Prediction of surgical outcomes in severe encapsulating peritoneal sclerosis using a computed tomography scoring system.

BACKGROUND/PURPOSE: Encapsulating peritoneal sclerosis (EPS) is a rare and potential lethal complication of peritoneal dialysis characterized by bowel obstruction. Surgical enterolysis is the only curative therapy. Currently, there are no tools for predicting postsurgical prognosis. This study aimed to identify a computed tomography (CT) scoring system that could predict mortality after surgery in patients with severe EPS. METHODS: This retrospective study enrolled patients with severe EPS who underwent surgical enterolysis in a tertiary referral medical center. The association of CT score with surgical outcomes including mortality, blood loss, and bowel perforation was analyzed. RESULTS: Thirty-four patients who underwent 37 procedures were recruited and divided into a survivor and non-survivor group. The survivor group had higher body mass indices (BMIs, 18.1 vs. 16.7 kg/m2, p = 0.035) and lower CT scores (11 vs. 17, p < 0.001) than the non-survivor group. The receiver operating characteristic curve revealed that a CT score of ≥15 could be considered a cutoff point to predict surgical mortality, with an area under the curve of 0.93, sensitivity of 88.9%, and specificity of 82.1%. Compared with the group with CT scores of <15, the group with CT scores of ≥15 had a lower BMI (19.7 vs. 16.2 kg/m2, p = 0.004), higher mortality (4.2% vs. 61.5%, p < 0.001), greater blood loss (50 vs. 400 mL, p = 0.007), and higher incidence of bowel perforation (12.5% vs. 61.5%, p = 0.006). CONCLUSION: The CT scoring system could be useful in predicting surgical risk in patients with severe EPS receiving enterolysis.

Feasibility of simultaneous development of laparoscopic and robotic pancreaticoduodenectomy.

Laparoscopic (LPD) and robotic pancreaticoduodenectomy (RPD) are both challenging procedures. The feasibility and safety of simultaneously developing LPD and RPD remain unreported. We retrospectively reviewed the data of patients undergoing LPD or RPD between 2014 and 2021. A total of 114 patients underwent minimally invasive pancreaticoduodenectomy (MIPD): 39 LPDs and 75 RPDs. The learning process of LPD and RPD were similar. The cutoff points of the learning curve were LPD, 13th patient (the 27th patient of MIPD), and RPD, 18th patient (the 31st patient of MIPD) according the cumulative sum analysis of operative time. A decrease in the operative time was associated with the case sequence (p < 0.001) but not with the surgical approach (p = 0.36). The overall surgical outcomes were comparable between both the LPD and RPD groups. When evaluating the learning curve impact on MIPD, LPD had higher major complication (≧ Clavien-Dindo grade III), bile leak and wound infection rates in the pre-learning curve phase than those in the after-learning curve phase, while RPD had similar surgical outcomes between two phases. Simultaneous development of LPD and RPD is feasible and safe for experienced surgeons, with similar learning process and comparable surgical outcomes.

Minimally invasive approaches increase postoperative complications in obese patients undergoing pancreaticoduodenectomy during the initial development period: a propensity score matching study.

BACKGROUND: Obesity increases surgical risks in various abdominal surgeries and its impact on open pancreaticoduodenectomy (OPD) and minimally invasive pancreaticoduodenectomy (MIPD) remains unknown. This study aimed to compare the surgical outcomes of OPD and MIPD in obese and non-obese patients by propensity score matching (PSM) analysis during the implementation of MIPD. METHODS: We retrospectively reviewed all pancreaticoduodenectomies from December 2014 to May 2021. Obesity was defined as body mass index > 25 kg/m2 according to World Health Organization International Obesity Task Force. PSM was used to minimize the selection bias of MIPD. RESULTS: Among 462 pancreaticoduodenectomies (339 OPDs, 123 MIPDs), there were 313 patients in the non-obese group (MIPD: 78, OPD: 235) and 149 patients in the obese group (MIPD: 45, OPD: 104). After PSM, there were 70 MIPD/106 OPD patients in the non-obese group and 38 MIPD/54 OPD patients in the obese group. The obese MIPD patients had more fluid collection (36.8% vs 9.8%, p = 0.002), a higher Clavien-Dindo (CD) grade (p = 0.007), more major complications (42.1% vs 14.8%, p = 0.004), and longer operative times (306 min vs 264 min, p < 0.001) than the obese OPD patients. The non-obese MIPD patients had lower CD grades (p = 0.02), longer operative times (294 vs 264 min, p < 0.001), and less blood loss (100 mL vs 200 mL) than the non-obese OPD patients. MIPD was a strong predictor of major complication (CD ≥ 3) in obese patients (odds ratio 3.11, 95% CI: 1.40-6.95, p = 0.005). CONCLUSIONS: Minimally invasive approaches deteriorate the CD grade, fluid collection, and major complications in obese patients undergoing pancreaticoduodenectomy during the initial development period. Non-obese patients may benefit from MIPD over OPD in terms of less blood loss and lower CD grades. The impact of BMI on MIPD should be considered when assessing the surgical risks.