Outcome of Very Early Onset Inflammatory Bowel Disease Associated With Primary Sclerosing Cholangitis: A Multicenter Study From the Pediatric IBD Porto Group of ESPGHAN

Abstract

Background

Whether primary sclerosing cholangitis related to inflammatory bowel disease (PSC-IBD) diagnosed before 6 years (ie, VEO-IBD) has a distinct phenotype and disease course is uninvestigated. We aimed to analyze the characteristics and natural history of VEO-PSC-IBD, compared with early and adolescent-onset PSC-IBD.

Methods

This is a multicenter, retrospective, case-control study from 15 centers affiliated with the Porto and Interest IBD group of ESPGHAN. Demographic, clinical, laboratory, endoscopic, and imaging data were collected at baseline and every 6 months. Inflammatory bowel disease–related (clinical remission, need for systemic steroids and biologics, and surgery) and PSC-related (biliary and portal hypertensive complications, need for treatment escalation and liver transplantation, cholangiocarcinoma, or death) outcomes were compared between the 2 groups.

Results

Sixty-nine children were included, with a median follow-up of 3.63 years (interquartile range, 1-11): 28 with VEO-PSC-IBD (23 UC [82%], 2 IBD-U [7%] and 3 [11%] CD), and 41 with PSC-IBD (37 UC [90%], 3 IBDU [7.5%] and 1 [2.5%] CD). Most patients with UC presented with pancolitis (92% in VEO-PSC-UC vs 85% in PSC-UC, P = .2). A higher number of patients with VEO-PSC-IBD were diagnosed with PSC/autoimmune hepatitis overlap syndrome than older children (24 [92%] vs 27 [67.5%] PSC-IBD, P = .03), whereas no other differences were found for PSC-related variables. Time to biliary strictures and infective cholangitis was lower in the VEO-PSC-IBD group (P = .01 and P = .04, respectively), while no difference was found for other outcomes. No cases of cholangiocarcinoma were reported.

Conclusions

Primary sclerosing cholangitis related to inflammatory bowel disease has similar baseline characteristics whether diagnosed as VEO-IBD or thereafter. A milder disease course in terms of biliary complications characterizes VEO-PSC-IBD.

Background

Hepatobiliary diseases are among the most frequent extraintestinal manifestations of inflammatory bowel diseases (IBD).^1,2 The most common of which (2%-5% of prevalence) is primary sclerosing cholangitis (PSC),^3,4 a chronic liver disease characterized by an idiopathic inflammatory process leading to progressive multifocal biliary fibrotic strictures.^5,6 In its form of PSC/autoimmune hepatitis (AIH) overlap syndrome, the diagnosis is based on features overlapping autoimmune hepatitis and PSC.^7–9 Such extraintestinal manifestations are associated with ulcerative colitis (UC) in 80% to 90% of cases.^3,10 Both adult and pediatric studies have confirmed a distinct phenotype of patients with IBD and PSC, namely more extensive involvement, relative rectal sparing, and backwash ileitis, as well as a milder luminal disease course that is not concordant with endoscopic disease activity. For this reason, such presentation is referred to as “PSC-IBD phenotype” by some authors.^10,11 Nevertheless, the natural history of pediatric PSC-IBD is poorly described.

Likewise, children diagnosed with IBD before 6 years of age (ie, very early onset IBD [VEO-IBD]) represent a specific group within the IBD spectrum, both in terms of disease location and natural history, with a higher prevalence of monogenetic etiologies than later-onset disease.¹² Additionally, VEO-IBD children with monogenic disorders present with a severe phenotype and disease course, while more favorable outcomes for patients with nonmonogenic VEO-IBD have been suggested.¹³ Whether a combination of VEO-IBD and VEO-PSC might account for specific clinical features and disease course remains uninvestigated. Therefore, we aimed to analyze the natural history and characteristics of VEO-PSC-IBD, comparing them to pediatric later-onset PSC-IBD.

Materials and Methods

Study Design and Participants

This is a retrospective, longitudinal, multicenter, case-control study performed in 15 centers affiliated with the Pediatric Porto IBD interest Group of ESPGHAN. Each participating center was asked to recruit 2 groups of patients diagnosed after the year 2010: (1) children with both IBD and PSC diagnosed before the age of 6 years (VEO-PSC-IBD) and (2) patients with both IBD and PSC diagnosed after the age of 6 and before the age of 18 years (PSC-IBD). The centers were asked to collect data of patients with VEO-PSC-IBD and PSC-IBD in a 1:2 ratio, although this was not mandatory. Patients with other associated enteropathies and genetic or metabolic liver diseases were excluded.

Data Collection

Data were collected from baseline, defined as the date of diagnosis of the first disease, either IBD or PSC, and every 6 months (±2 months) thereafter, until the data retrieval date (October 1, 2022). Data were recorded anonymously on a case report form, managed by the coordinating center Umberto I Hospital, Sapienza University of Rome, Italy.

The diagnosis of IBD was based on the commonly recognized clinical, endoscopic, and histologic findings.¹⁴ Information collected at baseline included age, gender, family history of IBD and liver disease, and anthropometric data (height and weight).

Information specific to IBD included date of diagnosis, type of IBD, disease location and behavior (according to the Paris classification¹⁵), other nonhepatic extraintestinal manifestations, and treatment received at diagnosis. Endoscopic disease activity, assessed through the Mayo score for UC^16,17 and the Simple Endoscopic Score for CD (SES-CD),¹⁸ was recorded at diagnosis and, when available, during follow-up. Clinical disease activity was defined according to the wPCDAI for CD and the PUCAI for UC and IBDU.^19,20

According to the European Association for the Study of the Liver,⁷ the diagnosis of PSC was based on magnetic resonance cholangiopancreatography (MRCP) and/or liver biopsy. The term “primary” was used to indicate the absence of specific causes of sclerosing cholangitis.

As for PSC, the following disease-specific information was collected: age at diagnosis, type of PSC, degree of liver fibrosis, risk classification, histological features (acute or chronic cholangitis associated with portal tract fibrosis, portal hepatitis, or cirrhosis, as well as concentric periductal fibrosis known as “onion skinning”), and imaging characteristics.

Primary sclerosing cholangitis was classified as “large duct” when duct abnormalities (dilation and/or stenosis) were evident on MRCP or hepatic ultrasound and as “small duct” when the duct abnormalities were evident only at the histology level with normal cholangiography.⁷

The PSC/AIH overlap syndrome was diagnosed if typical histological features of autoimmune hepatitis (AIH; interface hepatitis, lobular inflammation, lymphoplasmacytic infiltrate) were associated with histological and radiological features of sclerosing cholangitis.^9,21 The degree of liver fibrosis was staged with the METAVIR score.²² The SCOPE index was used to classify patients with PSC into low, medium, and high risk.²¹

Laboratory data, collected at baseline and every 6 months during follow-up, included hemoglobin, white blood cells, platelets (PLTs), C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), albumin, total serum proteins, aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma-glutamyl transferase (GGT), total and direct bilirubin, international normalized ratio (INR), serum sodium, total immunoglobulin G (IgG), anti-Saccharomyces Cerevisiae (ASCA), antineutrophil cytoplasmic antibodies (ANCA), antinuclear antibody (ANA), antismooth muscle antibody (ASMA), liver-kidney microsome antibody type 1 (LKM1), and fecal calprotectin (FC).

A FC score >250 mg/g was considered suggestive of active intestinal disease.²³ The normal ranges for CRP and ESR were <6 mg/L and <25 mm/h, respectively.²⁴

Concomitant medications and doses (including exclusive or partial enteral nutrition, corticosteroids [CS], 5-ASA, ursodeoxycholic acid, vancomycin, immunomodulators, biologics [anti-TNF-α agents, vedolizumab, ustekinumab]) were also recorded.

Outcomes

The following outcomes were evaluated:

1. For IBD: clinical relapse (defined as a PUCAI >10 or a wPCDAI >12.5); need for corticosteroids, biologics, and surgery; the occurrence of IBD-related complications (strictures, fistulae, and abscesses) and IBD-related hospitalizations.

2. For PSC: PSC-related complications included cholangitis, portal hypertension complications (defined as esophageal varices, hypersplenism, ascites, and hepatopulmonary syndrome), strictures of the biliary tract, need for dilatations of biliary strictures, need for treatment escalation, need for liver transplantation, and occurrence of cholangiocarcinoma. Treatment escalation was defined as the need for additional medical therapy based on a lack of biochemical response/remission with the use of the previous drug (ursodeoxycholic acid, vancomycin, azathioprine, corticosteroids).

Ethics committee approval

The Human Ethics Committee of Sapienza University of Rome approved the protocol in accordance with the Declaration of Helsinki (998/17). Written informed consent was obtained from patients and parents before enrollment in the study.

Statistical Analysis

Data were summarized and expressed mean ± standard deviation, or medians (interquartile range, IQR), as appropriate for the distribution normality. Categorical data were expressed as frequencies and percentages. Differences between groups were assessed by Student t test. Fisher’s correction χ² test was applied for categorical variables. A P value <0.05 was considered significant. The survival curve analysis was performed using the Kaplan–Meier method and log-rank test. We used time-to-event analysis for all outcomes accounting for the different follow-up periods of the patients. All statistical analyses were performed using IBM SPSS Statistics package (V.23.0 for Windows, IBM Armonk, New York, USA).

Results

Patient Characteristics

A total of 69 children with both IBD and PSC were included: 28 with VEO-PSC-IBD (40.5%) and 41 with PSC-IBD (59.5%). Magnetic resonance cholangiopancreatography was performed in 41 of 69 patients (59%), liver ultrasound in 28 of 69 (40.5%), and liver biopsy in 50 of 69 (72%). Thirty-five (50%) underwent both MRCP and liver biopsy. The median follow-up duration of the entire cohort was 3.63 years (6 years [IQR 1-11] for VEO-PSC-IBD and 3.25 years [IQR 1-9] for PSC-IBD, P = .047).

Ulcerative colitis was prevalent in both groups (23 [82%] in VEO-PSC-IBD and 37 [90%] in PSC-IBD, P = .35), with most patients showing pancolitis (23 [92%] vs 34 [85%] in VEO-PSC-IBD and PSC-IBD, respectively). A lower rate of VEO-PSC-IBD presented with a mild PUCAI than PSC-IBD (20% vs 47.5%, P = .03; Table 1).

A higher number of VEO-PSC-IBD children had a diagnosis of PSC/AIH overlap syndrome compared with PSC-IBD (24 of 28 [92%] vs 27 of 40 [67.5%], P = .03), with a higher SMA positivity rate in the VEO-PSC-IBD group (9 of 22 [40%] vs 1 of 16 [6%], P = .02). No other significant differences were found between the 2 groups for the other relevant PSC-related characteristics (Table 2).

Treatments received at the diagnosis of both IBD and PSC were comparable between the 2 groups (Table 3).

Outcomes Analysis

No differences were found between the 2 groups for time to IBD-related outcomes, including clinical relapse (P = .96), use of biological therapy (P = .55), and courses of systemic steroids (P = .83; Figure 1). During the follow-up study period, only 1 patient with PSC-IBD underwent surgery (1.4%).

Regarding PSC-related outcomes, patients with PSC-IBD had a shorter time to developing biliary stenosis (P = .012), strictures needing dilation (P = .016), and infective cholangitis (P = .04) than VEO-PSC-IBD ones (Figure 2). No differences were found for the time to treatment escalation (P = .09), portal hypertension (P = .24), and liver transplantation (P = .52). At 3-year follow-up, 1 of 28 (3.5%) and 3 of 41 (7.3%) VEO-PSC-IBD and older PSC-IBD children, respectively, underwent liver transplantation (P = .6). No deaths or malignancies were observed during the follow-up. We found no significant impact on the main outcomes based on the timing of PSC diagnosis vs IBD diagnosis (Supplementary Figure 1).

Discussion

Our study is the first to examine the clinical presentation and natural history of very early onset PSC-IBD. Our findings indicate that pediatric PSC-IBD shares similar baseline characteristics regardless of whether both diseases are diagnosed before or after the age of 6 years. Previous studies have highlighted that patients with VEO-IBD have specific disease characteristics, such as extensive disease in UC and colonic involvement in CD, compared with older children.^12,25 In our population, most patients had UC, and pancolitis with mild clinical symptoms was prevalent in the whole population. These findings align with previous studies that have identified a specific PSC-IBD phenotype characterized by pancolitis (with more severe involvement of the right colon), relative rectal sparing, and backwash ileitis with a relatively mild disease course that is commonly not concordant with endoscopic disease activity.^10,26

In our study, we demonstrated a higher occurrence of PSC/AIH overlap syndrome in VEO-PSC-IBD than in older children. Although the literature does not provide a clear definition of this entity, some authors suggest that it may represent an early phase in the PSC continuum.^10,11,27 Previous studies reported that up to one-third of children with PSC exhibit signs of autoimmunity,^27–30 commonly associated with IBD.^31,32 It is noteworthy that adults with PSC/AIH overlap tend to be diagnosed at a younger age than those with more typical manifestations of PSC.³¹ Thus, the higher prevalence of PSC/AIH overlap syndrome in our population of VEO-PSC-IBD population could be related to the younger age of disease onset and may represent an early stage of the disease continuum, characterized by a more inflammatory (rather than fibrotic) disease of the biliary tree.^10,27,33 Surprisingly, we did not observe any significant difference in treatment strategies between VEO and later-onset PSC-IBD at baseline, despite the expected higher use of immunosuppressive therapies based on the frequency of PSC/AIH overlap syndrome.^34,35 Indeed, according to certain pediatric and adult guidelines, there is widespread approval for using immunosuppressive treatments for PSC/AIH overlap syndrome based on a modest improvement in biochemical and parenchymal inflammation reported with these drugs.²¹

In line with previous data,³² we found a mild IBD course in the entire population, without age-related differences. Ricciuto et al³² found milder disease activity in a cohort of 74 children with PSC-UC/IBD compared with UC/IBD controls, with a lower median PUCAI at presentation, reduced rates of acute severe colitis, and milder endoscopic disease activity at IBD onset. During follow-up, these children required fewer steroids and biologics and had milder clinical activity, but there was discordance between clinical scores and endoscopic disease activity, suggesting the presence of subclinical persistent mucosal inflammation in the PSC-IBD phenotype. This may contribute to the increased risk of colorectal cancer in these patients.³² In our study, the lack of regular endoscopies might have underestimated the extent of endoscopic disease activity, and data on this parameter could not be obtained.

Despite an overall mild PSC disease course in terms of major outcomes (liver transplantation and portal hypertension complications), the probability of infective cholangitis and biliary strictures was higher in older children compared with VEO-IBD-PSC. It is interesting to note that the VEO-PSC-IBD children had a longer follow-up time than those with later disease onset, lasting for a median time of 6 years compared with 3.25 years. This could potentially indicate a less severe course of PSC in younger patients.

While the effect of PSC on IBD activity has been previously described, there has been little research on the effect of IBD on PSC natural history, and no PSC-distinct phenotype has been suggested in children with PSC-IBD compared with those with PSC only. Recently, Deneau et al⁶ analyzed the natural history of 781 children from the Pediatric PSC consortium in a 10-year follow-up period. The authors reported a high rate of complications in children with PSC but a more favorable prognosis in children with concomitant IBD. According to our results, analyzing the PSC-IBD phenotype from an age perspective, a very early onset disease might somehow represent a protective factor for the PSC disease course. Theoretically, the higher frequency of PSC/AIH overlap syndrome in younger children could have a role in the milder disease course of VEO-PSC-IBD due to a higher response to immunosuppressive therapies of portal inflammation related to the autoimmune features.²⁷ However, this explanation contradicts the findings of Deneau et al,⁶ who reported similar progression to adverse outcomes between patients with and without features of overlap syndrome.

We acknowledge that our study has some limitations. First, the number of patients enrolled might appear limited in absolute figures. Nevertheless, it is worth noting that VEO-PSC-IBD is a rare disease, and this is the largest population described in the literature so far. Moreover, the study’s retrospective design limited its results, as it did not include a standardized protocol for liver biopsies or imaging, and not all patients underwent a complete diagnostic work-up to diagnose the biliary disease subtypes. However, most patients received a complete laboratory autoimmunity evaluation and an imaging and/or histology evaluation, which allowed for the classification of patients with PSC. Additionally, the completeness of available data decreased during follow-up. Specifically patients with later-onset PSC-IBD had shorter median follow-up than those with VEO-PSC-IBD. For this reason, all the analyses were limited to a 5-year follow-up.

In conclusion, our study sheds light on the characteristics of VEO-IBD with concomitant PSC. Our data indicate that patients with VEO-PSC-IBD commonly present with autoimmune features and experience favorable PSC-related outcomes while sharing similar IBD disease course with older children. These findings suggest the need for future extensive studies with systematic, standardized protocols to explore potential genetic and pathophysiological factors specific to patients with VEO and develop more effective treatments for this vulnerable population.

Supplementary Data

Supplementary data is available at Inflammatory Bowel Diseases online.

Funding

No honorarium, grant, or other form of payment was given to anyone to write and produce the manuscript.

Conflicts of Interest

All the authors have no conflicts of interest related to the article to declare.

References

Author notes