Annals of Surgery Open (Sep 2022)

A Multicenter, Randomized, Double-Blinded, Clinical Trial Comparing Cattell-Warren and Blumgart Anastomoses Following Partial Pancreatoduodenectomy

  • Christopher M. Halloran, MD,
  • John P. Neoptolemos, MD,
  • Richard Jackson, PhD,
  • Kellie Platt, MSc,
  • Eftychia-Eirini Psarelli, MSc,
  • Srikanth Reddy, PhD,
  • Dhanwant Gomez, MD,
  • Derek A. O’Reilly, PhD,
  • Andrew Smith, MD,
  • Thomas M. Pausch, MD,
  • Andreas Prachalias, MD,
  • Brian Davidson, MD,
  • Paula Ghaneh, MD

DOI
https://doi.org/10.1097/AS9.0000000000000198
Journal volume & issue
Vol. 3, no. 3
p. e198

Abstract

Read online

Objective:. Whether a Blumgart anastomosis (BA) is superior to Cattell-Warren anastomosis (CWA) in terms of postoperative pancreatic fistula (POPF) following pancreatoduodenectomy. Importance:. Complications driven by POPF following pancreatic cancer resection may hinder adjuvant therapy, shortening survival. BA may reduce complications compared to CWA, improving the use of adjuvant therapy and prolonging survival. Methods:. A multicenter double-blind, controlled trial of patients undergoing resection for suspected pancreatic head cancer, randomized during surgery to a BA or CWA, stratified by pancreatic consistency and duct diameter. The primary end point was POPF, and secondary outcome measures were adjuvant therapy use, specified surgical complications, quality of life, and survival from the date of randomization. For a 10% POPF reduction, 416 patients were required, 208 per arm (two-sided α = 0·05; power = 80%). Results:. Z-score at planned interim analysis was 0.474 so recruitment was held to 238 patients; 236 patients were analyzed (112 BA and 124 CWA). No significant differences in POPF were observed between BA and CWA, odds ratio (95% confidence interval [CI]) 1·04 (0.58–1.88), P = 0.887, nor in serious adverse events. Adjuvant therapy was delivered to 98 (62%) of 159 eligible patients with any malignancy; statistically unrelated to arm or postoperative complications. Twelve-month overall survival, hazard ratio (95% CI), did not differ between anastomoses; BA 0.787 (0.713–0.868) and CWA 0.854 (0.792–0.921), P = 0.266, nor for the 58 patients with complications, median (IQR), 0.83 (0.74–0.91) compared to 101 patients without complications 0.82 (0.76–0.89) (P = 0.977). Conclusions:. PANasta represents the most robust analysis of BA versus CWA to date.