Abstract
Background:
Elevation in the serum immunoglobulin-G4 (IgG4) level has been used as a diagnostic marker to distinguish autoimmune pancreatitis (AIP) from pancreatic ductal adenocarcinoma (PDAC), but its true utility is ill-defined. This study evaluates the clinical utility of IgG4 in differentiating AIP from PDAC.
Methods:
All patients evaluated in the hepatopancreaticobiliary surgery clinics with measured serum IgG4 were included. Patients were divided into normal IgG4 (<135mg/dL) and elevated IgG4 (≥135mg/dL) groups. The final diagnosis was determined by operative pathology when available or by clinical outcome. The sensitivity, specificity, PPV, and NPV of IgG4 for diagnosing AIP was assessed.
Results:
Between 1997 and 2015, 298 patients were identified. Normal IgG4 levels were present in 85% of patients (254/298), while 15% (44/298) were elevated. The overall prevalence of AIP was 17% (52/298). The sensitivity and specificity of IgG4 for AIP was 67% and 96%, respectively; however, the PPV was only 80%, including a 9% occurrence of PDAC in patients with an elevated IgG4.
Conclusion:
In this study of selected patients who underwent IgG4 testing, 9% of elevated IgG4 patients had PDAC. The overreliance on IgG4 as diagnostic for AIP may lead to mis-diagnosis and delayed treatment for PDAC.
INTRODUCTION
Autoimmune pancreatitis (AIP) is an uncommon benign disease of the pancreas that can mimic pancreatic ductal adenocarcinoma (PDAC). Elevation in the serum immunoglobulin-G4 (IgG4) level has been used as a diagnostic marker of AIP, and has been recommended as a means to differentiate AIP from PDAC, which is not associated with elevated IgG4 levels(1, 2). The cut-off value used to define an elevated IgG4 level varies across different consensus groups. The 2013 Japanese consensus criteria used a cut-off value of ≥135 mg/dL, which has a sensitivity and specificity of 73-95% and 94%-97%, respectively, for identifying AIP(1, 3–5). In contrast, the 2011 international consensus diagnostic criteria (ICDC) advocated a cut-off of twice the upper limit of level (≥280 mg/dL), which has been reported to have a sensitivity and specificity of 53-81% and 90-99%, respectively(2, 6). Further confounding the picture, approximately 10% of AIP patients have normal IgG4 levels and are sub-classified as type 2 AIP, which has similar histologic manifestations to type I AIP of immune complex-mediated pancreatic duct and acini destruction but has no elevation in serological biomarkers(7–9). Type 2 AIP is more commonly reported in Western countries while type I AIP is predominantly found in Asian countries(10–12).
While significant advances in the understanding of AIP have occurred over the past 15 years, the wide range of clinical manifestations and the lack of any specific pathognomic characteristics continue to make AIP an extremely difficult entity to diagnose. The most important disease to distinguish from AIP is PDAC, as AIP can mimic a similar presentation of a mass-like lesion in the pancreatic head with obstruction of the biliary and/or pancreatic ductal system. IgG4 elevation has been proposed as a potential differentiating characteristic between AIP and PDAC as the latter has been rarely reported in the setting of elevated IgG4(3, 6, 13).
The aim of this study is to evaluate the clinical utility of IgG4 in diagnosing AIP and in differentiating AIP from PDAC.
METHODS
Patients were identified from the Memorial Sloan-Kettering Cancer Center (MSKCC) institutional database who were evaluated in the hepatopancreaticobiliary surgery clinics for pancreatic pathologies and had a measured serum IgG4 level. Searches included results from 1997 through 2015. Patients were excluded from the study if they had IgG4 levels evaluated for pathologies unrelated to the pancreas. We retrospectively identified 337 patients. Patients who did not return to MSKCC after their initial evaluation and thus did not have a definitive final diagnosis were excluded from the analysis (n=39). This study was approved by the MSKCC Institutional Review Board.
Patient demographics, presenting symptoms, serologic, radiographic, and pathologic results, and follow-up data were collected. The diagnosis was confirmed pathologically in the majority of patients using immunohistochemical staining to demonstrate increased IgG4-positive plasma cells. If a pathologic diagnosis was not available, a clinical diagnosis was made using all available serologic, radiographic, and steroid treatment response data. Data analysis was performed using IBM SPSS version 22 and Microsoft Excel 2007. Patients were divided into two groups for analysis, those with normal serum IgG4 (<135 mg/dL) and those with elevated IgG4 (≥135 mg/dL). Chi-squared test was used to compare categorical variables and Mann-Whitney U test to compare continuous variables between these two patient groups. A p-value <.05 was considered significant. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of IgG4 elevation for diagnosing AIP was assessed.
RESULTS
Between 1997 and 2015 (18 years), 298 patients were evaluated by the hepatopancreaticobiliary service at MSKCC and had serum IgG4 levels measured. Most patients were male (182/298, 61%), and the median age at initial evaluation was 65 years old (range 28-89 years). Nearly all patients (268/298, 90%) were symptomatic at presentation while 10% (30/298) had incidentally-noted pancreatic imaging abnormalities. The most common presenting symptoms included abdominal pain, weight loss, jaundice, and elevated liver function tests. The presence of a pancreatic mass was the most frequently reported radiographic abnormality (229/298, 77%), with associated pancreatic ductal dilation in 46% (136/298). On initial imaging, 37 patients (12%) were noted to have loss of normal pancreatic contour and 6 (2%) had a halo of inflammation surrounding the pancreas. Presence of a pancreatic mass and pancreatic ductal dilation were more common in patients with normal IgG4 (p=.003, p<.001), while loss of normal pancreatic contour and a halo of inflammation around the pancreas were more common in patients with elevated IgG4 (p<.001). Diabetes mellitus was present in 24% (72/298), and 6% of patients (19/298) had a history of other autoimmune disorders at initial presentation. Both history of diabetes and other autoimmune diseases were more common in patients with elevated IgG4 levels (diabetes: 39% vs. 22%, p = .015, other autoimmune disorders: 20% vs. 4%, p = .001). Other autoimmune disorders included ulcerative colitis (n=4), Crohn’s disease (n=2), systemic lupus erythematous (n=2), multiple sclerosis (n=2), immune-mediated thrombocytopenia (n=1), and Hashimoto’s thyroiditis (n=1).
The majority of patients (254/298, 85%) had IgG4 levels within normal limits (median 30 mg/dL, range 0-130 mg/dL), and 44 patients (15%) had elevated IgG4 levels (median 284 mg/dL, range 135-2300 mg/dL). The median pre-treatment cancer antigen 19-9 (CA 19-9) of all patients was 59 units/mL (range 0-17317 units/mL). There was no difference in pre-treatment CA 19-9 levels between patients with elevated IgG4 levels compared with patients with normal IgG4 levels (median 56 vs. 59 units/mL, p = .401). A comparison of patient characteristics and presentations between those with normal IgG4 levels and those with elevated IgG4 levels is presented in Table 1.
Table 1.
Patient characteristics and presentation.
| All N = 298 |
IgG4<135 N = 254 |
IgG4≥135 N = 44 |
P-value | |
|---|---|---|---|---|
| Gender | ||||
| Male | 182 (61%) | 145 (57%) | 37 | .001 |
| Age, years, median (range) | 65(28-89) | 64 (28-89) | 67 (39-81) | .149 |
| DM at initial visit | 72 (24%) | 55 (22%) | 17 | .015 |
| History of autoimmune disorders | 19 (6%) | 11 (4%) | 8 | .001 |
| Symptoms | ||||
| Abdominal pain | 135 (45%) | 117 (46%) | 18 | |
| Back pain | 39 (13%) | 36 (14%) | 3 | |
| Anorexia | 38 (13%) | 32 (13%) | 6 | |
| Weight loss | 123 (41%) | 99 (39%) | 24 | |
| Weight loss, lbs, median (range) | 15 (3-80) | 15 (3-80) | 17 (4-35) | |
| Jaundice | 122 (41%) | 95 (37%) | 27 | |
| Acute pancreatitis | 38 (13%) | 33 (13%) | 5 | |
| Chronic pancreatitis | 2 (1%) | 1 (.3%) | 1 | |
| New or worsening DM | 19 (6%) | 15 (6%) | 4 | |
| Elevated LFTs | 117 (39%) | 100 (39%) | 17 | |
| None | 30 (10%) | 29 (11%) | 1 | |
| Radiographic findings | ||||
| Pancreatic mass | 202 (80%) | 26 | .003 | |
| Loss of pancreatic contour | 24 (9%) | 13 | <.001 | |
| Halo of inflammation | 1 (.4%) | 5 | <.001 | |
| Pancreatic duct dilation | 128 (50%) | 8 | <.001 | |
| Suspected AIP at initial visit | 29 (10%) | 21 (8%) | 8 | |
| IgG4, mg/dL, median (range) | 34 (0-2300) | 30 (0-130) | 284 (135-2300) | |
| Max pre-treatment CA 19-9, units/mL, median (range) | 59 (0-17317)* | 59 (0-17317) | 56 (2-809) | .401 |
IgG4 indicates immunoglobulin G4; DM, diabetes mellitus; lbs, pounds; LFTs, elevated liver function tests; CA 19-9, cancer antigen 19-9.
CA 19-9 level was not available in 40 patients.
Pre-operative biopsies or brushings were obtained in 63% of patients (187/298) [Table 2]. Patients with an initial high-index of suspicion for PDAC or AIP did not undergo tissue sampling and proceeded directly to resection or underwent a steroid trial. Of the patients who underwent biopsies or brushings, malignant pathology was found in 42% (80/189), atypical or suspicious pathology in 32% (60/189), and no evidence of malignancy in 20% (38/189). All three patients who had biopsies or endoscopic ultrasound guided fine needle aspirations (EUS-FNA) consistent with AIP had a final diagnosis of AIP; these three patients also had elevated IgG4 levels as well (range 166-366 mg/dL). Two patients had pre-operative tissue sampling demonstrating malignancy and were only diagnosed with AIP in the operating room, one after a pancreatic biopsy, another after a pancreaticoduodenectomy. However, the majority of patients who demonstrated malignancy on pre-operative pathology had a final diagnosis of pancreatic cancer (78/80, 98%). Among the 60 patients with atypical or suspicious biopsy findings, there was a wide spectrum of final diagnoses, ranging from AIP to PDAC to pancreatitis of unknown etiology. However, patients with both atypical or suspicious biopsy findings and elevated IgG4 levels were more likely to have AIP (13/15) than PDAC, compared to those with atypical or suspicious biopsies and normal IgG4 (7/45). The remaining 38 patients had negative benign findings only on tissue sampling, of which nine had a final diagnosis of AIP and 16 had a final diagnosis of PDAC. Similar to those with atypical or suspicious biopsy findings, patients with negative biopsy findings and elevated IgG4 levels were more likely to have AIP (7/9) than PDAC, compared to those with normal IgG4 levels (2/28).
Table 2.
Patient treatment and final diagnoses.
| All N = 298 |
IgG4<135 N = 254 |
IgG4≥135 N = 44 |
|
|---|---|---|---|
| Pre-operative tissue diagnosis | 189 (63%) | 160 (63%) | 29 |
| AIP | 3 (2%) | 0 | 3 |
| Malignant | 80 (42%) | 79 (49%) | 1 |
| Atypical/Suspicious | 60 (32%) | 45 (18%) | 15 |
| Negative | 38 (20%) | 29 (11%) | 9 |
| Other | 8 (4%) | 7 3%) | 1 |
| Operation | 187 (63%) | 165 (65%) | 22 |
| Biopsy only | 43 (23%) | 40 (24%) | 3 |
| Bypass only | 24 (13%) | 19 (12%) | 5 |
| Pancreaticoduodenectomy | 96 (51%) | 83 (50%) | 13 |
| Distal pancreatectomy | 23 (12%) | 22 (13%) | 1 |
| Central pancreatectomy | 1 (.5%) | 1 (.6%) | 0 |
| Operative tissue diagnosis | |||
| AIP | 28 (15%) | 11 (7%) | 17 |
| PDAC | 109 (58%) | 106 (64%) | 3 |
| Other | 50 (27%) | 48 (29%) | 2 |
| Final diagnosis | |||
| AIP | 52 (17%) | 17 (7%) | 35 |
| PDAC | 170 (57%) | 166 (65%) | 4 |
| Other | 76 (26%) | 71 (28%) | 5 |
| Followup, months, median (range) | 18 (1-211) | 17 (1-172) | 37 (1-211) |
IgG4 indicates immunoglobulin G4; AIP, autoimmune pancreatitis; PDAC, pancreatic adenocarcinoma.
A steroid trial was administered as the first course of treatment in 19 patients (6.4%). These patients were treated with prednisone 40mg daily and re-evaluated every four to six weeks, followed by a gradual steroid taper. Fourteen of these patients were subsequently followed with serial serum IgG4 levels, of whom 12 demonstrated a decrease in IgG4 levels as well as improvement or resolution of symptoms. One patient continued to have elevated IgG4 levels (range 204-231 mg/dL) but demonstrated symptomatic and radiographic improvement with steroid treatment. Another patient had normal IgG4 levels at presentation and also had symptomatic improvement with steroids; this patient was also evaluated with EUS-obtained brushings and FNA that were both benign. Of the remaining five patients who were treated initially with steroids, serial IgG4 monitoring was not performed but all five had improved or complete resolution of symptoms. Repeat imaging was performed in three patients who had downtrending IgG4 levels that failed to normalize with steroid treatment and showed evidence of radiographic improvement, including restoration of normal pancreatic control and improvement of peripancreatic inflammation.
Resection was subsequently attempted in 187 patients (63%) [Table 2]. Resection was successful in 120 patients (64%); bypass (24/187, 13%) or open biopsy only (43/187, 23%) was performed in patients who were deemed unresectable in the operating room due to extensive vessel involvement or significant surrounding inflammation or fibrosis. Pancreaticoduodenectomy was the most common procedure (96/187, 51%), followed by distal pancreatectomy in 23 patients (12%), and then central pancreatectomy in one patient (.5%). The majority of patients with normal IgG4 levels proceeded to resection (165/254, 65%), compared with only half of patients (22/44) with elevated IgG4 levels who proceeded to resection [Figure 1]. Within the group of patients with normal IgG4 levels who underwent an operation, PDAC was the most common pathology, found in 106 patients (64%), while 11 patients (7%) were found to have pathology consistent with AIP. Other notable pathologies included 19 non-autoimmune pancreatitis, 18 pancreatic neuroendocrine tumors, eight intraductal papillary mucinous neoplasms, one mesenchymal neoplasm, one pseudo-papillary tumor, and one serous cystadenoma. Within the group of patients with elevated IgG4 levels who underwent an operation, 17 patients had pathology consistent with AIP, three patients had PDAC, and two patients had pancreatic neuroendocrine tumors. In total, there were 28 AIP patients who underwent an operation. All but one were presumed to have PDAC based on pre-operative imaging, biopsies, and clinical evaluation. One patient was known to have AIP and underwent resection for a presumed mucinous cystic neoplasm; however, pathology only demonstrated findings consistent with AIP.
Figure 1.
Patient flow chart by resection status and final diagnosis.
Overall, based on a complete clinical, radiographic, and pathologic evaluation, there were 170 patients (57%) diagnosed with PDAC, 52 patients (17%) diagnosed with AIP, and 76 patients (26%) diagnosed with a spectrum of other benign and malignant pancreatic pathologies. Within the group of patients with normal IgG4 levels, 65% (166/254) were diagnosed with PDAC and 7% (17/254) with AIP. Within the group of 44 patients with elevated IgG4 levels, 4 patients were diagnosed with PDAC and 35 with AIP. Of note, the median CA 19-9 level of patients was a final diagnosis of AIP was 59 units/mL (range 0-809 units/mL), including 12 patients with CA 19-9>100 units/mL.
The sensitivity and specificity of an elevated IgG4 level for diagnosing AIP was 67% and 96%, respectively [Table 3]. While the NPV was 93%, the PPV was only 80%, including a 9% (4/44) occurrence of PDAC in patients with an elevated IgG4 level. All four patients were male, with age of presentation ranging from 48 to 74 years old, and had a similar constellation of presenting symptoms including abdominal pain, weight loss, and jaundice. Three patients had a pancreatic mass visible on cross-sectional imaging; one patient did not have a discrete mass identified but did have dilation of the pancreatic duct. None of these patients had any history of other autoimmune disorders. The IgG4 levels ranged from 136 to 160 mg/dL. Two patients had elevated CA 19-9 levels (260 units/mL and 415 units/mL), while two had normal CA 19-9 levels (2 units/mL and 15 units/mL). All four patients had pre-operative brushings or biopsies; one patient had negative brushings only, two had atypical findings, and one demonstrated malignancy. The patient with malignant findings had unresectable disease and proceeded to chemotherapy. The three patients with negative or atypical pre-operative findings proceeded to resection, where a diagnosis of PDAC was confirmed on operative pathology.
Table 3.
Diagnostic performance of differing IgG4 cut-off levels for AIP
| IgG4 Cut-off | Sensitivity | Specificity | PPV | NPV | Number of cases with elevated IgG4 and PDAC |
|---|---|---|---|---|---|
| IgG4 ≥135 mg/dL | 67% | 96% | 80% | 93% | 4 |
| IgG4 ≥280 mg/dL | 42% | 100% | 96% | 96% | 0 |
AIP, autoimmune pancreatitis; IgG4 indicates immunoglobulin G4; PPV, positive predictive value; NPV, negative predictive value; PDAC, pancreatic adenocarcinoma
If the cut-off IgG4 level was raised from 135 to 280 mg/dL, as proposed by the ICDC guidelines, all four false-positive PDAC patients became negative, and 22 of the 23 patients with IgG4 levels ≥280 mg/dL had a final diagnosis of AIP and the remaining patient had a final diagnosis of pancreatitis of unspecified etiology. This significantly raises the PPV of IgG4 ≥280 mg/dL for diagnosing AIP to 96%. The specificity remains high at 99.6%; however, the sensitivity decreases to 42% [Table 3]. This reduced sensitivity increases the number of patients with a missed diagnosis of AIP.
DISCUSSION
Autoimmune pancreatitis is an uncommon diagnosis that can be very challenging to diagnose non-operatively(14–16). Clinically and radiographically, AIP can mimic PDAC, particularly when AIP presents as a mass-like lesion in the pancreatic head with symptoms of abdominal pain, weight loss, and jaundice(16–18). In many cases of diagnostic uncertainty, patients will undergo a malignancy workup and AIP becomes a diagnosis of exclusion once no definitive evidence of malignancy has been found(19).
Serum IgG4 levels have been proposed as a method of differentiating between AIP and PDAC, with elevated IgG4 levels having been reported to be associated with AIP(1, 2). However, the definition of an elevated IgG4 level remains an area of dispute, with some groups advocating for the use of ≥135 mg/dL as the cut-off, while others propose ≥280 mg/dL as the cut-off(4, 5, 20). The spectrum of cut-off values has demonstrated similar sensitivity and specificity for diagnosing AIP when validated across various patient populations. In the present study, we used 135 mg/dL as our cut-off value, which resulted in a sensitivity of 67% and specificity of 96% for diagnosing AIP, similar to previous studies(3). However, we also found that 9% of patients (4/44) who presented with elevated IgG4 levels in fact had PDAC, two of whom also had normal CA 19-9 levels.
Our rate of PDAC patients with elevated IgG4 is comparable to previous reports. In a study comparing 17 AIP patients presenting with a mass-like lesion in the pancreatic head and 31 patients with locally advanced pancreatic head adenocarcinoma, Kamisawa et al identified 14 patients (29%) with elevated IgG4 levels (≥135 mg/dL), of which two patients had PDAC, resulting in an overall rate of 6% of PDAC patients having elevated IgG4 levels(17). Chang et al identified a much higher percentage of PDAC among patients with elevated IgG4 levels(21). Using a cut-off of 140 mg/dL, 17% of patients with elevated IgG4 levels had PDAC (18/109). Even when the cut-off was increased to 280 mg/dL, as proposed by the ICDC guidelines, 10% of elevated IgG4 patients were still found to have PDAC (5/49). In a much larger series comprised of 548 PDAC patients with measured serum IgG4 levels, 10% of PDAC patients had elevated IgG4 levels when the upper limit of normal (121 mg/dL or 140 mg/dL depending on the laboratory standard) was used as the cut-off, compared with only 2.4% when twice the upper limit of normal was used as the cut-off(13).
In our study, if the cut-off IgG4 level was raised to 280 mg/dL, all four false-positive PDAC patients became negative, significantly raising the PPV of IgG4 ≥280 mg/dL for diagnosing AIP to 96%. However, this comes at the cost of reducing the sensitivity to 42%. As a result, more patients with a true diagnosis of AIP may be missed, potentially leading to an increased number of AIP patients undergoing unnecessary resections. Rather than using an absolute cut-off value, borderline elevated IgG4 levels should be interpreted with caution. None of the four false positive PDAC patients in our study had IgG4 levels elevated above 160 mg/dL. In these borderline elevated cases, further evaluation, such as with a thorough radiographic evaluation and core biopsy, should be pursued and AIP should remain a diagnosis of exclusion only after malignancy has been definitively ruled out.
For study patients who had pre-operative tissue sampling with EUS, if the sample demonstrated either AIP or malignancy, this was 98% concordant with the patient’s final diagnosis. However, one-third of patients (60/189, 32%) had atypical or suspicious findings and this was far less indicative of the patients’ final diagnoses. Even among patients with negative findings on tissue sampling, nine patients had a final diagnosis of AIP and 16 patients had a final diagnosis of PDAC. This demonstrates that EUS-obtained biopsies or brushings were a useful aid only in cases when AIP or malignancy was identified. However, this was only the case in less than half the patients who underwent EUS-guided biopsies or brushings (81/187). In the remaining cases where patients had a non-definitive or benign diagnosis on EUS-guided sampling, this was of limited diagnostic utility. In a prospective series of 78 patients with imaging characteristics concerning for AIP, Kanno et al applied the ICDC diagnostic histology criteria use for pancreatic core biopsies/resection specimens to EUS-FNA samples and found that 58% (45/78) met the ICDC criteria for AIP. However, even within this highly-selected patients for whom there was an initial high index of suspicious for AIP, 42% had non-diagnosis EUS-FNA findings(22)..Both the ICDC and Mayo HISORt criteria for AIP require either core biopsy or surgical specimens for a histologic diagnosis, and aspiration cytology by EUS alone is considered inadequate for a definitive diagnosis of AIP(2, 20). As endoscopic devices and techniques continue to evolve, the potential for using EUS-FNA for the diagnosis of AIP and differentiation from PDAC can be re-evaluated. However, until the diagnostic accuracy of EUS-FNA is established, a core pancreatic biopsy should be obtained in all patients undergoing diagnostic tissue sampling.
Limitations of this study include the inherent selection bias associated with a retrospective review and analysis of patients from a single institution. Our institution is also a cancer-specialized hospital so these findings may not be generalized to all institutions. There was also no standardized algorithm to determine for which patients serum IgG4 levels should be drawn; however, this appropriately reflects the real-life heterogeneity in patient presentation and diagnoses seen by a pancreatic specialist.
CONCLUSION
In conclusion, IgG4 elevation remains a highly specific test for AIP; however, a small but significant number of PDAC patients will also have elevated IgG4 levels. Particularly in borderline-elevated cases, the IgG4 level needs to be treated with caution as an overreliance on it as a diagnostic marker for AIP may lead to delayed and potentially missed identification of PDAC.
Acknowledgments
Support: Linda M. Pak, MD, was supported by the Clinical and Translation Science Center at Weill Cornell Medical Center and MSKCC Award Number UL1TR00457. This work was also supported in part by NIH/NCI P30 CA008748 Cancer Center Support Grant.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Previous communication: This study was presented as a Long Oral presentation at the 2017 Annual Meeting of the Americas Hepato-Pancreato-Biliary Association in Miami, FL on April 2, 2017.
REFERENCES
- 1.Hamano H, Kawa S, Horiuchi A, Unno H, Furuya N, Akamatsu T, et al. High serum IgG4 concentrations in patients with sclerosing pancreatitis. The New England journal of medicine. 2001;344(10):732–8. [DOI] [PubMed] [Google Scholar]
- 2.Chari ST, Takahashi N, Levy MJ, Smyrk TC, Clain JE, Pearson RK, et al. A diagnostic strategy to distinguish autoimmune pancreatitis from pancreatic cancer. Clinical gastroenterology and hepatology : the official clinical practice journal of the American Gastroenterological Association. 2009;7(10):1097–103. [DOI] [PubMed] [Google Scholar]
- 3.Choi EK, Kim MH, Lee TY, Kwon S, Oh HC, Hwang CY, 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(2):156–61. [DOI] [PubMed] [Google Scholar]
- 4.Okazaki K, Kawa S, Kamisawa T, Naruse S, Tanaka S, Nishimori I, et al. Clinical diagnostic criteria of autoimmune pancreatitis: revised proposal. Journal of gastroenterology. 2006;41(7):626–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Okazaki K, Kawa S, Kamisawa T, Ito T, Inui K, Irie H, et al. Amendment of the Japanese Consensus Guidelines for Autoimmune Pancreatitis, 2013 I. Concept and diagnosis of autoimmune pancreatitis. Journal of gastroenterology. 2014;49(4):567–88. [DOI] [PubMed] [Google Scholar]
- 6.Ghazale A, Chari ST, Smyrk TC, Levy MJ, Topazian MD, Takahashi N, et al. Value of serum IgG4 in the diagnosis of autoimmune pancreatitis and in distinguishing it from pancreatic cancer. The American journal of gastroenterology. 2007;102(8):1646–53. [DOI] [PubMed] [Google Scholar]
- 7.Okazaki K, Uchida K. Autoimmune Pancreatitis: The Past, Present, and Future. Pancreas. 2015;44(7):1006–16. [DOI] [PubMed] [Google Scholar]
- 8.Hart PA, Kamisawa T, Brugge WR, Chung JB, Culver EL, Czako L, et al. Long-term outcomes of autoimmune pancreatitis: a multicentre, international analysis. Gut. 2013;62(12):1771–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kanno A, Masamune A, Okazaki K, Kamisawa T, Kawa S, Nishimori I, et al. Nationwide epidemiological survey of autoimmune pancreatitis in Japan in 2011. Pancreas. 2015;44(4):535–9. [DOI] [PubMed] [Google Scholar]
- 10.Hart PA, Zen Y, Chari ST. Recent Advances in Autoimmune Pancreatitis. Gastroenterology. 2015;149( 1):39–51. [DOI] [PubMed] [Google Scholar]
- 11.Notohara K, Burgart LJ, Yadav D, Chari S, Smyrk TC. Idiopathic chronic pancreatitis with periductal lymphoplasmacytic infiltration: clinicopathologic features of 35 cases. The American journal of surgical pathology. 2003;27(8):1119–27. [DOI] [PubMed] [Google Scholar]
- 12.Kloppel G, Detlefsen S, Chari ST, Longnecker DS, Zamboni G. Autoimmune pancreatitis: the clinicopathological characteristics of the subtype with granulocytic epithelial lesions. Journal of gastroenterology. 2010;45(8):787–93. [DOI] [PubMed] [Google Scholar]
- 13.Ngwa T, Law R, Hart P, Smyrk TC, Chari ST. Serum IgG4 elevation in pancreatic cancer: diagnostic and prognostic significance and association with autoimmune pancreatitis. Pancreas. 2015;44(4):557–60. [DOI] [PubMed] [Google Scholar]
- 14.Kwon S, Kim MH, Choi EK. The diagnostic criteria for autoimmune chronic pancreatitis: it is time to make a consensus. Pancreas. 2007;34(3):279–86. [DOI] [PubMed] [Google Scholar]
- 15.van Heerde MJ, Buijs J, Rauws EA, de Buy Wenniger LJ, Hansen BE, Biermann K, et al. A comparative study of diagnostic scoring systems for autoimmune pancreatitis. Pancreas. 2014;43(4):559–64. [DOI] [PubMed] [Google Scholar]
- 16.Hoshimoto S, Aiura K, Tanaka M, Shito M, Kakefuda T, Sugiura H. Mass-forming type 1 autoimmune pancreatitis mimicking pancreatic cancer. Journal of digestive diseases. 2016;17(3):202–9. [DOI] [PubMed] [Google Scholar]
- 17.Kamisawa T, Imai M, Yui Chen P, Tu Y, Egawa N, Tsuruta K, et al. Strategy for differentiating autoimmune pancreatitis from pancreatic cancer. Pancreas. 2008;37(3):e62–7. [DOI] [PubMed] [Google Scholar]
- 18.Cao Z, Tian R, Zhang T, Zhao Y. Localized Autoimmune Pancreatitis: Report of a Case Clinically Mimicking Pancreatic Cancer and a Literature Review. Medicine. 2015;94(42):e1656. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Finkelberg DL, Sahani D, Deshpande V, Brugge WR. Autoimmune pancreatitis. The New England journal of medicine. 2006;355(25):2670–6. [DOI] [PubMed] [Google Scholar]
- 20.Shimosegawa T, Chari ST, Frulloni L, Kamisawa T, Kawa S, Mino-Kenudson M, et al. International consensus diagnostic criteria for autoimmune pancreatitis: guidelines of the International Association of Pancreatology. Pancreas. 2011;40(3):352–8. [DOI] [PubMed] [Google Scholar]
- 21.Chang MC, Liang PC, Jan S, Yang CY, Tien YW, Wei SC, et al. Increase diagnostic accuracy in differentiating focal type autoimmune pancreatitis from pancreatic cancer with combined serum IgG4 and CA19-9 levels. Pancreatology : official journal of the International Association of Pancreatology. 2014;14(5):366–72. [DOI] [PubMed] [Google Scholar]
- 22.Kanno A, Masamune A, Fujishima F, Iwashita T, Kodama Y, Katanuma A, et al. Diagnosis of autoimmune pancreatitis by EUS-guided FNA using a 22-gauge needle: a prospective multicenter study. Gastrointestinal endoscopy. 2016;84(5):797–804.e1. [DOI] [PubMed] [Google Scholar]