First-day Versus Early Drain Removal Following PD: a Randomized Controlled Trial

Introduction Pancreatic surgery is a complex discipline with a high risk of post-operative morbidity such as pancreatic fistula (POPF) with variable impact on clinical outcome, fluid collection and hemorrhage. For such reason the intraoperative placement of intra-abdominal drains (ID) to early detect, mitigate, and manage post-operative complications is still considered to be a routine practice.

Controversies on ID placement have emerged from both randomized and non-randomized clinical studies investigating its possible role in increasing POPF and postoperative morbidity as well as prolonged hospital stay. According to Conlon et al the presence of drains failed to reduce either the need for interventional radiologic drainage or surgical exploration for intraabdominal sepsis and was associated with the development of more intra-abdominal abscesses, collections, or fistulas. Some authors report significantly higher postoperative abdominal complications after late removal of ID. A prolonged period of drain insertion is associated with a higher rate of postoperative complications with increased hospital stay and costs, not to mention that dislocation of intra-abdominal drains is an early and frequent event after major pancreatic resection.

The optimal timing for drain removal after PD is still a subject of debate: most studies have shown that outcomes are best when ID are removed in postoperative day (POD) 3 when POPF is excluded (given a specific cut off for drain fluid amylase, DFA).

Even though many authors advocate early ID removal to prevent intra-abdominal infection or fistula, in clinical practice many surgeons may feel more comfortable to maintain IDs for a prolonged period (more than 3 days) even when criteria for POPF or other abdominal complications are not met.

In some studies correlation between drain fluid amylase (DFA) in POD1 after pancreatic resections and development of POPF has been examined, defining ideal cut-off points ranging between 5.000U/L and 90 U/l3-17-20-21 (depending on intraoperative risk stratification) that can be used to achieve the best timing for drain removal.

AIM In this non-inferiority study, we aim to compare postoperative surgical outcomes after PD in patients with low/medium risk for POPF (ISGPS risk class A-B-C), who undergo POD1 drains removal versus POD3 removal.

Primary aim: occurrence of one or more pancreas-specific complications, defined according to ISGPS as:1) grade B/C POPF; 2) post-pancreatectomy hemorrhage (PPH).

Secondary aims: occurrence of fluid collection, sepsis, SSI, need for reintervention, length of stay, CD>3, 90 days mortality, re-admission.

Patients and Methods:

Study Design:

This is a randomized, controlled, open-label study. The Study will be registered in the International Standard Randomized Controlled Trial Registry.

All patients encountering inclusion criteria will be randomized on POD1, using computer-generated randomization codes.

Group A: Drain removal on POD 1 (in case of POD 1 DFA< 300 U/L) Group B: Drain removal on POD 3 (in case of POD 1 DFA< 300 U/L)

Operative technique and drains positioning:

Experienced pancreatic surgeons will be performing a standard pancreaticoduodenectomy pylorus-preserving and a Whipple procedure if lesion infiltrates the duodenum. After resection, anastomoses will be constructed on a single jejunal loop. Management of pancreatic stump consisting in single- or double-layer, end-to-side pancreatojejunostomy (PJ) with non-absorbable interrupted sutures. Given the low risk for pancreatic fistula, no externalized trans-anastomotic stent will be placed. End-to-side hepaticojejunostomy will be performed 20 cm distally to PJ with absorbable continuous or interrupted sutures, and end-to side duodenojejunostomy with absorbable interrupted sutures. Two perianastomotic open passive penrose (Easyflow) drains are placed at the end of the procedure close to biliary anastomosis and close to PJ anastomosis15.

In conclusion, we aim, through this randomized trial, to demonstrate that POD1 drains removal after PD is not inferior to POD3 removal in low and medium risk patients and, moreover, it could prevent delayed removal-related complications, helping to guide post-operative care and improve patient outcomes.