Portal de Periódicos da CAPES

Long-Term Follow-Up of Autoimmune Pancreatitis: Characteristics of Chronic Disease and Recurrence

2009; Elsevier BV; Volume: 7; Issue: 11 Linguagem: Inglês

10.1016/j.cgh.2009.07.041

1542-7714

Shigeyuki Kawa, Hideaki Hamano, Yayoi Ozaki, Tetsuya Ito, Ryou Kodama, Yoshimi Chou, Mari Takayama, Norikazu Arakura,

Neuroendocrine Tumor Research Advances

Autoimmune pancreatitis is a unique disease, characterized by lymphoplasmacytic inflammation in the acute stages. However, the active clinical features are unlikely to persist for long periods. Through long-term follow-up, we investigated the disease course in 51 patients with autoimmune pancreatitis. We found recurrence in 21 (41%) patients and pancreatic stone formation in 9 (18%) patients. Pancreatic stone formation was significantly more frequent in the recurrence group (7/21, 33%), compared with the nonrecurrence group (2/30, 7%). Moreover, we found high serum immunoglobulin G4 concentrations in 13 of 175 (7.4%) patients with ordinary chronic pancreatitis. This suggested that pancreatic stone formation is closely associated with recurrence and that autoimmune pancreatitis might transform into ordinary chronic pancreatitis after several recurrences. We found that the immune complex level, with a cutoff value of 10 μg/dL, served as a good predictor of recurrence, with high sensitivity (61.9%), specificity (70.0%), and efficacy (66.7%). We also confirmed that HLA and cytotoxic T-lymphocyte antigen-4 polymorphisms were useful predictors for AIP recurrence. Autoimmune pancreatitis is a unique disease, characterized by lymphoplasmacytic inflammation in the acute stages. However, the active clinical features are unlikely to persist for long periods. Through long-term follow-up, we investigated the disease course in 51 patients with autoimmune pancreatitis. We found recurrence in 21 (41%) patients and pancreatic stone formation in 9 (18%) patients. Pancreatic stone formation was significantly more frequent in the recurrence group (7/21, 33%), compared with the nonrecurrence group (2/30, 7%). Moreover, we found high serum immunoglobulin G4 concentrations in 13 of 175 (7.4%) patients with ordinary chronic pancreatitis. This suggested that pancreatic stone formation is closely associated with recurrence and that autoimmune pancreatitis might transform into ordinary chronic pancreatitis after several recurrences. We found that the immune complex level, with a cutoff value of 10 μg/dL, served as a good predictor of recurrence, with high sensitivity (61.9%), specificity (70.0%), and efficacy (66.7%). We also confirmed that HLA and cytotoxic T-lymphocyte antigen-4 polymorphisms were useful predictors for AIP recurrence. Autoimmune pancreatitis (AIP) has been characterized by irregular narrowing of the main pancreatic duct (MPD) and sonolucent swelling of the parenchyma, both as a result of lymphoplasmacytic inflammation during the active stage of the disease. AIP has also been characterized by the absence of pancreatic stone formation.1Yoshida K. Toki F. Takeuchi T. et al.Chronic pancreatitis caused by an autoimmune abnormality: proposal of the concept of autoimmune pancreatitis.Dig Dis Sci. 1995; 40: 1561-1568Crossref PubMed Scopus (1249) Google Scholar These findings suggest that AIP is clearly different from chronic pancreatitis, including alcohol-induced chronic pancreatitis, and has a distinctive disease profile. However, these pathologic features are found in the acute stage of the disease; thus, it is unlikely that the inflammatory characteristic of this condition persists for long periods. In a chronic form of the disease, we might find different features from those now generally recognized. Furthermore, AIP is generally found in older people with suppressed immune surveillance systems; consequently, these patients might be susceptible to various malignant diseases. To clarify these issues, we conducted long-term follow-ups of AIP, and we investigated 4 topics: (1) outcome, recurrence, and pancreatic stone formation; (2) the ability of AIP to transform into chronic pancreatitis; (3) prediction of recurrence by serum markers; and (4) association of AIP with malignancies. We performed long-term follow-ups for 51 patients with AIP to assess outcomes.2Takayama M. Hamano H. Ochi Y. et al.Recurrent attacks of autoimmune pancreatitis result in pancreatic stone formation.Am J Gastroenterol. 2004; 99: 932-937Crossref PubMed Scopus (166) Google Scholar The observation periods were 24–178 months (mean, 72 months). Corticosteroid therapy was administered to 42 patients. During the long-term follow-up, 21 patients (41%) showed recurrences that required a second course of corticosteroid therapy (Figure 1). We previously reported that a high serum IgG4 concentration was frequently and specifically found in AIP, representing disease activity.3Hamano H. Kawa S. Horiuchi A. et al.High serum IgG4 concentrations in patients with sclerosing pancreatitis.N Engl J Med. 2001; 344: 732-738Crossref PubMed Scopus (2158) Google Scholar For these 51 patients, we found that the serum IgG4 concentration remained slightly high in more than 60% of patients, although they were in a clinically inactive state after corticosteroid therapy.4Kawa S. Hamano H. Kiyosawa K. High serum IgG4 concentrations in patients with sclerosing pancreatitis.N Engl J Med. 2001; 345: 147Crossref PubMed Scopus (22) Google Scholar This suggested that the active inflammatory process might persist even when patients are in a clinically inactive state; these conditions might facilitate recurrences. We found pancreatic stone formation in 9 of 51 patients (18%). Among the 21 patients with recurrence, 7 (33%) exhibited pancreatic stone formation; in contrast, pancreatic stones were found in only 2 of the 30 (7%) patients in the nonrecurrence group (Figure 1). Accordingly, pancreatic stone formation was judged to be closely associated with recurrence.2Takayama M. Hamano H. Ochi Y. et al.Recurrent attacks of autoimmune pancreatitis result in pancreatic stone formation.Am J Gastroenterol. 2004; 99: 932-937Crossref PubMed Scopus (166) Google Scholar Previous studies investigated the recurrence of AIP, but they reported lower recurrence rates, ranging from 6%–23%.5Wakabayashi T. Kawaura Y. Satomura Y. et al.Long-term prognosis of duct-narrowing chronic pancreatitis: strategy for steroid treatment.Pancreas. 2005; 30: 31-39PubMed Google Scholar, 6Kamisawa T. Okamoto A. Prognosis of autoimmune pancreatitis.J Gastroenterol. 2007; 42: 59-62Crossref PubMed Scopus (62) Google Scholar, 7Kamisawa T. Okamoto A. Wakabayashi T. et al.Appropriate steroid therapy for autoimmune pancreatitis based on long-term outcome.Scand J Gastroenterol. 2008; 43: 609-613Crossref PubMed Scopus (72) Google Scholar Although the exact reasons for these discrepancies are unclear, they might be related to the number of patients, the follow-up periods, and the type of corticosteroid therapy. Although a previous study reported that the absence of pancreatic stones is a characteristic of AIP,1Yoshida K. Toki F. Takeuchi T. et al.Chronic pancreatitis caused by an autoimmune abnormality: proposal of the concept of autoimmune pancreatitis.Dig Dis Sci. 1995; 40: 1561-1568Crossref PubMed Scopus (1249) Google Scholar the potential for forming pancreatic stones is not absent in AIP.2Takayama M. Hamano H. Ochi Y. et al.Recurrent attacks of autoimmune pancreatitis result in pancreatic stone formation.Am J Gastroenterol. 2004; 99: 932-937Crossref PubMed Scopus (166) Google Scholar, 6Kamisawa T. Okamoto A. Prognosis of autoimmune pancreatitis.J Gastroenterol. 2007; 42: 59-62Crossref PubMed Scopus (62) Google Scholar Incomplete obstruction of the MPD system and the stasis of pancreatic juice might give rise to the formation of pancreatic stones. The finding of irregular narrowing or stricture of the MPD in patients with AIP provided further support for this potential mechanism. In addition, recurrent attacks might have intensified incomplete obstruction of the duct system and caused pancreatic juice stasis, which could have facilitated stone formation; however, we lack evidence to confirm this hypothesis. One patient with AIP exhibited a pancreatic stone after several recurrences. In June 1996, a 55-year-old man was admitted to our hospital with epigastralgia. His serum amylase level was elevated to 3000 U/L, and he was diagnosed with acute pancreatitis. Next, obstructive jaundice appeared. His serum IgG4 concentration had increased to 486 mg/dL, and endoscopic retrograde pancreatography (ERP) showed irregular narrowing in the pancreatic head region (Figure 2A). He was diagnosed with AIP, and steroid therapy was administered. This resulted in amelioration of the ERP finding and lowered the IgG4 levels to 213 mg/dL. In August 1998, the patient showed jaundice and obstruction of the common bile duct by endoscopic retrograde cholangiopancreatography, but there was no swelling of the pancreatic parenchyma or irregular narrowing of the MPD. Serum IgG4 was dramatically elevated to 1135 mg/dL. He was diagnosed with recurrence, primarily in the bile duct lesion. Steroid therapy was administered, and this resulted in the amelioration of the bile duct obstruction. Ten years after the onset, in February 2006, the patient complained of epigastralgia, and ERP showed a narrowing in the body of the MPD and dilatation in the tail region (Figure 2B). At that time, he was prescribed prednisolone at a maintenance dose of 2.5 mg. An increase of prednisolone to 20 mg ameliorated the MPD narrowing and dilatation. However, after a reduction of the prednisolone dose, a pancreatic stone appeared in the body and tail regions of the MPD (Figure 2C). This case supported the hypothesis that multiple recurrences and the resulting MPD stenosis and pancreatic juice stasis might induce pancreatic stone formation. It also showed that high IgG4 concentrations correspond to recurrence and clearly indicated the active stage of AIP. On the other hand, we also examined a patient with AIP who had low serum IgG4 concentrations and no recurrences during 10 years of follow-up. In November 1997, a 65-year-old woman was admitted to an affiliated hospital, presenting with epigastralgia, obstructive jaundice, and pancreatic head swelling. She was diagnosed with pancreatic cancer. However, endoscopic retrograde cholangiopancreatography showed diffuse irregular narrowing of the MPD, suggesting a diagnosis of AIP (Figure 2D). Her serum IgG4 value was 42 mg/dL, and her serum antinuclear antibody level was ×160. She was diagnosed with AIP and given corticosteroid therapy, which ameliorated these findings. During the 10-year follow-up, she showed no serum elevation of IgG4 and no abnormal image findings including pancreatic swelling that would have suggested recurrences. Magnetic resonance imaging (MRI) (Figure 2E) and MRI of the pancreas (Figure 2F) also showed no progression or duct changes during those 10 years. Furthermore, she exhibited no pancreatic stone formation. Accordingly, this case suggested that a normal IgG4 concentration during long-term follow-up accurately represented an inactive state and no disease progression. This also suggested that low serum IgG4 concentrations might be considered an indication for the cessation of maintenance corticosteroid therapy. The results of this long-term follow-up suggested that some patients with AIP could develop pancreatic stones after several recurrent attacks. Further, this suggests that some cases of AIP might transform into ordinary chronic pancreatitis. If true, the next question was whether AIP was a precursor of ordinary chronic pancreatitis. We considered serum IgG4 elevation to be a serologic marker of AIP, even at chronic or advanced stages, because more than 60% of patients with AIP maintained high serum IgG4 concentrations after the clinical symptoms had resolved.4Kawa S. Hamano H. Kiyosawa K. High serum IgG4 concentrations in patients with sclerosing pancreatitis.N Engl J Med. 2001; 345: 147Crossref PubMed Scopus (22) Google Scholar To investigate whether AIP might result in ordinary chronic pancreatitis, we measured serum levels of IgG4 in 175 patients with chronic pancreatitis that had been diagnosed before 1995, the year that AIP was first described.1Yoshida K. Toki F. Takeuchi T. et al.Chronic pancreatitis caused by an autoimmune abnormality: proposal of the concept of autoimmune pancreatitis.Dig Dis Sci. 1995; 40: 1561-1568Crossref PubMed Scopus (1249) Google Scholar We found high serum IgG4 concentrations in 13 of 175 patients with ordinary chronic pancreatitis (7.4%) (12 men and 1 woman; mean age, 56 years; 9 alcoholic and 4 idiopathic). Of the 13 patients, 3 had been diagnosed for the first time with pancreatic cancer, and 1 had been recently diagnosed with AIP. The remaining 9 patients showed typical findings of ordinary chronic pancreatitis, including pancreatic stones or irregular dilation of the MPD (Figure 3). This suggested that an advanced stage of AIP might result in ordinary chronic pancreatitis. It did not rule out the possibility that AIP might represent an early stage of ordinary chronic pancreatitis.8Kawa S. Hamano H. Clinical features of autoimmune pancreatitis.J Gastroenterol. 2007; 42: 9-14Crossref PubMed Scopus (76) Google Scholar Consistent with our results, a previous study found that serum IgG4 was elevated in the sera of 11.9% of patients with ordinary chronic pancreatitis.9Choi E.K. Kim M.H. Lee T.Y. et al.The sensitivity and specificity of serum immunoglobulin G and immunoglobulin G4 levels in the diagnosis of autoimmune chronic pancreatitis: Korean experience.Pancreas. 2007; 35: 156-161Crossref PubMed Scopus (89) Google Scholar Various serum markers and genetic markers have been reported to be associated with the recurrence of AIP. We found that for monitoring AIP, both IgG4 and the immune complex (IC), determined by the monoclonal rheumatoid factor (mRF) method, were useful markers. In the clinical course of a 69-year-old woman, we found 2 recurrences in which serum elevations of IC and IgG4 preceded the overt appearance of clinical recurrence by several months. This indicated that, in addition to IgG4, IC can sensitively predict recurrence and represent disease activity.8Kawa S. Hamano H. Clinical features of autoimmune pancreatitis.J Gastroenterol. 2007; 42: 9-14Crossref PubMed Scopus (76) Google Scholar We then investigated various serum markers for their efficacy in predicting recurrences by comparing the levels of various markers in recurrence and nonrecurrence groups. We found that the IC value, as determined by the monoclonal rheumatoid method (IC-mRF), was significantly higher at onset in the recurrence group compared with the nonrecurrence group. With a cutoff value of 10 μg/dL, IC-mRF performed well in predicting recurrence, with good sensitivity (61.9%), specificity (70.0%), and efficacy (66.7%). The probability of recurrence was 60% when IC-mRF >10 mg/dL and 30% when IC-mRF <10 mg/dL. Complement factors C3 and C4 have also been reported as useful markers for monitoring disease activity or tissue damage. We found decreased serum C3 or C4 levels in 35% and 37% of AIP patients, respectively. This suggested that complement activation might play a role in the pathogenesis of AIP.10Muraki T. Hamano H. Ochi Y. et al.Autoimmune pancreatitis and complement activation system.Pancreas. 2006; 32: 16-21Crossref PubMed Scopus (138) Google Scholar We compared the serum levels of C3 and C4 in patients with high serum IC and those with normal IC levels. Serum C4 was significantly lower and serum C3 tended to be lower in the high IC group compared with the normal IC group.10Muraki T. Hamano H. Ochi Y. et al.Autoimmune pancreatitis and complement activation system.Pancreas. 2006; 32: 16-21Crossref PubMed Scopus (138) Google Scholar These results, together with the IC data, suggested that a classical pathway might be operating in some AIP patients. In turn, this suggested that high serum IC might be useful for predicting both tissue damage and the probability of recurrence. A report by Kubota et al11Kubota K. Iida H. Fujisawa T. et al.Clinical factors predictive of spontaneous remission or relapse in cases of autoimmune pancreatitis.Gastrointest Endosc. 2007; 66: 1142-1151Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar showed that high serum IgG, diffuse pancreatic swelling, and low bile duct stenosis were frequently observed in patients who had relapsed. A logistic regression analysis showed that diffuse pancreatic swelling was a predictor of recurrence. Specific HLA polymorphisms were also reported to predict the recurrence of AIP and acted as primary determinants of autoimmune hepatitis susceptibility or relapse.12Park do H. Kim M.H. Oh H.B. et al.Substitution of aspartic acid at position 57 of the DQbeta1 affects relapse of autoimmune pancreatitis.Gastroenterology. 2008; 134: 440-446Abstract Full Text Full Text PDF PubMed Scopus (129) Google Scholar Furthermore, substitution of an aspartic acid at position 57 of the HLA designated DQβ1 (DQβ1 57) was reported to affect the recurrence of AIP. Thus, Park et al12Park do H. Kim M.H. Oh H.B. et al.Substitution of aspartic acid at position 57 of the DQbeta1 affects relapse of autoimmune pancreatitis.Gastroenterology. 2008; 134: 440-446Abstract Full Text Full Text PDF PubMed Scopus (129) Google Scholar examined associations between the onset of AIP recurrence and the density of nonaspartic acids at DQβ1 57. They observed in patients who experienced a recurrence that homozygosity of nonaspartic acids was associated with a significantly earlier onset of recurrence compared with heterozygosity of nonaspartic acids. The cytotoxic T-lymphocyte antigen 4 (CTLA4) polymorphism has been reported to be another predictor of AIP recurrence.13Umemura T. Ota M. Hamano H. et al.Association of autoimmune pancreatitis with cytotoxic T-lymphocyte antigen 4 gene polymorphisms in Japanese patients.Am J Gastroenterol. 2008; 103: 588-594Crossref PubMed Scopus (112) Google Scholar CTLA4 is an inhibitory receptor expressed on the cell surface of activated-memory T cells and on regulatory T cells; it acts largely as a negative regulator of T-cell responses by modulating positive T-cell costimulatory signals on antigen-presenting cells. Single nucleotide polymorphisms (SNPs) in CTLA4, termed +49A/G, have been specifically associated with susceptibility to autoimmune diseases, including type 1 diabetes, autoimmune thyroid disease, autoimmune hepatitis, and primary biliary cirrhosis. To investigate whether CRLA4 was associated with AIP pathogenesis in our cohort, we determined the presence of 4 CTLA4 gene SNPs in patients with AIP and healthy controls. We detected SNP +6230G/G significantly more frequently in patients with AIP than in healthy subjects. In addition, we found that the +49A/A and +6230A/A genotypes were associated with an enhanced risk of recurrence.13Umemura T. Ota M. Hamano H. et al.Association of autoimmune pancreatitis with cytotoxic T-lymphocyte antigen 4 gene polymorphisms in Japanese patients.Am J Gastroenterol. 2008; 103: 588-594Crossref PubMed Scopus (112) Google Scholar AIP is generally found in older individuals with suppressed immune surveillance systems who consequently have increased susceptibility to various malignant diseases. To date, among 51 AIP patients, we found 11 malignant lesions in 9 patients (17.6%); these included malignant lymphoma, lung cancer, hepatocellular carcinoma, renal carcinoma, breast cancer, duodenal cancer, colon cancer, prostate cancer, and ovarian cancer. These findings were consistent with other reports that found various malignant lesions complicated with AIP.6Kamisawa T. Okamoto A. Prognosis of autoimmune pancreatitis.J Gastroenterol. 2007; 42: 59-62Crossref PubMed Scopus (62) Google Scholar, 14Nishino T. Toki F. Oyama H. et al.Long-term outcome of autoimmune pancreatitis after oral prednisolone therapy.Intern Med. 2006; 45: 497-501Crossref PubMed Scopus (150) Google Scholar, 15Kim T. Grobmyer S.R. Dixon L.R. et al.Autoimmune pancreatitis and concurrent small lymphocytic lymphoma: not just a coincidence?.J Gastrointest Surg. 2008; 12: 1566-1570Crossref PubMed Scopus (22) Google Scholar, 16Oh H.C. Kim J.G. Kim J.W. et al.Early bile duct cancer in a background of sclerosing cholangitis and autoimmune pancreatitis.Intern Med. 2008; 47: 2025-2028Crossref PubMed Scopus (41) Google Scholar Some malignancies might occur during or after maintenance therapy with corticosteroid, suggesting that a steroid-induced immunosuppressive state could induce malignant lesions. Because we had no age-matched controls, we could not determine whether AIP represented a significantly higher risk for malignant diseases. Incidentally, it is important to differentiate between AIP and pancreatobiliary malignancies. Bile duct cancer was reported to be associated with AIP16Oh H.C. Kim J.G. Kim J.W. et al.Early bile duct cancer in a background of sclerosing cholangitis and autoimmune pancreatitis.Intern Med. 2008; 47: 2025-2028Crossref PubMed Scopus (41) Google Scholar; thus, the co-occurrence of bile duct cancer should be investigated, even with a confirmed diagnosis of AIP. Twelve cases of pancreatic cancer associated with AIP have been previously reported.17Inoue H. Miyatani H. Sawada Y. et al.A case of pancreas cancer with autoimmune pancreatitis.Pancreas. 2006; 33: 208-209Crossref PubMed Scopus (102) Google Scholar, 18Ghazale A. Chari S. Is autoimmune pancreatitis a risk factor for pancreatic cancer?.Pancreas. 2007; 35: 376Crossref PubMed Scopus (68) Google Scholar, 19Kubota K. Iida H. Fujisawa T. et al.Clinical factors predictive of spontaneous remission or relapse in cases of autoimmune pancreatitis.Gastrointest Endosc. 2007; 66: 1142-1151Abstract Full Text Full Text PDF PubMed Scopus (122) Google Scholar, 20Fukui T. Mitsuyama T. Takaoka M. et al.Pancreatic cancer associated with autoimmune pancreatitis in remission.Intern Med. 2008; 47: 151-155Crossref PubMed Scopus (91) Google Scholar, 21Witkiewicz A.K. Kennedy E.P. Kennyon L. et al.Synchronous autoimmune pancreatitis and infiltrating pancreatic ductal adenocarcinoma: case report and review of the literature.Hum Pathol. 2008; 39: 1548-1551Abstract Full Text Full Text PDF PubMed Scopus (100) Google Scholar Among the 12 cases, 5 occurred concurrently with AIP, and 7 occurred from 3–5 years after the diagnosis of AIP. It has been estimated that, generally, two thirds of pancreatic cancers occur in the head region. Surprisingly, 9 of the 12 tumors (75%) associated with AIP were located in the body and tail regions of the pancreas, including 3 in the head, 5 in the body, 3 in the tail, and 1 in the body and tail regions. Accordingly, the occurrence of a tumor in the body and tail of the pancreas might be a characteristic finding of pancreatic cancer as a complication of AIP.22Tanaka S. Yoshida H. Ikegami A. et al.Therapeutic procedure and prognosis in patients with autoimmune pancreatitis [Japanese].Suizo. 2007; 22: 663-671Crossref Google Scholar In addition to the immunosuppressive state, a chronic inflammatory process similar to ordinary chronic pancreatitis might evoke pancreatic malignancy.23Lowenfels A.B. Maisonneuve P. Cavallini G. et al.Pancreatitis and the risk of pancreatic cancer: International Pancreatitis Study Group.N Engl J Med. 1993; 328: 1433-1437Crossref PubMed Scopus (1499) Google Scholar We believe that, although the exact cancer prevalence is unknown, a careful follow-up with tumor markers is mandatory. In conclusion, our findings showed that 40% of patients with AIP had recurrences during long-term follow-up. Many patients had been treated with prednisolone at onset because of the highly active state of the disease. We observed that although the disease appeared to be clinically inactive during or after maintenance therapy, relapse was likely to occur, and that long-term therapy was required. To prevent clinical relapse, corticosteroid therapy should be restarted at an early subclinical stage of relapse, and reliable parameters must be identified for predicting relapse. Among several possible serum markers, we found that the IC-mRF value showed significant elevation in the recurrence group, suggesting that IC might be useful for predicting the recurrence of AIP. Furthermore, a normal IgG4 value represented the inactive state and might be an indicator for the cessation of maintenance corticosteroid therapy. Some patients with AIP exhibited complications with pancreatic stones after several attacks of recurrence. Thus, AIP appeared to transform into a form of chronic pancreatitis. AIP appears to be associated with a variety of malignant lesions. To date, 12 cases of pancreatic cancer associated with AIP have been reported, with a preponderance in body and tail regions of the pancreas. Although AIP should be carefully differentiated from pancreatic cancer, we recommend monitoring for the co-occurrence of pancreatic cancer even when a diagnosis of AIP is confirmed.