An infographic for this article is available at:https://links.lww.com/MPG/B863.
What Is Known/What Is New
What Is Known
- Primary sclerosing cholangitis and inflammatory bowel disease share pathogenic mechanisms along the gut-Liver axis.
- Vedolizumab is a monoclonal antibody used to treat both ulcerative colitis and Crohn disease that may also be active in the liver.
- Liver and bowel outcomes on vedolizumab for primary sclerosing cholangitis- inflammatory bowel disease in children are unknown
What Is New
- Vedolizumab did not improve hepatobiliary outcomes.
- Inflammatory bowel disease activity improved on vedolizumab at rates similar to non- primary sclerosing cholangitis inflammatory bowel disease patients.
- Bowel disease remission was correlated with stabilization of γ-glutamyltransferase levels.
Primary sclerosing cholangitis (PSC) is a disorder of chronic and progressive hepatobiliary inflammation. Half of children diagnosed with PSC will have disease complications within 10 years, including dominant biliary strictures, recurring cholangitis, cirrhosis, end-stage liver disease, liver transplantation requirement, or cholangiocarcinoma (1). There is currently no treatment proven to delay disease progression.
Most PSC co-occurs with inflammatory bowel disease (IBD) (2,3). PSC and IBD share pathogenic mechanisms along the gut-liver axis. Vedolizumab (VDZ) is a monoclonal antibody used to treat both Crohn disease (CD) and ulcerative colitis (UC) (4–7). VDZ targets the lymphocyte α4β7 integrin and prevents integrin binding to gut endothelial Mucosal Addressin Cellular Adhesion Molecule-1 (MAdCAM-1), blocking lymphocyte trafficking to the gastrointestinal tract (8–10). Although MAdCAM-1 is not expressed in normal liver tissue, it is inducible in the portal tract endothelium of chronically inflamed liver tissue (10,11). Activated T-cells expressing the α4β7 integrin may be responsible for the progressive biliary inflammation seen in PSC because of upregulated MAdCAM-1 in the inflamed biliary tract (11–14). Therefore, VDZ may reduce gut-lymphocyte trafficking and liver inflammation.
VDZ has not shown hepatobiliary benefit in multiple studies in adults with PSC-IBD (8,13,15). Children, however, tend to have an earlier stage of PSC with less fibrosis. This may equate to increased therapeutic benefit from VDZ compared with adults. To date, there is no data on the effect of VDZ in pediatric patients with PSC-IBD. We investigated changes in biochemical, histopathologic, radiographic, and IBD activity on VDZ therapy in children with PSC-IBD using a multicenter, international cohort of patients.
The Pediatric PSC Consortium is an active research registry involving 55 sites throughout Europe, North and South America, the Middle East, and Asia. We evaluated known cases of pediatric-onset PSC-IBD who initiated VDZ therapy at our sites. Patients initiated VDZ therapy between February 2015 and January 2019. We collected labs within 3 months before initiation of VDZ, and for up to 9 to 12 months on treatment including: hemoglobin, platelet count, total bilirubin, albumin, aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), γ-glutamyltransferase (GGT), erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and AST-to-platelet ratio index (APRI). As bloodwork was obtained at different timepoints in all patients, we included initial baseline laboratory studies in addition to studies within 3 to 6 months, +/- 2 weeks as a mid-point, and 9 to 12 months, +/- 2 weeks as a final point for comparison. Wherever multiple values were present within the time period, median values were used. Baseline pre-VDZ and post-VDZ magnetic resonance cholangiopancreatography (MRCP), liver histology, Mayo endoscopic subscore from colonoscopy, physician global assessment (PGA) of IBD activity (16), pediatric ulcerative colitis activity index (PUCAI), and pediatric Crohn disease activity index (PCDAI) were also collected.
We recorded VDZ-dosing regimens, age at initiation, duration of therapy, reasons for discontinuation, ursodeoxycholic acid use, and concurrent immunosuppressive medications. Hospitalizations, liver transplantation, malignancies, cholangitis, ascites, variceal bleeding, encephalopathy, and infections after VDZ introduction were all documented.
A biochemical response was defined as a 75% or greater reduction in initial GGT from baseline, or a GGT that fell to <50 IU/L after completing up to 12 months of VDZ therapy; as this has been shown to predict favorable outcomes in children with PSC (17). Additionally, we assessed changes in other biochemical parameters on therapy, change in Metavir fibrosis stage on serial liver histologic samples, and radiologist's interpretation of improvement in degree of biliary stricturing when serial MRCP was available.
Inflammatory Bowel Disease Outcomes
IBD clinical activity was assessed with serial PGA, PUCAI, and PCDAI as available. Clinical remission was defined as a PUCAI or PCDAI less than 10 or a PGA of zero. Additionally, wherever available, serial Mayo endoscopic subscores of colonoscopies were collected to assess for endoscopic remission (Mayo 0), endoscopic treatment response (decrease of at least 1 category) or no response (no change in category or worsening). We defined an IBD clinical/treatment response as either improved PGA (decrease of at least 1 category), PUCAI (minimum decrease, 20 points), and PCDAI (minimum decrease, 12.5 points) or improved Mayo endoscopic subscore (decrease of at least 1 category) if endoscopy was available, after up to 9 to 12 months of VDZ therapy.
Categorical data were compared using the chi-squared test or Fisher exact test as appropriate. Continuous data were compared using the Wilcoxon rank sum test. Paired observations were compared using the sign rank test. Calculations were done using Stata version 16 (StataCorp, College Station, Texas, USA). Institutional review board and research ethics approval was granted at all participating sites.
We identified 40 patients with pediatric-onset PSC-IBD treated with VDZ from 19 centers. Cohort characteristics are summarized in Table 1. The median age was 16 years [IQR 15–18] at VDZ initiation. IBD and PSC were present for a median of 3.5 years [IQR 2.4–5.5] and 2.8 years [IQR 1.5–4.9] before VDZ initiation. Prior TNF-α failure had occurred in 76%. 24% of patients were prescribed ursodeoxycholic acid while receiving VDZ. Nine patients had a baseline GGT <50 IU/L. Three patients were excluded because of starting VDZ after receiving a liver transplant, without evidence of recurrent PSC (rPSC) in the graft. One patient started VDZ after liver transplant for rPSC and was included. Thus 37 patients were included in our analysis.
Liver biochemical response to VDZ was achieved in 22% and not achieved in 78% in their first year on the IBD treatment. GGT before and during treatment is shown in Figure 1A. There was a mild increase in median GGT after initiation of VDZ. Most patients returned to pre-VDZ levels by 9 to 12 months as seen in Figure 1B. Overall, median GGT before treatment and after up to 1 year after VDZ initiation was unchanged (P = 0.708). Total bilirubin was similarly unaffected by VDZ therapy (P = 0.768) as shown in Figure 1C.
Pretreatment and posttreatment MRCPs were available in 13 patients: relative stricturing was worse in 20% and unchanged in 80% post therapy based on individual radiologist interpretations of written reports. Serial liver biopsies were available in 7 patients. Metavir fibrosis score was unchanged from baseline in all patients.
IBD phenotype was UC in 59%, CD in 30%, and IBD unclassified in 11%. Prior TNF-α failure had occurred in 76%. VDZ was discontinued before 12 months in 22% because of lack of IBD treatment response. Other concurrent IBD medications, which were prescribed at the start of VDZ induction included: azathioprine 22%, methotrexate 6%, aminosalicylates 19%, and budesonide 3%.
Sixteen patients had serial endoscopies performed; 75% showed endoscopic improvement, (P = 0.065) as shown in Figure 2. 19% of patients with serial endoscopies achieved endoscopic remission. Of the 19%, all patients had a Mayo 1 endoscopy sub score before achieving remission. Four patients at follow-up endoscopy achieved a Mayo 1 endoscopy subscore, 75% of these patients had a subscore of 2 at previous endoscopy. Using clinical activity scores in patients in whom serial endoscopy was not performed, 6-month follow-up IBD activity data was available in 34 patients. All had active disease at baseline: 32% achieved clinical remission, 30% had a clinical response but not remission, and 38% had no response or worsening, including 4 patients who required colectomy within 14 months of initiation of VDZ. Three patients included in the analysis had received a colectomy before initiation of VDZ. Of the 3 patients, 2 had CD, and 1 with UC complicated by colonic mucosal dysplasia before colectomy, sacroiliitis, and pouchitis.
Liver inflammation was correlated with gastrointestinal inflammation in this study. Liver biochemistry had a worsening trend in patients with IBD unresponsive to VDZ. Final GGT after up to 9 to 12 months of VDZ was 51 [IQR 28–71] in patients with IBD in remission versus 127 [IQR 63–226] in patients with active IBD on VDZ (P = 0.066).
Adverse liver events were uncommon over the 1 year of follow-up. One patient developed acute bacterial cholangitis requiring hospitalization 3 months after VDZ initiation and subsequently underwent liver transplantation 9 months later. One patient underwent liver transplantation 2 years after starting VDZ. No other hospitalizations, surgeries, liver transplantations, malignancies, cholangitis, ascites, variceal bleeding, encephalopathy, or other infections after VDZ introduction were documented.
Predictors of Response
We compared baseline pre-VDZ characteristics of patients who did or did not achieve a liver biochemical response. Additionally, we compared baseline pre-VDZ characteristics of patients who did or did not achieve IBD remission while on VDZ, shown in Table 2. Serum albumin of liver biochemical responders was slightly higher than liver biochemistry of nonresponders. More patients who achieved IBD remission had features of overlap with AIH compared with those who did not achieve IBD remission. None of the patients with an IBD-U phenotype (n = 3) achieved either a liver biochemical response or IBD remission.
We conducted the first multicenter study evaluating the efficacy of VDZ in the treatment of pediatric-onset PSC-IBD. There were 3 major findings from this study. First, this is an important negative study showing that VDZ was not effective in treating pediatric-onset PSC based on serum GGT levels. Second, we verified VDZ was associated with favorable clinical and endoscopic IBD outcomes. Third, we showed that patients with a VDZ-refractory IBD phenotype had more severe biochemical liver disease, indicating that hepatobiliary and colonic inflammation may be pathogenically linked.
Despite a promising mechanism of blocking lymphocyte migration into the colon and liver simultaneously, VDZ did not improve hepatobiliary biochemistry, histopathology, or radiographic imaging in pediatric-onset PSC-IBD after up to 12 months of therapy in this retrospective analysis. These results are consistent with 3 retrospective studies of VDZ therapy in adult patients with PSC-IBD. No improvements were seen in biochemistry, radiographic studies, or elastography after 1.5 to 13 months of therapy (8,13,18).
It is unknown if MAdCAM-1 hepatobiliary upregulation is critical to the pathophysiology of PSC (10), or whether it is an epiphenomenon. MAdCAM-1 expression in the portal venous epithelium is variable and may play a role in biochemical improvement. Indeed, MAdCAM-1 expression has been detected in a minority of liver biopsy samples from PSC patients (11,19). MAdCAM-1 tissue assays were not done as part of this study, so it is unknown, which patients actively expressed MAdCAM-1.
VDZ pharmacokinetics remain incompletely understood (20,21). Additionally, trough and antibody concentrations are not routinely used in pediatric clinical practice and were not collected in this study, making it difficult to ascertain, which patients achieved therapeutic levels throughout the review. In 1 study, VDZ was found to be relatively slow to achieve its peak effect, Showing restoration of normal colonic immune gene expression microarray profiles at 52 weeks in UC patients on VDZ. In contrast, restoration of similar gene expression microarray profiles occurred between 4 and 6 weeks with infliximab-treated UC patients (22). The differences in normalization of immune gene profiles is likely because of the many downstream effects of infliximab compared with VDZ and its gut-selective integrin mechanism (22). These findings may suggest a similar mucosal restoration mechanism between infliximab and VDZ, although different rates at which healing occurs. This data suggests that PSC-IBD patients may experience a hepatobiliary biochemical response after 52 weeks when the mucosal integrity and immune gene expression of the lumen has been restored. Suggesting that our cohort of patients received an appropriate treatment duration to observe a clinical response; however, this was not observed.
In this cohort, VDZ was found to be generally effective for the colonic disease of PSC-IBD. 62% of patients achieved a clinical and/or endoscopic IBD response to VDZ, with over half of these achieving complete endoscopic or clinical remission. This response rate in pediatric PSC-IBD is similar to the 35% to 60% rates of clinical response or clinical remission from clinical trials in IBD patients without PSC (4–7,23). Studies have investigated predictors of clinical response in IBD when treated with VDZ (20,21). From these studies, it was identified that lower trough concentrations, higher body mass, elevated CRP, lower albumin, and more severe lumenal disease are predictors of poor response to VDZ. In this analysis, lower albumin was found to be a predictor of poor liver biochemical response. Features of overlap with AIH patients, however, experienced higher rates of remission of IBD. Perhaps as they were on more aggressive immunosuppressive regimens; or possibly as they have heightened expression of MAdCAM-1 in their liver (24), and may have heightened expression in the lumen leading them to be more responsive to VDZ.
There are no prospective assessments of liver biochemistry of patients with PSC-IBD with sustained endoscopic/histologic IBD remission (25). We showed that patients with VDZ-refractory IBD had higher GGT levels than those with VDZ-responsive IBD. Posttreatment GGT levels were lower amongst patients with colonic disease responsive to VDZ. Similarly, this was extrapolated from a different study in patients who had a colectomy before their PSC diagnosis; these patients had better liver prognosis and outcomes without underlying colonic inflammation present, implying some aspect of colonic disease driving biliary disease (26).
GGT is a biomarker for hepatobiliary disease of pediatric PSC (17). In our study, however, GGT was found to be a possible marker of mucosal healing in PSC-IBD. As patients with colonic inflammatory improvement also saw generally lower GGT levels compared with the active IBD subset of patients. In other studies, however, fecal calprotectin and GGT measurements over time were not correlated well in children (27), and the severity of ductular disease was not strongly associated with the severity of colitis in adults (28).
Persistent colonic inflammation may drive increased hepatobiliary inflammation because of the shared lymphocyte integrin mechanism along the gut-liver axis. It is possible that the subset of patients who achieved IBD remission or an IBD response with VDZ were more likely to express MAdCAM-1 in the liver, as demonstrated by their lower GGT. This liver biochemical response, however, may have occurred via other mechanisms along the gut-liver axis known to play a role in PSC-IBD pathogenesis, such as dysbiosis (29), disturbances in bile acid metabolism, toxicity of bile acids (30), and abnormal intestinal permeability (31).
The main strength of this study is the diverse patient population from secondary and tertiary referral centers. The main weaknesses are our reliance on retrospective data and lack of a control group. Though we pulled patients from many centers, this remains a relatively small cohort and type II error is possible. Patients were diagnosed and managed per local institutional protocols with varied clinical/pharmaceutical management and did not have the same evaluations with the same follow-up timing. As such, patients who underwent follow-up endoscopies, liver biopsies or imaging may have been more likely to be exhibiting symptoms, limiting these subanalyses to a more severe group. It is possible that 1 year was not long enough to see changes in liver histology and biliary imaging. VDZ trough and antibody concentrations were not evaluated throughout this study, and dosing of VDZ varied from center to center.
We present the first outcomes of pediatric-onset PSC-IBD patients treated with VDZ. VDZ generally did not improve or alter liver outcomes. VDZ was, however, effective for treating the IBD component of this unique PSC-IBD phenotype. IBD response rates were similar to those observed in other IBD patients without PSC. In patients who achieved IBD remission while on VDZ, stabilization of GGT levels was observed compared with patients with refractory IBD. This association helps support the theory that colonic inflammation may correlate with persistent hepatobiliary inflammation in pediatric PSC-IBD. Effective treatments for PSC are still needed.
1. Deneau MR, El-Matary W, Valentino PL, et al. The natural history of primary sclerosing cholangitis in 781 children: a multicenter, international collaboration. Hepatology
2. Weismuller TJ, Trivedi PJ, Bergquist A, et al. International PSC Study Group. Patient age, sex, and inflammatory bowel disease phenotype associate with course of primary sclerosing cholangitis. Gastroenterology
3. Sorensen JO, Nielsen OH, Andersson M, et al. Inflammatory bowel disease with primary sclerosing cholangitis: a Danish population-based cohort study 1977–2011. Liver Int
4. Feagan BG, Rutgeerts P, Sands BE, et al. GEMINI 1 Study Group. Vedolizumab as induction and maintenance therapy for ulcerative colitis. N Engl J Med
5. Sandborn WJ, Feagan BG, Rutgeerts P, et al. GEMINI 2 Study Group. Vedolizumab as induction and maintenance therapy for Crohn's disease. N Engl J Med
6. Sands BE, Peyrin-Biroulet L, Loftus EV Jr, et al. VARSITY Study Group. Vedolizumab versus adalimumab for moderate-to-severe ulcerative colitis. N Engl J Med
7. Dulai PS, Singh S, Jiang X, et al. The real-world effectiveness and safety of vedolizumab for moderate-severe crohn's disease: results from the US VICTORY Consortium. Am J Gastroenterol
8. Christensen B, Micic D, Gibson PR, et al. Vedolizumab in patients with concurrent primary sclerosing cholangitis and inflammatory bowel disease does not improve liver biochemistry but is safe and effective for the bowel disease. Aliment Pharmacol Ther
9. Lobaton T, Vermeire S, Van Assche G, et al. Review article: anti-adhesion therapies for inflammatory bowel disease. Aliment Pharmacol Ther
10. de Krijger M, Wildenberg ME, de Jonge WJ, et al. Return to sender: Lymphocyte trafficking mechanisms as contributors to primary sclerosing cholangitis. J Hepatol
11. Hillan KJ, Hagler KE, MacSween RN, et al. Expression of the mucosal vascular addressin, MAdCAM-1, in inflammatory liver disease. Liver
12. Grant AJ, Lalor PF, Salmi M, et al. Homing of mucosal lymphocytes to the liver in the pathogenesis of hepatic complications of inflammatory bowel disease. Lancet
13. Caron B, Peyrin-Biroulet L, Pariente B, et al. Vedolizumab therapy is ineffective for primary sclerosing cholangitis in patients with inflammatory bowel disease: a GETAID multicentre cohort study. J Crohns Colitis
14. Eksteen B, Grant AJ, Miles A, et al. Hepatic endothelial CCL25 mediates the recruitment of CCR9+ gut-homing lymphocytes to the liver in primary sclerosing cholangitis. J Exp Med
15. Lynch KD, Chapman RW, Keshav S, et al. Effects of vedolizumab in patients with primary sclerosing cholangitis and inflammatory bowel diseases. Clin Gastroenterol Hepatol
16. Samaan MA, Mosli MH, Sandborn WJ, et al. A systematic review of the measurement of endoscopic healing in ulcerative colitis clinical trials: recommendations and implications for future research. Inflamm Bowel Dis
17. Deneau MR, Mack C, Abdou R, et al. Gamma glutamyltransferase reduction is associated with favorable outcomes in pediatric primary sclerosing cholangitis. Hepatol Commun
18. Tse CS, Loftus EV Jr, Raffals LE, et al. Effects of vedolizumab, adalimumab and infliximab on biliary inflammation in individuals with primary sclerosing cholangitis and inflammatory bowel disease. Aliment Pharmacol Ther
19. Ala A, Brown D, Khan K, et al. Mucosal addressin cell adhesion molecule (MAdCAM-1) expression is upregulated in the cirrhotic liver and immunolocalises to the peribiliary plexus and lymphoid aggregates. Dig Dis Sci
20. Dreesen E, Verstockt B, Bian S, et al. Evidence to support monitoring of vedolizumab trough concentrations in patients with inflammatory bowel diseases. Clin Gastroenterol Hepatol
21. Rosario M, Dirks NL, Gastonguay MR, et al. Population pharmacokinetics-pharmacodynamics of vedolizumab in patients with ulcerative colitis and Crohn's disease. Aliment Pharmacol Ther
22. Arijs I, De Hertogh G, Lemmens B, et al. Effect of vedolizumab (anti-alpha4beta7-integrin) therapy on histological healing and mucosal gene expression in patients with UC. Gut
23. Schreiber S, Dignass A, Peyrin-Biroulet L, et al. Systematic review with meta-analysis: real-world effectiveness and safety of vedolizumab in patients with inflammatory bowel disease. J Gastroenterol
24. Grant AJ, Lalor PF, Hubscher SG, et al. MAdCAM-1 expressed in chronic inflammatory liver disease supports mucosal lymphocyte adhesion to hepatic endothelium (MAdCAM-1 in chronic inflammatory liver disease). Hepatology
25. Laborda TJ, Jensen MK, Kavan M, et al. Treatment of primary sclerosing cholangitis in children. World J Hepatol
26. Nordenvall C, Olen O, Nilsson PJ, et al. Colectomy prior to diagnosis of primary sclerosing cholangitis is associated with improved prognosis in a nationwide cohort study of 2594 PSC-IBD patients. Aliment Pharmacol Ther
27. Shiau H, Ihekweazu FD, Amin M, et al. Unique inflammatory bowel disease phenotype of pediatric primary sclerosing cholangitis: a single-center study. J Pediatr Gastroenterol Nutr
28. Stockbrugger RW, Olsson R, Jaup B, et al. Forty-six patients with primary sclerosing cholangitis: radiological bile duct changes in relationship to clinical course and concomitant inflammatory bowel disease. Hepatogastroenterology
29. Vaughn BP, Kaiser T, Staley C, et al. A pilot study of fecal bile acid and microbiota profiles in inflammatory bowel disease and primary sclerosing cholangitis. Clin Exp Gastroenterol
30. Sabino J, Vieira-Silva S, Machiels K, et al. Primary sclerosing cholangitis is characterised by intestinal dysbiosis independent from IBD. Gut
31. Karlsen TH, Folseraas T, Thorburn D, et al. Primary sclerosing cholangitis - a comprehensive review. J Hepatol