How to Improve Diagnostic Accuracy in Suspicious Ampullary Lesions?

See "Diagnostic Accuracy of the Initial Endoscopy for Ampullary Tumors" by Hee Seung Lee, Jong Soon Jang, Seungho Lee, et al., on page 239-246

In cases of ampullary lesions, differential diagnosis between tumors and inflammation by endoscopic findings is sometimes challenging. Even if operators have experience in endoscopic techniques, diagnostic accuracy by gross morphology is only 67%. Although histologic evaluation of ampullary lesions with forceps biopsy is performed during endoscopy, there is still a risk of false negativity for tumorous lesions. It is well known that one or two forceps biopsies are not sufficient for diagnosis during upper or lower endoscopy, and the diagnostic accuracy of the first forceps biopsy was as low as 67% in this study. Although the results did not reach statistical significance, there was a trend toward improved diagnostic accuracy with more biopsies (56% by two biopsies vs. 75% by three biopsies). Therefore, multiple ampullary biopsies should be considered as a means of overcoming this problem. However, acute pancreatitis can be a major concern of multiple forceps biopsies. Less invasive and more accurate diagnostic modalities are necessary for the detection and differential diagnosis of ampullary lesions. There have been many developments in diagnostic endoscopy, including endoscopic ultrasound (EUS), intraductal ultrasonography (IDUS), narrow band imaging (NBI), and confocal laser endomicroscopy (CLE). EUS is useful for the detection and differential diagnosis of ampullary tumors, and can also provide more information about the depth of tumor invasion.1 IDUS is likely to be superior to EUS in terms of T staging, but sometimes tends to overestimate tumor invasion into the bile duct.2,3 A recent retrospective study has reported that NBI was useful for the differential diagnosis of small ampullary lesions.4 The usefulness of CLE for the diagnosis of ampullary lesions has not been demonstrated, and is still under investigation since there has been no standardization of CLE image criteria for ampullary tumors.5

Although the accuracy of endoscopic imaging and forceps biopsy each was low, the diagnostic accuracy was 91% when endoscopic diagnoses and biopsy results were concordant. Therefore, the authors conclude that if there is discordance between endoscopic findings and biopsy results, re-evaluation with side-viewing endoscopy after resolution of inflammation is recommended.6

In summary, the diagnostic accuracy of first endoscopy for ampullary tumors is reliable if both endoscopic findings and biopsy results are concordant. When both results are repeatedly in disagreement, further evaluation with other diagnostic modalities including EUS, IDUS, and NBI should be considered.