Elevated liver enzymes of newly diagnosed pediatric celiac patients—a prospective-observational study

Celiac disease clinical presentation is constantly changing. We set to determine the prevalence of elevated transaminases in newly diagnosed celiac patients and to evaluate this sub-group of patients for associated clinical and laboratory findings and assess their natural course of disease following therapeutic diet initiation. We conducted a prospective-observational study of all newly diagnosed pediatric celiac patients, between August 2016 and April 2018, in a pediatric gastroenterology clinic. Clinical data, anthropometrics, and blood test results were recorded at diagnosis and at 3, 6, and 12 months, respectively, of follow-up. We compared patients with normal and elevated transaminases at diagnosis. ALT threshold was set at 24 U/l. Of 125 newly diagnosed celiac patients, 31 (24.8%) had elevated ALT at diagnosis; two (1.6%) with over 3 × ULN. Patients with elevated ALT at diagnosis were significantly younger (mean age 5.5 (SD 3.4) vs. 7.3 (SD 3.7) years, p < 0.01) and more commonly presented with diarrhea (32.3% vs. 14.9%, p = 0.03). Eighty percent of patients with elevated ALT levels normalized their ALT within 3 months and all within 1 year. Following gluten-free diet initiation, patients with elevated ALT had similar clinical course, growth, serology normalization rate, and laboratory results, compared to patients with normal ALT over a 1-year follow-up. A single patient was simultaneously co-diagnosed with celiac disease and autoimmune hepatitis. Conclusion: Clinically significant ALT abnormalities are rare among newly diagnosed pediatric celiac patients. Significant elevations failing to normalize on a gluten-free diet should raise concern of a concomitant primary liver disease and warrant further investigations. What is Known: • Elevated liver enzymes may be an extra-intestinal manifestation of celiac disease. • Reported prevalences of ALT elevations among children with a new diagnosis of celiac disease ranges between 5 and 40%. What is New: • ALT elevations are present in 25% of children with a new diagnosis of celiac disease. • Significant elevations (>3 × ULN) are rare (1.6%). • Elevated liver enzymes are associated with earlier age at diagnosis. • The natural history of patients with elevated liver enzymes at diagnosis is comparable to those without. What is Known: • Elevated liver enzymes may be an extra-intestinal manifestation of celiac disease. • Reported prevalences of ALT elevations among children with a new diagnosis of celiac disease ranges between 5 and 40%. What is New: • ALT elevations are present in 25% of children with a new diagnosis of celiac disease. • Significant elevations (>3 × ULN) are rare (1.6%). • Elevated liver enzymes are associated with earlier age at diagnosis. • The natural history of patients with elevated liver enzymes at diagnosis is comparable to those without.


Introduction
Celiac disease (CD) is an autoimmune disorder, characterized by typical serological and histological findings, triggered by intestinal exposure to gluten in genetically susceptible individuals affecting mainly the small bowel [1]. The disorder is common with a prevalence of up to 1% in the general population [2][3][4]. CD classical presentation of severe and prolonged diarrhea, weight loss, wasting, and abdominal distention is rare nowadays [5]. More commonly, celiac patients present with non-specific abdominal symptoms, extra-intestinal manifestations, and occasionally patients are asymptomatic altogether and are diagnosed following screening [6]. The main extra-intestinal manifestations include malabsorption and related disorders, impaired growth and weight gain, fatigue, skin rash, and liver involvement among others [7].
Liver involvement in the setting of pediatric CD may be differentiated to celiac-associated elevated liver enzymes (ELE), occasionally referred to as hypertransaminasemia, and more rarely to other autoimmune and/or inflammatory liver diseases i.e., autoimmune-hepatitis (AIH), sharing a common background with CD [8][9][10]. Data regarding the nature of celiac-associated ELE in children is limited. Past studies, using different thresholds of normal alanineaminotransferase (ALT) levels, reported prevalence ranging between 4 and 40% among newly diagnosed celiac patients. Clinical features, including higher Marsh scores, presence of malabsorption, and lower hemoglobin levels, were described in association with ELE; however, without consistency [11][12][13][14][15][16][17][18]. Furthermore, little is known about the natural course of ELE-resolution and potential associations with clinical parameters. Finally, the clinical significance of ELE in newly diagnosed celiac patients is controversial, specifically, whether liver involvement may advance, independent of gluten exposure, to chronic hepatitis with resultant chronic liver disease complications [19].
In this prospective-observational study, we followed a cohort of newly diagnosed CD patients for 1 year. We describe the prevalence of ELE, clinical associations, and the natural course during the follow-up period; upper-normal transaminase levels were set, as now commonly accepted, according to the thresholds found in the NHANE survey [20,21]. Finally, we review past results in the literature and discuss comparison to our present findings.

Methods
We conducted a prospective-observational study of children (< 18 years) diagnosed with CD at the Pediatric Gastroenterology Clinic of the Tel Aviv Medical Center-a tertiary medical center, between August 2016 and April 2018. Patients were thereafter followed for 1 year. The celiac diagnosis was based on the North-American and European Societies for Gastroenterology, Hepatology, and Nutrition (ESPGHAN 2012, NASPGHAN 2016) guidelines [22,23]. In short, diagnosis was based on positive anti-tissue transglutaminase (aTTG) serology and histology compatible with CD (Marsh ≥ 2), or alternatively, on suggestive symptoms, highly elevated (> 10 × upper limit of normal (ULN)) aTTG serology, positive anti-endomysial antibody, and clinical improvement following gluten-free diet (GFD) initiation. We collected data including patient demographics, personal and family medical history, anthropometrics, clinical data, and blood test results prior to the diagnosis, and at 3, 6, and 12 months of followup. We excluded patients diagnosed elsewhere and patients with unavailable pre-diagnostic ALT levels and uncertain CD diagnosis. None of the patients included in the study had had a second intestinal or a known primary liver disease.
Weight, height, and body mass index (BMI) were recorded as percentiles for sex and age according to the Centers of Disease Control and Prevention (CDC) growth charts https:// www. cdc. gov/ growt hchar ts/ clini cal_ charts. htm). ULN of ALT was based on the findings of the third American National Health and Nutrition Examination Survey (NHANES) 1999-2006 [20,21], i.e., male/female average of 24 U/l. ULN of aspartate-aminotransferase (AST)-40 U/l and gamma-glutamyl-transpeptidase (GGT)-28 U/l were based on local laboratory thresholds. aTTG was categorized as normal, elevated, and highly elevated (> 10 × ULN).
Categorical variables were reported as number and percentage. Continuous variables were evaluated for normal distribution using histogram and Q-Q plot. Normally distributed continuous variables were reported as mean and standard deviation (SD), and skewed variables were reported as the median and interquartile range (IQR). Categorical variables were compared between patients with and without elevated ALT using the chi-square test or Fisher's exact test, and continuous variables were compared using independent sample T-test or Mann-Whitney U test. Generalized estimating equations (GEE) models were applied for repeated measure analysis. GEE models included age, sex, ALT level at presentation, and time. In further analysis, we also included the interaction between ALT levels at presentation and time. All statistical tests were two-sided and p < 0.05 was considered statistically significant. SPSS software was used for all statistical analyses (IBM SPSS Statistics for Windows, Ver. 26; IBM Corp., Armonk, NY, USA, 2019). The study was reviewed and approved by the institutional ethics committee.

Study cohort
During the study period, we identified in our clinic 389 patients with a new diagnosis of CD; 125 patients, 73 (57.6%) females, had a biochemical profile including liver enzymes at diagnosis, constituting the main cohort. The mean age at diagnosis was 6.8 (SD 3.7) years ranging between 1.6 and 17.4 years. Seven patients (5.6%) had prior medical history of other autoimmune (AI) disorders and four (3.2%) had immune deficiencies (three (2.4%) with IgA deficiency, one (0.8%) with common variable immune deficiency). Family history of AI/inflammatory disorders was recorded in 38 (30.4%) patients, including 27 (21.6%) with a family history of CD and five (4%) patients with a family history of inflammatory bowel disease (Table 1).

Liver enzymes
Thirty-six (28.8%) children had elevated liver enzymes: 31 (24.8%) with elevated ALT levels (> 24 U/l), of Table 1 Characteristics of 125 patients with celiac disease at presentation AI autoimmune, SD standard deviation, IQR interquartile range, BMI body mass index, IU international unit, ALT alanine aminotransferase, AST aspartate aminotransferase, GGT gamma-glutamyl transpeptidase, ALP alkaline-phosphatase, ULN upper limit of normal, aTTG anti-tissue transglutaminase * Patients with normal aTTG were IgA deficient and were diagnosed using anti-deamidated gliadin protein (aDGP) IgG antibodies whom two had significant elevation of > 3 × ULN; 19 (22.1%) with elevated AST, and one with elevated GGT (Table 2). Children with elevated ALT at presentation were significantly younger (mean age 5.5 (SD 3.4) years   (Table 3). A single patient with elevated GGT level at diagnosis was in the range of 1-2 × ULN and normalized the GGT in the following blood test; no other patients had cholestaticenzyme abnormalities.

Natural history of CD patients with ELE
Patients were followed for up to 1-year post-diagnosis; 60 patients completed a full follow-up of 12 months (fifteen (25%) with elevated liver enzymes). All patients experienced clinical improvement following the introduction of GFD. Eighty percent of patients with elevated ALT levels normalized their ALT within 3 months of follow-up, and all patients within 1 year (Fig. 1). Liver enzyme decrement was associated with, and paralleled, normalization of aTTG values. AST levels showed a similar trend; however, 3 patients (9.4%) had persistent elevated AST levels. Patients with elevated ALT had similar levels of aTTG at diagnosis and similar decrease rates compared to celiac patients with normal liver enzymes during the follow-up period. Likewise, hemoglobin and albumin levels and anthropometrics at 1-year post-diagnosis were all comparable between the two groups (Fig. 2). Two patients had highly elevated liver enzymes, with ALT > × 3 ULN. One patient, with a maximal ALT level of 91 U/l, had a quick resolution, and ALT was normal at the 3-month follow-up visit. The other patient was simultaneously co-diagnosed with celiac disease and autoimmune hepatitis. Family screening, for celiac diagnosis in a sibling, demonstrated a positive celiac serology and highly elevated liver enzymes with maximal ALT of 1570 U/l. Further investigations including autoimmune serology, immunoglobulin levels, gastroscopy, and liver biopsy were positive for the final co-diagnoses. None of the other patients with elevated ALT was subsequently diagnosed with a primary liver disease during active follow-up.

Discussion
In this prospective-observational study of 125 newly diagnosed celiac patients, 31 (24.8%) had elevated ALT at diagnosis. However, only two patients had significant elevations of over 3 × ULN, one of whom was later co-diagnosed with autoimmune hepatitis. Patients with elevated ALT at diagnosis were younger, had associated elevated AST levels (but not cholestatic enzymes), and tended to present more Height percentiles for age and sex (median, IQR). C Weight percentiles for age and sex (median, IQR). D BMI percentiles for age and sex (median, IQR). E Hemoglobin levels (mean, SD). F Albumin levels (mean, SD). aTTG, anti-tissue transglutaminase; ELE, elevated liver enzymes; ALT, alanine-aminotransferase; SD, standard deviation; IQR, interquartile range commonly with diarrhea, compared to the remaining of the cohort. Their natural history was similar to that of normal-ALT patients, including clinical improvement rates on GFD, growth, aTTG decrease rate, and basic laboratory results' changes, over 1 year of follow-up. Elevated liver enzymes, most notably ALT, were previously described as one of the extra-intestinal manifestations of celiac disease, ranging between 4 and 40% of newly diagnosed patients (Table 4; [11][12][13][14][15][16][17][18]). Studies in children demonstrate a downward trend in the incidence of elevated ALT at celiac diagnosis over time. Recent reports demonstrated a relatively low prevalence of 5-15% [16][17][18], compared to older studies in pediatrics with a prevalence of up to ~ 40%; similar to the prevalence in adult CD patients [11][12][13][14][15]24]. In our study, using a relatively low threshold, we found 25% of patients with some degree of ALT elevation. However, considering the commonly used 40 IU threshold, only 5% of patients had ALT elevations, in line with the results of recent studies in children. These temporary changes, we presume, reflect the net effects of increased awareness of the celiac disease in the public and among treating physicians, with a resultant quicker diagnosis for some, and an overall increase in celiac diagnosis in general, specifically of patients with a less severe phenotype of the disease.
Our finding of an association between elevated ALT at presentation and young age at diagnosis was previously described by others: Farre et al., Lee et al., and Benelli et al. [13,17,18]. Other associations, however, were less consistent: Benelli et al. described the association of elevated ALT with a complex of symptoms and laboratory findings, designated as "malabsorption," including diarrhea, failure to thrive, low hemoglobin, and ferritin levels at presentation [18]. Furthermore, other associations described by others, such as higher Marsh scores, lower hemoglobin, ferritin, and thyroidstimulating hormone levels at diagnosis, were not significant in our present study [16,18]. Considering these findings together, we postulate that those patients who are diagnosed younger, on the background of the overall global trend of increasing age at celiac diagnosis [25,26], have a more aggressive phenotype of disease resulting in earlier diagnosis, a more "classical" symptomatic gastrointestinal presentation (i.e., diarrhea) and extraintestinal manifestations.
Nevertheless, despite the high number of patients with elevated ALT, the clinical significance of this finding is limited. The vast majority of patients (80%) normalized their ALT once started on GFD within 3 months, and all by 1 year of follow-up. Furthermore, the natural history of patients with elevated ALT, including clinical improvement, weight gain, linear growth, and basic laboratory studies, did not differ from the remaining of the cohort (Fig. 2).
A key element affecting the primary endpoint of our study, the prevalence of elevated liver enzymes among newly diagnosed celiac patients, is the threshold beyond which a result is considered abnormal. Most past studies set the upper limit of the norm at ALT values of 30-45 U/l (Table 4). We chose to rely on the cutoff of 24 U/l, an average value for males and females, as was found in the national health and nutritional survey (NHANES, [20,21]). This threshold is now widely accepted as the "true" upper limit of normal in lean and healthy children. Setting the bar at a low threshold increased the sensitivity of our study and therefore increased the prevalence we found; however, most of these abnormal findings were not clinically significant, and ultimately, only two patients had substantial elevations. Analyzing AST levels yielded similar results to ALT analysis (data is partially shown); however, the usage of AST in children is hampered by a lack of specificity to the liver, especially in the face of the commonly encountered increased AST levels caused by technical problems in blood drawing resulting in hemolysis. Cholestatic enzymes' elevations were not noted in previous reports, and apart from a single patient with a one-time-only GGT elevation, neither did we find such patients. Finally, our ability to assess alkaline-phosphatase in the cohort was hindered by lack of specificity of the enzyme to the liver and the fact that children have a high bone turnover rate as part of their natural growth, causing elevated alkaline-phosphatase levels of bone origin.
Over time, celiac disease is associated with other autoimmune disorders, most notably, diabetes mellitus type 1 with a co-prevalence of ~ 10% [27]. In our cohort, we found only 7 (6%) patients with other autoimmune diseases, lower than expected. This may reflect the relatively short followup that failed to fully appreciate the autoimmune burden in this cohort. On the contrary, 38 (30%) patients had a positive autoimmune and/or inflammatory family history, underscoring the strong genetic background of this disease.
One patient, presenting with significantly elevated ALT, was co-diagnosed with autoimmune hepatitis. It is likely that the hepatitis evident as highly ELE was driven mainly by specific autoantibodies targeting hepatic epitopes, rather than the non-specific autoantibodies milieu produced by the CD pathogenesis. Nevertheless, it is difficult to assess the true separate contribution of each of the autoimmune processes. Celiac disease prevalence in children with AIH is 3.5-6.4%, and AIH prevalence in patients with celiac disease is 1.4% [28][29][30] further data describing the temporary relationships and interplay of the co-diagnoses is lacking.
The observational-prospective nature of our study enabled us to collect detailed demographic, anthropometric, and clinical data during the diagnostic process of the participants and up to 1 year of follow-up. The majority of patients normalized their ALT levels quickly once the gluten-free diet was introduced, and all, within the 1-year follow-up timeframe. Therefore, prolonging the follow-up duration was not likely to produce further prognostication results regarding the patients with ELE. The main limitation of our study is the moderate size of the cohort and a consequently small number of participants with significant ALT elevations, preventing further sub-group analysis. Patients with significant ALT elevations may very well present a specific disease phenotype; however, our study was not large enough to detect it.
In conclusion, setting a low threshold, ALT abnormalities were common (~ 25%) among newly diagnosed celiac patients and were typically associated with young age at diagnosis; however, significant ALT elevations are uncommon (1.6%). CD should be considered in the differential diagnosis of any unexplained ELE. The natural history of celiac patients with elevated ALT is similar to that of patients with normal liver enzymes. While for most newly diagnosed celiac patients, the existence of elevated liver enzymes is clinically insignificant; significant elevations (> 3 × ULN) failing to normalize on gluten-free diet should raise concern of a co-existing primary liver disease and warrant further investigations.
Authors' contributions Dr. Regev collected the clinical data, analyzed the data and wrote the first version of the manuscript, and approved the final manuscript as submitted. Dr. Ben-Tov participated in the design of the study, reviewed the manuscript, and approved the final manuscript as submitted. Dr. Yerushalmy-Feler reviewed the manuscript and approved the final manuscript as submitted. Dr. Weintraub reviewed the manuscript and approved the final manuscript as submitted. Dr. Moran-Lev reviewed the manuscript and approved the final manuscript as submitted. Prof. Cohen participated in the data analysis, reviewed the manuscript, and approved the final manuscript as submitted. Dr. Amir conceptualized and designed the study, participated in data collection and analysis, wrote the manuscript, and approved the final manuscript as submitted.
Availability of data and material All data is available by request from the author.

Declarations
Ethics approval The study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.
Consent to participate Informed consent was obtained from legal guardians.

Consent for publication
The authors affirm that human research participants provided informed consent for publication of the images in Figs. 1 and 2.

Conflict of interest
The authors declare no competing interests.