Prophylactic active irrigation drainage reduces the risk of post-operative pancreatic fistula-related complications in patients undergoing limited pancreatic resection.

OBJECTIVE: The objective of this study is to evaluate the efficacy of prophylactic active irrigation drainage in preventing post-operative pancreatic fistula (POPF) and POPF-related complications in patients undergoing limited pancreatic resection (LPR).
MATERIALS AND METHODS: Patients who underwent LPR for benign or borderline pancreatic lesions between February 2014 and March 2019 were enroled in this retrospective study. Patients were divided into two groups according to the type of intraperitoneal drainage used: closed-suction drainage (CSD) or continuous active irrigation drainage (CAID). Data regarding the outcomes and complications of surgery were collected and analysed.
RESULTS: A total of 50 patients (33 women; age, 50.1 ± 10.8 years) were included in this study. Twenty-nine patients were treated with CSD, and 21 patients were treated with CAID. Clinically relevant POPF and POPF-related complications occurred in 11 patients in the CSD group and in two patients in the CAID group ( P = 0.024). Patients in the CSD group demonstrated a longer tube indwelling time than those in the CAID group (17.1 ± 10.2 days vs. 13.7 ± 7.5 days; P = 0.044). Mean post-operative hospital stay was also longer in the CSD group than in the CAID group (20.6 ± 7.9 days vs. 16.1 ± 6.3 days; P = 0.039).
CONCLUSIONS: Prophylactic CAID appears to be an effective alternative for the management of POPF and POPF-related complications in patients undergoing LPR.

INTRODUCTION

Post-operative pancreatic fistula (POPF) is a common complication after pancreatic surgery, even limited pancreatic resection (LPR); research has shown that approximately 60%–91% of patients undergoing LPR experience POPF.[123456] Clinically relevant POPF can lead to severe complications such as haemorrhage or sepsis, with a mortality rate as high as 40%; indeed, POPF-related complications are the main cause of mortality after pancreatic surgery.[7]

To reduce the incidence of POPF and POPF-related complications after LPR, various methods have been proposed, such as prophylactic Roux-en-Y pancreaticojejunostomy after enucleation,[8] the use of fibrin sealants to seal the pancreatic wounds[9] and the use of somatostatin analogues to reduce pancreatic secretion.[10] However, no consensus has been reached regarding the optimal method for preventing the occurrence of POPF and POPF-related complications.[110111213]

Another method used to prevent and treat POPF and POPF-related complications after pancreatic surgery is intraperitoneal drainage.[11] This drainage is usually achieved through the use of single-lumen drainage tubes with or without negative pressure.[2] However, these drainage tubes are prone to blockage, leading to incomplete drainage and potentially to the occurrence of POPF or POPF-related complications.[3] Active irrigation drainage using multiple tubes may prevent this blockage. However, to our knowledge, no study has assessed the use of active irrigation drainage after LPR. In this study, we, therefore, sought to evaluate the efficacy of prophylactic continuous active irrigation drainage (CAID) versus that of closed-suction drainage (CSD) in preventing POPF and POPF-related complications after LPR.

MATERIALS AND METHODS

Data collection

This study was approved by our Institutional Review Board (No. 2018022). Patients who underwent LPR between February 2014 and March 2019 were searched. Patients were included in the study if they had benign or borderline pancreatic lesions and a solid pseudopapillary tumour, had good organ function, had complete medical data available and had their procedure performed by one given team led by two experienced senior surgeons. Patients were excluded if they had organ dysfunction, had incomplete medical data, had their procedure performed by a different team or underwent an LPR that was converted to pancreaticoduodenectomy or distal pancreatectomy.

Between February 2014 and December 2016, CSD was routinely performed for patients undergoing LPR at our institution; however, a high rate of POPF and POPF-related complications was observed when CSD was used. Thus, our institution switched to the use of CAID between January 2017 and March 2019. For this study, we divided the cases into two groups based on the type of drainage used (CSD group and CAID group).

Definitions

POPF was defined as peripancreatic drainage persisting for more than 3 weeks and amylase activity in the drainage fluid more than three times above the upper limit of normal serum value according to the International Study Group on Pancreatic Surgery.[7] POPF-related complications were defined as serious problems resulting from POPF, including intraabdominal haemorrhage, intraabdominal infection, abscess, organ failure, re-operation and/or death.[7]

Procedures

LPR procedures were performed according to a general consensus. For enucleation to be performed, the tumours had to be located at least 3 mm away from the main pancreatic duct.[456] All tumours were completely resected with or without the inclusion of some adjacent pancreatic tissues. The main pancreatic duct was carefully preserved, and the tissue defects were covered with omental flaps. For the Beger procedure, tissue from the pancreatic head was resected between the portal vein and the intrapancreatic segment of the common bile duct, with a left pancreaticojejunostomy performed and a side-to-end anastomosis created between the jejunal loop and the remainder of the pancreatic head.[14] For central pancreatectomy, a middle segment resection and a left pancreaticojejunostomy were performed. The proximal pancreatic stump was closed with mechanical devices or with manual suturing.

In the CSD group, one 20 French single-lumen silicone tube with multiple side holes (Aiyuan, Taizhou, China) was placed near the pancreatic defects or anastomoses. The tube was connected to a drainage bag, which was kept at constant negative pressure (2–4 kPa). In the CAID group, a handmade device called a double catheterisation cannula was placed near the pancreatic anastomoses or defects [Figure 1]. This device consisted of a 22 French outer catheter with multiple side holes (Aiyuan), a 10 French inner drainage tube with multiple side holes (Yaxin Medical, Suzhou, China) and an 8 French irrigation tube (Yaxin Medical). Normal saline was perfused into the abdominal cavity at a rate of 1 mL/min through the 8 French afferent tube. The 10 French efferent tube was connected to a drainage device that maintained negative pressure (2–4 kPa) of the tube.

Figure 1

Continuous irrigation. A single-lumen irrigation tube was placed near the pancreatic anastomoses or wound (blue), and a multiside-hole drainage tube (orange) was placed into a multiside-hole casing (yellow) that was positioned next to the irrigation tube. (a) The hand-made device called was made up of a French 22 multiside-hole outer catheter, a French 10 multiside-hole inner drainage tube and a French 8 irrigation tube. (b) Working principle schematic diagram of the device. (c-e) Schematic diagram illustrating continuous irrigation for enucleation procedure, Berger procedure and central pancreatectomy

Post-operative care

All patients were closely monitored for procedure-related complications and were treated with guideline-based therapy after the procedure.[15] The amylase levels in the drainage fluid were measured every other day. Plain computed tomography (CT) scan was routinely performed 5 or 7 days after surgery to evaluate whether there was fluid collection or infection in the abdominal cavity, especially around the pancreatic jejunum anastomosis. If any POPF-related complications were identified, the appropriate treatment was immediately initiated. In both the groups, the drainage tube was removed when no obvious abdominal collection was seen on CT with the amylase level of the drainage fluid less than three times the upper limit of normal serum value.

In CAID group, the actual amylase value was calculated as follows: actual amylase value = measured amylase value × (volume of measured drainage fluid/actual volume of drainage fluid).[16] If the amylase level was less than three times the upper limit of normal serum value, irrigation was stopped and constant negative pressure as in the CSD group was adopted; otherwise, irrigation was continued.

Statistical analysis

Comparisons were performed using an independent sample t-test or Chi-square test or Fisher's exact test. The SPSS software version 17.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis.

RESULTS

Patients

A total of 50 patients (17 men and 33 women; age: 50.1 ± 10.8 years [range, 31–69 years]) were included in the study. The diagnosis for all patients was based on rapid intraoperative pathological analysis during surgery; the diagnosis was confirmed by routine paraffin histological or immunohistochemical examination after surgery. Of the 50 patients, 29 (58%) were treated with CSD, and 21 (42%) were treated with CAID. There were no significant differences in patient characteristics between the two groups [Table 1]. Mean drainage tube indwelling time and mean post-operative hospital stay were significantly shorter in the CAID group than in the CSD group [P < 0.05; Table 1].

Table 1

Patient characteristics

Characteristics	CSD group (n=29)	CAID group (n=21)	P
Age (years)	60.1±13.0	54.9±15.3	0.125
Sex
Male	10	7	0.933
Female	19	14
Body mass index	20.3±4.8	21.1±5.2	0.835
Serum laboratory value
Total bilirubin (µmol/L)	13.9±5.3	15.5±5.8	0.405
ALT (U/L)	32.9±9.4	30.4±11.7	0.692
AST (U/L)	27.0±10.8	29.3±12.9	0.437
Albumin (g/L)	37.9±6.2	36.8±4.6	0.585
Pancreatic lesion location
Head	16	13	0.892
Neck	8	5
Body	5	3
Diameter of the lesion (mm)	34.1±12.5	31.8±6.3	0.381
Procedure
Berger procedure	7	6	0.923
Enucleation	14	10
Central pancreatectomy	8	5
Pathology
Serous cystadenoma	4	4	0.977
Mucinous cystadenoma	6	5
Intraductal papillary mucinous neoplasm	7	3
True pancreatic cyst	1	1
Insulinoma	3	2
Solid pseudopapillary tumour	4	2
Chronic pancreatitis	3	2
Neuroendocrine tumour	1	2
Procedure time (min)	201.7±89.5	186.0±51.2	0.490
Intraoperative blood loss (mL)	155.4±130.2	146.1±122.6	0.857
Drainage tube indwelling time (days)	17.1±10.2	13.7±7.5	0.044
Postoperative hospital stay (days)	20.6±7.9	16.1±6.3	0.039

ALT: Alanine aminotransferase, AST: Aspartate aminotransferase, CAID: Continuous active irrigation drainage, CSD: Closed-suction drainage

Post-operative pancreatic fistula and post-operative pancreatic fistula-related complications and outcomes

Clinically relevant POPF and POPF-related complications occurred in 11 patients (37.9%) in the CSD group and in two patients (9.5%) in the CAID group (P < 0.05); this reduced the incidence of POPF and POPF-related complications in the CAID group was primarily due to a reduced incidence of Grade B complications (P = 0.041) [Table 2]. A total of three patients suffered from POPF-related haemorrhage in the CSD group; two patients were successfully treated with interventional embolisation (Grade B) and one patient was treated with reoperation (Grade C). Three patients of POPF-related abdominal abscess, including two patients in the CSD group and 1 in the CAID group, were successfully managed with percutaneous drainage (Grade B). No inhospital deaths occurred in either group. All patients were successfully treated and discharged.

Table 2

Post-operative pancreatic fistula and post-operative pancreatic fistula-related complications

Complications	CSD group (n=29)	CAID group (n=21)	P*
Grade B	10	2
Drainage persisting for >3 weeks	6	1	0.041
POPF-related haemorrhage	2	0
POPF-related abdominal abscess	2	1
Grade C
Reoperation for POPF-related haemorrhage	1	0	0.580
Grade B and C, n (%)	11 (37.9)	2 (9.5)	0.024

*One-sided Fisher’s exact test. CAID: Continuous active irrigation drainage, CSD: Closed-suction drainage, POPF: Post-operative pancreatic fistula

DISCUSSION

Intraoperative placement of drains is common among patients undergoing pancreatic resection. Simple passive drainage and CSD are frequently used; simple passive drainage uses gravity alone (with intraabdominal pressure) to evacuate fluids, whereas CSD uses active suction through a negative pressure device. These techniques lead to similar post-operative outcomes.[2] With both systems, the maintenance of drain patency is essential to achieve complete evacuation of pancreatic fluid collections. Unfortunately, the presence of non-liquid materials such as necrotic tissues, thick pus and absorbable or non-absorbable sutures can lead to the obstruction of these drains, eventually resulting in inadequate drainage or drainage failure.

The double catheter irrigation-suction drainage system was first introduced approximately 50 years ago for the management of enterocutaneous fistulas.[17] This device consists of a large catheter pipe, an efferent suction pipe and an irrigation afferent pipe [Figure 1]. It is not complex for surgeons to make this device. Continuous irrigation with sterilised saline and continuous suction through low negative pressure keeps the drain system unimpeded, and collections such as pancreatic fluid and non-liquid materials can be effectively suctioned out following the irrigating water stream. This device has been recommended as a more effective method of drainage than simple passive drainage or CSD, leading to a reduced incidence of intraabdominal infection, abscess, haemorrhage and pancreatic fistula.[31819]

In this study, we found that the incidence of POPF and POPF-related complications was 37.9% in patients receiving CSD after LPR and 9.5% in those receiving CAID after LPR. Patients receiving CAID also demonstrated significant reductions in drainage tube indwelling time and length of post-operative hospital stay. A number of factors have previously been shown to increase the post-operative length of stay, including the occurrence of POPF and POPF-related complications.[20] The reduction in the incidence of clinically relevant POPF and POPF-related complications and decreased drainage tube indwelling time we observed with CAID in this study were likely contributors to this reduced length of hospital stay.

We also found that fewer patients in the CAID group had drainage persisting for more than 3 weeks. The improved drainage seen with CAID likely promotes the healing of POPF arising from either the pancreatic wound or pancreatoenteric anastomosis. However, the small sample size of this study did not allow us to determine whether CAID can significantly reduce the incidence of POPF-related abscess and haemorrhage. Further research regarding this question is ongoing at our institution.

Two concerns regarding the irrigation method deserve mention. With irrigation, there is a risk of diffusion of the irrigation saline to uninvolved abdominal areas, which may result in secondary liquid collection and infection.[16] Another concern is the risk of retrograde contamination through the irrigation route, which may lead to retrograde abdominal infection. Although no irrigation-related concerns were observed in this study, the possibility of fluid collection or contamination through the irrigation tubes must be considered, with preventive measures taken in advance, such as maintaining unobstructed drainage and application of the filter for inflow air.

This study had several limitations. First, the device for irrigation drainage was hand made by surgeons during the operation. Second, this was a single-centre retrospective analysis, which may limit the generalisability of our results. Finally, this study is limited by its small sample size; however, this study includes one of the largest patient cohorts reported to date, and it is the first to assess the value of continuous irrigation in patients undergoing LPR. Nevertheless, further study is needed to fully evaluate the use of CAID after LPR.

CONCLUSIONS

Our results demonstrate that continuous irrigation after LPR can reduce the incidence of clinically relevant POPF and POPF-related complications. The use of continuous irrigation may also shorten the drainage tube indwelling time and post-operative length of stay. In light of these findings, CAID can be considered an effective alternative for the management of POPF and POPF-related complications after LRP.

Financial support and sponsorship

This study was financially supported by the Social Development Foundation of Science and Technology of Jiangsu (BE2016658); the Changzhou Sci and Tech Program (CE20165020) and the High-Level Medical Talents Training Project of Changzhou (2016CZLJ007). The funder had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.

Conflicts of interest

There are no conflicts of interest.