examination showed a girl with a generalized distribution of ponderal excess. Height was 139 cm (>75th percentile), weight was 55.2 kg (>95th percentile), which was 160% of her ideal body weight. Puberty evaluation showed B2 and P2 stages according to Tanner. Clinical examination was normal. In particular, no clinical evidence of endocrine or genetic obesity was present, and liver and spleen sizes were within normal limits.

Liver function tests showed a moderate increase of serum aspartate aminotransferase and alanine aminotransferase with normal values of other indexes of liver disease (bilirubin, gamma glutamyl transpeptidase, alkaline phosphatase, bile acids, protein electrophoresis, and prothrombin time). Also, blood glucose levels, lipids, and common nutritional laboratory parameters were normal. Abdominal ultrasonography showed a bright liver pattern (Fig. 1A). Muscular disease, viral hepatitides B and C, autoimmunity, α1 antitrypsin deficiency, inborn errors of amino acid metabolism, and Wilson disease were ruled out by means of appropriate laboratory tests.

FIG. 1.: Ultrasonographic images of the liver showing severe (A), moderate (B), and mild (C) hyperechogenicity of the liver before, during, and after hypocaloric diets, respectively. D: Disappearance of the liver hyperechogenicity after the patient began a gluten-free diet. L, liver; K, kidney.

Balanced hypocaloric diet (1,500 calories, i.e., approximately 30 calories/kg per day; carbohydrate 55%, fat 30%, protein 15%) was prescribed along with a moderate exercise daily program, with a target of 500 g of weight loss per week to reach 120% of ideal body weight.

Despite an acceptable compliance to diet, liver enzymes and ultrasonographic findings of fatty liver levels did not reach normal (Table 1;Fig. 1B,C). Therefore, a further search for less frequent causes of fatty liver was undertaken. This led to test for the presence of antigliadin and antiendomisium antibodies, the test results for which repeatedly were positive (Table 1). Human leukocyte antigen pattern was consistent with celiac disease (HLA-DR3,DQ2 aplotype) and was confirmed by the finding of subtotal atrophy of intestinal mucosa with severe cellular infiltration of lamina propria. To evaluate better the type and degree of hepatic damage, needle liver biopsy was performed, which showed slight micro–macro vesicular steatosis with mild, sparse lobular inflammation and mild portitis as well (Fig. 2).

TABLE 1: Clinical and laboratory findings of the patient during follow-up

FIG. 2.: Needle liver biopsy taken when the patient was on a gluten-containing hypocaloric diet. A: Note residual mild micro–macro vacuolar steatosis, sparse lobular inflammation, and mild portitis (stain, hematoxylin–eosin; original magnification, ×100). B: Detail of the mild inflammatory infiltrate of the portal tract (stain, hematoxylin–eosin; original magnification, ×400).

A gluten-free diet was prescribed, and the girl was regularly followed-up. After 3 months, liver enzyme levels became normal for the first time, and antiendomisium antibodies tested negative. Three months later, liver enzyme levels were still normal, antiendomisium antibodies tested negative, and liver ultrasonography showed disappearance of the bright pattern (Table 1 and Fig. 1C). One year later, she gained excessive weight, which led to a relapse of bright liver at ultrasonography; reduction of caloric intake resulted in renormalization of the radiologic pattern (data not shown).

DISCUSSION

Common sense may lead one to discard celiac disease as a condition associated with obesity, and consequently also as a factor contributing toward liver disease in the initial evaluation of an obese individual with long-standing hypertransaminasemia and ultrasonographic hepatic steatosis. However, excessive weight has already been reported in association with celiac disease not only after ⁽⁸⁾ but also before ^(5–7,9,10) withdrawal of gluten from diet. Moreover, both obesity ^(1–3) and celiac disease ^(4,11–13) are well-recognized causes of liver function abnormalities, hepatic steatosis, or both. Nevertheless, the latter, has hitherto been described only in cases of celiac disease with severe malnutrition ^(11–13).

Liver damage of our obese patient only partially responded to a weight reduction diet. Despite a reasonable weight loss, liver function tests, sonograms, and biopsy still showed puzzling mild residual abnormalities. Other common causes of underlying liver disease had been ruled out by appropriate tests. Only gluten withdrawal prescribed after the diagnosis of celiac disease allowed complete normalization of her long-standing liver impairment. It is difficult to establish whether minor histologic changes found in her liver biopsy depended exclusively on untreated celiac disease or were a residual consequence of the previous weight excess. Portal inflammation was consistent both with the usual liver pathologic features found in children with elevated serum aminotransferase activity as an early manifestation of atypical celiac disease ⁽⁴⁾ and with inflammation seen in the livers of individuals with obesity-related liver disease ^(1–3).

We do not know whether the mild histologic steatosis and ultrasonographic bright liver pattern of our patient were dependent only on subtle, undefined nutritional deficiencies induced by atypical celiac disease or they were also a sequel of the recent excess of weight. Fatty liver described in celiac disease in fact regarded only adult malnourished patients, and it reasonably depended on severe malnutrition itself rather than on specific hepatotropic effects of celiac disease ^(11–13). Because a subsequent regain of weight while following a gluten-free diet was followed by the reappearance of radiologic abnormalities that reverted to normal after reduction of caloric intake, it may be argued that in our patient, both conditions played an equal role in causing liver abnormalities.

Although abnormalities of intestinal mucosa may theoretically depend on causes other than celiac disease, subclinical celiac disease appears to be the most likely explanation of the intestinal pathologic features in our patient, according to human leukocyte antigens pattern and to presence of antiendomisium antibodies while following a regular diet followed by their loss after withdrawal of gluten from diet. Silent celiac disease resulting from an underlying patchy pattern of damaged intestinal mucosa still may have not been sufficient to halt an adequate absorption of nutrients, and it is conceivable that an excessive caloric intake may have resulted in obesity despite the existing histologic damage.

Celiac disease was an unexpected factor contributing to liver abnormalities in an obese girl with weight-loss–resistant obesity-related liver disease. We recommend that silent celiac disease be considered among the possible causes of liver impairment also in obese children, especially in those whose biochemical, ultrasonographic, or histologic liver abnormalities, or a combination thereof, fail to respond completely to an adequate loss of weight. Early recognition of celiac disease in these patients may avoid the performance of more invasive and expensive investigations or resorting to inadequate treatments ^(14,15). Additionally, it may prevent progression of hepatic steatosis ⁽³⁾ and may lower the risks related to undiagnosed celiac disease ⁽⁷⁾.