Influence of a novel classification of the papilla of Vater on the outcome of needle-knife fistulotomy for biliary cannulation

Abstract

Abstract Background Existing proposed classification systems for the Papilla of Vater (PV) suboptimally account for all relevant, encountered PV appearances, are too complex or have not been assessed for intra- or interobserver variability. We proposed a novel endoscopic classification system for PV, determined its inter- and intraobserver rates and used the classification system to assess whether the success and complications of needle-knife fistulotomy (NKF) are influenced by the morphology of the PV. Methods The classification system was developed by expert endoscopists. To evaluate the inter- and intraobserver agreement, an online questionnaire was sent to 20 endoscopists from several countries (10 experts and 10 nonexperts) that included 50 images of papillae of Vater divided among various categories. Four weeks later, a second survey, with the images from the first questionnaire randomly reordered, was sent to the same endoscopists. The inter- and intraobserver agreements among the experts and nonexperts was calculated. Using the proposed classification system, all 361 consecutive patients who underwent NKF for biliary access to a naïve papilla were prospectively enrolled in the study. Results The novel classification system comprises 7 categories: type I, flat type, lacking an oral protrusion; type IIA, prominent tubular nonpleated type, with an oral protrusion and < 1 transverse fold over the oral protrusion; type IIB, prominent tubular pleated type, with an oral protrusion and > 2 transverse folds over the oral protrusion; type IIC: prominent bulging type, with an enlarged and bulging oral protrusion; type IIIA, diverticular-intradiverticular type, with a papillary orifice inside the diverticulum; type IIIB: diverticular-diverticular border type, with a papillary orifice less than 2 cm from the diverticular border; type IV: unclassified papilla, with no morphology classified in the other categories. The interobserver agreement between experts was substantial (K = 0.611, 95% CI 0.498–0.709) and was higher than that between nonexperts (K = 0.516; 95% CI 0.410–0.636). The intraobserver agreement was substantial among both experts (K = 0,651; 95% CI 0.586–0.715) and nonexperts (K = 0.646, 95% CI 0.615–0.677). In a multivariate model, type IIIA and IIIB were the only independent risk factors for difficult rescue NKF biliary cannulation (P = 0.003 and P = 0.019, respectively), and type I and type IIB were the only independent risk factors for a prolonged cannulation time using NKF (P < 0.001 and P = 0.005, respectively). Conclusions The novel endoscopic classification system for PV is highly reproducible among experienced ERCPists according to the substantial level of agreement between experts. However, nonexperts require further training in its use. Using the novel classification system, we identified different types of papillae significantly associated with a lower efficacy of NKF and a prolonged time to obtain successful biliary cannulation using NKF.