Inflammatory bowel disease (IBD) is in fact a range of diseases, which includes Crohn disease (CD), ulcerative colitis (UC), and indeterminate colitis (IC) (1). The extraintestinal manifestations of IBD include pathology in the liver, the biliary tree, and the pancreas. This pathology consists of nonspecific irregularities of liver function tests (LFTs) (2), fatty liver (3), cholelithiasis (3), chronic pancreatitis (4), autoimmune hepatitis (AIH) (5), and cholangiocarcinoma, probably due to the coexistence of underlying primary sclerosing cholangitis (PSC) (3).
PSC is the most adequately described extraintestinal manifestation of IBD referring to the above organs, with a relatively high prevalence reported (∼4.5% of patients with IBD, ∼5% of patients with UC, ∼3.6% of patients with CD) (6,7). PSC is a chronic inflammatory disease of intrahepatic (IHD) and extrahepatic ducts (EHD). It is of an unknown cause, characterized by occlusive fibrosis resulting in the emergence of cholestasis that leads to biliary cirrhosis and liver failure (8). It is noteworthy that in patients with PSC, the pancreatic ducts are occasionally affected as well (0%–77%) (9). The prognosis of PSC is considered dramatic because no medication has been effective (8), and because PSC is often (∼20%) complicated by the manifestation of incurable and lethal cholangiocarcinoma (8). Liver transplantation is considered to be the only treatment in the final stage of PSC (10). In adults, the mean time from PSC diagnosis to death or transplantation is estimated to be only 8 years (8).
Notably, pediatric PSC presents with peculiarities and has been more sparsely described than adult PSC. Prognosis is moderate; up to one-third of children shall need liver transplantation by the early years of adult life (8). Timely diagnosis of PSC is essential to achieve regular follow-up and treatment of patients, as well as to queue for a transplant in time.
Magnetic resonance cholangiopancreatography (MRCP) is a noninterventional method comparable with criterion standard endoscopic retrograde cholangiopancreatography (ERCP) in diagnosing PSC (11). Patients undergo MRCP without the need of either sedation/anesthesia (if cooperative) or exposure to ionizing radiation or contrast media (8,12).
Using MRCP as a diagnostic tool, the aims of the present study were to estimate the frequency of PSC-type lesions in pediatric patients with IBD, and to investigate the association among demographic data, simple laboratory tests, and magnetic resonance enterography (MRE) findings on one hand and the presence of such lesions in these patients on the other hand.
A total of 73 children (43 girls and 30 boys; age range 7–17 years at the time of visiting our department, median age 12 years) were included in the study. They were referred to us by our associated pediatric gastroenterology clinic between January 2006 and September 2010; all of the children underwent MRCP, as well as MRE in the majority of cases (66/73 or 90.4%). All of the study subjects were given a histologically proven initial IBD diagnosis on the ground of compatible IBD clinical and laboratory profile (CD: 49/73 or 67.1%; IC: 19/73 or 26.0%; UC: 5/73 or 6.8%), almost always before (71/73 or 97.3%) or—in 2 cases (2.7%)—after the magnetic resonance (MR) imaging. The general policy of our pediatricians was to refer patients independently of clinical or laboratory data, ordering MRCP usually as a complement to requested assessment of bowel disease by MRE.
At the time of the initial diagnosis of IBD and MRCP, specific laboratory data—taken under standardized conditions—were recorded for each patient regarding blood serum levels of aspartate transaminase (AST), alanine transaminase (ALT), alkaline phosphatase (ALP), γ-glutamyl transferase (γ-GT), direct bilirubin (DBil), total bilirubin (TBil), C-reactive protein (CRP), and the value of erythrocyte sedimentation rate (ESR). The value of the above variables was considered abnormal when exceeding the upper normal limit (UNL) set by our associated laboratory department. Additionally, we recorded the child's age at the date of initial diagnosis of IBD, its age at the date of MRCP (DMRCP), the time interval between the date of initial diagnosis of IBD and the DMRCP, and the priorly established diagnosis which was indicative of proceeding to MRCP.
Although fulfilling the inclusion criteria (ie, available MRCP plus an IBD histologically established diagnosis), 2 additional patients were excluded from the study because their medical history at DMRCP indicated causes of possible development of secondary sclerosing cholangitis (cystic fibrosis, previous biliary surgery, respectively) (13). Furthermore, 2 of our 73 patients were excluded from statistical analysis concerning the bilirubin blood serum levels, after having been diagnosed as having Gilbert syndrome (14).
All of the parents provided informed consent before their child's inclusion in the study. The study protocol was approved by the ethics committee of our hospital.
MR Imaging Protocol
Nil by mouth was ordered to all of the patients on the day of the MR imaging. Cleansing preparation of the small intestine was saved for patients who underwent MRE in addition to MRCP (and its accompanying tissues sequences); this consisted of low-residue diet from 3 days before the examination (ample fluids, no dairy products).
MRCP was undertaken after placing the children in the supine position, without the need of administering any drug or contrast means.
MR examinations were performed on a 1.5-T MR scanner (Intera release 11, Philips, Best, The Netherlands) with a phased-array coil. In all of the pulse sequences, the field-of-view ranged from 30 to 40 cm depending on the body type of the child.
As far as the MRCP and its survey tissues sequences are concerned, the following pulse sequences were used: Half Fourier Acquisition Turbo Spin Echo (HASTE)/TE80 (coronal plane, repetition time [TR] = 750 ms, echo time [TE] = 80 ms, slice thickness [ST] = 7 mm); Half Fourier Acquisition Turbo Spin Echo (HASTE)/TE80 (coronal plane, TR = 750 ms, TE = 80 ms, ST = 7 mm); Half Fourier Acquisition Turbo Spin Echo (HASTE) with fat suppression (axial plane, TR = 810 ms, TE = 80 ms, ST = 8 mm); 3D T1 spoiled gradient echo, with fat suppression (axial plane, TR = 3.7 ms, TE = 1.8 ms, ST = 5 mm); double echo spoiled gradient echo (DUAL/FFE) (axial plane, TR = 200 ms, TE = 2.3 ms, ST = 7 mm); diffusion weighted images (HASTE/BB) (axial plane, TR = 3470 ms, TE = 60 ms, ST = 7 mm); sDW_HASTE_b600 (axial plane, TR = 1410 ms, TE = 60 ms, ST = 8 mm); steady-state gradient echo (B-FFE) (coronal plane, TR = 3.6 ms, TE = 1.8 ms, ST = 4 mm); steady-state gradient echo (B-FFE) (axial plane, TR = 3.2 ms, TE = 1.6 ms, ST = 7 mm); rapid acquisition relaxation enhancement (RARE) (radial, TR = 8000 ms, TE = 800 ms, ST = 25 mm); rapid acquisition relaxation enhancement (RARE) (coronal plane, TR = 10600 ms, TE = 400 ms, ST = 4 mm). The total examination time for MRCP plus its accompanying tissues sequences was about 15 minutes. If MRE was to follow, this was undertaken according to the local protocol, which included a dynamic study with the use of contrast medium.
All of the images were assessed by an experienced radiologist, who was unaware of any sort of data included in the patients’ medical history. An interactive workstation was used for the imaging evaluation.
For the purposes of the present study, sole MRCP was regarded as the criterion standard method to detect PSC-type lesions, using the Ponsioen classification for PSC (modified for MRCP studies) (15) as a radiological diagnostic guide. On the basis of this, we recorded whether the IHD and/or the EHD were affected in each patient and then we divided our patients into 2 groups: the PSC group (1 ≤ Ponsioen combined score ≤ 5) and the non-PSC group (Ponsioen combined score = 0). Furthermore, for the whole study population, we recorded whether imaging of pancreatic ducts was indicative of pancreatitis (especially type II autoimmune pancreatitis).
Survey Tissues Sequences Accompanying MRCP
Our standard MRCP protocol included gross estimation of the upper abdomen with basic tissue sequences, without any significant additional examination time. We qualitatively recorded pathological findings per anatomical structure: liver along with gallbladder and biliary tract wall, pancreas, spleen; such findings included hepatomegaly, splenomegaly, parenchymal MR signal disorders, and so on. We then coded these findings as indicative of nonspecific pathology located to the structures mentioned above. Portal lymphadenopathy and fluid accumulations were recorded, as well.
We recorded whether each anatomical segment of the bowel (stomach, duodenum, jejunum, ileum, cecum, ascending colon, transverse colon, descending colon, sigmoid colon, rectum) presented with more or less typical features of IBD (eg, bowel wall edema, ulcerations, bowel stenoses) and, if yes, whether the imaging characteristics (ie, especially the MRE dynamic studies in the arterial, portal venous, and equilibrium phase) were indicative of high or low disease activity (16). We further defined the colon disease extent as proctitis/proctosigmoiditis, left-sided colitis (up to the splenic flexure), substantial colitis/pancolitis (above the splenic flexure), or regional colitis; rectal spearing was also documented. Moreover, the presence of local lymphadenopathy, fluid accumulations, and fibrofatty proliferation between and surrounding the intestinal loops was recorded.
Frequencies and percentages were used for the description of qualitative variables, whereas description of quantitative variables was based on the median and the first and third quartile. Associations between the various demographic, laboratory, and MRE/MR survey tissues sequences data and the MRCP-based diagnosis were evaluated by applying the Fisher exact test. The criterion of statistical significance was P < 0.05. All of the statistical analyses were performed using the statistical package STATA 9 (StataCorp LP, College Station, TX).
In the majority of cases (65/73 or 89%), sole IBD (CD: 42/73 or 57.5%; IC: 16/73 or 21.9%; UC: 7/73 or 9.6%) was the established diagnosis, which served as an indication of proceeding to MRCP, due to its association with PSC. In few cases (6/73 or 8.2%), IBD diagnosis was accompanied by other pathologies (ie, UC plus AIH: 1/73 or 1.4%; IC plus Hashimoto disease: 2/73 or 2.7%; CD plus idiopathic thrombocytopenic purpura: 2/73 or 2.7%; CD plus celiac disease: 1/73 or 1.4%), all of which are also related to potential pancreaticobiliary tract lesions (8,13). Finally, in 2 cases (2.7%), no certain pathology was yet proven before the MRCP (which, interestingly, revealed PSC-type lesions), and the referral of these subjects was apparently based on indicative clinical and laboratory data. Taking into consideration these 2 patients who were given a CD diagnosis within 1 month from the time of MRCP, 47 children were considered as CD cases, 18 children as IC cases, and 8 children as UC cases (CD:IC:UC = 64.4%:24.7%:11%) around the time of MRCP. The rest demographic data of the 73 patients are demonstrated in Table 1.
At initial diagnosis of IBD, the median values of the investigated laboratory variables for the whole study group were recorded as follows: AST: 22 IU/dL (interquartile range [IQR]: 17–27 IU/dL, UNL: 60 IU/dL); ALT: 16 IU/dL (IQR: 12–23 IU/dL, UNL: 45 IU/dL); ALP: 176 IU/dL (IQR: 133–246 IU/dL, UNL: 240 IU/dL); γ-GT: 11 IU/dL (IQR: 9–15.5 IU/dL, UNL: 50 IU/dL); DBil: 0.11 mg/dL (IQR: 0.09–0.15 mg/dL, UNL: 0.3 mg/dL); TBil: 0.4 mg/dL (IQR: 0.3–0.4 mg/dL, UNL: 1 mg/dL); CRP: 7.06 mg/L (IQR: 1.5–18 mg/L, UNL: 10 mg/L); ESR: 45 mm/hour (IQR: 29–58 mm/hour; UNL: prior to puberty—13 mm/hour, after puberty—girls: 20 mm/hour, boys: 15 mm/h). At the time of MRCP, the respective median values were recorded as follows: AST: 19 IU/dL (IQR: 17–30 IU/dL); ALT: 19 IU/dL (IQR: 13–31 IU/dL); ALP: 166 IU/dL (IQR: 107–239 IU/dL); γ-GT: 13 IU/dL (IQR: 10–21 IU/dL); DBil: 0.11 mg/dL (IQR: 0.09–0.16 mg/dL); TBil: 0.4 mg/dL (IQR: 0.3–0.5 mg/dL); CRP: 3.49 mg/L (IQR: 1.3–9.13 mg/L); ESR: 40 mm/hour (IQR: 25–65 mm/hour).
Among patients with the respective laboratory abnormalities, the recorded median values at initial diagnosis of IBD were as follows: AST: 118.5 IU/dL (IQR: 84–141 IU/dL); ALT: 122 IU/dL (IQR: 69–184 IU/dL); ALP: 273 IU/dL (IQR: 250–377 IU/dL); γ-GT: 162 IU/dL (IQR: 82–477 IU/dL); DBil: 0.31 mg/dL (IQR: 0.31–0.38 mg/dL); TBil: 1.3 mg/dL (IQR: 1.25–1.3 mg/dL); CRP: 29.1 mg/L (IQR: 11.7–62.25 mg/L); ESR: 45 mm/hour (IQR: 32–61 mm/hour). At the time of MRCP, the respective median values were recorded as follows: AST: 157.5 IU/dL (IQR: 114–187 IU/dL); ALT: 62.5 IU/dL (IQR: 48–165.5 IU/dL); ALP: 312 IU/dL (IQR: 277–348 IU/dL); γ-GT: 98.5 IU/dL (IQR: 59–261 IU/dL); DBil: 0.35 mg/dL (IQR: 0.33–0.46 mg/dL); TBil: 1.15 mg/dL (IQR: 1–1.43 mg/dL); CRP: 12.45 mg/L (IQR: 11.18–17.75 mg/L); ESR: 42.5 mm/hour (IQR: 30–67 mm/hour). Furthermore, Table 2 shows the differences of LFTs’ abnormality proportions between the PSC and the non-PSC group.
A total of 11 children (6 boys; median age at DMRCP: 14 years; IQR: 11–16 years) presented with PSC-type lesions at MRCP (15.1%, 95% CI 7.8%–25.4%) (Fig. 1). Around the time of cholangiography, 6 of them were considered to be CD cases, including the 2 subjects with PSC whose MRCP preceded their CD diagnosis (6/47 or 12.8%); 3 children were diagnosed as having IC (3/18, or 16.7%) and the remaining 2 with UC (2/8 or 25%). Furthermore, the prevalence of PSC-type lesions by IBD initial diagnosis subtype somewhat deviated (CD: 7/49 or 14.3%; IC: 3/19 or 15.8%; UC: 1/5 or 20%). The IHD were solely involved in 3 patients (27.3%), the EHD were solely involved in another 3 (27.3%), whereas the IHD and the EHD were both affected in the remaining 5 patients (45.5%). The Ponsioen classification was applied to the PSC group (data not shown). No ductal signs of pancreatitis were detected in all of the 73 children.
On survey tissues sequences, liver along with gallbladder and biliary tract wall was abnormally imaged in 8 cases (3 of whom belong to the PSC group), spleen in 3 cases (none of whom belongs to the PSC group), and pancreas in none of the cases. Portal lymphadenopathy was detected in 2 cases (one of whom belongs to the PSC group), whereas pleural fluid accumulation was incidentally reported in 1 non-PSC case.
On MRE, small intestine pathology was seen in 36 cases (54.5%); bowel disease activity was high in 24 cases (66.7%) and low in 12 cases (33.3%). Abnormal imaging findings were present in the large intestine in 37 cases (56.1%); the proportions of patterns regarding the colonic involvement extent were 33.3% (22/66), 10.6% (7/66), 7.6% (5/66), and 4.5% (3/66) for regional colitis, substantial colitis/pancolitis, left-sided colitis, and proctosigmoiditis, respectively. Rectal sparing was recognized in 7 of 66 cases (10.6%). Large bowel disease activity was high in 25 cases (67.6%) and low in 12 cases (32.4%) of colon pathology. Among the different segments of the gastrointestinal tract, ileum was affected more often (34/66 = 51.5%), whereas duodenum pathology was not detected in any case (data not shown). Both high and low bowel disease activities were more frequently reported for ileum: 23/66 = 34.8% and 11/66 = 16.7%, respectively (data not shown). Ascites was reported in 9 cases (1 of whom belongs to the PSC group) or 13.6%, local lymphadenopathy in 33 cases (50.0%), and fibrofatty proliferation in 19 cases (28.8%).
It is noteworthy that the data concerning demographic features, CRP, ESR, MRE, and MR survey tissues sequences findings were found statistically unrelated to either study group (except a shown trend toward a more frequent imaging of nonspecific pathology concerning the liver in the PSC group, P = 0.094).
In our study, PSC-type lesions were detected by MRCP in 15.1% of patients with IBD. Whereas the PSC prevalence has been estimated in numerous reports among adult patients with IBD (∼4.5%) (6), only few studies have focused on pediatric populations (1.5%–3%) (17–20). The differences in definition of PSC result in prevalence variability in the respective studies (6). Moreover, a study of large pediatric population revealed a PSC prevalence of 0.5% before or at the time of IBD diagnosis (1% among patients with UC and IC, respectively; none among patients with CD) (21). The main reason for the rather high MRCP-based PSC prevalence reported in this article is that subjects were deliberately unselected in terms of other sort of data—including LFTs's abnormalities (19) and results provided by ERCP, liver biopsy, immunological tests, and so on—to reduce selection bias; however, we should take into account that our associated pediatric gastroenterology clinic is a tertiary national center, a fact that may have caused such bias. Furthermore, regarding the PSC group, 2 patients were diagnosed as having IBD after MRCP, implying a referral bias, although minimized by assigning MRCP diagnosis to a blinded radiologist. In addition, underlying hepatic cirrhosis could lead to false-positive PSC cases at MRCP (22), although the present study was not designed to exclude such possibility and no supportive clinical or other evidence of cirrhosis was present in our patients. Interestingly, because MRCP is characterized by reduced spatial resolution (23), possible underreporting of PSC may have occurred in our study.
The distribution of biliary lesions was IHD:(IHD + EHD):EHD = 27%:46%:27%. Previous studies (24–28) revealed a respective type of (16%–53%):(31%–84%):(0%–16%), which suggests that sclerosing alterations initially present at the IHD and then spread to the EHD (24). Due to the reduced spatial resolution of MRCP, we may have overreported EHD against (IHD + EHD). Furthermore, imaging of pancreatic ducts was normal in all of the patients, which is in line with the 0.9% prevalence of pancreatitis previously reported in pediatric IBD (20).
The study of demographic data revealed no significant differences, probably due to our relatively small study size, as well as questions concerning methodology (eg, definition of PSC and its onset, determination of the actual—subclinical onset of IBD) (6). In pediatric PSC, bibliography favors a male preponderance (1.7:1) along with a preteen age at PSC onset (8,29). Furthermore, age at initial diagnosis of UC is suggested to be unrelated to PSC development, at least in adults (30). What is more, PSC is considered a surrogate of relatively long-standing colitis (6).
As far as subtype of IBD is concerned, UC is more prevalent than CD among pediatric patients with IBD-PSC (73% vs 16%) (29), whereas children with UC are 3 times more likely to develop PSC than those with CD (20). Nevertheless, our results were of no statistical significance; perhaps the preponderance of CD over other subtypes of IBD in children (31), combined with the small size of our PSC group, may have resulted in a greater prevalence of CD than expected in this group. Moreover, due to the fact that IBD diagnosis often fluctuates in pediatric patients (32), the proportions of each IBD subtype may be eventually altered toward a distribution favored by present bibliography. Furthermore, a reason for overreporting CD (and IC) against UC in patients with IBD-PSC is colonic disease with rectal sparing, a feature that interestingly characterizes this syndrome (33,34); this may be the case for at least 2 of our patients with PSC.
The study of MRE features revealed no significant results. This is in contrast to other work suggesting an extensive and mild colitis—often characterized by rectal sparing—accompanying PSC, with or without small intestine involvement (7,33,34). In past studies, other, perhaps more appropriate, means (eg, intestinal endoscopy and histology) have been used to evaluate the extent and activity of bowel disease. Although MRE is not considered optimal in examining the large intestine, it is a revolutionary approach for the assessment of the small intestine.
According to our study design, the laboratory data were recorded at 2 stages: the initial diagnosis of IBD and the time of performing MRCP. At these 2 stages, the abnormality prevalences of 3 variables (AST, ALT, γ-GT) were significantly higher in the PSC group. Less consistent results were documented for bilirubin and ALP, especially at initial diagnosis of IBD. Despite the above-shown relations, the LFTs's abnormality prevalences were only low to moderate (14.3%–50%) in the PSC group, even at the time of MRCP; this is mostly in contrast to present view (29). Thus, cholangiographic detection of PSC-type lesions is possible, even if the pattern of LFTs is hardly indicative of this probability. Fortunately, the LFTs's abnormality prevalences in the non-PSC group were bearable (1.7–28%); such abnormalities may be attributed to flare-ups of bowel disease (6,35).
Based on our results, γ-GT was proven a more reliable “cholestatic” marker than ALP in this age group, as supported by previous work (25,28). ALP is indeed characterized by a wide range of normal values, owing to elevated bone ALP isoenzyme during growth (25). Interestingly, compared with other LFTs, ALP demonstrated the highest abnormality prevalence in the non-PSC group (21.3%–28%), implying reduced specificity for hepatobiliary injury. Furthermore, our results referring to liver transaminases were surprisingly similar to those referring to γ-GT; this may be partly attributed to coexistence of AIH features in at least 2 of our patients with PSC because AIH/PSC overlap syndrome is rather prevalent in this age group (∼30%) (25,28).
Undoubtedly, the clinical and laboratory profile of a number of our PSC patients does not raise high suspicions of this still inhomogeneously defined syndrome. Having in mind that we provide care to children, invasive methods such as liver biopsy and ERCP have been carefully reserved for cases of strong clinical indication. In fact, ERCP has yet to be applied in any of our 73 patients. Hepatic biopsy has confirmed the MRCP-based PSC diagnosis in 1 case and modified an MRCP-based PSC diagnosis to an AIH/PSC overlap syndrome diagnosis in another case. Furthermore, the sole priorly, histologically proven subject with IBD/AIH was given a second liver biopsy after his MRCP-based PSC diagnosis; this confirmed the AIH/PSC overlap syndrome diagnosis. It should be noted that a follow-up MRCP was given to 1 patient from the PSC group, showing no substantive changes of radiological features. Moreover, ursodiol—which is the medication most widely used for PSC, although controversial (36,37)—has been applied only to 4 convincing subjects with PSC (including the 3 cases that underwent a liver biopsy); it has been well tolerated as well as efficient in correcting abnormal biochemical indices. Ordinarily treating the accompanying IBD has been our policy in the remaining cases. Finally, none of our PSC patients has undergone biliary surgery or drainage as a treatment.
As we have already implied, the cross-sectional character and the relatively small population size have been the main drawbacks of the present study. Moreover, the simple concept and low-cost study design, as well as the suggested extremely high specificity (∼100%) of MRCP for the diagnosis of PSC in children (23), discouraged us from using liver-healthy controls, although this would have provided greater statistical power to our results. Furthermore, we should notice that local policy denied IBD treatment before the histological establishment of such diagnosis, thus reducing the bias of the data at initial IBD diagnosis; however, this was not the case for laboratory variables and imaging features (eg, bowel inactivity, hepatomegaly) regarding the time of the subsequent MRCP. Due to relatively small study population, we could not reach exact conclusions for the potential role of drugs (along with total parenteral nutrition) (19,38,39) as a confounding factor. Nevertheless, because at time of raising the question for an MRCP referral, most of these patients have already been exposed to IBD treatment, checking for the possibly confounding role of such treatment may be of limited interest.
Prompt diagnosis of PSC benefits patients with IBD-PSC in terms of scheduling surveillance for colorectal neoplasia, especially in adolescents, because the risk of it increases with the presence of PSC (33,36); and timely queuing for possible liver transplantation (36). It should be mentioned that relative risk measures could help the attending pediatricians of patients with IBD to decide necessity and set priority for MRCP referral, on the basis of using easily applicable demographic, clinical, and laboratory data. Although accurate estimation of such measures was not feasible here due to small sample size, our study could trigger further research on larger study populations upon this aspect.
The authors thank Giorgos Bakoyannis for contribution to the statistical analysis and Stavros Xenophontos and Constantinos Tryfonos for editorial assistance.
1. Geboes K, Colombel JF, Greenstein A, et al. Crohn's disease, ulcerative colitis or indeterminate colitis—how important is it to differentiate? Acta Gastroenterol Belg
2. Mendes FD, Levy C, Enders FB, et al. Abnormal hepatic biochemistries in patients with inflammatory bowel disease. Am J Gastroenterol
3. Memon MI, Memon B, Memon MA. Hepatobiliary manifestations of inflammatory bowel disease. HPB Surg
4. Heikius B, Niemelä S, Lehtola J, et al. Pancreatic duct abnormalities and pancreatic function in patients with chronic inflammatory bowel disease. Scand J Gastroenterol
5. Floreani A, Rozzotto ER, Ferrara F, et al. Clinical course and outcome of autoimmune hepatitis/primary sclerosing cholangitis overlap syndrome. Am J Gastroenterol
6. Loftus EV, Sandborn WJ, Lindor KD, et al. Interactions between chronic liver disease and inflammatory bowel disease. Inflamm Bowel Dis
7. Saich R, Chapman R. Primary sclerosing cholangitis, autoimmune hepatitis and overlap syndromes in inflammatory bowel disease. World J Gastroenterol
8. LaRusso NF, Shneider BL, Black D, et al. Primary sclerosing cholangitis: summary of a workshop. Hepatology
9. Heikius B, Niemelä S, Lehtola J, et al. Hepatobiliary and coexisting pancreatic duct abnormalities in patients with inflammatory bowel disease. Scand J Gastroenterol
10. Brandsaeter B, Isoniemi H, Broome U, et al. Liver transplantation for primary sclerosing cholangitis; predictors and consequences of hepatobiliary malignancy. J Hepatol
11. Berstad AE, Aabakken L, Smith HJ, et al. Diagnostic accuracy of magnetic resonance and endoscopic retrograde cholangiography in primary sclerosing cholangitis. Clin Gastroenterol Hepatol
12. Vitellas KM, Keogan MT, Spritzer CE, et al. MR cholangiopancreatography of bile and pancreatic duct abnormalities with emphasis on the single-shot fast spin-echo technique. Radiographics
13. Abdalian R, Heathcote EJ. Sclerosing cholangitis: a focus on secondary causes. Hepatology
14. Berk PD, Jones EA, Howe RB, et al. Disorders of bilirubin metabolism. In: Bondy PK, Rosenberg LE, eds. Metabolic Control and Disease
, 8th ed. Philadelphia: Saunders; 1980:1009.
15. Petrovic BD, Nikolaidis P, Hammond NA, et al. Correlation between findings on MRCP and gadolinium-enhanced MR of the liver and a survival model for primary sclerosing cholangitis. Dig Dis Sci
16. Gourtsoyiannis N, Papanikolaou N, Grammatikakis J, et al. MR eneteroclysis: technical considerations and clinical applications. Eur Radiol
17. Nemeth A, Ejderhamm J, Glaumann H, et al. Liver damage in juvenile inflammatory bowel disease. Liver
18. Ong JC, O’Loughlin EV, Kamath KR, et al. Sclerosing cholangitis in children with inflammatory bowel disease. Aust N Z J Med
19. Hyams JS, Markowitz J, Treem W, et al. Characterization of hepatic abnormalities in children with inflammatory bowel disease. Inflamm Bowel Dis
20. Dotson JL, Hyams JS, Markowitz J, et al. Extraintestinal manifestations of pediatric inflammatory bowel disease and their relation to disease type and severity. J Pediatr Gastroenterol Nutr
21. Timmer A, Behrens R, Buderus S, et al. Childhood onset inflammatory bowel disease: predictors of delayed diagnosis from the CEDATA German-language pediatric inflammatory bowel disease registry. J Pediatr
22. Lewin M, Vilgrain V, Ozenne V, et al. Prevalence of sclerosing cholangitis in adults with autoimmune hepatitis: a prospective magnetic resonance imaging and histological study. Hepatology
23. Ferrara C, Valeri G, Salvolini L, et al. Magnetic resonance cholangiopancreatography in primary sclerosing cholangitis in children. Pediatr Radiol
24. Gregorio GV, Portmann B, Karani J, et al. Autoimmune hepatitis/sclerosing cholangitis overlap syndrome in childhood: a 16 year prospective study. Hepatology
25. Feldstein AE, Perrault J, El-Youssif M, et al. Primary sclerosing cholangitis in children: a long-term follow-up study. Hepatology
26. Batres LA, Russo P, Mathews M, et al. Primary sclerosing cholangitis in children: a histologic follow-up study. Pediatr Dev Pathol
27. El-Shabrawi M, Wilkinson ML, Portmann B, et al. Primary sclerosing cholangitis in childhood. Gastroenterology
28. Wilschanski M, Chait P, Wade JA, et al. Primary sclerosing cholangitis in 32 children: clinical, laboratory, and radiographic features, with survival analysis. Hepatology
29. Erickson NI, Balistreri WF. In: Suchy FJ, Sokol RJ, Balistreri WF, eds. Liver Disease in Children
, 3rd ed. Cambridge, UK: Cambridge University Press; 2007:459–77.
30. Olsson R, Danielsson Å, Järnerot G, et al. Prevalence of primary sclerosing cholangitis in patients with ulcerative colitis. Gastroenterology
31. Koletzko S. Epidemiology in pediatric inflammatory bowel disease. In: Walker-Smith JA, Lebenthal E, Branski D, eds. Pediatric Inflammatory Bowel Disease: Perspective and Consequences
. Vol. 14. Basel: Karger; 2009:19–28.
32. Heyman MB, Kirschner BS, Gold BD, et al. Children with early-onset inflammatory bowel disease (IBD): analysis of a pediatric IBD consortium registry. J Pediatr
33. Faubion WA, Loftus EV, Sandborn WJ, et al. Pediatric “PSC-IBD”: A descriptive report of associated inflammatory bowel disease among pediatric patients with PSC. J Pediatr Gastroenterol Nutr
34. Loftus EV, Harewood GC, Loftus CG, et al. PSC-IBD: a unique form of inflammatory bowel disease associated with primary sclerosing cholangitis. Gut
35. Broomé U, Glaumann H, Hellers G, et al. Liver disease in ulcerative colitis: an epidemiological and follow up study in the county of Stockholm. Gut
36. Chapman R, Fevery J, Kalloo A, et al. Diagnosis and management of primary sclerosing cholangitis. Hepatology
37. Eaton JE, Silveira MG, Pardi DS, et al. High-dose ursodeoxycholic acid is associated with the development of colorectal neoplasia in patients with ulcerative colitis and primary sclerosing cholangitis. Am J Gastroenterol
38. Gisbert JP, Luna M, González-Lama Y, et al. Liver injury in inflammatory bowel disease: Long-term follow-up study of 786 patients. Inflamm Bowel Dis
39. Bengoa JM, Hanauer SB, Sitrin MD, et al. Pattern and prognosis of liver function test abnormalities during parenteral nutrition in inflammatory bowel disease. Hepatology