Colonial Wig Pancreaticojejunostomy

Summary

We describe a new technique for pancreaticojejunostomy reconstruction after pancreaticoduodenectomy that is associated with a very low rate of postoperative pancreatic fistula.

Abstract

Postoperative pancreatic fistula (POPF) is one of the most problematic complications after pancreaticoduodenectomy (PD). We describe a series of 48 pancreatic-head resections from our institution, in which we compare a new technique to create the pancreaticojejunostomy (PJ) reconstruction with standard techniques. The goal is to achieve a lower rate of POPF. This new PJ is termed the "Colonial Wig" (CW) PJ due to the novel appearance of the jejunum wrapping around the pancreas, resembling a Colonial wig wrapping around the head of a Colonial Whig (e.g., George Washington). In our consecutive series, 22 cases were performed using the new CW technique to perform the PJ and were compared to 26 traditional PDs with traditional reconstruction. There was an incidence of clinically relevant POPF of 0% in the CW group, compared to 15% in 26 conventional PJs. Our proposed CW PJ reconstruction is associated with a lower the incidence of POPF following PD, and hence may be a way to improve outcomes after PD.

Introduction

Postoperative pancreatic fistula (POPF) is described as the Achilles' heel of pancreaticoduodenectomy (PD) with an incidence rate ranging between 4–36%^1,2,3. The goal of the presently described method of pancreaticojejunostomy (PJ), termed the "Colonial Wig" (CW), is to lower the rate of POPF following PD.

The morbidity of POPF is variable and it can range from being asymptomatic (Grade A, or clinically insignificant biochemical leak) to being symptomatic, causing deviation in the postoperative management, requiring percutaneous, endoscopic or angiographic interventions (Grade B) or requiring operative interventions, causing organ failure or death (Grade C)⁴. Multiple risk factors have been described to be associated with increased POPF, including soft pancreatic texture, small diameter of pancreatic duct, and increased intraoperative blood loss, and a validated 10-point scoring system has been described to predict the risk of POPF in patients undergoing PD^5,6.

To reduce the incidence and mitigate the severity of POPF, several PJ reconstruction techniques have been described in the literature with a variable POPF incidence and severity. In this paper, we describe a novel PJ reconstructive technique, the CW PJ, which has the advantage of combining what we assess to be the best aspects of the best and most common techniques of the PJ. We compared the risk factors and outcomes of patients undergoing the new CW PJ technique versus standard techniques.

Protocol

This study was approved by the Ethics Committee (Institutional Review Board) of Saint Agnes Hospital (No. 2016-020).

1. Preparations

1. Preoperatively, give 5,000 units of heparin subcutaneously to prevent deep venous thrombosis prophylaxis, and give antibiotics per the hospital's surgical care improvement project (SCIP) policy, such as 2 g of cefazolin and 500 mg of metronidazole.
2. Place an epidural as discussed with the anesthesia team for optimal postoperative pain control and to promote enhanced recovery after surgery (ERAS).
3. Intraoperatively, perform a classic or pylorus-preserving PD in a standard fashion.⁷

2. Transection of the Pancreatic Neck

1. Prior to transection of the pancreatic neck, place four full-thickness, transpancreatic, 3-0 silk, stay sutures on the superior and inferior edges of the pancreas and then divide the pancreatic neck between these stay sutures. Place a Crile clamp on each of the four silk sutures.
2. Proceed to completion of the PD resection in standard fashion.

3. Preparing the Jejunum and Pancreas for Anastomosis

1. After removal of the specimen, dissect the posterior surface of the pancreas free from the retroperitoneum for several centimeters.
2. Bring the stapled end of the jejunum into position in preparation for anastomosis.

4. Suture Placement:

NOTE: The following sutures are placed in the following order to create the anastomosis (as shown in Figure 1).

1. Two 3-0 silk CW sutures (cw)
   1. For each of these, take a full-thickness bite through the pancreas a few centimeters from the cut surface, one at the superior border and one at the inferior border of the pancreas, each passing through a generous seromuscular bite of jejunum, as shown in Figure 1. The bites through the jejunum should be approximately 6 cm from each other (to allow for 2 cm of jejunum on either side of the jejunotomy, which is typically ~2 cm in length, but will vary with the thickness of the pancreatic neck).
   2. Leave these sutures untied, as they will later join the inferior border of the pancreas to the antimesenteric border of the proximal jejunum, and the superior border to the more distal antimesenteric border of the jejunum, wrapping the jejunum around the sides of the pancreatic remnant covering the corners of the anastomosis and giving the final appearance of a colonial wig (Figure 1).
2. Two 3-0 absorbable (e.g., Vicryl^TM) U-sutures (u)
   1. Place these two stitches with a straight(ened) needle. For the first one, starting at the proximal aspect of the jejunotomy and traveling anterior to posterior, take a full-thickness bite through the anterior wall of the jejunotomy, then a full thickness bite through the pancreas, about 1 cm from the cut surface, and just a few mm from the superior border, and then take a full-thickness bite through the posterior wall of the proximal jejunotomy.
   2. Now the needle is at the bottom of the "U," so turn 180 degrees and reverse the path, travelling posterior-to-anterior, taking a full-thickness bite through the posterior jejunum, then pancreas, then anterior wall of the jejunum (Figure 1A, 1B). These U-stitches are used to compress the small ducts (similar to the Blumgart anastomosis¹¹) and to keep the pancreas securely invaginated in the jejunotomy (similar to the "dunking PJ" anastomosis⁸). Place the second U-suture in a similar fashion, but at the distal half of the jejunotomy and inferior to the pancreatic duct (Figure 1A, 1B).
   3. Place a metallic probe (e.g., a Garrett dilator) in the pancreatic duct while taking the pancreatic bites close to the main pancreatic duct, to make sure that the suture does not go through the duct. Each of these should encompass most of the width of pancreatic parenchyma on either side of the pancreatic duct.
3. Two silk 3-0 stay sutures (s)
   1. Attach a French-eye needle to the previously placed stay stitches on the pancreatic remnant, and take a full-thickness bite, in-to-out, through the jejunum 1 cm away from the jejunotomy. The purpose of these stitches is to secure the invagination of the corners of the pancreatic remnant deep into the jejunotomy.
   2. After placing stitches #1-3, pull taut on the s and u stitches to invaginate the pancreatic remnant into the jejunotomy, then proceed with tying them in the following order: u then s then cw (Figure 1C, 1D). The jejunum should now look much like a Colonial wig fitted snugly around the sides of a Colonial Whig's head.
4. Several interrupted 3-0 silk sutures provide a final, outer layer (o)
   1. Place these stitches very closely together in a vertical fashion between the anterior border of the pancreatic remnant and the cut edge of the jejunum to hermetically seal the redundant cuff of jejunum remaining after tying the u sutures to pancreatic capsule (Figure 1). Two of these stitches may be placed posteriorly as well, typically in a horizontal fashion, either now or earlier, e.g., after placing the sutures in steps 1.2 above.

5. Placement of Omental Wrap

1. Wrap the PJ anastomosis with a harvested tongue of healthy omentum. Place two, 19-F, round, fluted (e.g., Blake) drains near, but not touching, the anastomosis (the omental flap serves in part to protect the PJ from the drains).
2. Perform the remainder of the reconstruction, viz., the hepaticojejunostomy and the gastro- or duodenojejunostomy, as previously described.⁷

6. Optional Adjuncts

1. In cases with high-risk features, such as soft pancreas parenchyma, consider decompressing the bilopancreatic limb, e.g., with the creation of a Braun enteroenterostomy between the afferent and efferent limbs of the gastro- or duodenojejunostomy or between the afferent and efferent limbs of the hepaticojejunostomy. The administration of somatostatin analogues is also recommended and is supported by prospective, randomized data.

7. Postoperative Management

1. Postoperatively, extubate the patient once stable from a hemodynamic and respiratory standpoint. Admit patients to the intensive care unit for overnight close monitoring, and follow local ERAS pathways.
2. Initiate enteric feeds via an intraoperatively placed nasojejunal tube immediately postoperatively at a rate of 10 mL/h and advance to goal once there is evidence of return of small-bowel function. If starting somatostatin analogues pre- or intraoperatively, continue postoperatively.
3. For high-risk cases, use pasireotide at 900 µg twice daily for a week. For medium-risk cases and some low-risk cases, use octreotide at 100 µg three times daily until the day of discharge.
4. Check serum and drain amylase daily to evaluate for the presence of POPF. Remove drains on postoperative day #3, depending on the amylase level.

Representative Results

Perioperative data are found in our original publication on this procedure.⁹ Briefly, the POPF rate for the first 26 (control) PDs was 27%. There were 3 (12%) grade-A (clinically insignificant) fistulas, 4 (15%) grade-B, and 0 grade-C fistulas. This clinically relevant POPF (CR-POPF) rate (grade B + grade C) was 15%. In the next 22 CW PJs, however, the CR-POPF rate was 0 (P = 0.052, by Chi-square test) among eligible cases. There was one grade-A POPF in the CW group (5%, Table 1).

The lower POPF rate in the CW group was not due to the presence of lower-risk glands in the CW group (Table 2). Similarly, the two groups were similar regarding important parameters such as gland texture, pancreatic duct diameter, distribution of pathologies, and estimated blood loss.

Figure 1: Schematic of suture placement. (A) The U sutures (u) are used to secure the invagination of the pancreatic remnant deep in the jejunotomy, while the stay sutures (s), which were placed before division of the pancreatic neck, are used to secure the corners of the remnant pancreas. (B-D) "Colonial Wig" sutures (cw) are used to bury the corners of the PJ under the jejunal serosa, which makes the jejunum resemble a traditional Colonial wig, which covers the tops of the ears like the jejunum covers, protects, and seals the corners of the PJ. Finally, the outer-layer sutures (o) are placed to provide a more hermetic anastomosis (D). This figure has been reproduced from our original publication on this procedure⁹. Please click here to view a larger version of this figure.

ISGPS grade	Conventional PD	CWPJ
None	19 (73%)	19 (95%)
A	3 (12%)	1 (5%)
B	4 (15%)	0
C	0	0

Table 1: Comparison of conventional PD and CWPJ cases by occurrence and grade of POPF. PD = pancreaticoduodenectomy; CWPJ = "Colonial Wig" pancreaticojejunostomy; POPF = postoperative pancreatic fistula; ISGPS = International Study Group of Pancreatic Surgery.

FRS parameter	Conventional PD	CWPJ	P value
Soft pancreas texture (n)	14/24* (58%)	10/20 (50%)	0.58
Pancreatic duct diameter (mm)	3 ± 2.3	4 ± 2.5	0.28
PDAC/pancreatitis (n)	12/26 (46%)	10/20 (50%)	0.8
Estimated blood loss (mL)	500 ± 539	500 ± 316	0.33

Table 2: Comparison of conventional PD and CWPJ cases by FRS parameter shows that the groups were similar regarding other known risk factors for POPF. *Two cases were missing data for gland texture. Two deaths early in the postoperative period were excluded since they precluded assessment of POPF. PD = pancreaticoduodenectomy; CWPJ = "Colonial Wig" pancreaticojejunostomy; FRS = fistula risk score; POPF = postoperative pancreatic fistula; PDAC = pancreatic ductal adenocarcinoma.

Discussion

There are many descriptions of novel PJs reported in the literature. It is generally true that the more ways that exist to perform a given task, the less likely there is a single perfect way to do it. This is likely accurate for PJ reconstruction as well. Each of the multiple different PJ techniques reported reports a low incidence of POPF. Nevertheless, POPF continues to be considered the "Achilles heel" of PD and more work is therefore needed to find a better way to construct this anastomosis.

Our current design for PJ reconstruction developed after studying widely used anastomoses and evaluating the likely technical sources of failure of the PJ anastomosis, such as leaks from the small ducts on the cut surface of the pancreatic neck, from the corners of the PJ anastomosis, and from the suture lines on the anterior and posterior surfaces of the PJ.

Failure to address these potential leak points likely increases the risk of POPF. In two of the most widely used PJ reconstruction methods, Cameron's duct-to-mucosa invaginating PJ and the Blumgart's PJ, some compression of the small ducts on the pancreatic cut surface is provided, but these techniques do not provide hermetic sealing of the corners of the anastomosis. This is provided by the cw sutures in the CW anastomosis. Our reconstruction also addresses the potential for leak from small ducts on the cut surface of the pancreatic by further compressing the parenchyma using the U-stitches, which further decreases the leak risk by also serving to deeply invaginate the pancreas within the jejunotomy. The s sutures ensure that the corners of the pancreatic remnant, which are prone to slip out to the jejunotomy, instead stay securely fixed within the jejunum. The o sutures provide further protection by providing a hermetic coverage of the anterior and posterior aspect of the PJ.

To mitigate the risk of POPF even further, the anastomosis is treated with several adjunctive measures. First, the PJ is wrapped with a vascularized omental pedicle, which was used universally when available. Second, in high-risk patients a Braun enteroenterostomy was created between the afferent and efferent loop to decompress the pancreaticobiliary limb.  This limb may also be decompressed with a tube or with an enteroenterostomy between the afferent and efferent limb of the hepaticojejunostomy. Finally, as described above, somotostatin analogues are selectively used. Our low incidence of leak could be explained by the combination of all these measures, as well as simply by reduced tension across the anastomosis owing to the effect of the cw sutures.

Future applications of this technique may include comparison to other techniques in prospective randomized trials.

In conclusion, the existence of many PJ anastomotic techniques suggests none is ideal for all surgeons. Therefore, the best technique for now may be the one most familiar to the surgeon. However, this novel "Colonial Wig" anastomosis is easy to learn and may be a safe and effective way to lower POPF rates after PD.

Materials

Name	Company	Catalog Number	Comments
French eye needle, tapered	Anchor Products Co Inc, Addison, IL	1861-2dc
Garrett dilator	Medline, Northfield, IL	MDS2040030
Octreotide	Sagent, Schaumburg, IL	2055879
Pasireotide	Curascript SD, Grove City , OH	246492