This retrospective study collected 635 periampullary carcinoma patients receiving PD treatment between June 2012 and January 2018, including demographic data, surgical variables, postoperative outcomes, pathological results, and extended follow-up. Twenty-three patients with inadequate baseline data or missing primary outcome data were excluded. A total of 612 patients were included in this study. Data were collected on patient characteristics, surgical details, morbidity and mortality, postoperative hospital stay, and pathological outcomes. Percutaneous transhepatic cholangial drainage (PTCD) were performed for all patients with elevated bilirubin preoperatively, and surgical treatment was performed after the bilirubin returned to normal. Preoperative examination included appropriate imaging diagnosis to exclude distant metastases. Preoperative characteristics included age, sex, complications, body mass index (BMI), and American society of anesthesiologists (ASA) score. Surgical details include operative time (from incision to wound closure), estimated blood loss, and transfusion volume (obtained from anesthesia records). Postoperative follow-up was conducted for 3 months (90 days), and all complications were recorded according to the Clavien-Dindo (CD) system score. Because no objective criteria to recommend laparoscope or open PD in periampullary carcinoma, the surgical method depended on preoperative evaluation and patient’s choice of will. Vascular involvement or prior surgical history or diaphanous variation were taken into account, but were not absolute contraindications to LPD. All patients were informed of the need for surgery and possible complications, as well as the pros and cons of laparoscopic versus open surgery. All patients receiving OPD surgery were performed by 5 experienced doctors in the center, and LPD was performed by Dr. Qin alone. All patients were followed up according to tumor grade to monitor tumor recurrence. The patients were regularly followed up one month after the operation, and then every three months for two years. Follow-up was conducted every six months thereafter. The follow-up of these included patients ended in December 2019. Overall survival (OS) is defined as the time interval between the date of surgery and the date of death.

All OPD surgery were performed by 5 experienced surgeons and LPD was mainly performed by one experienced surgeon (Dr. Qin, director of department of Biliary–Pancreatic Surgery). OPD is performed in the supine position following standard Whipple procedures. With regard to LPD, our center for the first time proposed an “individualized surgical strategy” for LPD. Conventional “five-hole method” can meet the needs of surgical operation. Different from the viewpoint of the first assistant and the dominant surgeon advocated by many centers, our center adopts the fixed strategy of the dominant surgeon, which reduces the change of surgical staff by constantly adjusting body position to cooperate with the exposure of surgical field.

According to whether the tumor invaded or compressed the peripheral vascular system and the degree of invasion, we divided the tumors around the head of the pancreas into five types (11), and conducted specific individualized surgical strategies for each type of pancreatic cancer based on surgical experience and other literature. The most commonly used preferred route of the common retroperitoneal artery in our center had been described in detail in previous articles (12). The advantages of this approach are as follows: firstly, the resectability of the tumor was defined; Secondly, through the retroperitoneal route, the root of all the blood supply arteries on the head of the pancreas was preferentially severed, which reduced the amount of blood loss, prevented massive intraoperative bleeding and ensured safety. In addition, the R0 resection rate of carcinoma of the head of the pancreas was increased, and the vascular resection length was reduced. The reconstruction of digestive tract was Child type. For pancreas-intestinal anastomosis, the center uses our own original “implantable anastomosis” which had been described in detail in previous articles (13).

PH was measured by arterial blood drawn by the anesthesiologist after specimen removal was completed during the operation. Operative time was the time from the skin incision or trocar to the complete skin closure. Intraoperative estimated blood loss was recorded the anesthesiologist through a vacuum system. Postoperative hospital stay was defined as the days from surgery to discharge. Morbidity and mortality were defined as any complications or deaths that occurred during hospitalization or within 30 days of discharge after surgery. Readmission within 30 days after surgery was considered unplanned. OS was defined as the time interval between the date of surgery and the date of death. Deaths during hospitalization were excluded in the calculation of postoperative hospital stay. Postoperative complications were evaluated according to the CD classification system (14) and including postoperative pancreatic fistula (15), delayed gastric emptying (16), bile leakage (17), postoperative bleeding (18).

Continuous variables are reported as mean with standard deviation or median with interquartile range, and categorical data are presented as frequency. Comparisons between two groups were conducted using Student’s t-test or the Mann–Whitney test for quantitative data and the chi-square test or Fisher’s exact test for qualitative data. Survival analysis was performed by the Kaplan–Meier method, and the log-rank test was used to compare the different survival curves. The cut-off values of pH for predicting 18-month survival were determined using a receiver operating characteristic (ROC) curve, and the pH was then dichotomized into low and high groups. Univariate Cox regression analysis was conducted to estimate the risk factors for OS, and variables with a P value of <0.1 were then entered into the multivariate Cox regression model. The hazard ratio and 95% confidence interval were calculated, and the lower limit of the 95% confidence interval above zero was considered a significant risk factor. All statistical analyses were performed using SAS statistical software (SAS Institute, Inc., Cary, NC, USA), and two-sided hypothesis testing with a predetermined level of P < 0.05 was considered statistically significant.

From June 2012 to January 2018, 612 patients with periampullary adenocarcinoma underwent PD in our center. The last follow-up was on 15 December 2019, and the median follow-up period was 18 months (IQR, 6–42). The pH was recorded before suturing peritoneum, and the median pH was 7.37 (range, 6.69–7.64). As shown in Figure 1 , the patients were divided into a high and low pH group according to the cut-off value of the ROC curve (7.34) (a sensitivity of 80.8% and a specificity of 92.2%). The area under the curve (AUC) of pH was 0.923 (95% CI, 0.897–0.952, P < 0.001). The low pH group (pH ≤ 7.34) comprised 421 patients, and the high pH group (pH >7.34) comprised 191 patients.

The relationships between the pH and the patients’ clinical parameters are shown in Table 1 . There was no significant correlation between pH and age, BMI, sex, blood group, diabetes mellitus, family history, tumor size, alkaline phosphatase, gamma-glutamyl transpeptidase, cholesterol, triglyceride, lactate dehydrogenase, total protein, albumin, white blood cells, preoperative total bilirubin, preoperative biliary drainage, ASA, TNM stage, histopathological diagnosis, or serum tumor markers.

As shown in Table 2 , low pH was associated with a longer operative time (median, 363 vs 307 min; P < 0.001), a higher intraoperative bleeding volume (mean, 445 vs 259 ml; P < 0.001), and a higher intraoperative blood transfusion volume (mean, 1.62 vs 0.98 U; P = 0.015). There was no significant difference in the R0 removal rate between the low and high pH groups (median, 84.56% vs 87.43%; P = 0.350). Similarly, there was no significant difference in the number of lymph nodes cleared between the low and high pH groups (P = 0.476). There was a significant difference in the operation method between the low and high pH groups (P < 0.001).

The patients’ postoperative factors are shown in Table 3 . The rate of total complications was significantly higher in the low pH group than high pH group (31.83% vs 18.32%, P = 0.026). Specifically, postoperative pancreatic fistula (19.95% vs 10.99%, P = 0.031), renal failure (2.14% vs 0.00%, P = 0.016), infection (5.24% vs 1.05%, P = 0.021), and delayed gastric emptying (26.37% vs 17.28%, P = 0.024) were more common in the low pH group. There was no significant difference in the rate of positive lymph nodes sent for examination between the low and high pH groups (P = 0.385). There was also no difference in the 30-day unplanned readmission rate after surgery (4.75% vs 4.71%, P = 0.163); however, the low pH group had a longer postoperative hospital stay (23 vs 19 days, P = 0.043). The reoperation rate was similar between the low and high pH groups (2.14% vs 0.52%, P = 0.144). The 90-day mortality rate was significantly higher in the low than high pH group (4.99% vs 0.52%, P < 0.001).

As shown in Table 4 , we performed univariate and multivariate survival analyses for OS to determine whether pH was an independent prognostic factor for periampullary carcinoma. Single-factor analysis showed that pH, operation method, ASA, tumor stage, R state, and pancreas texture were significantly correlated with OS (P < 0.001). To determine the independent prognostic factors, the important factors in the univariate analysis were included in the multivariate analysis. The multivariate analysis showed that pH, R state and tumor stage were independent prognostic factors for OS (P < 0.001). Besides, the Kaplan–Meier survival curve suggested that low-pH was associated with low OS, and the difference in survival rate between the low-pH group and the high-pH group was statistically significant (P < 0.001, Figure 2 ).

As shown in Table 5 , the average pH in the LPD group was 7.41, which was significantly higher than that in the OPD group (7.34) (P < 0.001). The operation time was significantly shorter in the LPD than OPD group (305 vs 374 min, P < 0.001), as was the blood loss volume (205 vs 512 ml, P < 0.001). The intraoperative red blood cell infusion volume was also significantly lower in the LPD than OPD group (0.42 vs 2.12 U, P < 0.001). The postoperative hospital stay was significantly shorter in the LPD than OPD group (20 vs 21 days, P = 0.005). There were no statistically significant differences in postoperative complications, 90-day mortality, 30-day unplanned hospitalization, or the R0 resection rate between the two groups.