Robotic versus Open Pancreatoduodenectomy (PORTAL): multicentre, single masked, phase 3, non-inferiority randomised controlled trial
Jiabin Jin, Kai Qin, Georgios Gemenetzis, Zheng Wang, Chenxi Zheng, Wei Li, Zehua Wu, Yusheng Shi, Mengmin Chen, Jianfeng Qian, Yuanchi Weng, Wei Chen, Weiwei Jin, Renan Jin, Jian Li, Leshan Liu, Yuntao Chen, Bing Han, Xiao Liang, Chuandong Sun, Zheng Wu, Yiping Mou, Xiaoyu Yin, Xiaxing Deng, Chenghong Peng, Tianlin He, Yanyan Song, Baiyong ShenAbstract
Objective
To determine whether robotic pancreatoduodenectomy (RPD) is non-inferior to open pancreatoduodenectomy (OPD) in terms of postoperative functional recovery, without compromising safety or oncological quality.
Design
Multicentre, single masked, phase 3, non-inferiority randomised controlled trial.
Setting
Seven tertiary high volume pancreatic centres in China, 15 June 2020 to 28 November 2024.
Participants
268 adults with resectable pancreatic or periampullary disease.
Interventions
Participants were randomised to receive standardised RPD (n=142) or OPD (n=126), with enhanced recovery pathways.
Main outcome measures
The primary outcome was time from surgery to postoperative functional recovery, defined as adequate pain control without parenteral analgesia, ≥50% oral intake without intravenous fluids, independent mobilisation, and absence of active intra-abdominal infection. The restricted mean event time (RMET) within 40 days was a summary for time from surgery to postoperative functional recovery. Secondary outcomes included operative metrics, disease related outcomes, length of stay, postoperative morbidity, including complications of Clavien-Dindo grade II or higher (defined as complications requiring drug treatment or more intensive intervention), and hospital admission costs.
Results
Overall, 254 of 268 randomly assigned participants (mean age 62 years; 172 (64.2%) men) underwent surgery, completed follow-up, and were included in the modified intention-to-treat population; 14 did not undergo surgery. In the modified intention-to-treat population, the RMET was 12.1 days (95% confidence interval (CI) 11.2 to 13.1) in the RPD group and 16.0 days (14.5 to 17.5) in the OPD group (difference -3.9 days, 95% CI -5.6 to -2.2; P<0.001). Operative time was longer in the RPD group (300 minutes (interquartile range (IQR) 240-360 minutes) versus 270 (210-300) minutes in the OPD group, P<0.001) but postoperative length of stay was shorter in the RPD group (13 (IQR 11-16) days v 16 (13-20) days, P<0.001). Overall postoperative morbidity was 31.1% (41/132) in the RPD group versus 36.1% (44/122) in the OPD group and incidence of any complications of Clavien-Dindo grade II or higher was 23.5% (31/132) versus 34.4% (42/122), respectively. 90 day mortality was 0.8% (1) in the RPD group and 2.5% (3) in the OPD group. Median total costs of hospital admission (including readmission cost) were higher in the RPD group than in the OPD group (¥130 905 (£14 369; $19 351; €16 628) (IQR ¥114 853-¥152 547) v ¥108 071 (¥92 134-¥128 035), difference ¥22 834 (95% CI ¥16 744 to ¥30 522); P<0.001).
Conclusions
In high volume centres with credentialled surgeons, RPD met the non-inferiority margin for time from surgery to postoperative functional recovery, with comparable disease related outcomes and overall burden from postoperative complications. To generate wider system level efficiency gains, the implementation of RPD should take account of institutional expertise, procedural volume, acquisition of robotic surgical platforms and maintenance costs, and the potential for shorter hospital stay.
Trial registration
ClinicalTrials.gov