Early Drain Removal Versus Standard Drain Management After Distal Pancreatectomy (Early-Dist)

RATIONALE AND RESEARCH QUESTION Distal pancreatectomy (DP) is defined as the resection of the left portion of the pancreas extending from the neck to the tail of the gland. DP is burdened by the occurrence of clinically-relevant postoperative pancreatic fistula (CR-POPF) in a significant proportion of patients.Traditionally, during DP drains were placed intraoperatively in proximity of the pancreatic stump to mitigate the potential occurrence of POPF. Postoperatively, drains were usually kept in place until a POPF was ruled out by the presence of drain fluid amylase (DFA) values lower than 3-fold the upper limit of normal serum amylase, and a non "sinister" appearance of drain fluid in accordance with the International Study Group on Pancreatic Surgery (ISGPS) definition. For this reason, drains were usually maintained in place after surgery for at least 5 - 7 days. Although there is some preliminary evidence that early drain removal in the subgroup of patients with DFA1 < 2,000 U/L, no prospective study has yet evaluated the impact of an early drain removal strategy compared to standard management. Therefore, the overall aim of this study is to contribute evidence about the effect of an early drain removal policy on postoperative outcomes following distal pancreatectomy.

The primary research question is to evaluate to what extent does early postoperative drain removal according to a validated DFA1 impact on clinically-relevant POPF rate after distal pancreatectomy in comparison to standard drain management. The primary (confirmatory) hypothesis is that, early drain removal will result in a reduced proportion of patients experiencing grade B-C POPF according to ISGPS definition.

Furthermore the exploratory hypotheses that, in comparison to standard care, early drain removal will: (1) reduce time to postoperative functional recovery, (2) reduce number and severity of postoperative complications, (3) reduce postoperative surgical site infections, (4) improve self-reported physical function status, (5) improve generic health-related quality of life and (6) be cost-effective, will be tested.

STUDY DESIGN AND PARTICIPANTS The study is designed as a two-group, assessor-blind, randomized trial. Participants will be randomly assigned with a 1:1 ratio into one of two groups: (1) standard drain management or (2) early drain removal strategy. The intervention will be provided during hospital stay. Adult people with pancreatic body or tail diseases (i.e. pancreatic cancer, cystic neoplasms, pancreatic neuroendocrine tumors, etc.) planned for distal pancreatectomy with or without splenectomy at the San Raffaele Hospital IRCCS (HSR) will be considered for inclusion.

RECRUITMENT PROCESS

The step-by-step recruitment process is described below:

1. Patients scheduled for elective distal pancreatectomy who meet the inclusion/exclusion criteria will be informed about the study by a surgeon at their first office consultation
2. If the patient has interest in participating in the study, at the preoperative clinic visit or at the time of hospital admission, a trial investigator will discuss the study in detail with the patient
3. The potential participant or their legal representative will read the consent form. Once the informed consent is signed and the patient enrolled in the study, preoperative assessment will be undertaken

SURGICAL TECHNIQUE The standard surgical procedure for DP will be performed and will not be influenced by this study protocol. The choice of minimally invasive versus open technique will be at surgeon's discretion. As a general rule, patients with benign lesions or neuroendocrine tumors will be approached laparoscopically, while pancreatic cancer cases are discussed on an individual basis according to tumor location and vessel proximity. Splenectomy and standard lymphadenectomy will be routinely performed in cancer patients. A spleen preserving procedure with preservation of splenic vessels will be attempted only in non-neoplastic cases at surgeon's discretion. Pancreatic stump transection will be performed with a stapler in laparoscopic procedures, whereas in open procedures the pancreas will be cut using the scalpel followed by selective suturing of the pancreatic duct when visible.

To evaluate the association between pancreatic parenchymal structure and occurrence of POPF, in addition to standard pathology assessment, the pancreatic specimen will undergo a specific pathological evaluation. Haematoxylin and eosin (H\&E) stained sections (between 1 and 4 for each case) of the pancreatic resection margin will be reviewed by two expert pancreatic pathologists, blinded to all clinical information. All the definitive slides will be assessed for their acinar, fat and fibrosis content as a proportion of the total surface area. Acinar, fat and fibrosis scores will then calculated for each patient so that their sum is equal to was 100%. The final score for each patient will be the result of two pathologists reached a consensus agreement by the two pathologists.on the scoring of each patient.

DRAIN MANAGEMENT

At the end of the operation, after reassessing exclusion criteria, participants will be randomized into one of two groups:

1. Standard postoperative drain management: Participants randomized to this group will receive the current standard of care at our institution. The drain will be maintained at least until postoperative day 5. If DFA on POD 5 is lower than 3-fold the upper limit of normal serum amylase value, and/or the drain fluid does not appear "sinister" (i.e. varying from dark brown to greenish fluid to milky water to clear "spring water" pancreatic juice like appearance) according to the clinical team, the drain will be removed. If the drain is maintained in place beyond POD 5, it will be removed inhospital or after discharge when the daily amount is lower than 10 ml and there are no signs of ongoing infection.
2. Early drain removal: Participants randomized to this group, will have an early drain removal (before POD 3) if DFA on POD1 is lower than 2000 U/L and/or the drain fluid does not appear "sinister" (i.e. varying from dark brown to greenish fluid to milky water to clear "spring water" pancreatic juice like appearance) according to the clinical team. If DFA is greater than 2000 U/L, the drain will be maintained in place. If the sample on POD1 DFA is not available (e.g. drainage fluid is viscous and amylase value cannot be evaluated) or the sample could not be analyzed (e.g. the fluid collection tube is lost or not sealed) DFA will be evaluated on POD2, and drain removal will be considered if DFA is lower than 2000 U/L. If DFA on POD 3 is lower than 300 U/L, that is 3-fold the upper limit of normal serum amylase value, and/or the drain fluid does not appear "sinister" (i.e. varying from dark brown to greenish fluid to milky water to clear "spring water" pancreatic juice like appearance) according to the clinical team, the drain will be removed. Otherwise, the drain will be kept in place and .the patient will then follow standard postoperative drain management as described above.

All patients will measure serum amylase and drain fluid amylase along with other routine postoperative laboratory tests on POD 1 - 3 - 5 as the current standard of care.

Furthermore, as drain fluid microbial contamination is considered to play a potential role for the POPF occurrence, a drain fluid culture for aerobic and /anaerobic microorganisms will be performed on POD1 (baseline) and POD5, in order to evaluate the predictive ability of early bacterial contamination in the abdominal fluid for clinically relevant POPF.

MEASUREMENTS AND OUTCOMES Demographic data, medical history, laboratory values and information relevant to the surgical procedure will be recorded. The main (confirmatory) outcome of interest will be the occurrence of CF-POPF at 90 days after surgery, defined as grade B or C POPF according to the 2016 ISGPS definition.

The proportion of patients discharged home without an abdominal drain represents the explanatory outcome measure. It will be evaluated at time of hospital discharge.

Time to functional recovery (TFR) will be measured by subtracting the date of surgery from the date when participants achieves standardized criteria (tolerance of oral intake, recovery of lower gastrointestinal function, adequate pain control on oral analgesia, ability to mobilize and self-care and no evidence of untreated medical problems).

Postoperative complications occurring during the index admission or in a subsequent readmission, when related to surgery, will be recorded up to 90 days after surgery and graded by severity using the Dindo-Clavien classification. Additionally, the Comprehensive Complication Index will be calculated.

Postoperative morbidity at 90 days after surgery will include:

• Surgical site infections classified as superficial incisional, deep incisional or organ-space according to the definition by the Center for Disease Control and Prevention
• Post-pancreatectomy hemorrhage and delayed gastric empting as defined by the ISGPS Self-reported activity status will be measured using the Duke Activity Status Index (DASI) and Generic heath related quality of life will be measured using the Patient Reported Outcome Measure Information System (PROMIS)-29+2 Profile v2.1. These questionnaires will be completed before surgery, at 15, 30 and 90 days after surgery. Cost-effectiveness will be estimated by the ratio of differences in costs to differences in quality-adjusted life years (QALYs) between the two groups.

BLINDED ASSESSMENT An examiner blinded to the participants' treatment allocation will obtain all measures. The code for group allocation will not be revealed until all analyses are completed. The effectiveness of blinding will be estimated by asking the assessor to guess the participant's group allocation at the completion of the 90 days assessment. It is not possible to blind the participants to their group assignment.

STATISTICAL CONSIDERATIONS AND TIMELINES Recent studies carried at the HRS Division of Pancreatic Surgery showed that, when standard drain management is used, the proportion of patients developing CR-POPF is approximately 46%. Sample size requirement for the present study was estimated for an α level of 0.05 and 80% power to detect a 23% reduction in this proportion (risk ratio 0.23/0.46=0.5). According to this estimate, a sample of 66 participants per group is considered sufficient for our analysis. A sample size of 75 participants per group (total sample of 150) is targeted to account for possible dropouts. Regression models will be used to test the main hypothesis related to the superiority of early drain removal based on a reduction of CR-POPF at 90-days after surgery. The randomization list will be created with the "Block randomization" rule, in order to reduce bias and achieve balance in the allocation of participants to the two arms.

The Division of Pancreatic Surgery at HSR performs about 90 distal pancreatectomies annually. Based on previous trial experience, approximately 90% of these patients will be eligible and agree to participate. Therefore, 150 participants could be feasibly enrolled in 21 months. With an additional 90 days follow-up required for the primary outcome, the timeline for this study is 24 months.

SAFETY Previously completed studies have shown that early drain removal in patients with DFA1 values lower than a validated threshold in patients undergoing pancreatic head resection is not only safe but beneficial, as it reduces the incidence of CR-POPF and the risk of further complications due to the persistence of a foreign body