Inflammatory bowel disease in the setting of ... - Wiley Online Library

ORIGINAL ARTICLE

Inflammatory Bowel Disease in the Setting of Autoimmune Pancreatitis Karthik Ravi, MD,* Suresh T. Chari, MD,† Santhi S. Vege, MD,† William J. Sandborn, MD,† Thomas C. Smyrk, MD,‡ and Edward V. Loftus, Jr, MD†

Background: Despite scattered case reports, the prevalence of inflammatory bowel disease (IBD) in patients with autoimmune pancreatitis (AIP) is unknown. We sought to better characterize the putative association between the conditions.

Methods: Medical records of 71 patients meeting accepted criteria for AIP were reviewed to identify those with endoscopic and histological evidence of IBD. Colon samples in patients with both AIP and IBD were immunostained to identify IgG4-positive cells.

Results: Four patients with AIP (5.6%) had a diagnosis of IBD: 3 had ulcerative colitis (UC) and 1 had Crohn’s disease (CD). The diagnosis of IBD preceded or was simultaneous to that of AIP. Two AIP-UC patients treated for AIP with prednisone had a recurrence of AIP, and 1 required 6-mercaptopurine for long-term corticosteroidsparing treatment. Two AIP-IBD patients underwent Whipple resections, and 1 had recurrent AIP. All 3 patients with UC presented with pancolitis, and 2 required colectomy. Colon samples from 1 patient with UC and 1 patient with CD were available for review. Increased numbers of IgG4-positive cells (10 per high-power field) were noted on the colon sample from the patient with UC. Conclusions: Almost 6% of patients with proven AIP had a diagnosis of IBD, compared to a prevalence of ⬇0.4%– 0.5% in the general population, potentially implying a 12–15-fold increase in risk. Patients with both AIP and IBD may have increased extent and severity of IBD. The finding of IgG4-positive cells on colon biopsy suggests that IBD may represent an extrapancreatic manifestation of AIP. (Inflamm Bowel Dis 2009;15:1326 –1330) Key Words: autoimmune pancreatitis, inflammatory bowel disease, risk, Crohn’s disease, ulcerative colitis

Received for publication January 7, 2009; Accepted January 13, 2009. From the *Department of Internal Medicine, †Miles & Shirley Fiterman Center for Digestive Diseases, ‡Department of Pathology, Mayo Clinic, Rochester, Minnesota. Presented in part at the 72nd Annual Meeting of the American College of Gastroenterology, October 12–17, 2007, Philadelphia, Pennsylvania (Am J Gastroenterol. 2007;102(S2):S483). Reprints: Edward V. Loftus, Jr., MD, Division of Gastroenterology & Hepatology, Mayo Clinic, 200 First St., SW, Rochester, MN 55905 (e-mail: [email protected]). Copyright © 2009 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1002/ibd.20898 Publishedonline23February2009inWileyInterScience(www.interscience. wiley.com).

1326

A

utoimmune pancreatitis (AIP) was first described by Sarles et al in 1961,1 but not until the mid-1980s and early 1990s was it formally recognized as a clinical entity. AIP can involve not only the pancreas but other organs including bile ducts, gallbladder, salivary glands, retroperitoneum, lymph nodes, pleura, lungs, and kidneys. Affected organs have a lymphoplasmacytic infiltrate and fibrosis.2– 6 Histological hallmarks include diffuse irregular narrowing of the main pancreatic duct and stenosis of the common bile duct.7–11 The clinical presentation is variable, ranging from completely asymptomatic patients to those with jaundice, nausea, vomiting, and weight loss. Radiographic findings are similarly nonspecific. Interestingly, elevated serum immunoglobulin G4 (IgG4) levels have been correlated with AIP. Furthermore, tissue immunostaining of affected organs reveals abundant infiltration with IgG4-positive plasma cells.12,13 Effective treatment response to oral prednisone has been well documented, and surgical resection has been reserved for refractory cases.10 Previous studies have suggested associations between AIP and other disorders, such as primary sclerosing cholangitis (PSC), Sjo¨gren’s syndrome, and retroperitoneal fibrosis. More recent studies have suggested that sclerosing inflammation mimicking these conditions may represent systemic manifestations of AIP. This idea is supported by the fact that the infiltrate tends to be rich in IgG4-positive plasma cells regardless of the organ or site involved.2– 6 Furthermore, some manifestations are responsive to corticosteroid therapy, whereas the disorders they mimic typically are not.12,13 Ball et al14 were the first to describe a relationship between IBD and chronic pancreatitis. In this postmortem study, chronic interstitial pancreatitis was detected in 46 of 86 ulcerative colitis (UC) patients. Most subsequent reports examining the relationship between UC and chronic pancreatitis have been limited to case reports.15–19 This has led to uncertainty as to the true nature of this putative association and the clinical features of inflammatory bowel disease (IBD) in this setting. Barthet et al,15 in a small case series and literature review, estimated the incidence of chronic pancreatitis in patients with IBD to be 1.2%. This was 250 times greater than the incidence of chronic pancreatitis found in the general population.20 However, identification of pancreatitis was based largely on symptoms, with histological evidence confirming the diagnoses in only 2 cases. Inflamm Bowel Dis

●

Volume 15, Number 9, September 2009

Inflamm Bowel Dis

●

IBD and AIP


TABLE 1. HISORt Criteria for Autoimmune Pancreatitis Category

Criteria

Histology

At least 1 of the following: 1) Periductal lymphoplasmacytic infiltrate with obliterative phlebitis and storiform fibrosis (lymphoplasmacytic sclerosing pancreatitis). 2) Lymphoplasmacytic infiltrate with storiform fibrosis showing abundant (ⱖ10 cells/highpowered field) IgG4-positive cells. Typical: diffusely enlarged gland with delayed (rim) enhancement; diffusely irregular, attenuated main pancreatic duct. Others: Focal pancreatic mass/enlargement; focal pancreatic duct stricture; pancreatic atrophy; pancreatic calcification; or pancreatitisa. Elevated serum IgG4 level (normal, 8-140 mg/dL). Hilar/intrahepatic biliary strictures, persistent distal biliary stricture, parotid/lacrimal gland involvement, mediastinal lymphadenopathy, retroperitoneal fibrosis. Resolution/marked improvement of pancreatic/extrapancreatic manifestation with steroid therapy.

Pancreatic imaging

Serology Other organ involvementb Response to corticosteroid therapyc

Adapted from Chari, et al.7,8 a With negative work-up for known etiologies for pancreatic disease, especially pancreaticobiliary cancer. b Radiologic evidence of organ involvement can be confirmed by biopsy showing lymphoplasmacytic infiltrate with abundant IgG4-positive cells or its resolution/improvement with corticosteroid therapy. c Corticosteroid therapy should be given only to patients with negative work-up for known etiologies for pancreatic disease and only to those in whom response can be objectively assessed. It should not be used as a substitute for a thorough search for etiology.

Overall, previous studies do not provide clear answers as to the relationship between AIP and IBD. Specifically, the frequency of concomitant symptomatic IBD in patients with AIP is unknown. In those with AIP and IBD, it is unclear if IBD is merely an associated immune-mediated disease or a true extrapancreatic manifestation of AIP. If the latter, one would expect the colitis in AIP to be associated with an infiltrate rich in IgG4-positive cells. It has been shown that IgG4 infiltration in the pancreas is specific for AIP and is not seen in chronic pancreatitis. Thus, if IBD were an extrapancreatic manifestation of AIP, we hypothesize that it would be associated with infiltration of IgG4-positive cells in the affected small intestine or colon which would not be seen in usual IBD. Our aim was to examine the association between AIP and IBD and to attempt to answer the question as to whether IBD represents an extrapancreatic manifestation of AIP.

MATERIALS AND METHODS The study was approved by the Mayo Foundation Institutional Review Board. A retrospective review of 71 patients identified from a database of patients meeting the recently published HISORt criteria7,8 for AIP (Table 1) was conducted. Medical records were reviewed to identify patients with a diagnosis of IBD. A clinical diagnosis of IBD required confirmation by either endoscopy or surgery and histology. Patients with AIP and IBD were compared to those with AIP alone. Demographic data, clinical presentation, and serological markers including serum IgG4, antinuclear antibodies (ANA), rheumatoid factor (RF), and carbohydrate antigen 19-9 (CA 19-9) were reviewed.

Elevated concentrations of IgG4, ANA, RF, and CA 19-9 were defined by values greater than 140 mg/dL, 1.0 U, 15 IU/mL, and 40 U/mL, respectively. A history of other organ involvement such as PSC-like changes in the biliary tree, tubulointerstitial nephritis, and retroperitoneal fibrosis was noted. Finally, response to treatment, whether medical or surgical, was evaluated in both groups. Available colon tissue samples from patients with AIP and IBD were reviewed by a single pathologist (T.C.S.). Tissue immunostaining, utilizing monoclonal antihuman IgG4 antibody, was then performed using standard immunohistochemical techniques. The number of immunohistochemically identified IgG4-positive plasma cells per highpower field (HPF) was counted in each specimen.

RESULTS Four of the 71 AIP patients (5.6%; 95% confidence interval [CI], 1.8%–14.0%) had a confirmed diagnosis of IBD. The median age at diagnosis of AIP in these patients was 54 years (range, 44 –70 years), and all 4 were Caucasian males (Table 2). In contrast, the median age at diagnosis in the 67 patients with AIP alone was 64 years, with 52 male patients (78%). Race was identified in the medical records of 57 patients, with 47 (87.7%) of these being Caucasian. There were no statistically significant differences in age, gender, or race between the 2 groups. Clinical presentation was also similar in both groups. Obstructive jaundice was the most common symptom at presentation, occurring in 46 patients with AIP alone (69%) and 2 patients with AIP and IBD (50%). Abdominal pain was noted by 31 patients with AIP

1327

Inflamm Bowel Dis

Ravi et al

TABLE 2. Demographics and Symptoms at Presentation of Patients with AIP Alone and AIP in Conjunction with IBD Characteristics Median age, years (range) Males, n (%) Caucasian, n (%) Obstructive jaundice, n (%) Abdominal pain, n (%) New onset diabetes, n (%) Steatorrhea, n (%)

AIP (n ⫽ 67)

AIP-IBD (n ⫽ 4)

54 (44–70) 52 (77.6%) 47 (87.7%) 46 (68.7%) 31 (46.3%) 11 (16.4%) 6 (9.0%)

64 (18–86) 4 (100%) 4 (100%) 2 (50%) 2 (50%) 2 (50%) 1 (25%)

AIP, autoimmune pancreatitis; IBD, inflammatory bowel disease.

alone (46%) and 2 patients with both AIP and IBD (50%), with new-onset diabetes and steatorrhea seen less commonly (Table 2). Elevated serum IgG4 was seen in the majority of patients. Thirty-one patients with AIP alone (74% of those tested) had elevated IgG4, compared to 1 of 2 patients with AIP and IBD (Table 3). Fewer patients had abnormal ANA and RF levels, with elevated concentrations of ANA and RF seen in 5 out of 23 (21.7%) and 3 of 6 (50%), respectively, in patients with AIP alone. No patients with AIP and IBD had ANA or RF concentrations obtained. Interestingly, 14 out of 41 patients with AIP alone (34.1%) and 1 out of 2 patients with AIP and IBD (50%) had elevated CA 19-9 concentrations. However, 10 of these patients (9 with AIP alone and 1 with AIP and IBD) had jaundice with elevated hepatic biochemical tests at presentation, potentially confounding the finding of an elevated CA 19-9. Thirteen of 16 patients with AIP alone (81.3%) and 1 of 2 patients with AIP and IBD (50%) had positive IgG4 immunostaining (greater than or equal to 10 IgG4-positive cells per HPF) in pancreatic biopsies. Evidence of extrapancreatic involvement was seen in 18 patients with AIP alone (27%) and 1 patient with AIP and

TABLE 3. Clinical Profile of Patients with AIP Alone and Those with Both IBD and AIP Clinical Profile

AIP, n/N tested (%)

AIP-IBD, n/N tested (%)

Elevated serum IgG4a Elevated serum CA 19-9b Positive IgG4 immunostain Extrapancreatic involvement Recurrence of AIP

31/42 (73.8%) 14/41 (34.1%) 13/16 (81.3%) 18/67 (26.9%) 18/61 (29.0%)

1/2 (50%) 1/2 (50%) 1/2 (50%) 1/4 (25%) 3/4 (75%)

AIP, autoimmune pancreatitis; IgG4, immunoglobulin G subtype 4; CA 19-9, carbohydrate antigen 19-9. a Serum IgG4 ⬎ 140 mg/dL. b Serum CA 19-9 ⬎ 40 U/mL.

1328

●


IBD (25%). The AIP-IBD patient had involvement of the proximal extrahepatic bile duct. In those with AIP alone, involved organs included the proximal extrahepatic and intrahepatic bile ducts (9 patients), retroperitoneum (3 patients), kidneys (3 patients), salivary glands (2 patients), and mediastinum (1 patient). Of the 4 patients with AIP and IBD, 3 were diagnosed with UC and 1 with Crohn’s disease (CD). All 3 UC patients presented with pancolitis, while the patient with CD had ileocolonic involvement. The diagnosis of UC preceded that of AIP in all 3 patients, while the diagnoses of CD and AIP were simultaneous in the other. Surgery for refractory colitis was performed prior to diagnosis of AIP in 2 of the 3 UC patients. Colon samples from 2 patients were available for review, 1 from the patient with CD and the other from a patient with UC. IgG4 immunostaining revealed 10 IgG4positive cells per HPF on the biopsy sample from the patient with UC (Fig. 1), while no IgG4-positive cells were seen on the sample from the patient with CD. Two patients with AIP and UC received a standard course of prednisone 40 mg daily for 4 weeks followed by a taper. Both patients had subsequent recurrence of AIP. One patient required corticosteroid-sparing maintenance therapy with 6-mercaptopurine. The other patient was treated effectively with a second 4-week course of prednisone 40 mg daily followed by a taper. The third patient with AIP and UC underwent surgical resection with a Whipple procedure without recurrence. A Whipple procedure was also performed in the patient with AIP and CD; however, there was a recurrence of AIP symptoms in this case. Of the 61 patients with AIP alone who had available follow-up, 18 patients (29%) had recurrence, compared to 3 out of 4 with AIP and IBD (75%). Thirty-three patients with AIP alone (54.1%) were treated with a standard 4-week course of prednisone followed by a taper. Fifteen of these 33 patients (45%) had recurrence of symptoms.

DISCUSSION Autoimmune pancreatitis is a newly described clinical entity. The diagnosis is based on several factors, including histological findings, radiographic findings, serology, extrapancreatic involvement, and response to corticosteroid therapy.7,8 Several sets of diagnostic criteria for AIP have been established. Although different criteria are typically utilized in Japan,21,22 the recently published HISORt criteria represent an acceptable tool utilized in the North American population. As the recognition of AIP has increased, several important clinical associations have become apparent. Associations with other autoimmune conditions such as PSC, retroperitoneal fibrosis, and sclerosing sialadenitis have previously been described.2– 4 Kamisawa et al23 demonstrated moderate to severe IgG4-positive plasma cell infiltration in affected extrapancreatic tissue in patients with confirmed AIP. Peripancreatic tissue, gallbladder, bile duct, colonic

Inflamm Bowel Dis

●


IBD and AIP

FIGURE 1. IgG4-positive colon sample from a patient with AIP and IBD. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

mucosa, salivary gland, and bone marrow in AIP patients had greater than 10 IgG4-positive cells per HPF, compared to less than 3 IgG4-positive cells per HPF at the same sites in controls. These findings suggest that AIP is an IgG4-related systemic disease, and associated extrapancreatic conditions could represent extensive organ involvement mimicking autoimmune conditions. Furthermore, the finding suggests that infiltration of extrapancreatic organs with a significant number of IgG4-positive cells is specific for AIP. In this preliminary study of a cohort of 71 patients with confirmed AIP, 4, or roughly 6%, also carried a diagnosis of IBD. The majority of patients were Caucasian males, with no differences in demographic or clinical characteristics between those with AIP and IBD and those with AIP alone. Roughly one-quarter of patients had evidence of extrapancreatic involvement of AIP. All 4 patients with IBD had extensive colitis, with 2 patients requiring surgery for refractory colitis. Similarly, both patients treated with a conventional course of

prednisone for AIP failed treatment, with 1 requiring longterm treatment with 6-mercaptopurine. One of 2 patients who underwent Whipple resection also had recurrence of AIP. Therefore, the recurrence rate of AIP was 75% in this small population, elevated compared to the 29% rate seen in those patients with AIP alone. IgG4 immunostaining of colon samples from 1 of 2 patients with AIP-IBD revealed 10 IgG4positive cells per HPF, which is diagnostic for AIP. It may very well be that the pancreatitis observed in some of these patients did not represent AIP. Indeed, both acute and chronic pancreatitis in the setting of IBD have been well described, with associated etiologies including gallstones, alcohol, duodenal Crohn’s involvement, and medications.14 –19,24 Further, there is also a suggestion that pancreatitis in this setting may represent an extraintestinal manifestation of IBD. However, in this patient cohort, all cases of pancreatitis met the HISORt diagnostic criteria, and were therefore consistent with a diagnosis of AIP.7,8

1329

Ravi et al

The relationship between AIP and IBD remains poorly defined. Previous studies have shown an increased rate of chronic pancreatitis in patients with IBD.15 However, whether chronic pancreatitis in these cases represented AIP is unclear. Moreover, the frequency of IBD in patients with confirmed AIP is unknown. Using the recently published HISORt criteria,7,8 our study found that 5.6% of patients in a cohort of 71 patients with AIP carried a diagnosis of IBD, with a 95% CI of 1.8%–14.0%. The prevalence of IBD in the general population is estimated to be 0.4%– 0.5%.24,25 Therefore, this represents a statistically significant increased prevalence of IBD in patients with AIP. Moreover, these findings potentially imply a greater than 10-fold increase in the frequency of IBD in patients with AIP. Although the possibility of “Berkson’s bias” (patients with 2 uncommon conditions are more likely to be referred to a tertiary medical center than patients with just 1 such condition)26 could have inflated the magnitude of this association to some extent, it still seems likely that the 2 conditions in this instance are related to a degree. Immunohistochemical IgG4 staining of the colon sample of a patient with AIP and IBD found 10 IgG4-positive cells per HPF. Interestingly, all 3 patients with a diagnosis of UC and AIP presented with pancolitis, and 2 required total colectomy for refractory colitis. In pancreatic tissue samples, a value of 10 IgG4-positive cells per HPF or greater has been considered strongly supportive of AIP.7,8 This value has also been used as a cutoff in sites of suspected extrapancreatic involvement.23 This finding therefore suggests that, in at least some cases, patients with AIP have IgG4 colitis mimicking IBD. Clinical presentation of AIP in terms of symptoms and serology was not significantly different regardless of the presence of IBD. Both groups had similar rates of extrapancreatic involvement, not including the colon. This finding is in accordance with the concept that AIP represents an IgG4related systemic process. However, recurrence of AIP occurred in 75% of patients with AIP with IBD compared to only 29% in patients with AIP alone. This is a preliminary study suggesting a potential association between AIP and IBD. The primary limitations of this study are its retrospective design and the relatively small number of patients. These factors limit the interpretation of our findings, and should be considered hypothesis-generating. However, AIP is a rare condition, and our cohort of 71 patients represents the largest single-center experience to date. In conclusion, AIP is a manifestation of a systemic IgG4-related condition. Retroperitoneal fibrosis, sclerosing sialadenitis, sclerosing cholangitis, and other conditions have previously been shown to be extrapancreatic manifestations. Our study establishes that patients with AIP also have an increased risk of IBD. This finding may represent an extrapancreatic manifestation of AIP rather than an associated immune-mediated condition in some instances. Further stud-

1330

Inflamm Bowel Dis

●


ies are needed to delineate the relationship between AIP and IgG4 colitis.

REFERENCES 1. Sarles H, Sarles JC, Muratore R, et al. Chronic inflammatory sclerosis of the pancreas: an autonomous pancreatic disease? Am J Dig Dis. 1961; 6:688 – 698. 2. Kamisawa T, Funata N, Hayashi Y, et al. A new clincopathological entity of IgG4-related autoimmune disease. J Gastroenterol. 2003;38: 982–984. 3. Kamisawa T, Funata N, Hayashi Y. Lymphoplasmacytic sclerosing pancreatitis is a pancreatic lesion of IgG4-related systemic disease. Am J Surg Pathol. 2004;24:1114. 4. Kamisawa T, Egawa N, Nakajima H. Autoimmune pancreatitis is a systemic autoimmune disease. Am J Gastroenterol. 2003;98:2811–2812. 5. Nishino T, Toki F, Oyama H, et al. Biliary tract involvement in autoimmune pancreatitis. Pancreas. 2005;30:76 – 82. 6. Kamisawa T, Nakajima H, Hishima T. Close correlation between chronic sclerosing sialadenitis and immunoglobulin G4. Intern Med J. 2006;36:527–529. 7. Chari ST, Smyrk TC, Levy MJ, et al. Diagnosis of autoimmune pancreatitis: the Mayo Clinic experience. Clin Gastroenterol Hepatol. 2006; 4:1010 –1016. 8. Chari ST. Diagnosis of autoimmune pancreatitis using its five cardinal features: introducing the Mayo Clinic’s HISORt criteria. J Gastroenterol. 2007;42(suppl 18):39 – 41. 9. Plaza JA, Colonna J, Vitellas KM, et al. Lymphoplasmacytic sclerosing pancreatitis. Ann Diagn Pathol. 2005;9:298 –301. 10. Ketikoglou I, Moulakakis A. Autoimmune pancreatitis. Dig Liver Dis. 2005;37:211–215. 11. Zamboni G, Lu¨ttges J, Capelli P, et al. Histopathological features of diagnostic and clinical relevance in autoimmune pancreatitis: a study on 53 resection specimens and 9 biopsy specimens. Virchows Arch. 2004; 445:552–563. 12. Kamisawa T, Okamoto A, Funata N. Clinicopathological features of autoimmune pancreatitis in relation to elevation of serum IgG4. Pancreas. 2005;31:28 –31. 13. Hamano H, Kawa S, Horiuchi A, et al. High serum IgG4 concentrations in patients with sclerosing pancreatitis. N Engl J Med. 2001;344:732–738. 14. Ball WP, Baggenstoss AH, Bargen JA. Pancreatic lesions associated with chronic ulcerative colitis. Arch Pathol. 1950;50:345–358. 15. Barthet M, Hastier P, Bernard JP, et al. Chronic pancreatitis and inflammatory bowel disease: true or coincidental association? Am J Gastroenterol. 1999;94:2141–2148. 16. Lysy J, Goldin E. Pancreatitis in ulcerative colitis. J Clin Gastroenterol. 1992;15:336 –339. 17. Forbes D, Scott RB, Trevenen C, et al. Chronic pancreatitis associated with ulcerative colitis. Clin Invest Med. 1987;10:321–324. 18. Seyrig JA, Jian R, Modigliani R, et al. Idiopathic pancreatitis associated with inflammatory bowel disease. Dig Dis Sci. 1985;30:1121–1126. 19. Okano A, Hajiro K, Takakuwa H, et al. Pseudotumorous pancreatitis associated with ulcerative colitis. Intern Med. 2001;40:1205–1208. 20. Faivre J, Bernades P. Prevalence and incidence of chronic pancreatitis in France. A population based study. Gut. 1994;335:A224. 21. Okazaki K, Uchida K, Matsushita M, et al. How to diagnose autoimmune pancreatitis by the revised Japanese clinical criteria. J Gastroenterol. 2007;42:32–38. 22. Okazaki K, Chiba T. Autoimmune related pancreatitis. Gut. 2002;51:1– 4. 23. Kamisawa T, Funata N, Hayashi Y, et al. A new clinicopathological entity of IgG4-related autoimmune disease. J Gastroenterol. 2003;38: 982–984. 24. Moolsintong P, Loftus EV, Chari ST, et al. Acute pancreatitis in patients with Crohn’s disease: clinical features and outcomes. Inflamm Bowel Dis. 2005;11:1080 –1084. 25. Loftus EV. The burden of inflammatory bowel disease in the United States: a moving target? Clin Gastroenterol Hepatol. 2007;5:1383– 1384. 26. Berkson J. Limitations of the application of fourfold tables to hospital data. Biometr Bull. 1946;2:47–53.