Page 14 - Read Online

Page 8 of 15 Shannon et al. Mini-invasive Surg 2023;7:32 https://dx.doi.org/10.20517/2574-1225.2023.83

study from Zheng et al. looking at over 1,000 patients undergoing open vs. minimally invasive pancreas
surgery for PNET [88-94] . These studies reveal that a robotic approach to PD is safe and feasible and allows a
larger pool of surgeons to perform a minimally invasive PD as compared to a laparoscopic approach.

Future trials on minimally invasive pancreaticoduodenectomy
There are multiple ongoing clinical trials examining minimally invasive PD, as shown in Table 2. The
PORTAL trial is a multicenter phase III patient-blinded trial with the primary outcome of time to functional
recovery in patients with benign or malignant head of pancreas disease among patients undergoing open vs.
robot-assisted PD. Secondary outcomes include overall complication rates, mortality, oncologic outcomes,
cost, and quality of life. Patient recruitment is expected to be complete in 2022, with published results in
[95]
2024 .

The TJDBPS01 trial (NCT03138213) is a multicentre, prospective, and randomized controlled trial that
compares total laparoscopic PD to open PD. The authors hypothesize that laparoscopic PD has equal or
better safety and post-operative recovery compared to an open approach with a primary outcome of
hospital LOS . Similarly, the TJDBPS07 trial (NCT03785743) is another ongoing randomized controlled
[96]
trial, which will assess five-year overall and disease-free survival, 90-day mortality, and complication rates
between laparoscopic and open PD . Lastly, an ongoing trial from Johns Hopkins University
[97]
(NCT04171440) is a randomized trial comparing minimally invasive PD to open PD. Patients with both
benign and malignant pathologies are included with a primary outcome of time to functional recovery.
[68]
Recruitment is currently ongoing, with results expected in 2024 .

DISCUSSION
Based on a comprehensive review of prospective randomized trials, minimally invasive pancreatectomy is
safe and has equivalent oncologic outcomes and morbidity compared to traditional open approaches when
in experienced hands. The traditional open PD is a complex operation with numerous steps, requiring a
mastery of abdominal anatomy and its variations and demanding technical excellence from the surgeon.
The main limitations for minimally invasive pancreatectomy are the cost and increased operative times
associated with the procedure. Additionally, the hesitation for universal adoption of minimally invasive
pancreatectomy remains due to its associated learning curve required to achieve equivalent and optimal
outcomes.

High intraoperative cost has been one of the main limitations to the adoption of minimally invasive surgery,
specifically, the robotic approach. Much of the cost is related to increased operating room time and the cost
of disposables and equipment used . However, the benefits of minimally invasive approaches compared to
[98]
open ones in terms of decreased LOS may counteract these costs . A propensity score-matched analysis
[98]
comparing robotic pancreatic surgery to open surgery showed a higher intraoperative cost with the robot;
however, it also demonstrated a decreased overall hospital LOS, which improved overall cost-
effectiveness . Nevertheless, much of the literature evaluating cost is retrospective in nature, and
[99]
prospective data is limited to single-center studies. Consequently, there is a need for multicenter prospective
trials assessing the cost-effectiveness of minimally invasive approaches to pancreatic surgery.

The “learning curve” for surgeons is a well-known phenomenon with an inverse relationship between
[100]
surgeon volume and mortality . Much has been studied on the learning curve associated with minimally
invasive pancreas surgery. A systematic review by Chan et al. analyzed the number of cases needed to
surmount the learning curve and showed no difference in laparoscopic vs. robotic PD (34.1 cases vs. 36.7
[101]
cases, respectively) . Another cohort study showed that the learning curve for proficiency was around