Assessing Endoscopic Ultrasound-Guided Radiofrequency Ablation in Unresectable Pancreatic Ductal Adenocarcinoma: A Single-Center Historic Cohort Study.

BACKGROUND: /aims: Endoscopic ultrasound-guided radiofrequency ablation (EUS-RFA) has emerged as an alternative for the local treatment of unresectable pancreatic ductal adenocarcinoma (PDAC). We aim to assess the feasibility and safety of EUS-RFA in patients with unresectable PDAC. METHODS: The following was a historic cohort compounded by locally advanced (LA) and metastatic (m) PDAC naïve patients, who underwent EUS-RFA between October 2019 to March 2022. EUS-RFA was performed with a 19-g needle electrode with a 10 mm active tip for energy delivery. Study primary endpoints were feasibility, safety, and clinical follow-up; secondary endpoints were performance status (PS), local control (LC) and overall survival (OS). RESULTS: Twenty-six patients were selected: 15/26 LA-PDAC and 11/26 mPDAC. Technical success was achieved in all patients with no major adverse events. Six months after EUS-RFA, OS was 11/26 (42.3%), with significant PS improvement (P=.03). LC was achieved, with tumor reduction from 39.5 to 26 mm (P=.04). Post-treatment hypodense necrotic area was observed at six-month follow-up in 11/11 alive cases. Metastatic disease was a significant factor for OS worsening (HR 5.021; IC 95% 1.589 - 15.87; P=.004) CONCLUSIONS: EUS-RFA of pancreatic adenocarcinoma is a minimally invasive and safe technique that may have an important role as targeted therapy for local treatment of unresectable cases, as well as an alternative for poor surgical candidates. Also, RFA may play a role in downstaging cancer with potential OS increase in non-metastatic cases. Large prospective cohorts are required to evaluate this technique in clinical practice.

Artificial intelligence for diagnosing neoplasia on digital cholangioscopy: development and multicenter validation of a convolutional neural network model.

BACKGROUND: We aimed to develop a convolutional neural network (CNN) model for detecting neoplastic lesions during real-time digital single-operator cholangioscopy (DSOC) and to clinically validate the model through comparisons with DSOC expert and nonexpert endoscopists. METHODS: In this two-stage study, we first developed and validated CNN1. Then, we performed a multicenter diagnostic trial to compare four DSOC experts and nonexperts against an improved model (CNN2). Lesions were classified into neoplastic and non-neoplastic in accordance with Carlos Robles-Medranda (CRM) and Mendoza disaggregated criteria. The final diagnosis of neoplasia was based on histopathology and 12-month follow-up outcomes. RESULTS: In stage I, CNN2 achieved a mean average precision of 0.88, an intersection over the union value of 83.24 %, and a total loss of 0.0975. For clinical validation, a total of 170 videos from newly included patients were analyzed with the CNN2. Half of cases (50 %) had neoplastic lesions. This model achieved significant accuracy values for neoplastic diagnosis, with a 90.5 % sensitivity, 68.2 % specificity, and 74.0 % and 87.8 % positive and negative predictive values, respectively. The CNN2 model outperformed nonexpert #2 (area under the receiver operating characteristic curve [AUC]-CRM 0.657 vs. AUC-CNN2 0.794, P < 0.05; AUC-Mendoza 0.582 vs. AUC-CNN2 0.794, P < 0.05), nonexpert #4 (AUC-CRM 0.683 vs. AUC-CNN2 0.791, P < 0.05), and expert #4 (AUC-CRM 0.755 vs. AUC-CNN2 0.848, P < 0.05; AUC-Mendoza 0.753 vs. AUC-CNN2 0.848, P < 0.05). CONCLUSIONS: The proposed CNN model distinguished neoplastic bile duct lesions with good accuracy and outperformed two nonexpert and one expert endoscopist.