Article Text
Statistics from Altmetric.com
- Endoscopic retrograde cholangio pancreatography
- IgG4
- autoimmune pancreatitis
- cholangitis
- differential diagnosis
- autoimmune disease
- imaging
- immunoregulation
- pancreatitis
Autoimmune pancreatitis (AIP) is a puzzling disease. Although its cardinal features were first reported in Europe,1 it seemed, for a long time an Asian phenomenon which many Western clinicians felt they could largely ignore.2 3 As more case series were reported from Europe and America it became increasingly clear that AIP affects patients from many ethnic backgrounds. AIP shares features with two other disorders of the pancreas from which a distinction is critical and determines appropriate treatment. The first is chronic pancreatitis of either the environmentally induced (alcohol, tobacco smoke) or the hereditary variety.4–6 The most important difference in terms of treatment, prognosis and, to a lesser degree, diagnosis is that AIP rapidly responds to the administration of steroids7 and other varieties of pancreatitis do not.
Reports from Japan suggesting that AIP can be distinguished from chronic pancreatitis by serological markers alone, most prominently serum IgG4 levels,8 were greeted with great enthusiasm. Unfortunately, clinical chemistry was quickly found to be much less reliable in Caucasian patients.9 The reason behind this difference lies in two subtypes of AIP with different prevalence in Europe and Asia. The first subtype, and by far the most common in Asia, has recently been termed lymphoplasmacytic sclerosing pancreatitis (LPSP or type I AIP) according to its histological features.10 It is commonly associated with immunological changes such as elevated IgG4 serum levels or various autoantibodies of lesser diagnostic value. A second disease variety named idiopathic duct-centric pancreatitis (IDCP or type II AIP), that barely exists in Japan, accounts for a significant number of Caucasian patients, displays often none of the immunological changes and is characterised histologically by granulocytic epithelial lesions.11 This limits diagnostic options for type II AIP to either histology from resection specimens, core biopsies obtained by endoscopic ultrasound, or cross sectional diagnostic imaging. The features of diagnostic imaging, however, are not nearly as well defined or universally established as one would wish.
This brings us to the second disease AIP needs to be distinguished from which is pancreatic cancer. The characteristic imaging appearance of AIP has been reported to include diffuse swelling of the entire organ (the ‘sausage shaped’ pancreas) and a diffuse narrowing of the pancreatic duct, often combined with similar changes in the bile ducts when systemic IgG4-associated autoimmune disease is present.12 The latter appearance can easily mimic the ‘double duct’ sign of pancreatic cancer.13 When inflammatory infiltrates manifest themselves as focal pancreatic enlargement a distinction between cancer and AIP becomes all the more difficult. A focal enlargement was present in no less than 40% of AIP patients in the trial discussed below. This difficulty in distinguishing between pancreatic cancer and AIP—particularly of the IDCP, type II or European variety—remains the principal reason why many patients with AIP still undergo unnecessary pancreatic resection, only to learn after histological examination that surgery was neither required nor will it cure their disease. Until specific immunological tests become available for type II AIP antigens such as UBR214 or trypsin,15 16 or until better imaging modalities will permit a safer distinction between pancreatitis and cancer the surgical resection rate in patients with AIP will probably not decline.
With the latter solution in mind, Sugumar and co-workers17 started their study published in this issue of Gut (see page 666). They assembled a panel of experts from Asia, Europe and the USA and made them read endoscopic retrograde pancreatography (ERP) films of confirmed cases of AIP (n=20), chronic pancreatitis (n=10) and pancreatic cancer (n=10). The first phase of the study resulted in a rather poor sensitivity for ductal changes, an acceptable specificity of 92%, and a dismal inter-observer agreement between experts. Phase I of the trial could, however, identify four fairly reliable features for differentiating cancer from AIP. The first was the length of the stricture in the pancreatic duct. Long strictures involving more than one third of the duct length, strictures that did not result in an upstream dilatation of the duct, or strictures from which side branches arose were more likely to be caused by AIP than by cancer. The fourth characteristic was the presence of multiple strictures in the duct.
The authors went on to generate a teaching module for ERP reading, trained the same 21 experts and asked them to reevaluate the same ERP films 3 months later in a blinded fashion. Apparently the teaching model had improved sensitivity to 71% without compromising specificity but the inter-observer agreement remained poor (0.40).
A number of important points can be concluded not only from the study results but also from its limitations:
We do not learn from the study what the bile ducts of included patients looked like. Most patients in the study must have undergone endoscopic retrograde cholangio-pancreatography (ERCP) because 71% had presented with painless jaundice. Viewing ERC films side by side with corresponding ERP films may have increased the sensitivity, specificity and inter-observer agreement greatly. It does not reflect clinical practice to limit the evaluating physician to seeing only the latter.
The study shows that Asian physicians have much more expertise in diagnosing AIP than European or US-American pancreas experts. One explanation is that the incidence of AIP in Japan and Korea is between ten- and 100-fold higher. Another reason is that type I AIP, the variety most often diagnosed using IgG4 levels and without need for histology, accounts for >95% of patients in Asia. Asian physicians therefore have a diagnostic test at hand to immediately confirm or refute their ERCP interpretation. In the present study 30% of patients were included without diagnostic IgG4 levels. Unfortunately, we do not learn from the paper how many of these were recruited before IgG4 became the marker of choice of type I AIP or whether they all represented type II (IDCP) cases.
The current paradox is the following. The Japanese consensus guidelines have made ERCP a mandatory diagnostic criterion although an alternative test (serum IgG4) would be diagnostic in the majority of Japanese patients.18 Conversely, the HISORt criteria from the Mayo Clinic,19 derived from experience with mostly Caucasian patients, do not include ERCP as a mandatory diagnostic test although Caucasian patients with AIP often present without diagnostic lab tests and are much more likely to profit from diagnostic ERCP because core biopsy is the only alternative test for confirming the diagnosis (short of obtaining resection specimens) in the prevalent type II AIP. Not surprisingly, a study with mostly Western patients relied for differentiation between cancer and AIP on a protocol that included ERCP20 whereas a Japanese study addressing the same issue made do without it.21 What remains to be addressed in future studies is how well the ERP criteria of Sugumar et al work specifically in type II autoimmune pancreatitis (IDCP) because this is the subgroup which potentially benefits the most from diagnostic ERCP and in which ERCP may prevent most of the unnecessary pancreatic resections. The studies also need to clarify whether concerns about post-ERCP pancreatitis in patients with suspected AIP are unfounded.
The good news of the study is that non-Asian pancreas experts can greatly improve their skills in reading ERP-films if given an appropriate teaching module. Moreover, ER(C)P was confirmed as a powerful tool to diagnose AIP and to distinguish it from either chronic pancreatitis or pancreatic cancer. The study supports the Japanese consensus which strongly recommends ERCP for the diagnosis of AIP18 and argues against the Mayo view19 which does not make it a requirement. If the ductal changes reported here can be confirmed in their specificity for type II autoimmune pancreatitis, as well as the superiority of ERCP over MRCP in detecting them, then AIP will side with primary sclerosing cholangitis as a condition in which diagnostic ERCP cannot yet be replaced by less invasive diagnostic test. Whether ERCP can serve in a similar role for monitoring treatment response in AIP will also have to be addressed in future trials.
References
Footnotes
Linked articles 207951.
Funding The author's own work is supported by the Alfried-Krupp-von-Bohlen-und-Hallbach-Foundation (Graduate Schools Tumour Biology and Free Radical Biology), the Deutsche Krebshilfe/ Dr. Mildred-Scheel-Stiftung (109102), the Deutsche Forschungsgemeinschaft (DFG GRK840-E3/E4, MA 4115/1-2/3, NI 1297/1-1), the Federal Ministry of Education and Research (BMBF GANI-MED 03152061A and BMBF 0314107) and the European Union (EU-FP-7: EPC-TM and EU-FP7-REGPOT-2010-1).
Competing interests None declared.
Patient consent Obtained.
Provenance and peer review Commissioned; externally peer reviewed.