Secondary Logo

Journal Logo

Short Communication: Gastroenterology: Inflammatory Bowel Disease

Evaluation of Nonalcoholic Fatty Liver Disease in Pediatric Patients With Inflammatory Bowel Disease

Cohen, Mallory E.; Deepak, Parakkal; Khanna, Geetika†,‡; Samson, Charles M.

Author Information
Journal of Pediatric Gastroenterology and Nutrition: April 2021 - Volume 72 - Issue 4 - p 574-578
doi: 10.1097/MPG.0000000000003023

Abstract

What Is Known/What Is New

What Is Known

  • Metabolic risk factors, such as obesity, are known to be associated with nonalcoholic fatty liver disease.
  • There is a coexistence of inflammatory bowel disease and nonalcoholic fatty liver disease in adults without traditional metabolic risk factors.

What Is New

  • Pediatric patients with inflammatory bowel disease and nonalcoholic fatty liver disease lack the same risk factors as adult counterparts.
  • Prevalence of nonalcoholic fatty liver disease in pediatric patients with inflammatory bowel disease is similar to general pediatric population.

Nonalcoholic fatty liver disease (NAFLD) is one of the most common liver disorders in the United States, resulting from excessive fat accumulation within the liver and can progress to nonalcoholic steatohepatitis (1,2). The prevalence of childhood NAFLD is increasing, with estimates as high as 10% (3). Recent studies demonstrate the coexistence of NAFLD and inflammatory bowel disease (IBD); however, these patients lack traditional metabolic NAFLD risk factors (4–6). Moreover, there has been a rise of obesity within pediatric IBD populations, which is ordinarily a risk factor for NAFLD (7). There is a paucity of data examining the prevalence and risk factors of NAFLD in children with IBD.

Here, we sought to establish the prevalence of NAFLD in a single-center pediatric IBD cohort, and confirm or refute the presence of classic risk factors.

METHODS

Study Design and Population

This study was approved by the Washington University School of Medicine Human Research Protection Office. We enrolled children with IBD who underwent abdominal magnetic resonance enterography (MRE) as part of routine care between October 2018 and September 2019 at St Louis Children's Hospital. Exclusion criteria consisted of age older than 21 years, known liver disease, and use of sedation for the MRE.

MRE was performed on Siemens 1.5 T scanner (Aera). Fat percentage in the liver was estimated by proton density fat fraction (PDFF) technique using a multiecho DIXON vibe sequence (Siemens LiverLab). An experienced pediatric radiologist (G.K.) trained one of the authors (M.E.C.) to draw regions of interest (ROIs) over the liver to estimate fat percentage. ROIs were drawn on 4 slices using a free-hand method through a mid-section of the liver, following the contour of the liver and avoiding all major vessels and biliary structures (Figure, Supplemental Digital Content 1, https://links.lww.com/MPG/C134). The mean of the 4 ROIs was used as the PDFF of the liver. The authors were blinded to clinical data at the time of drawing ROIs. For the first 7 cases, PDFF was estimated by the radiologist and trained author to determine interobserver correlation. For this study, we classified NAFLD as mean PDFF >5% (8).

Data Collection

Demographic data, medical and surgical history, medications used, total parental nutrition (TPN) use, comorbidities, disease classification/extent, family history, and growth were collected from the electronic medical record. The short pediatric Crohn disease activity index and pediatric ulcerative colitis activity disease activity index was determined from the most recent clinic visit (9,10). Laboratory data obtained as part of routine care in the 30 days before or after the MRE were also collected.

Statistics

Statistical analyses were performed using IBM SPSS Statistics version 25 (IBM, Armonk, NY). Categorical variables were analyzed using Fischer exact test for variables with 2 classifications and χ2 analysis for variables with >2 classifications. Continuous variables were analyzed using the Mann-Whitney U test. Correlations between continuous variables were analyzed using Pearson correlation coefficient. Results were considered statistically significant for 2-tailed P values ≤0.05.

RESULTS

Eighty-three patients with IBD underwent an MRE examination during the study interval. The study cohort median age was 14.8 (13.0–16.7) years. Fifty-seven (69%) had Crohn disease, 23 (28%) had ulcerative colitis, and 3 (3%) had indeterminate colitis. Forty (48%) MRE examinations were obtained as part of the initial diagnostic evaluation, 24 (29%) to evaluate acute symptoms after diagnosis, and 19 (23%) to evaluate for radiologic improvement in patients in clinical remission. PDFF estimates ranged from 0.1% to 9.9%, with a median of 1.4% (Fig. 1A). Five (6%) of the 83 patients had PDFF estimated >5% on MRE. Subjective assessment by the radiologist did not show evidence of hepatic fibrosis in any patients. There was no difference in IBD diagnosis of patients with and without NAFLD (P = 0.06) or in disease activity scores (P = 0.54). The ages of the children with and without NAFLD were similar (P = 0.68), as was sex (P = 0.64). The overall interval between diagnosis of IBD and performance of the MRE was 29 (1–94) months in the NAFLD group and 3 (0–44) months in the group without NAFLD (P = 0.24) (Table 1). There was no correlation between disease duration and PDFF (R = 0.16, P = 0.16) (Fig. 1B).

F1
FIGURE 1:
Correlations of PDFF. PDFF (%) was determined by calculating the mean of 4 regions of interests on magnetic resonance imaging (MRI). Linear fit-line slopes are shown (B–D) and Pearson correlation coefficients determined. A, Histogram of PDFF (%). B, PDFF was correlated with the BMI (kg/m2) at the clinical encounter before MRI; R = 0.35, P < 0.001. C, PDFF was correlated with ALT (IU/L); R = 0.10, P = 0.40. D, PDFF was correlated with disease duration at the time of the MRI; R = 0.16, P = 0.16. ALT = alanine aminotransferase; BMI = body mass index; PDFF = proton density fat fraction.
TABLE 1 - Patient characteristics
NAFLD (n = 5) Non-NAFLD (78) Overall P
Age median (IQR) 13.6 (12.5–17.1) 14.8 (13.0–16.7) 14.8 (13.0–16.7) 0.68
Male sex (%) 4 (80) 46 (59) 50 (60) 0.64
Family history IBD, yes (%) 3 (60) 21 (27) 24 (29) 0.14
Ethnicity (%) 0.74
 White 5 (100) 62 (80) 67 (81)
 African American 0 (0) 12 (15) 12 (14)
 Asian 0 (0) 3 (4) 3 (4)
 Native American 0 (0) 1 (1) 1 (1)
Family history liver disease, yes (%) 0 1 (1) NAFLD 1 (1) 1.0
Diagnosis (%) 0.06
 Crohn disease 4 (80) 53 (68) 57 (69)
 UC 0 (0) 23 (29) 23 (28)
 Indeterminate 1 (20) 2 (5) 3 (4)
Disease location (Crohn and indeterminate) (%) 0.87
 SB only 1 (20) 15 (27) 16 (27)
 Colonic only 1 (20) 3 (6) 4 (7)
 SB + colon 3 (6) 37 (67) 40 (67)
Disease activity score (%) 0.54
 Remission 1 (20) 31 (40) 32 (39)
 Mild 3 (60) 28 (36) 31 (37)
 Moderate/severe 1 (20) 19 (24) 20 (24)
Indication for MRE (5) 0.64
 Initial diagnostic evaluation 2 (40) 38 (49) 40 (48)
 Acute symptom evaluation 1 (20) 23 (29) 24 (29)
 Mucosal healing evaluation 2 (40) 17 (22) 19 (23)
 Median time between diagnosis and MRE (IQR), mo 29 (1–93.5) 3 (0–44) 4 (1–45) 0.24
Current medication (%)
 Steroids 2 (40) 35 (45) 37 (44) 1.0
 Biologic 4 (80) 18 (23) 22 (26) 0.06
ALT concentration units/L, median (IQR) (number assessed) 12 (7.5–26.5) (n = 5) 11 (9–17) (n = 74) 11 (9–17) (n = 79) 0.84
Median BMI (IQR) 21.4 (19.6–29.2) 19.25 (16.5–22.2) 19.4 (16.6–22.3) 0.29
Weight status (%) 0.30
 Underweight 0 (0) 13 (17) 13 (16)
 Normal 3 (60) 53 (68) 56 (68)
 Overweight 1 (20) 9 (12) 10 (12)
 Obese 1 (20) 3 (4) 4 (5)
ΔWAZ over prior 6 mo (%) 0.25
 Decrease 3 (60) 26 (33) 29 (35)
 No change 0 (0) 27 (35) 27 (32)
 Increase 2 (40) 25 (32) 27 (32)
ΔBMIZ over prior 6 mo (%) 0.84
 Decrease 2 (40) 30 (39) 32 (39)
 No change 1 (10) 24 (31) 25 (30)
 Increase 2 (40) 23 (30) 25 (30)
ALT = alanine aminotransferase; BMI = body mass index; BMIZ, body mass index z score; IBD = inflammatory bowel disease; IQR, interquartile range; MRE = magnetic resonance enterography; NAFLD = nonalcoholic fatty liver disease; SB, small bowel; UC, ulcerative colitis; WAZ, weight z score.

Of the 5 patients with NAFLD, 3 had normal body mass index (BMI) at time of imaging. There was a statistically significant positive correlation between BMI and PDFF estimate (R = 0.35, P < 0.001) (Fig. 1C). Twelve of the 78 non-NAFLD patients were overweight (BMI >85th percentile) or obese (BMI >95th percentile), 13 were underweight (BMI < 5th percentile), and 53 had a normal BMI. Three of the patients with NAFLD had recently lost weight and 2 had gained weight. The weight loss of one of the patients was insufficient to reduce the BMI. The other 2 had ΔBMI of −0.2 over 5 months and −0.4 over 2 months, whereas the 2 patients who gained weight had ΔBMI of +0.6 over 3 months and +0.3 over 6 months, respectively (Table, Supplemental Digital Content 2, https://links.lww.com/MPG/C135). Median BMI in the NAFLD group was 21.4 (19.6–29.2) versus 19.25 (16.5–22.2) in the non-NAFLD group (Table 1).

Two of the 5 patients with NAFLD were on steroids at the time of imaging, versus 35 of 78 in the non-NAFLD group (P = 1.0). Four of the 5 NAFLD patients were receiving anti-tumor necrosis factor (TNF) biologics at the time of imaging (P = 0.06); the fifth was receiving mesalamine (Table, Supplemental Digital Content 2, https://links.lww.com/MPG/C135). There were no patients in the study on TPN.

Serum alanine aminotransferase (ALT) concentrations among patients with IBD with and without NAFLD were 12 U/L (7.5–26.5) and 11 U/L (9–17), respectively (P = 0.84). As noted by similar values, ALT concentration was not bimodal. There was not a statistically significant correlation between ALT concentration and PDFF estimate (R = 0.10, P = 0.40) (Fig. 1D).

Mean PDFF determinations between the 2 readers were within 0.1% with an intraclass correlation coefficient of 0.99 (P < 0.001) (Figure, Supplemental Digital Content 3, https://links.lww.com/MPG/C136). The patient with the highest PDFF score is an 18.9-year-old boy with Crohn disease who underwent MRE 77 months after diagnosis. He was on maintenance anti-TNF therapy with infliximab, and had not received steroids for at least 2 years before the MRE was performed. He was obese at the time of imaging (BMI = 35.7) despite recent weight loss. His serum ALT concentration at the time of imaging was 31 U/L but was elevated to 71 U/L, 1 month earlier (males: mild elevation >26 U/L, high elevation >50 U/L) (11); his serum aspartate aminotransferase was normal. He reported recent alcohol use. He was the only patient in this study to undergo bowel surgery before his MRE, having an ileocecal resection 6 years before his MRE. He also underwent fecal transplant 5 years before MRE for recurrent Clostridium difficile infection. He had a family history of Crohn disease but not of liver disease.

Compared to the population with IBD and no NAFLD, none of the evaluated risk factors were statistically significant (Table 1).

DISCUSSION

Although the co-existence of IBD and NAFLD has been well-recognized in adults and associated with specific risk factors, this study shows for the first time that children with both NAFLD and IBD lack some of the same associations.

The percentage prevalence of NAFLD in this IBD cohort is similar to recent reports in the general pediatric population. Schwimmer et al (12) showed a 9.6% prevalence of NAFLD in the general pediatric population in San Diego and a recent meta-analysis found a worldwide prevalence of 7.6% (13). Metabolic comorbidities are known to increase the prevalence of NAFLD in the general pediatric population (3). Despite this, we found that in our IBD population, patients with NAFLD did not have identifiable metabolic risk factors. This has implications on treatment challenges this population as management of pediatric NAFLD largely targets lifestyle modifications (14). We also had a large population of patients with IBD with metabolic risk factors who did not have NAFLD.

One factor that often sparks investigation in the general childhood populations is elevated transaminases. These concentrations are often monitored in patients with IBD, to screen for hepatobiliary disease, and nonpersisting ALT elevations in patients with IBD are common and often mild (15). Our findings reinforce these observations, as many patients had significant ALT elevations without associated liver abnormalities on MRE.

The duration of time since IBD diagnosis is a common risk factor for NAFLD in adults with IBD (6). While in general, pediatric patients with IBD will have had shorter intervals since diagnosis before their first MRE, we did not observe a correlation between time since diagnosis and NAFLD. A shorter duration of disease may, however, explain the lower prevalence of NAFLD in pediatric patients with IBD compared to adults with IBD and NAFLD.

Other cited risk factors in the adult population are prior small bowel surgery and current steroid use (5). Despite just less than half of our patients being on steroids at the time of diagnosis and more than half having used steroids in the 6 months before imaging, we found no significant increase in NAFLD in the children receiving steroids. Our evaluation of prior surgery as a risk factor was limited in that only 1 study patient had undergone surgery before obtaining MRE.

Contrary to reports that suggest NAFLD develops in patients with IBD of younger or older age than those without NAFLD, our study did not demonstrate a large effect of age on NAFLD development (16). This could also be due to by definition of our population. All patients in this study were younger than 22 years. The timing of detection should not be confused with timing of onset, and NAFLD may have been detected earlier, if MREs were performed at equivalent points of illness after diagnosis. On the contrary, the prevalence of NAFLD in this IBD population is not different from that in the general non-IBD population.

Our study used MRE as our imaging modality as it is the preferred imaging modality for patients with IBD as it lacks ionizing radiation and has the capacity to examine mucosal and extraluminal findings with high diagnostic reliability (17). We used PDFF maps which are quantitative markers of MRI visible hepatic fat content to diagnose NAFLD as they have high diagnostic accuracy to classify and predict histologic steatosis grade in children with NAFLD (18).

NAFLD is defined as excessive fat accumulation in the liver with >5% of hepatocytes affected, making PDFF >5 the common threshold for NAFLD in pediatrics (8). There has been discussion within adult literature of a higher cutoff, as PDFF values do not demonstrate a bimodal distribution, which more easily designate findings as categorically normal or abnormal (18). This was similarly demonstrated in our population.

Liver biopsy is the current criterion standard to diagnosis NAFLD. Although safe, it typically is reserved to identify alternative etiologies for hepatic steatosis (14). Therefore, imaging has been used as a screening tool for pediatric NAFLD. Although readily available, routine ultrasonography has low sensitivity and specificity (19). Although MRI-PDFF has high diagnostic accuracy as discussed above, it does not provide quantitative assessment of fibrosis and cost-effectiveness for only NAFLD screening needs to be addressed (20).

There were several limitations to our study, mainly its small sample size and performance at a single center. The lack of patients with the risk factors identified in adult studies limited our ability to evaluate if they would apply to the pediatric population as well. Nonetheless, our data do demonstrate differences in characteristics between adults and children with IBD and presence of NAFLD. In addition, there may be other possible confounders, such as alcohol use which was not systematically collected, which we did not identify in this study.

In summary, we demonstrate the overall low prevalence and occult nature of NAFLD in pediatric patients with IBD. The prevalence is not at variance with general teenage populations. In addition, we could not identify risk factors including those which are seen in general pediatric populations. Further investigation is warranted to understand the natural history of NAFLD within the pediatric IBD population and determine screening recommendations.

Acknowledgments

The authors thank Phillip Tarr, MD for his comments on our manuscript.

REFERENCES

1. Wegermann K, Suzuki A, Mavis AM, et al. Tackling NAFLD: three target populations. Hepatology 2020; [Epub ahead of print].
2. Loomba R, Chalasni N. The hierarchical model of NAFLD: prognostic significance of histologic features in NASH. Gastroenterology 2015; 149:278–281.
3. Nobili V, Piotr S. Pediatric nonalcoholic fatty liver disease: current thinking. J Pediatr Gastroenterol Nutr 2018; 66:188–192.
4. McHenry S, Sharma Y, Tirath A, et al. Crohn's disease is associated with an increased prevalence of nonalcoholic fatty liver disease: a cross-sectional study using magnetic resonance proton density fat fraction mapping. Clin Gastroenterol Hepatol 2019; 17:2816–2818.
5. Lin A, Roth H, Anyane-Yeboa A, et al. Prevalence of nonalcoholic fatty liver disease in patients with inflammatory bowel disease: a systematic review and meta-analysis. Inflamm Bowel Dis 2020; [Epub ahead of print].
6. Glassner K, Malaty HM, Abraham BP. Epidemiology and risk factors of nonalcoholic fatty liver disease among patients with inflammatory bowel disease. Inflamm Bowel Dis 2017; 23:998–1003.
7. Long MD, Crandall WV, Leibowitz IH, et al. Prevalence and epidemiology of overweight and obesity in children with inflammatory bowel disease. Inflamm Bowel Dis 2011; 17:2162–2168.
8. Mouzaki M, Trout AT. Virtual reality: new insights regarding the prevalence of nonalcoholic fatty liver disease in children and adolescents with obesity using magnetic resonance imaging. J Pediatr 2019; 207:8–10.
9. 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; 17:112–117.
10. Turner D, Otley AR, Mack D, et al. Development, validation, and evaluation of a pediatric ulcerative colitis activity index: a prospective multicenter study. Gastroenterol 2007; 133:423–432.
11. Schwimmer JB, Dunn W, Norman GJ, et al. SAFETY study: alanine aminotransferase cutoff values are set too high for reliable detection of pediatric chronic liver disease. Gastroenterology 2010; 138:1357–1364.
12. Schwimmer JB, Deutsch R, Kahen T, et al. Prevalence of fatty liver in children and adolescents. Pediatrics 2006; 118:1388–1393.
13. Anderson EL, Howe LD, Jones HE, et al. The prevalence of non-alcoholic fatty liver disease in children and adolescents: a systematic review and meta-analysis. PLoS One 2015; 10:e0140908.
14. Vos MB, Abrams SH, Barlow SE, et al. NASPGHAN clinical practice guideline for the diagnosis and treatment of nonalcoholic fatty liver disease in children: recommendations from the expert committee on NAFLD (ECON) and the North American society of pediatric gastroenterology, Hepatology and Nutrition (NASPGHAN). J Pediatr Gastroenterol Nutr 2017; 64:319–334.
15. Cappello M, Randazzo C, Bravatà I, et al. Liver function test abnormalities in patients with inflammatory bowel diseases: a hospital-based survey. Clin Med Insights Gastroenterol 2014; 7:25–31.
16. Zou ZY, Shen B, Fan JG. Systematic review with meta-analysis: epidemiology of nonalcoholic fatty liver disease in patients with inflammatory bowel disease. Inflamm Bowel Dis 2019; 25:1764–1772.
17. Kilcoyne A, Kaplan JL, Gee MS. Inflammatory bowel disease imaging: current practice and future directions. World J Gastroenterol 2016; 22:917–932.
18. Middleton MS, Van Natta ML, Heba ER, et al. Diagnostic accuracy of magnetic resonance imaging hepatic proton density fat fraction in pediatric nonalcoholic fatty liver disease. Hepatology 2018; 67:858–872.
19. Awai HI, Newton KP, Sirlin CB, et al. Evidence and recommendations for imaging liver fat in children, based on systematic review. Clin Gastroenterol Hepatol 2014; 12:765–773.
20. Ko JS. New perspectives in pediatric nonalcoholic fatty liver disease: epidemiology, genetics, diagnosis and natural history. Pediatr Gastroenterol Hepatol Nutr 2019; 22:501–510.
Keywords:

inflammatory bowel disease; magnetic resonance enterography; nonalcoholic fatty liver disease; proton density fat fraction

Supplemental Digital Content

Copyright © 2020 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition