Abstract:Background: Pancreatic carcinoma is often inoperable, carries a poor prognosis and is commonly complicated by malignant biliary obstruction. Phase I/II studies have demonstrated good safety and early stent patency using endoscopic biliary radiofrequency ablation (RFA) as an adjunct to self-expanding metal stent (SEMS) insertion for biliary decompression. Aim: To analyse the clinical efficacy of endobiliary RFA. Methods: Retrospective case-control analysis of 23 patients with surgically unresectable pancreatic carcinoma and malignant biliary obstruction undergoing endoscopic RFA and SEMS insertion, and 46 controls (SEMS insertion alone) in a single tertiary care centre. Controls were stringently matched for age, sex, metastases, ASA/co-morbidities. Survival, morbidity, and stent patency rates were assessed. Thank you for considering our manuscript and for the helpful reviews.Please find enclosed a revised version of the manuscript for reconsideration, together with a detailed point by point response to the reviewer comments.We have taken independent statistical advice on the manuscript as suggested.
RESPONSES TO REVIEWER COMMENTSReviewer #1: The authors present a retrospective series of 23 pts with unresectable panc ca treated with RFA and uncovered SEMS vs. 46 with uncovered alone and matched for chemo, mets, ASA, etc. It it novel data but the main premise for RFA is that it induces a coagulative necrosis -how it improves survival if it doesn't improve stent patency is not hypothesized. Stent patency was similar between the twogroups. Are the authors suggesting that RFA may be able to reduce tumor burden? Several comments to help the authors improve the submission. I do feel that the manuscript would benefit from more clarity and defined goals for its work. What are the primary and secondary endpoints for the study. These should be clearly stated in the Methods.The primary and secondary endpoints of the study are survival and stent patency/procedure safety, respectively. This has been highlighted in the methods section (cf point 13, below). Please see the response to point 14 below, re putative beneficial effects of RFA on patients with pancreatic cancer.Other comments: 1) Those patients who underwent chemo was higher in the RFA group (70% vs. 52%) but not statistically sig't. This may be a Type 2 error given the small sample size and this limitation should be clearly highlighted in the discussion. In fact, this seems to be the most plausible explanation as to the survival benefit in the RFA group unless the authors can provide other possible hypothesis.The possibility of a type 2 error is now mentioned in the relevant discussion section.2) The improved median survival of 226D vs. 124D has a CI that goes through 1. Please explain the limitations of this analysis and include in the discussion.The upper limit of the 95% CI is 1.06. This is represented in the fact that the two survival curves on Kaplan-Meier analysis cross at late time points, when the number of surviving patients is small. Our statistica...