Table 3.
Most relevant studies on the diagnostic role of cholangioscopy in biliary disease.
| First Author | Year of Pubblication | Number of Patients | Typology of Study | Main Results of the Study |
|---|---|---|---|---|
| Sethi A. [23] | 2022 | 40 | Two-phase validation study | Validation of the Monaco Classification for the use of DSOC in BS. Eight visual criteria were identified and global diagnostic accuracy was 70%. |
| Robles-Medranda C. [24] | 2018 | 171 | Observational, analytical, case-crossover, ambispective, diagnostic study | Validation of a novel classification for the use of DSOC in BS. High reproducibility (k > 80%) and sensitivity of 96.3% for neoplastic diagnosis. |
| Kahaleh M [25] | 2022 | 50 | Validation study | Validation of the Mendoza classification for the use of DSOC in BS. Five revised visual criteria were identified. Agreement was almost perfect and the overall accuracy was 77%. |
| De Oliveira P [28] | 2020 | 283 | Systematic review and meta-analysis | The overall pooled sensitivity and specificity of DSOC in the visual interpretation of malignant BS were 94% and 95%. |
| Fugazza A [30] | 2022 | 381 | Prospective, multicenter | This real-world study evaluated the efficacy of POC for evaluation of BS and treatment of DBS. Overall procedure success was 96.7%. Sensitivity of visual impression for malignant BS was superior to cholangioscopy-guided biopsies (88.5% vs. 80.2%). |
| De Vries A.B [31] | 2020 | 80 | Retrospective, single center | This study highlighted the limited diagnostic use of POC in specific situations such as PSC and previous placement of biliary stent. |
| Gerges C [32] | 2020 | 60 | Randomized controlled trial | DSOC-guided biopsy samples were superior to ERCP-guided brushing (68.2% vs. 21.4%) for diagnosis of malignant hilar stricture. |
| Saraiva MM [40] | 2022 | 85 | Pilot validation study | The authors developed, trained, and validated a CNN based on DSOC images. The model had an overall accuracy of 94.9% for differentiating malignant BS from benign BS. |
| Robles-Medranda C [41] | 2023 | Phase 1: 48 patients Phase 2: 116 patients |
Two-stage validation study | Validation of a CNN model for identification of neoplasia in indeterminate BS. The model showed a good accuracy in distinguishing neoplastic lesions. |
| Zhang X [43] | 2023 | 150 | Multicenter diagnostic study | Validation of a novel AI model, MBSDeiT, which accurately identified 92.3% of malignant BS in prospective testing videos. |
| Anderloni A [44] | 2019 | 36 | Retrospective, single center | This study showed the efficacy of D-POC for the evaluation of complete clearance of CBD after removal of DBS. |
| Arnelo U [45] | 2015 | 47 | Prospective, single center | Study performed on patients with PSC showing the limited sensitivity of DSOC for malignant BS. |
DSOC, digital single-operator cholangioscopy; BS, biliary stricture; POC, peroral cholangioscopy; CNN, convolutional neural network; CBD, common bile duct; DBSs, difficult bile stones; PSC, primary sclerosing cholangitis.