Recent randomized controlled trials in pancreaticoduodenectomy

Objectives: To analyze and summarize the recent randomized controlled trials (RCTs) investigating pancreaticoduodenectomy (PD). Methods: A MEDLINE search was performed to identify prospective RCTs on PD published during the last decade. Eligible RCTs were analyzed using the following items: publication year, geographical area, study theme, sample size, and multicenter study. Moreover, the quality of each RCT was evaluated. Results: Thirty-four articles were eligible for review. One to 6 RCTs have been carried out annually during the recent 10 years. Geographically, 15 trials were performed in Europe, 10 trials in North America, and 9 in Asia. Studies concerning postoperative complications in the early postoperative period such as pancreatic fistula and delayed gastric emptying have been most frequent. Randomized controlled trials comparing anastomotic procedures for the remnant pancreas, standard PD versus PD with extended lymphadenectomy, and PD versus pylorus-preserving PD follow in descending order. The average sample size has been 117, and 10 RCTs had sample size less than 50. The rate of multicenter studies among all RCTs is 21%, with the rate in the most recent 5 years having increased 2-fold compared with that in the earlier period. Concerning the quality of RCTs, calculation of sample size was described in only 14 RCTs and intention to treat analysis was performed in 26 RCTs. Conclusions: This study reviewed 34 RCTs on PD performed all over the world. Although the quality of every RCT was not satisfactory, high-grade evidence obtained by these RCTs should be applied in clinical settings to improve surgical quality and quality of life for each patient.