Back to 2025 Abstracts
Results of a Phase III Randomized Controlled Trial of Pyloroplasty Versus No Pyloroplasty in Patients Undergoing Minimally Invasive Esophagectomy or Robot Assisted Minimally Invasive Esophagectomy
James Luketich, *Inderpal Sarkaria, *Ryan Levy, *Evan Alicuben, *William Gooding, *Omar Awais, *Neil Christie, *Renee Levesque, *Julie Ward, *John Ryan, *Matthew Schuchert, *Arjun Pennathur, University of Pittsburgh/UPMC MIE RAMIE Pyloroplasty Trial Study Group
Cardiothoracic Surgery, University of Pittsburgh, Pittsburgh, PA
Objective: Esophagectomy is a complex operation and many reports have included the addition of pyloroplasty. There are pros and cons to performing a pyloric drainage procedure, but few studies have included level one evidence to support or refute this step, especially in the era of minimally invasive esophagectomy (MIE) and robotically assisted MIE (RAMIE). We conducted a prospective Phase III randomized controlled trial (RCT) to assess the value of the addition of a pyloroplasty procedure vs. no pyloroplasty during the performance of MIE or RAMIE (NCT03740542).
Methods: An adaptive randomization trial design was utilized to maximize the number of patients treated on the more effective therapy and conversely to minimize accrual to a less effective procedure. The trial was designed to proceed until one arm is established as superior or until a total of 140 patients have been treated equally and deemed evaluable for response. Based on 2000 trial simulations with binomial means for each arm equal to the prior probabilities (.13 or .26), the adaptive randomization trial was designed to have 86% power to assert the superiority of the pyloroplasty arm with a 14% type I error. The primary endpoints of the study were pneumonia and anastomotic leak requiring surgery within 30 days of surgery.
Results: Over a four-year period, 143 patients were randomly assigned to pyloroplasty vs. no pyloroplasty, and among these 134 patients were evaluable (Table). The greater likelihood of success for pyloroplasty throughout the trial resulted in more patients randomized towards pyloroplasty (n= 90) vs. no pyloroplasty (n= 44). The primary endpoints occurred in 16 of 90 (18%) in the pyloroplasty arm compared to 12 of 44 (27%) in the no pyloroplasty arm. The posterior probability that the pyloroplasty arm is superior was continuously monitored, and stopping criteria were met when this posterior probability reached 90%.
Conclusions: This RCT used an adaptive trial design to compare outcomes of pyloroplasty vs. no pyloroplasty in patients undergoing RAMIE or MIE. The results and design of this trial led to early stopping since the short-term results indicated that outcomes in the pyloroplasty arm were superior to the non-pyloroplasty arm. This RCT provides evidence for short-term benefits of the addition of pyloroplasty to MIE or RAMIE. The long-term outcomes and quality of life measures continue to be monitored and will be included in future updates.
Patient and Operative Characteristics
Back to 2025 Abstracts
