Impact of external drainage of the main pancreatic duct and common bile duct on postoperative pancreatic fistula following pancreatoduodenectomy: protocol for a multicentre randomized clinical trial
Department of Pancreatic Surgery, Zhongshan Hospital, Fudan University, Shanghai 200032, China
Cancer Center, Zhongshan Hospital, Fudan University, Shanghai 200032, China
Department of General Surgery, Zhongshan Hospital, Fudan University, Shanghai 200032, China
Department of Biostatistics, Zhongshan Hospital, Fudan University, Shanghai 200032, China
Department of Hepatobiliary Surgery, First People’s Hospital of Yunnan Province, Kunming 650032, China
Department of Hepato-Biliary-Pancreatic Surgery, Huadong Hospital Affiliated to Fudan University, Shanghai 200040, China
Department of Hepatobiliary Surgery, First Affiliated Hospital of Xi’an Jiaotong University, Xi’an 710061, China
Department of Biliary-Pancreatic Surgery, Shanghai Jiao Tong University School of Medicine Affiliated Renji Hospital, Shanghai 200001, China
Department of General Surgery, Nanjing University Medical School Affiliated Nanjing Drum Tower Hospital, Nanjing 210008, China
Correspondence to: Liang Liu, Department of Pancreatic Surgery, Zhongshan Hospital, Fudan University, 180 Feng Lin Road, Shanghai 200032, China. (e-mail: liu.liang@zs-hospital.sh.cn)Abstract
Background
According to a retrospective study at Zhongshan Hospital, external drainage of the main pancreatic duct (MPD) and common bile duct (CBD) is potentially superior over internal drainage. As yet there is no consensus regarding the optimal drainage strategy, and previous studies have not adequately addressed risk stratification for postoperative pancreatic fistula (POPF). The aim of this study is to determine the clinical advantage of external drainage of the MPD and CBD over internal drainage during pancreatoduodenectomy.
Methods
This multicentre randomized clinical superiority study is designed to compare the effects of external and internal drainage of the MPD and CBD on the incidence of postoperative complications for patients at intermediate or high risk of POPF. In all, 322 eligible patients will be recruited across six pancreatic centres and randomly assigned 1 : 1 to either an external or internal drainage group. The primary outcome is the incidence of clinically relevant POPF (Grade B/C) within 90 days after surgery. The anticipated duration of enrolment is 1 year, along with a minimum follow-up period of 2 years, with follow-up visits every 3 months.
Conclusion
This trial will provide evidence for the efficacy of simultaneous external drainage of the MPD and CBD in the management of pancreatoduodenectomy, optimizing drainage strategies for pancreatoduodenectomy and facilitating the adoption of advanced drainage technologies. Registration number: NCT06322680 (http://www.clinicaltrials.gov); ChiCTR2400086321 (https://www.chictr.org.cn).
Teaser
This article describes the protocol for a randomized clinical trial comparing the clinical benefit of external versus internal drainage techniques in pancreatoduodenectomy. A patented drainage device for external drainage was designed and the trial plans to enrol patients at moderate to high risk of postoperative pancreatic fistula (POPF) because retrospective data suggest that external drainage potentially reduces the incidence of POPF.
Introduction
Pancreatoduodenectomy (PD) remains the only potential curative treatment for tumours located in the pancreatic head, lower bile duct, duodenum, and ampulla of Vater1, with a postoperative complication rate of approximately 54.7%2. Although technical advances have reduced mortality to below 5.0%, postoperative complications such as postoperative pancreatic fistula (POPF), delayed gastric emptying, postpancreatectomy haemorrhage, and bile leakage persist as major challenges, even in high-volume centres3,4.
In recent years, pancreatic duct stenting during pancreaticojejunostomy, using either internal or external drainage techniques, has emerged as a strategic intervention to mitigate the incidence of severe complications, especially clinically relevant POPF (CR-POPF)5–7. No consensus has been established regarding the comparative efficacy of external versus internal pancreatic duct drainage. A randomized equivalence trial led by Jang et al.8 demonstrated a numerical trend favouring internal drainage, although equivalence was not established. A retrospective review of 316 patients who underwent PD with either internal or external drainage of the main pancreatic duct (MPD) revealed that the severity of the fistula was reduced with the use of an external versus internal drainage pattern9. Meta-analysis results10 ,11 also support the potential superiority of external drainage in reducing the incidence of high-grade POPF. Preliminary retrospective analyses12–14 suggest the efficacy of external transanastomotic stenting in mitigating biliary complications and POPF. The rationale for concurrent external drainage of the MPD and common bile duct (CBD) is to prevent perianastomotic fluid accumulation through luminal decompression during the early postoperative phase, and to segregate pancreaticobiliary secretions to avert bile-mediated enzyme activation. Thus, this multicentre randomized clinical trial is being undertaken to determine whether combined external drainage of the MPD and CBD can further reduce postoperative complications compared with internal drainage.
Methods
Study design
This trial is a prospective multicentre randomized clinical superiority study conducted in China that is recruiting patients scheduled for PD due to any related disease at participating research institutes. The risk of POPF is defined according to the Alternative Fistula Risk Score (A-FRS)15, which combines body mass index, pancreatic duct size (assessed on the most recent computed tomography scan or magnetic resonance imaging), and pancreatic texture (soft or not) evaluated during surgery15. The study will only enrol patients with an intermediate (5.0–20.0%) or high (≥20.0%) risk of POPF, who will be allocated in ratio of 1 : 1 to either the external drainage group or the internal drainage group at (Fig. 1). This protocol is version 2.1 and was modified on 8 January 2025. The study protocol adheres to the SPIRIT guidelines, as detailed in Table S1. The study has been registered with ClinicalTrials.gov (ID NCT06322680) and the Chinese Clinical Trial Register (ID ChiCTR2400086321).
Eligibility criteria
Inclusion criteria
Patients are eligible for inclusion in the study if they:
- are ≥ 18 and ≤80 years of age;
- have a preoperative clinical diagnosis including, but not limited to, tumours located in the pancreatic head, periampullary region, duodenum, and distal bile duct (according to clinical symptoms, imaging examinations, tumour markers.) requiring PD, regardless of previous neoadjuvant treatment history;
- are at intermediate or high risk of POPF according to the A-FRS;
- have an Eastern Cooperative Oncology Group performance status of 0 or 1; have a life expectancy of ≥ 12 weeks;
- have an American Society of Anaesthesiologists (ASA) score ≤ 2;
- do not have any serious dysfunctions of the blood, heart, lung, or autoimmune systems defined as white blood cell count ≥ 3 × 109/L, an absolute neutrophil count ≥ 1.5 × 109/L, a platelet count ≥ 100 × 109/L, haemoglobin levels ≥ 90 g/L, aspartate aminotransferase, serum glutamic oxaloacetic transaminase/alanine aminotransferase, and serum glutamic pyruvate transaminase ≤ 2.5 × the institutional upper limit of normal (ULN), total bilirubin ≤ ULN, and creatinine ≤ 1.5 × the ULN; and
- have provided written informed consent before treatment.
Patients requiring PD for other conditions, such as traumatic duodenal rupture, are also eligible for inclusion in the study.
Exclusion criteria
Patients will be excluded from the study if they:
- have metastatic disease or a locally advanced tumour (invasion of the superior mesenteric artery, celiac trunk, inferior vena cava, or abdominal aorta);
- have a history of endoscopic retrograde cholangiopancreatography or percutaneous transhepatic cholangiodrainage (PTCD) in the internal drainage control group, although patients with pre-existing PTCD randomized to the external drainage intervention group remained eligible;
- have active infection (greater than Grade 2 according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 4.0)16;
- are pregnant or lactating;
- have undergone major surgery within 4 weeks of starting the trial or have not recovered from the side effects of such surgery;
- underwent curative treatment for another malignancy < 3 years ago;
- with upper gastrointestinal bleeding within 4 weeks before surgery;
- Poor adherence, or inability/unwillingness to provide informed consent; and
- are unable or are unwilling to provide written informed consent.
Surgical intervention and perioperative management
In the external drainage group, approximately 2 cm of the pancreatic stump is preserved and anchored to the jejunal seromuscular layer using interrupted 3/0 polypropylene U-shaped sutures. An appropriately sized external drainage tube is retrogradely passed through two enterotomies, advanced along the jejunal lumen, and exteriorized at the pancreatic transection surface. The posterior wall of the pancreatic duct and the jejunal mucosa are sutured using 4/0 Prolene interrupted sutures, followed by suturing of the anterior wall of the pancreatic duct and the jejunal mucosa, again using interrupted sutures. The anastomosis is reinforced using 3/0 Prolene interrupted sutures through the anterior pancreatic wall and anterior jejunal wall. The outward end of the external drainage tube is secured using purse-string sutures. Referring to instructions provided with the drainage device, the drainage tube is exteriorized through the abdominal wall and connected to a drainage bag. For the hepaticojejunostomy, an enterotomy is created 10 cm distal to the pancreatic anastomosis. Interrupted 4-0 absorbable sutures are used for duct-to-jejunal mucosa approximation. The external drainage tube for the CBD is placed, secured, and exteriorized in a similar way to that described for external drainage of the MPD (Fig. S1).
In the internal drainage group, both the end-to-end invagination and the duct-to-mucosa anastomotic approach are permitted when performing pancreaticojejunostomy. A transanastomotic stent is placed to span the MPD and jejunal lumen without external diversion. Drainage tube placement is omitted during the hepaticojejunostomy, using direct mucosa-to-mucosa approximation with 4/0 absorbable sutures (Fig. 2).
For patients in the external drainage group with a past history of PTCD, the hepaticojejunostomy is performed at the same as for patients in the internal drainage group. This double external pancreatic–biliary drainage device has been authorized by the State Intellectual Property Office of China (Patent no.: ZL202222162808.2).
During the perioperative stage, prophylactic administration of broad-spectrum antibiotics and octreotide is indicated. Serum and drainage fluid amylase levels are monitored on postoperative days 1, 3, and 5. Routine abdominal computed tomography is performed on postoperative day 6. Detailed perioperative information and treatment information is recorded in the case report form (CRF). Surgical complication rates are reassessed at 90 days after the procedure. In the absence of complications, both external drainage tubes can be removed in the outpatient setting at this time. Tumour recurrence is assessed every 3 months for 2 years. Subsequent follow-up for survival outcomes is conducted via telephone.
Results
Primary outcome
The primary outcome of the study is the incidence of CR-POPF (grade B/C) within the 90-day postoperative period, assessed according to the stratified grading system proposed in 2016 by the International Study Group on Pancreatic Surgery17.
Secondary outcomes
Secondary outcomes include the incidence of other surgical complications, nutritional status, 90-day postoperative mortality and morbidities, and patient-reported outcomes, assessed via the 3-level version of the EQ-5D™ (EQ-5D-3L) questionnaire. Although the primary endpoint focuses on the incidence of CR-POPF, other complications related to PD, including bile leakage, chyle leak, gastrointestinal/intestinal anastomotic fistula, postpancreatectomy haemorrhage, delayed gastric emptying, intra-abdominal infection, postpancreatectomy acute pancreatitis, deep vein thrombosis of the lower limbs, pulmonary thromboembolism, myocardial infarction, cerebral infarction, and incisional complications, are considered as part of the secondary outcomes18–23 and are re-evaluated 90 days after surgery. The definitions and grading systems for the postpancreatectomy complications are listed in Table S2. The overall severity of surgical complications is assessed using the Clavien–Dindo classification system. Overall survival and disease-free survival related to malignant tumours are also measured as secondary outcomes. Overall survival is defined as the time from intervention to death from any cause; disease-free survival is calculated as the time from the intervention until tumour recurrence or death due to any cause.
Participant timeline
The study is estimated to span 3 years, with enrolment activities projected to be completed within 1 year at six designated pancreatic centres, followed by a minimum 2-year follow-up period. After confirmation of patient eligibility and obtaining written informed consent, patients will undergo comprehensive baseline assessment. Eligible patients with a moderate or high risk of POPF determined during surgery will be randomly allocated to the external or internal drainage group. A representation of the participant timeline is presented in Table 1.
| Screening | Surgery | Re-evaluation 90 days after surgery | Long-term follow-up | |
|---|---|---|---|---|
| Demographics | × | |||
| Medical history and physical examination | × | × | × | |
| Blood chemistry, tumour markers, testing of hepatic and renal function | × | × | × | |
| Imaging examination | × | × | × | |
| Cardiopulmonary function | × | |||
| Preoperative treatment | × | |||
| ECOG scores | × | |||
| Urine pregnancy test | × | |||
| ctDNA | × | |||
| Informed consent | × | |||
| Laboratory testing of blood, urine, and faeces | × | |||
| Surgical information | × | |||
| Pathological information | × | |||
| Complications monitoring and recording | × | |||
| Costs and length of hospitalization | × | |||
| Adjuvant therapy | × | |||
| Pathological examination before second-line treatment | × | |||
| Second-line therapy | × | |||
| Confirmation of tumour recurrence/metastasis status | × | |||
| Confirmation of tumour progression and related survival | × |
Sample size
The trial is structured as a superiority study with the primary objective being to demonstrate that the external drainage pattern used in pancreatoduodenectomy is superior to the internal drainage pattern in preventing POPF. Based on preliminary clinical data from Zhongshan Hospital collected at an earlier stage, the incidence rate of CR-POPF in the control group in which internal drainage is used is 25%, and it is predicted that this will decrease to 10% in the intervention (external drainage) group. To determine the minimum sample size necessary for detecting a significant difference, Fisher’s exact test is used, with a significance level (α) of 5% and a power (β) of 90%. Using these parameters and PASS 2021 statistical software (US, Utah, Kaysville), it was determined that 292 patients are required. Considering a potential dropout rate of 10%, the final recruitment target is set at 322 patients, with an equal allocation of 161 patients to the control and intervention groups. A total of 322 patients will be enrolled from across six different research institutes, with Zhongshan Hospital accounting for 172 patients.
To meet the recruitment goal, the enrolment period will extend for a full year. Neither the participants nor researchers will be provided with any form of financial remuneration.
Randomization and blinding
Drainage (internal or external) assignment is conducted using a simple randomization without blocks and stratification. To reduce selection bias at the preoperative stage, a random number table is generated first by an independent statistician using statistical software (SPSS® 23.0; IBM, Armonk, NY, USA) at the leading unit. The random numbers and interventions are then sealed in sequentially numbered opaque envelopes. Following intraoperative exploration and assessment of pancreatic texture to confirm eligibility, the primary surgeon contacts the research assistant. The assistant then opens the sealed envelope to disclose the randomly assigned pancreato-intestinal and bilio-intestinal anastomotic approaches to the surgical team. This process ensures that investigators responsible for patient screening and recruitment have no access to the allocation sequence, and the statistician and research assistant are not involved in clinical decision-making and patient care. Within the study, participants and investigators are unable to predict allocations at the preoperative stage, but are unavoidably unblinded to the exact drainage pattern after surgery due to the visible presence of external drainage tubes intrinsic to the intervention. Data collectors and outcome assessors will also be unblinded after the procedure, because the surgical details and outpatient records are not concealed. Data analysts are kept blinded to minimize analytical bias and to ensure objectivity of study outcomes. This partial blinding is justified because the endpoints—definite grading of complications and survival status—are objective measures, minimizing the assessment bias despite non-blinding.
Data collection and management
All data in this study will be collected using an electronic CRF, which will be reviewed, approved, and signed through an electronic data capture (EDC) system. The researchers are responsible for ensuring the timely, accurate, complete, and legible entry of data. All entries, modifications, and replacements in the CRF must be completed by researchers or other authorized personnel. The principal researchers or their designated representatives should complete and sign the CRF for each subject, including those who fail the screening process. If a subject decides to withdraw from the study, the reasons for withdrawal should be documented in the CRF. According to the study timeline, researchers should document all observations of note in both the original medical record and the CRF at the time of each planned visit.
Statistical analysis
Statistical analyses will be performed using SPSS® 23.0 and R version 4.2.1 (R Foundation for Statistical Computing, Vienna, Austria), whereas the sample size will be calculated using PASS 2021. A two-tailed P value of < 0.05 is considered statistically significant. With the use of the EDC system, researchers or a trained and authorized clinical coordinator must promptly input the data required by the trial protocol into the EDC system. All statistical analyses are conducted on the intention-to-treat population. Continuous variables will be presented as the mean with standard deviation, and categorical variables will be presented as the number and proportion of patients in each category. Missing values will be addressed via multiple imputation by chained equations under the missing-at-random assumption, with sensitivity analyses to assess robustness. All between-group comparisons of continuous variables will be made using Welch’s t test, which does not assume homogeneity of variances, with adjusted degrees of freedom explicitly reported. Non-parametric Mann–Whitney U tests will be used in scenarios where either the sample size is insufficient (< 20 patients per group) or severe deviations from normality are objectively confirmed via Q-Q plots or Shapiro–Wilk tests (P < 0.05 for non-normality). Categorical variables will be compared using χ2 tests or the Fisher’s exact test. Effect sizes for assessed outcomes, primarily the incidence of complications, will be reported as absolute risk differences with 95% confidence intervals. To address multiplicity concerns, a hierarchical fixed-sequence testing procedure will be prospectively implemented. The primary endpoint evaluating the 90-day incidence of CR-POPF will be tested at a two-sided α = 0.05. The sequential analysis of secondary endpoints, including postoperative complications and nutritional status, proceeds only if the primary endpoint achieves statistical significance (P < 0.05) following a prespecified testing hierarchy. The procedure terminates at any non-significant result (P ≥ 0.05) in the sequence, requiring no additional multiplicity adjustments. Stratified analyses using the Cochran–Mantel–Haenszel test will be conducted, with adjustments for the predefined risk stratification and study centre effects. Prespecified subgroup analyses are stratified according to pancreatic texture (soft versus hard), pancreatic duct size (< 5 versus ≥ 5 mm), and body mass index categories (< 24 versus ≥24 kg/m2). To identify potential risk factors for POPF, univariable and multivariable logistic regression models will be implemented, incorporating clinically relevant covariates.
Ethical considerations
The study protocol and written informed consent form have been reviewed and were approved by the Medical Ethics Committee of Shanghai Zhongshan Hospital (Approval no. B2023-401R) in November 2023. The study protocol has also received ethics approval from another five centres. Researchers from the six participating centres will meet annually to compile an annual multicentre research report, which will be submitted to the respective ethics committees. The researchers must provide written notification to the ethics committees upon discontinuation and/or completion of the study. Any changes in the practical work, such as revisions to the protocol and informed consent, must be reported to the ethics committees and cannot proceed without approval from the ethics committees, unless they are intended to eliminate obvious and direct risks to subjects.
Discussion
There has been a growing focus on developing effective strategies to mitigate postoperative complications after PD, along with increased research in recent years, involving modified anastomotic methods used in pancreaticojejunostomy, the placement of ductal stents24, drainage patterns, pancreatogastrostomy25, and the appropriate use of somatostatin analogues26, among others27–29. In terms of drainage patterns, internal pancreatic stenting provides advantages by reducing digestive fluid loss, enhancing functional recovery, and simplifying surgical procedures, particularly in laparoscopic PD29. Studies stratifying POPF risk according to the Fistula Risk Score (FRS) have demonstrated30,31 that externalized stents significantly reduce the incidence of POPF in intermediate- and high-risk patients. A prospective trial30 (540 patients) stratified by the original 10-point FRS revealed that externalized stents without malfunction significantly reduced the mean complication burden in patients at intermediate risk and the rate of CR-POPF in the high-risk cohort. External stent application in patients with an FRS ≥ 4 resulted in a 19.4% reduction in the incidence of CR-POPF compared with the internal stent group, as evidenced by a retrospective study31. The trial described herein will be the first to use a novel and patented device for external drainage of bile and the pancreatic duct. This device provides three alternatives for tube diameter (6, 10, and 14 Fr) and a blunt end to minimize the risk of puncture and bleeding.
The proposed study has several limitations. Full blinding was not possible due to the inherent nature of the surgical interventions, resulting in an unavoidable open-label design for surgeons and patients. The exclusion of low-risk POPF patients (predicted POPF probability < 5.0% by the A-FRS) precludes assessment of drainage benefits in this subgroup. Future studies should evaluate the potential benefit versus adverse effects of external drainage in low-risk cohorts. The exclusive inclusion of pancreatic centres from China may limit the generalizability of the efficacy of external drainage to other countries. Multicentre international validation of the results is warranted.
The study results will provide compelling evidence for the efficacy of simultaneous external drainage of the MPD and CBD in the management of PD, particularly in reducing the incidence of CR-POPF in high-risk patients. These findings have the potential to optimize drainage strategies of PD and facilitate the adoption of advanced drainage technologies.
Supplementary Material
Funding
This work was supported by grants from Zhongshan Hospital (2023ZSCX21), the National Natural Science Foundation of China (82473459, 82303443, 82273382, 82272929), Program of Shanghai Academic/Technology Research Leader (23XD1400600), Shanghai Science and Technology Commission Innovative Pharmaceutical Products Application Demonstration Project (24SF1900300), Shanghai Municipal Health Commission (20244Y0023, 201940019), Shanghai Magnolia Talent Plan Pujiang Programme (24PJD012), Shanghai Anticancer Association EYAS PROJECT (SACA-CY24B09), Beijing Xisike Clinical Oncology Research Foundation (Y-2022METAZQN-0003, Y-Gilead2024-PT-0002, Y-HR2022MS-0251, Y-HR2022QN-0085), Clinical Research of Zhongshan Hospital (ZSLCYJ202329), and Liu Liang Expert Workstation of Yunnan Province (202305AF150148). The funding agencies had no role in the study design, data collection and analyses, decision to publish, or preparation of the manuscript.
Disclosure
The authors declare no conflict of interest.
Supplementary material
Supplementary material is available at BJS Open online.