Endoscopic retrograde pancreatography criteria to diagnose autoimmune pancreatitis: an international multicentre study.

BACKGROUND

Characteristic pancreatic duct changes on endoscopic retrograde pancreatography (ERP) have been described in autoimmune pancreatitis (AIP). The performance characteristics of ERP to diagnose AIP were determined.

METHODS

The study was done in two phases. In phase I, 21 physicians from four centres in Asia, Europe and the USA, unaware of the clinical data or diagnoses, reviewed 40 preselected ERPs of patients with AIP (n=20), chronic pancreatitis (n=10) and pancreatic cancer (n=10). Physicians noted the presence or absence of key pancreatographic features and ranked the diagnostic possibilities. For phase II, a teaching module was created based on features found most useful in the diagnosis of AIP by the four best performing physicians in phase I. After a washout period of 3 months, all physicians reviewed the teaching module and reanalysed the same set of ERPs, unaware of their performance in phase I.

RESULTS

In phase I the sensitivity, specificity and interobserver agreement of ERP alone to diagnose AIP were 44, 92 and 0.23, respectively. The four key features of AIP identified in phase I were (i) long (>1/3 the length of the pancreatic duct) stricture; (ii) lack of upstream dilatation from the stricture (<5 mm); (iii) multiple strictures; and (iv) side branches arising from a strictured segment. In phase II the sensitivity (71%) of ERP significantly improved (p<0.05) without a significant decline in specificity (83%) (p>0.05); the interobserver agreement was fair (0.40).

CONCLUSIONS

The ability to diagnose AIP based on ERP features alone is limited but can be improved with knowledge of some key features.