ORIGINAL ARTICLE: GASTROENTEROLOGY
Incidence, Pattern, and Etiology of Elevated Liver Enzymes in Pediatric Inflammatory Bowel Disease Antoinette J. Pusateri, Sandra C. Kim, Jennifer L. Dotson, Jane P. Balint, Carol J. Potter, Brendan M. Boyle, and Wallace V. Crandall ABSTRACT Objectives: Patients with inflammatory bowel disease (IBD) often develop elevated liver enzymes (ELE), which are frequently a benign, transient finding, but may be related to treatment or IBD-associated liver diseases. Distinguishing benign from pathologic ELE is crucial for focused diagnostic and therapeutic interventions. We sought to characterize the incidence, character, chronicity, degree, and etiology of ELE in children with IBD. Methods: Institutional review board–approved retrospective review of all of the patients with IBD (2–21 years) seen between October 2009 and October 2012 with >9 months of follow-up were included in the study. We examined body mass index, disease activity, extent, phenotype, concurrent medications, and character, chronicity, degree of enzyme elevation, and final diagnosis. Results: A total of 219 of 514 patients with IBD had 1 episode of ELE. Five patients were excluded for preexisting liver disease, leaving 214 patients (Crohn disease [CD]: 14.8 3.5 years, 46% girls; ulcerative colitis [UC]: 14.4 4.2 years, 37% girls). One hundred forty-eight patients (69%) had a hepatic, 17 (8%) cholestatic, and 49 (23%) mixed character of ELE. There were no significant differences in character, chronicity, or degree of ELE between CD and UC (P ¼ 0.71, P ¼ 0.58, P > 0.33). Of the 128 patients with sufficient data to determine chronicity, 98 (77%) had transient elevations, (CD: n ¼ 66, 75% and UC: n ¼ 32, 80%). Episodes of ELE were idiopathic in 87% of patients with IBD. A final diagnosis of idiopathic ELE was associated with a lower degree of ELE elevation (P < 0.0001). Conclusions: Pediatric patients with IBD commonly experience transient, idiopathic ELE. Our findings suggest that higher degrees of ELE, specifically alanine aminotransferase, are associated with an etiology that requires more extensive evaluation. Key Words: Crohn disease, inflammatory bowel disease, liver enzyme, pediatrics, transaminase, ulcerative colitis
(JPGN 2015;60: 592–597)
H
epatobiliary complications of inflammatory bowel disease (IBD) are well recognized and may contribute to the morbidity, mortality, and medication choices of these patients. Potential Received August 5, 2014; accepted December 4, 2014. From the Division of Pediatric Gastroenterology, Hepatology and Nutrition, Nationwide Children’s Hospital, Columbus, OH. Address correspondence and reprint requests to Wallace Crandall, MD, Nationwide Children’s Hospital, 700 Children’s Dr, Columbus, OH 43205 (e-mail: [email protected]). This project was supported by the Glynn Family Honors Program of the University of Notre Dame (A.J.P.). The authors report no conflicts of interest. Copyright # 2015 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition DOI: 10.1097/MPG.0000000000000672
complications, examined primarily in adult patients, include nonalcoholic fatty liver disease (1%–55%), cholelithiasis (11%–34%), primary sclerosing cholangitis (PSC, 1%–5%), and likely autoimmune hepatitis (AIH) (1). Furthermore, medications used for the treatment of IBD can be hepatotoxic. Therefore, to provide optimal patient care, elevated liver enzymes (ELE) in patients with IBD must be evaluated carefully, given that they may represent significant liver disease or treatment-related toxicity. It has also, however, been recognized that hepatic biochemistries are often elevated without a clear underlying etiology. A study in adults found that approximately 30% of adult patients with IBD experienced an elevation of hepatic biochemistries during a 1-year period, with only 6% having a defined chronic liver disease (2). It has been proposed that active IBD may also be an etiology for ELE. Although some studies have observed that the ELE are correlated with active IBD (3–5), others have shown no association (2). Specifically, Mendes et al (2) found that in patients with abnormal hepatic biochemistries, there was a higher prevalence of patients with inactive IBD, and when patients with chronic liver disease were excluded, there was no significant difference in IBD activity between those with abnormal and those with normal hepatic biochemistries. The effects of disease duration and extent on ELE have also been evaluated. A study by Riegler et al (4) suggests there is more liver damage in the first year after diagnosis, although other studies found no correlation between ELE and disease duration or disease extent (6,7). Data regarding the prevalence, characteristics, and etiology of ELE in pediatric IBD are even more limited. Therefore, we sought to characterize the incidence, character, degree, chronicity, and etiology of ELE in pediatric patients with IBD, and to determine the factors associated with definable liver disease.
METHODS We conducted a retrospective chart review of patients, ages 2 to 21 years, with an established diagnosis of Crohn disease (CD) or ulcerative colitis (UC), who were seen in the Nationwide Children’s Hospital gastroenterology clinic between October 2009 and October 2012. Patients were included if they had 9 months of follow-up (to ensure adequate time for observation and evaluation of persistently abnormal values) and at least 1 repeated measurement of liver enzymes after the first ELE during the study period. As this was a retrospective review, the frequency of repeated liver enzyme measurement was not standardized and reflected the treating physician’s usual practice. Patients were excluded if they had ELE before their IBD diagnosis or if they had known chronic liver disease. Patients with >1 episode of ELE during the study period were only included once, based on the initial
592 JPGN Volume 60, Number 5, May 2015 Copyright 2015 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
JPGN
Volume 60, Number 5, May 2015
Elevated Liver Enzymes in Pediatric IBD
episode. This study was approved by the institutional review board of the Nationwide Children’s Hospital. Demographic data including age, sex, race, and ethnicity were collected on all of the patients with IBD seen during the period of study. An ‘‘episode’’ of ELE was defined as the date of the first ELE to the date of the first normal liver enzymes (that then remained normal for a period of at least 1 year, regardless of the number of times that liver enzymes were measured during that year of follow-up). For patients with at least 1 episode of ELE who met the inclusion/exclusion criteria, we recorded the start date and the end date of the episode, if there was one within the study window. We also recorded the body mass index, IBD diagnosis and disease activity (physician global assessment [PGA] and short Pediatric Crohn’s Disease Activity Index [sPCDAI] (8) or Pediatric Ulcerative Colitis Activity Index [PUCAI] (9)), duration, extent, and phenotype of IBD (Paris Classification (10)), and medications (specifically metronidazole, ciprofloxacin, corticosteroids, sulfasalazine, mesalamine, thiopurines, and antitumor necrosis factor agents) at the time of first ELE. The peak level of alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase, gamma-glutamyl transpeptidase (GGT), and direct bilirubin for each episode was recorded. The definitions of character, chronicity, and degree of ELE were determined before data collection (Table 1). For patients in group 2 (transient, prolonged, consistent), they were subsequently divided (postdata collection and review) into 2a and 2b to distinguish patients who had only had 2 measurements (both elevated) during the study period by which to define ‘‘consistent.’’ The final diagnosis, as determined by the treating physician, was also recorded. Final diagnoses were grouped as ‘‘idiopathic’’ when no definable liver disease was found, or as ‘‘defined’’ when a specific etiology was identified. All of the liver enzyme measurements were reviewed to determine the patterns of character, chronicity, and degree of elevation. The highest value was recorded to determine degree of elevation. Character, chronicity, and degree of ELE were compared between CD and UC patients, as well as against the final diagnoses. Univariate logistic regression was performed comparing multiple predictor variables including the degree of elevation of each hepatic biochemistry, IBD type, PGA, sPCDAI, PUCAI, IBD
phenotype, extent, and duration with the binary outcome of final diagnosis (idiopathic vs defined). Significant factors from the univariate analysis were selected to fit a multivariable logistic regression. Backward model selection was used to select the final model. Categorical data were compared between the groups using likelihood ratio x2 test or Fisher exact test as appropriate. Continuous data were compared using nonparametric Wilcoxon 2-sample test or t test. A P value 1 episode (each of them had 2 episodes) of ELE and were included only once in the analysis. Demographic data are shown in Table 2. There were no significant differences in the demographics or IBD disease activity between those patients with CD and those with UC. There was, however, a significant difference in medications between patients with CD and patients with UC, specifically for prednisone (CD 30.9%, UC 45.2%, P < 0.05), mesalamine (CD 22.4%, UC 62.9%, P < 0.0001), and antitumor necrosis factor-a (CD 39.5%, UC 12.9%, P < 0.0001).
Frequency of Laboratory Testing Patients with a longer duration of transaminase elevation or a higher peak elevation had more liver enzyme measurements. Specifically, patients with a chronicity pattern of 0 had a median of 2 measurements, pattern 1 had 5, pattern 2a had 4, pattern 2b had 2 (by definition), pattern 3 had 9, and pattern 4 had 7. Similarly, patients with a peak ALT elevation of degree 2 had a median of 5.5 measurements, degree 3 had 9, and degree 4 had 10.
TABLE 1. Definitions of pattern, chronicity, and degree of ELE Pattern Hepatic Cholestatic Mixed Chronicity 0 1 2a 2b 3 4 Degree 1 2 3 4
Elevated ALT and/or AST; normal AP, GGT, and DB Elevated AP, GGT and/or DB; normal ALT and AST Elevated ALT and/or AST and elevated AP, GGT, and/or DB Transient, brief (all of the liver enzyme values normalized within 30 days) Transient, prolonged, intermittent (all liver enzyme values normalized between 30 and 180 days, and not every value was elevated in the time period) Transient, prolonged, consistent (all liver enzyme values normalized between 30 and 180 days, every value was elevated in time period) Transient, prolonged, consistent (all liver enzyme values normalized between 30 and 180 days; however, only 2 values were obtained within the time period, ie, both were elevated) Chronic, intermittent (liver enzyme elevation lasted >180 days, not every value was elevated in time period) Chronic, consistent (liver enzyme elevation lasted >180 days, every value was elevated in time period) Peak Peak Peak Peak
liver liver liver liver
enzyme enzyme enzyme enzyme
elevation elevation elevation elevation
was was was was
0–1ULN (ie, ULN) >1–2ULN >2–4ULN >4ULN
ALT ¼ alanine transaminase; AP ¼ alkaline phosphatase; AST ¼ aspartate transaminase; DB ¼ direct bilirubin; ELE ¼ elevated liver enzymes; GGT ¼ gamma-glutamyl transpeptidase; ULN ¼ upper limit of normal.
www.jpgn.org
593
Copyright 2015 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
Crandall et al
JPGN
Volume 60, Number 5, May 2015
TABLE 2. Demographics of patients with ELE at time of first elevation (n ¼ 214) Category No. patients with IBD in study (n) Age (mean SD), y Sex: male, n (%) Ethnicity: non-Hispanic, n (%) Race, n (%) White African American Asian Other BMI at first elevation (mean SD), kg/m2 Disease duration (mean SD), y CD location, n (%) L1 Terminal ileal limited cecal disease L2 Colonic L3 Ileocolonic CD location (L4), n (%) Yes CD behavior, n (%) B1 Nonstricturing nonpenetrating B2 stricturing B3 penetrating CD perianal disease, n (%) UC extent, n, (%) E1 Ulcerative proctitis E2 Left sided E3 Extensive UC severity, n (%) S0 Clinical remission S1 Mild S2 Moderate Medications, n (%) Metronidazole Ciprofloxacin Prednisone Budesonide Methylprednisone Sulfasalazine Mesalamine Thiopurine (6MP/AZA) Methotrexate Anti-TNF-a PGA score (mean SD) Remission Mild Moderate Short PCDAI (mean SD) PUCAI score (mean SD)
n
CD
UC
P
214 214 121 212
152 14.8 3.49 82 (53.9) 151 (99.3)
62 14.4 4.25 39 (62.9) 61 (98.4)
n/a 0.923 0.229 0.497
183 11 3 17 164 214
129 (84.9) 7 (4.6) 2 (1.3) 14 (9.2) 21.2 4.7 2.5 2.5
54 (87.1) 4 (6.5) 1 (1.6) 3 (4.8) 22.0 5.4 1.8 3.0
0.654
20 23 106
20 (13.5) 23 (15.4) 106 (71.1)
n/a n/a n/a
n/a
96
96 (64.9)
n/a
122 11 17 27
122 (81.3) 11 (7.3) 17 (11.3) 27 (18)
n/a n/a n/a n/a
n/a
1 7 51
n/a n/a n/a
1 (1.7) 7 (11.9) 51 (86.4)
n/a
34 22 4
n/a n/a n/a
34 (56.7) 22 (36.7) 4 (6.7)
n/a
15 2 75 4 2 4 73 130 31 68
11 (7.2) 2 (1.3) 47 (30.9) 4 (2.6) 2 (1.3) 2 (1.3) 34 (22.4) 90 (59.2) 24 (15.8) 60 (39.5)
4 (6.5) 0 (0) 28 (45.2) 0 (0) 0 (0) 2 (3.2) 39 (62.9) 40 (64.5) 7 (11.3) 8 (12.9)
1.000 1.000 4ULN had less chance of an idiopathic diagnosis as compared with patients with normal results (odds ratio [OR] 0.098, 95% confidence interval [CI] 0.017–0.559, P ¼ 0.009), as did patients with elevations of 2 to 4ULN (OR 0.095, 95% CI 0.019–0.467, P ¼ 0.004) and patients with elevations of 1 to 2ULN (OR 0.156, 95% CI 0.030–0.822, P ¼ 0.028).
A total of 187 (87.4%) patients had an idiopathic etiology for the ELE (Fig. 1). For patients with defined liver disease, the most common identifiable etiology was medication-related, as determined by the treating physician. There was no association between final diagnosis and type of IBD, disease extent, activity, or duration.
Prior research, performed primarily in adults, suggests that transient elevation of liver enzymes is a common phenomenon, occurring in 14% to 40% of patients (2,3,5,11). A more limited
DISCUSSION
TABLE 3. Chronicity of ELE for patients with CD and patients with UC (n ¼ 128) Chronicity
CD n (%)
UC n (%)
0 Transient, brief ( < 180, not every value) 2a Transient, prolonged, consistent (30 > < 180, every value) 2b 2a except only 2 values taken in this time period 3 Chronic, intermittent (>180 days, not every value) 4 Chronic, consistent (>180 days, all values) Total
27 5 11 23 18 4 88
8 5 7 12 7 1 40
(30.7) (5.7) (12.5) (26.1) (20.5) (4.6)
(20) (12.5) (17.5) (30) (17.5) (2.5)
Total n 35 10 18 35 25 5 128
Pattern of chronicity between patients with CD and patients with UC, P ¼ 0.58. CD ¼ Crohn disease; ELE ¼ elevated liver enzymes; UC ¼ ulcerative colitis.
www.jpgn.org
595
Copyright 2015 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
Crandall et al
JPGN
TABLE 5. Degree of ELE for patients with CD and patients with UC
ALT
AST
AP
GGT
DB
Category
Total
CD, n (%)
UC, n (%)
P
0–1ULN >1–2ULN >2–4ULN >4 ULN 0–1ULN >1–2ULN >2–4ULN >4ULN 0–1ULN >1–2ULN >2–4ULN >4ULN 0–1ULN >1–2ULN >2–4ULN >4ULN 0–1ULN >1–2ULN >2–4ULN >4ULN
81 66 42 25 60 107 38 9 160 42 2 1 50 4 2 4 183 15 2 2
60 (39.5) 50 (32.9) 27 (17.8) 15 (9.9) 42 (27.6) 79 (52) 26 (17.1) 5 (3.3) 117 (79.6) 27 (18.4) 2 (1.4) 1 (0.7) 34 (81) 3 (7.1) 1 (2.4) 4 (9.5) 130 (90.3) 10 (6.9) 2 (1.4) 2 (1.4)
21 (33.9 (2) 16 (25.8) 15 (24.2) 10 (16.1) 18 (29) 28 (45.2) 12 (19.4) 4 (6.5) 43 (74.1) 15 (25.9) 0 (0) 0 (0) 16 (88.9) 1 (5.6) 1 (5.6) 0 (0) 53 (91.4) 5 (8.6) 0 (0) 0 (0)
0.334
0.674
0.534
0.623
0.941
ALT ¼ alanine transaminase; AP ¼ alkaline phosphatase; AST ¼ aspartate transaminase; CD ¼ Crohn disease; DB ¼ direct bilirubin; ELE ¼ elevated liver enzymes; GGT ¼ gamma-glutamyltranspeptidase; UC ¼ ulcerative colitis; ULN ¼ upper limit of normal.
pediatric literature suggests similar rates of ELE. A small study by Nemeth et al (6) studied 46 children with IBD for up to 12 years and found ‘‘pathological liver function tests’’ in approximately 50% of patients. Hyams et al (5) subsequently reviewed 555 pediatric patients with IBD from 2 centers, and found approximately 15% to have an elevation of ALT 80. Although estimates of ELEs will vary by study based on the study definition of elevation, the frequency with which laboratory reports were obtained, which hepatic biochemistries were examined, and the study duration, the rate of ELE seen in our study (41.6%) is consistent with that
Volume 60, Number 5, May 2015
previously described in both adults and children, and adds to a sparse pediatric-specific literature. Taken together, this literature suggests that ELE are a common occurrence in pediatric patients with IBD. The pattern of elevation of hepatic biochemistries has been even less well defined than the overall prevalence. Prior studies suggest that transient elevations occur more frequently than persistent elevations (3,5), that elevations are most commonly mild (2,5), and that a hepatic picture occurs more often than a cholestatic one (3). Similarly, we found that most patients with ELE had transient, low-level elevations, most commonly with a hepatic character. This would be consistent with mild, benign, transient elevations as was most commonly seen. In examining which of these factors (character, chronicity, and degree) was associated with defined liver disease, only the degree of elevation was, however, statistically associated with outcomes. Specifically, lower degrees of liver enzyme elevation were more commonly seen with benign, idiopathic etiologies. Of note, in our study, the measurement of AST in patients with ELE did not add to the value of ALT alone in predicting identifiable liver disease. Some authors have suggested that ELE may be related to disease activity (3–5), whereas others have found no such association (2,6). Other studies suggest there is no correlation between ELE and IBD extent or duration (6,7), although Riegler et al suggest there is more liver damage in the first year after diagnosis (4). In our patients, there was no association between defined liver disease and disease activity, extent, or duration. The etiology of ELE in patients with IBD includes thiopurine-related toxicity (11), concurrent corticosteroid and 5-aminosalicylic acid–related toxicity (4), nonalcoholic fatty liver disease, and cholelithiasis (1,3,4,11), in addition to PSC and AIH (1–3,5,7). In our study, of those patients with a diagnosed reason for their ELE, over half had elevations suspected to be secondary to toxicity of medications related to the treatment of their IBD. Definable infections, including Epstein-Barr virus, cytomegalovirus, and histoplasmosis, were next most common, followed by patients with AIH and AIH/PSC overlap. Of the 5 patients with preexisting liver disease who were excluded from the analysis, 1 patient had PSC, giving an overall rate of AIH, PSC, and AIH/PSC overlap of 1.4% in our population. Viral Histoplasmosis, 3.7% CMV, 3.7% EBV, 7.4%
Vascular malformation, 3.7% NAFLD, 11.1%
Medication related Tylenol, 3.7% MTX, 3.7% Idiopathic, 87.4%
Other, 12.6%
Cholelithiasis, 3.7%
6MP, 51.9% AIH, 3.7%
PSC/AIH overlap, 3.7%
FIGURE 1. Final diagnosis for patients with IBD with ELE. AIH ¼ autoimmune hepatitis; CMV ¼ cytomegalovirus; IBD ¼inflammatory bowel disease; EBV ¼ Epstein-Barr virus; ELE ¼ elevated liver enzymes; 6MP ¼ 6-mercaptopurine; MTX ¼ methotrexate; NAFLD ¼ nonalcoholic fatty liver disease; PSC ¼ primary sclerosing cholangitis.
596
www.jpgn.org
Copyright 2015 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
JPGN
Volume 60, Number 5, May 2015
Most patients (87%) with ELE in our study did not ultimately have a defined etiology; rather the episode was considered idiopathic. It is possible that, in patients with brief and/or low-level elevations, this is a result of a less aggressive evaluation, which could result in underdiagnosis of transient events such as viral infections. It seems, however, unlikely that patients with chronic liver disease would be significantly underdiagnosed. Specifically, all of the patients in our study had follow-up after the first episode of ELE of at least 9 months. Although variation certainly exists, it is common in our practice to proceed with a laboratory and ultrasonography evaluation after a few months of low-level ELE (sooner if more significantly elevated), and to consider liver biopsy after 6 months. Hence, it is likely that patients with ELE for >6 months underwent an extensive evaluation. Furthermore, only a small number of patients had a second episode of ELE, and of those patients, 14 of 15 were attributed to an idiopathic cause. As with any retrospective review, our study is limited in that there was no standardization in the frequency or type of hepatic biochemistry measurement, the evaluation of ELE, or the criteria for assigning causation (etiology). Given our center’s practice of obtaining hepatic biochemistries regularly on patients with IBD, the long study duration, and the substantial follow-up period, it is, however, unlikely that there are numerous patients in our practice with definable chronic liver disease that have not been identified, and our data are likely a reasonable estimate of the rate, pattern, and etiology of ELE in children with IBD. There may, however, be specific circumstances such as in newly diagnosed patients or patients treated with enteral therapy, rather than medication in which our data do not apply. In 2010, Schatorje and Hoekstra (12) reported that in 11 children with newly diagnosed IBD and normal liver enzymes, initiation of 6 weeks of enteral therapy resulted in elevation of AST and ALT levels that peaked at approximately 3 weeks of treatment, and normalized by 12 weeks. Goyal et al (7) reported on 1569 patients from the Pediatric IBD Collaborative Research Group who were newly diagnosed as having IBD and had liver enzymes (ALT, GGT) measured within 90 days of diagnosis. Of these patients, 1.8% was diagnosed as having IBD-related liver disease (PSC, AIH, or PSC/AIH overlap syndrome). In patients in whom both ALT and GGT were elevated during the first 90 days, 21 of 42 (50%) were subsequently found to have IBD-related liver disease. GGT was not, however, routinely obtained by all centers/providers, suggesting that elevation of GGT in the first 90 days may be the result of confounding by indication (ie, it was obtained as part of the evaluation of patients in whom liver disease was already suspected). As with the Goyal et al study, we had limited numbers of patients for whom a GGT was obtained, as it is not a part of our routine liver panel. We therefore have the same risk of confounding by indication were we to include GGT in the multivariate analysis.
www.jpgn.org
Elevated Liver Enzymes in Pediatric IBD Therefore, we chose not to include GGT in our model given the limited number of measurements in our population. We would, however, recommend that future studies of liver enzyme elevations include GGT routinely. In conclusion, ELE is a common finding in children with IBD, occurring in >40% of patients in 3 years in our center. The vast majority of episodes of ELE (87%) are transient events without an identified etiology. Low-grade elevations are associated with a transient, undefined etiology, suggesting that careful observation with repeat measurement of hepatic biochemistries, particularly ALT, may be appropriate in patients lacking other evidence of chronic liver disease, before undertaking a more extensive evaluation. Prospective, population-based studies are needed to definitively explore this issue.
REFERENCES 1. Gizard E, Ford AC, Bronowicki JP, et al. Systematic review: the epidemiology of the hepatobiliary manifestations in patients with inflammatory bowel disease. Aliment Pharmacol Ther 2014;40:3–15. 2. Mendes FD, Levy C, Enders FB, et al. Abnormal hepatic biochemistries in patients with inflammatory bowel disease. Am J Gastroenterol 2007;102:344–50. 3. Yamamoto-Furusho JK, Sanchez-Osorio M, Uribe M. Prevalence and factors associated with the presence of abnormal function liver tests in patients with ulcerative colitis. Ann Hepatol 2010;9:397–401. 4. Riegler G, D’Inca R, Sturniolo GC, et al. Hepatobiliary alterations in patients with inflammatory bowel disease: a multicenter study. Caprilli & Gruppo Italiano Studio Colon-Retto. Scand J Gastroenterol 1998; 33:93–8. 5. Hyams J, Markowitz J, Treem W, et al. Characterization of hepatic abnormalities in children with inflammatory bowel disease. Inflamm Bowel Dis 1995;1:27–33. 6. Nemeth A, Ejderhamn J, Glaumann H, et al. Liver damage in juvenile inflammatory bowel disease. Liver 1990;10:239–48. 7. Goyal A, Hyams JS, Lerer T, et al. Liver enzyme elevations within three months of diagnosis of inflammatory bowel disease and likelihood of liver disease. J Pediatr Gastroenterol Nutr 2014;59:321–3. 8. Kappelman MD, Crandall WV, Colletti RB, et al. Short pediatric Crohn’s disease activity index for quality improvement and observational research. Inflamm Bowel Dis 2011;17:112–7. 9. Turner D, Otley AR, Mack D, et al. Development, validation, and evaluation of a pediatric ulcerative colitis activity index: a prospective multicenter study. Gastroenterology 2007;133:423–32. 10. Levine A, Griffiths A, Markowitz J, et al. Pediatric modification of the Montreal classification for inflammatory bowel disease: the Paris classification. Inflamm Bowel Dis 2011;17:1314–21. 11. Gisbert JP, Luna M, Gonzalez-Lama Y, et al. Liver injury in inflammatory bowel disease: long-term follow-up study of 786 patients. Inflamm Bowel Dis 2007;13:1106–14. 12. Schatorje E, Hoekstra H. Transient hypertransaminasemia in paediatric patients with Crohn disease undergoing initial treatment with enteral nutrition. J Pediatr Gastroenterol Nutr 2010;51:336–40.
597
Copyright 2015 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
Incidence, Pattern, and Etiology of Elevated Liver Enzymes in Pediatric Inflammatory Bowel Disease Antoinette J. Pusateri, Sandra C. Kim, Jennifer L. Dotson, Jane P. Balint, Carol J. Potter, Brendan M. Boyle, and Wallace V. Crandall ABSTRACT Objectives: Patients with inflammatory bowel disease (IBD) often develop elevated liver enzymes (ELE), which are frequently a benign, transient finding, but may be related to treatment or IBD-associated liver diseases. Distinguishing benign from pathologic ELE is crucial for focused diagnostic and therapeutic interventions. We sought to characterize the incidence, character, chronicity, degree, and etiology of ELE in children with IBD. Methods: Institutional review board–approved retrospective review of all of the patients with IBD (2–21 years) seen between October 2009 and October 2012 with >9 months of follow-up were included in the study. We examined body mass index, disease activity, extent, phenotype, concurrent medications, and character, chronicity, degree of enzyme elevation, and final diagnosis. Results: A total of 219 of 514 patients with IBD had 1 episode of ELE. Five patients were excluded for preexisting liver disease, leaving 214 patients (Crohn disease [CD]: 14.8 3.5 years, 46% girls; ulcerative colitis [UC]: 14.4 4.2 years, 37% girls). One hundred forty-eight patients (69%) had a hepatic, 17 (8%) cholestatic, and 49 (23%) mixed character of ELE. There were no significant differences in character, chronicity, or degree of ELE between CD and UC (P ¼ 0.71, P ¼ 0.58, P > 0.33). Of the 128 patients with sufficient data to determine chronicity, 98 (77%) had transient elevations, (CD: n ¼ 66, 75% and UC: n ¼ 32, 80%). Episodes of ELE were idiopathic in 87% of patients with IBD. A final diagnosis of idiopathic ELE was associated with a lower degree of ELE elevation (P < 0.0001). Conclusions: Pediatric patients with IBD commonly experience transient, idiopathic ELE. Our findings suggest that higher degrees of ELE, specifically alanine aminotransferase, are associated with an etiology that requires more extensive evaluation. Key Words: Crohn disease, inflammatory bowel disease, liver enzyme, pediatrics, transaminase, ulcerative colitis
(JPGN 2015;60: 592–597)
H
epatobiliary complications of inflammatory bowel disease (IBD) are well recognized and may contribute to the morbidity, mortality, and medication choices of these patients. Potential Received August 5, 2014; accepted December 4, 2014. From the Division of Pediatric Gastroenterology, Hepatology and Nutrition, Nationwide Children’s Hospital, Columbus, OH. Address correspondence and reprint requests to Wallace Crandall, MD, Nationwide Children’s Hospital, 700 Children’s Dr, Columbus, OH 43205 (e-mail: [email protected]). This project was supported by the Glynn Family Honors Program of the University of Notre Dame (A.J.P.). The authors report no conflicts of interest. Copyright # 2015 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition DOI: 10.1097/MPG.0000000000000672
complications, examined primarily in adult patients, include nonalcoholic fatty liver disease (1%–55%), cholelithiasis (11%–34%), primary sclerosing cholangitis (PSC, 1%–5%), and likely autoimmune hepatitis (AIH) (1). Furthermore, medications used for the treatment of IBD can be hepatotoxic. Therefore, to provide optimal patient care, elevated liver enzymes (ELE) in patients with IBD must be evaluated carefully, given that they may represent significant liver disease or treatment-related toxicity. It has also, however, been recognized that hepatic biochemistries are often elevated without a clear underlying etiology. A study in adults found that approximately 30% of adult patients with IBD experienced an elevation of hepatic biochemistries during a 1-year period, with only 6% having a defined chronic liver disease (2). It has been proposed that active IBD may also be an etiology for ELE. Although some studies have observed that the ELE are correlated with active IBD (3–5), others have shown no association (2). Specifically, Mendes et al (2) found that in patients with abnormal hepatic biochemistries, there was a higher prevalence of patients with inactive IBD, and when patients with chronic liver disease were excluded, there was no significant difference in IBD activity between those with abnormal and those with normal hepatic biochemistries. The effects of disease duration and extent on ELE have also been evaluated. A study by Riegler et al (4) suggests there is more liver damage in the first year after diagnosis, although other studies found no correlation between ELE and disease duration or disease extent (6,7). Data regarding the prevalence, characteristics, and etiology of ELE in pediatric IBD are even more limited. Therefore, we sought to characterize the incidence, character, degree, chronicity, and etiology of ELE in pediatric patients with IBD, and to determine the factors associated with definable liver disease.
METHODS We conducted a retrospective chart review of patients, ages 2 to 21 years, with an established diagnosis of Crohn disease (CD) or ulcerative colitis (UC), who were seen in the Nationwide Children’s Hospital gastroenterology clinic between October 2009 and October 2012. Patients were included if they had 9 months of follow-up (to ensure adequate time for observation and evaluation of persistently abnormal values) and at least 1 repeated measurement of liver enzymes after the first ELE during the study period. As this was a retrospective review, the frequency of repeated liver enzyme measurement was not standardized and reflected the treating physician’s usual practice. Patients were excluded if they had ELE before their IBD diagnosis or if they had known chronic liver disease. Patients with >1 episode of ELE during the study period were only included once, based on the initial
592 JPGN Volume 60, Number 5, May 2015 Copyright 2015 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
JPGN
Volume 60, Number 5, May 2015
Elevated Liver Enzymes in Pediatric IBD
episode. This study was approved by the institutional review board of the Nationwide Children’s Hospital. Demographic data including age, sex, race, and ethnicity were collected on all of the patients with IBD seen during the period of study. An ‘‘episode’’ of ELE was defined as the date of the first ELE to the date of the first normal liver enzymes (that then remained normal for a period of at least 1 year, regardless of the number of times that liver enzymes were measured during that year of follow-up). For patients with at least 1 episode of ELE who met the inclusion/exclusion criteria, we recorded the start date and the end date of the episode, if there was one within the study window. We also recorded the body mass index, IBD diagnosis and disease activity (physician global assessment [PGA] and short Pediatric Crohn’s Disease Activity Index [sPCDAI] (8) or Pediatric Ulcerative Colitis Activity Index [PUCAI] (9)), duration, extent, and phenotype of IBD (Paris Classification (10)), and medications (specifically metronidazole, ciprofloxacin, corticosteroids, sulfasalazine, mesalamine, thiopurines, and antitumor necrosis factor agents) at the time of first ELE. The peak level of alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase, gamma-glutamyl transpeptidase (GGT), and direct bilirubin for each episode was recorded. The definitions of character, chronicity, and degree of ELE were determined before data collection (Table 1). For patients in group 2 (transient, prolonged, consistent), they were subsequently divided (postdata collection and review) into 2a and 2b to distinguish patients who had only had 2 measurements (both elevated) during the study period by which to define ‘‘consistent.’’ The final diagnosis, as determined by the treating physician, was also recorded. Final diagnoses were grouped as ‘‘idiopathic’’ when no definable liver disease was found, or as ‘‘defined’’ when a specific etiology was identified. All of the liver enzyme measurements were reviewed to determine the patterns of character, chronicity, and degree of elevation. The highest value was recorded to determine degree of elevation. Character, chronicity, and degree of ELE were compared between CD and UC patients, as well as against the final diagnoses. Univariate logistic regression was performed comparing multiple predictor variables including the degree of elevation of each hepatic biochemistry, IBD type, PGA, sPCDAI, PUCAI, IBD
phenotype, extent, and duration with the binary outcome of final diagnosis (idiopathic vs defined). Significant factors from the univariate analysis were selected to fit a multivariable logistic regression. Backward model selection was used to select the final model. Categorical data were compared between the groups using likelihood ratio x2 test or Fisher exact test as appropriate. Continuous data were compared using nonparametric Wilcoxon 2-sample test or t test. A P value 1 episode (each of them had 2 episodes) of ELE and were included only once in the analysis. Demographic data are shown in Table 2. There were no significant differences in the demographics or IBD disease activity between those patients with CD and those with UC. There was, however, a significant difference in medications between patients with CD and patients with UC, specifically for prednisone (CD 30.9%, UC 45.2%, P < 0.05), mesalamine (CD 22.4%, UC 62.9%, P < 0.0001), and antitumor necrosis factor-a (CD 39.5%, UC 12.9%, P < 0.0001).
Frequency of Laboratory Testing Patients with a longer duration of transaminase elevation or a higher peak elevation had more liver enzyme measurements. Specifically, patients with a chronicity pattern of 0 had a median of 2 measurements, pattern 1 had 5, pattern 2a had 4, pattern 2b had 2 (by definition), pattern 3 had 9, and pattern 4 had 7. Similarly, patients with a peak ALT elevation of degree 2 had a median of 5.5 measurements, degree 3 had 9, and degree 4 had 10.
TABLE 1. Definitions of pattern, chronicity, and degree of ELE Pattern Hepatic Cholestatic Mixed Chronicity 0 1 2a 2b 3 4 Degree 1 2 3 4
Elevated ALT and/or AST; normal AP, GGT, and DB Elevated AP, GGT and/or DB; normal ALT and AST Elevated ALT and/or AST and elevated AP, GGT, and/or DB Transient, brief (all of the liver enzyme values normalized within 30 days) Transient, prolonged, intermittent (all liver enzyme values normalized between 30 and 180 days, and not every value was elevated in the time period) Transient, prolonged, consistent (all liver enzyme values normalized between 30 and 180 days, every value was elevated in time period) Transient, prolonged, consistent (all liver enzyme values normalized between 30 and 180 days; however, only 2 values were obtained within the time period, ie, both were elevated) Chronic, intermittent (liver enzyme elevation lasted >180 days, not every value was elevated in time period) Chronic, consistent (liver enzyme elevation lasted >180 days, every value was elevated in time period) Peak Peak Peak Peak
liver liver liver liver
enzyme enzyme enzyme enzyme
elevation elevation elevation elevation
was was was was
0–1ULN (ie, ULN) >1–2ULN >2–4ULN >4ULN
ALT ¼ alanine transaminase; AP ¼ alkaline phosphatase; AST ¼ aspartate transaminase; DB ¼ direct bilirubin; ELE ¼ elevated liver enzymes; GGT ¼ gamma-glutamyl transpeptidase; ULN ¼ upper limit of normal.
www.jpgn.org
593
Copyright 2015 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
Crandall et al
JPGN
Volume 60, Number 5, May 2015
TABLE 2. Demographics of patients with ELE at time of first elevation (n ¼ 214) Category No. patients with IBD in study (n) Age (mean SD), y Sex: male, n (%) Ethnicity: non-Hispanic, n (%) Race, n (%) White African American Asian Other BMI at first elevation (mean SD), kg/m2 Disease duration (mean SD), y CD location, n (%) L1 Terminal ileal limited cecal disease L2 Colonic L3 Ileocolonic CD location (L4), n (%) Yes CD behavior, n (%) B1 Nonstricturing nonpenetrating B2 stricturing B3 penetrating CD perianal disease, n (%) UC extent, n, (%) E1 Ulcerative proctitis E2 Left sided E3 Extensive UC severity, n (%) S0 Clinical remission S1 Mild S2 Moderate Medications, n (%) Metronidazole Ciprofloxacin Prednisone Budesonide Methylprednisone Sulfasalazine Mesalamine Thiopurine (6MP/AZA) Methotrexate Anti-TNF-a PGA score (mean SD) Remission Mild Moderate Short PCDAI (mean SD) PUCAI score (mean SD)
n
CD
UC
P
214 214 121 212
152 14.8 3.49 82 (53.9) 151 (99.3)
62 14.4 4.25 39 (62.9) 61 (98.4)
n/a 0.923 0.229 0.497
183 11 3 17 164 214
129 (84.9) 7 (4.6) 2 (1.3) 14 (9.2) 21.2 4.7 2.5 2.5
54 (87.1) 4 (6.5) 1 (1.6) 3 (4.8) 22.0 5.4 1.8 3.0
0.654
20 23 106
20 (13.5) 23 (15.4) 106 (71.1)
n/a n/a n/a
n/a
96
96 (64.9)
n/a
122 11 17 27
122 (81.3) 11 (7.3) 17 (11.3) 27 (18)
n/a n/a n/a n/a
n/a
1 7 51
n/a n/a n/a
1 (1.7) 7 (11.9) 51 (86.4)
n/a
34 22 4
n/a n/a n/a
34 (56.7) 22 (36.7) 4 (6.7)
n/a
15 2 75 4 2 4 73 130 31 68
11 (7.2) 2 (1.3) 47 (30.9) 4 (2.6) 2 (1.3) 2 (1.3) 34 (22.4) 90 (59.2) 24 (15.8) 60 (39.5)
4 (6.5) 0 (0) 28 (45.2) 0 (0) 0 (0) 2 (3.2) 39 (62.9) 40 (64.5) 7 (11.3) 8 (12.9)
1.000 1.000 4ULN had less chance of an idiopathic diagnosis as compared with patients with normal results (odds ratio [OR] 0.098, 95% confidence interval [CI] 0.017–0.559, P ¼ 0.009), as did patients with elevations of 2 to 4ULN (OR 0.095, 95% CI 0.019–0.467, P ¼ 0.004) and patients with elevations of 1 to 2ULN (OR 0.156, 95% CI 0.030–0.822, P ¼ 0.028).
A total of 187 (87.4%) patients had an idiopathic etiology for the ELE (Fig. 1). For patients with defined liver disease, the most common identifiable etiology was medication-related, as determined by the treating physician. There was no association between final diagnosis and type of IBD, disease extent, activity, or duration.
Prior research, performed primarily in adults, suggests that transient elevation of liver enzymes is a common phenomenon, occurring in 14% to 40% of patients (2,3,5,11). A more limited
DISCUSSION
TABLE 3. Chronicity of ELE for patients with CD and patients with UC (n ¼ 128) Chronicity
CD n (%)
UC n (%)
0 Transient, brief ( < 180, not every value) 2a Transient, prolonged, consistent (30 > < 180, every value) 2b 2a except only 2 values taken in this time period 3 Chronic, intermittent (>180 days, not every value) 4 Chronic, consistent (>180 days, all values) Total
27 5 11 23 18 4 88
8 5 7 12 7 1 40
(30.7) (5.7) (12.5) (26.1) (20.5) (4.6)
(20) (12.5) (17.5) (30) (17.5) (2.5)
Total n 35 10 18 35 25 5 128
Pattern of chronicity between patients with CD and patients with UC, P ¼ 0.58. CD ¼ Crohn disease; ELE ¼ elevated liver enzymes; UC ¼ ulcerative colitis.
www.jpgn.org
595
Copyright 2015 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
Crandall et al
JPGN
TABLE 5. Degree of ELE for patients with CD and patients with UC
ALT
AST
AP
GGT
DB
Category
Total
CD, n (%)
UC, n (%)
P
0–1ULN >1–2ULN >2–4ULN >4 ULN 0–1ULN >1–2ULN >2–4ULN >4ULN 0–1ULN >1–2ULN >2–4ULN >4ULN 0–1ULN >1–2ULN >2–4ULN >4ULN 0–1ULN >1–2ULN >2–4ULN >4ULN
81 66 42 25 60 107 38 9 160 42 2 1 50 4 2 4 183 15 2 2
60 (39.5) 50 (32.9) 27 (17.8) 15 (9.9) 42 (27.6) 79 (52) 26 (17.1) 5 (3.3) 117 (79.6) 27 (18.4) 2 (1.4) 1 (0.7) 34 (81) 3 (7.1) 1 (2.4) 4 (9.5) 130 (90.3) 10 (6.9) 2 (1.4) 2 (1.4)
21 (33.9 (2) 16 (25.8) 15 (24.2) 10 (16.1) 18 (29) 28 (45.2) 12 (19.4) 4 (6.5) 43 (74.1) 15 (25.9) 0 (0) 0 (0) 16 (88.9) 1 (5.6) 1 (5.6) 0 (0) 53 (91.4) 5 (8.6) 0 (0) 0 (0)
0.334
0.674
0.534
0.623
0.941
ALT ¼ alanine transaminase; AP ¼ alkaline phosphatase; AST ¼ aspartate transaminase; CD ¼ Crohn disease; DB ¼ direct bilirubin; ELE ¼ elevated liver enzymes; GGT ¼ gamma-glutamyltranspeptidase; UC ¼ ulcerative colitis; ULN ¼ upper limit of normal.
pediatric literature suggests similar rates of ELE. A small study by Nemeth et al (6) studied 46 children with IBD for up to 12 years and found ‘‘pathological liver function tests’’ in approximately 50% of patients. Hyams et al (5) subsequently reviewed 555 pediatric patients with IBD from 2 centers, and found approximately 15% to have an elevation of ALT 80. Although estimates of ELEs will vary by study based on the study definition of elevation, the frequency with which laboratory reports were obtained, which hepatic biochemistries were examined, and the study duration, the rate of ELE seen in our study (41.6%) is consistent with that
Volume 60, Number 5, May 2015
previously described in both adults and children, and adds to a sparse pediatric-specific literature. Taken together, this literature suggests that ELE are a common occurrence in pediatric patients with IBD. The pattern of elevation of hepatic biochemistries has been even less well defined than the overall prevalence. Prior studies suggest that transient elevations occur more frequently than persistent elevations (3,5), that elevations are most commonly mild (2,5), and that a hepatic picture occurs more often than a cholestatic one (3). Similarly, we found that most patients with ELE had transient, low-level elevations, most commonly with a hepatic character. This would be consistent with mild, benign, transient elevations as was most commonly seen. In examining which of these factors (character, chronicity, and degree) was associated with defined liver disease, only the degree of elevation was, however, statistically associated with outcomes. Specifically, lower degrees of liver enzyme elevation were more commonly seen with benign, idiopathic etiologies. Of note, in our study, the measurement of AST in patients with ELE did not add to the value of ALT alone in predicting identifiable liver disease. Some authors have suggested that ELE may be related to disease activity (3–5), whereas others have found no such association (2,6). Other studies suggest there is no correlation between ELE and IBD extent or duration (6,7), although Riegler et al suggest there is more liver damage in the first year after diagnosis (4). In our patients, there was no association between defined liver disease and disease activity, extent, or duration. The etiology of ELE in patients with IBD includes thiopurine-related toxicity (11), concurrent corticosteroid and 5-aminosalicylic acid–related toxicity (4), nonalcoholic fatty liver disease, and cholelithiasis (1,3,4,11), in addition to PSC and AIH (1–3,5,7). In our study, of those patients with a diagnosed reason for their ELE, over half had elevations suspected to be secondary to toxicity of medications related to the treatment of their IBD. Definable infections, including Epstein-Barr virus, cytomegalovirus, and histoplasmosis, were next most common, followed by patients with AIH and AIH/PSC overlap. Of the 5 patients with preexisting liver disease who were excluded from the analysis, 1 patient had PSC, giving an overall rate of AIH, PSC, and AIH/PSC overlap of 1.4% in our population. Viral Histoplasmosis, 3.7% CMV, 3.7% EBV, 7.4%
Vascular malformation, 3.7% NAFLD, 11.1%
Medication related Tylenol, 3.7% MTX, 3.7% Idiopathic, 87.4%
Other, 12.6%
Cholelithiasis, 3.7%
6MP, 51.9% AIH, 3.7%
PSC/AIH overlap, 3.7%
FIGURE 1. Final diagnosis for patients with IBD with ELE. AIH ¼ autoimmune hepatitis; CMV ¼ cytomegalovirus; IBD ¼inflammatory bowel disease; EBV ¼ Epstein-Barr virus; ELE ¼ elevated liver enzymes; 6MP ¼ 6-mercaptopurine; MTX ¼ methotrexate; NAFLD ¼ nonalcoholic fatty liver disease; PSC ¼ primary sclerosing cholangitis.
596
www.jpgn.org
Copyright 2015 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
JPGN
Volume 60, Number 5, May 2015
Most patients (87%) with ELE in our study did not ultimately have a defined etiology; rather the episode was considered idiopathic. It is possible that, in patients with brief and/or low-level elevations, this is a result of a less aggressive evaluation, which could result in underdiagnosis of transient events such as viral infections. It seems, however, unlikely that patients with chronic liver disease would be significantly underdiagnosed. Specifically, all of the patients in our study had follow-up after the first episode of ELE of at least 9 months. Although variation certainly exists, it is common in our practice to proceed with a laboratory and ultrasonography evaluation after a few months of low-level ELE (sooner if more significantly elevated), and to consider liver biopsy after 6 months. Hence, it is likely that patients with ELE for >6 months underwent an extensive evaluation. Furthermore, only a small number of patients had a second episode of ELE, and of those patients, 14 of 15 were attributed to an idiopathic cause. As with any retrospective review, our study is limited in that there was no standardization in the frequency or type of hepatic biochemistry measurement, the evaluation of ELE, or the criteria for assigning causation (etiology). Given our center’s practice of obtaining hepatic biochemistries regularly on patients with IBD, the long study duration, and the substantial follow-up period, it is, however, unlikely that there are numerous patients in our practice with definable chronic liver disease that have not been identified, and our data are likely a reasonable estimate of the rate, pattern, and etiology of ELE in children with IBD. There may, however, be specific circumstances such as in newly diagnosed patients or patients treated with enteral therapy, rather than medication in which our data do not apply. In 2010, Schatorje and Hoekstra (12) reported that in 11 children with newly diagnosed IBD and normal liver enzymes, initiation of 6 weeks of enteral therapy resulted in elevation of AST and ALT levels that peaked at approximately 3 weeks of treatment, and normalized by 12 weeks. Goyal et al (7) reported on 1569 patients from the Pediatric IBD Collaborative Research Group who were newly diagnosed as having IBD and had liver enzymes (ALT, GGT) measured within 90 days of diagnosis. Of these patients, 1.8% was diagnosed as having IBD-related liver disease (PSC, AIH, or PSC/AIH overlap syndrome). In patients in whom both ALT and GGT were elevated during the first 90 days, 21 of 42 (50%) were subsequently found to have IBD-related liver disease. GGT was not, however, routinely obtained by all centers/providers, suggesting that elevation of GGT in the first 90 days may be the result of confounding by indication (ie, it was obtained as part of the evaluation of patients in whom liver disease was already suspected). As with the Goyal et al study, we had limited numbers of patients for whom a GGT was obtained, as it is not a part of our routine liver panel. We therefore have the same risk of confounding by indication were we to include GGT in the multivariate analysis.
www.jpgn.org
Elevated Liver Enzymes in Pediatric IBD Therefore, we chose not to include GGT in our model given the limited number of measurements in our population. We would, however, recommend that future studies of liver enzyme elevations include GGT routinely. In conclusion, ELE is a common finding in children with IBD, occurring in >40% of patients in 3 years in our center. The vast majority of episodes of ELE (87%) are transient events without an identified etiology. Low-grade elevations are associated with a transient, undefined etiology, suggesting that careful observation with repeat measurement of hepatic biochemistries, particularly ALT, may be appropriate in patients lacking other evidence of chronic liver disease, before undertaking a more extensive evaluation. Prospective, population-based studies are needed to definitively explore this issue.
REFERENCES 1. Gizard E, Ford AC, Bronowicki JP, et al. Systematic review: the epidemiology of the hepatobiliary manifestations in patients with inflammatory bowel disease. Aliment Pharmacol Ther 2014;40:3–15. 2. Mendes FD, Levy C, Enders FB, et al. Abnormal hepatic biochemistries in patients with inflammatory bowel disease. Am J Gastroenterol 2007;102:344–50. 3. Yamamoto-Furusho JK, Sanchez-Osorio M, Uribe M. Prevalence and factors associated with the presence of abnormal function liver tests in patients with ulcerative colitis. Ann Hepatol 2010;9:397–401. 4. Riegler G, D’Inca R, Sturniolo GC, et al. Hepatobiliary alterations in patients with inflammatory bowel disease: a multicenter study. Caprilli & Gruppo Italiano Studio Colon-Retto. Scand J Gastroenterol 1998; 33:93–8. 5. Hyams J, Markowitz J, Treem W, et al. Characterization of hepatic abnormalities in children with inflammatory bowel disease. Inflamm Bowel Dis 1995;1:27–33. 6. Nemeth A, Ejderhamn J, Glaumann H, et al. Liver damage in juvenile inflammatory bowel disease. Liver 1990;10:239–48. 7. Goyal A, Hyams JS, Lerer T, et al. Liver enzyme elevations within three months of diagnosis of inflammatory bowel disease and likelihood of liver disease. J Pediatr Gastroenterol Nutr 2014;59:321–3. 8. Kappelman MD, Crandall WV, Colletti RB, et al. Short pediatric Crohn’s disease activity index for quality improvement and observational research. Inflamm Bowel Dis 2011;17:112–7. 9. Turner D, Otley AR, Mack D, et al. Development, validation, and evaluation of a pediatric ulcerative colitis activity index: a prospective multicenter study. Gastroenterology 2007;133:423–32. 10. Levine A, Griffiths A, Markowitz J, et al. Pediatric modification of the Montreal classification for inflammatory bowel disease: the Paris classification. Inflamm Bowel Dis 2011;17:1314–21. 11. Gisbert JP, Luna M, Gonzalez-Lama Y, et al. Liver injury in inflammatory bowel disease: long-term follow-up study of 786 patients. Inflamm Bowel Dis 2007;13:1106–14. 12. Schatorje E, Hoekstra H. Transient hypertransaminasemia in paediatric patients with Crohn disease undergoing initial treatment with enteral nutrition. J Pediatr Gastroenterol Nutr 2010;51:336–40.
597
Copyright 2015 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
