Mesh-reinforced pancreaticojejunostomy versus conventional pancreaticojejunostomy after pancreaticoduodenectomy: a retrospective study of 126 patients

Pancreaticoduodenectomy (PD) has, for a long time, been used as the standard surgical procedure for the treatment of patients with malignant or benign diseases of the pancreatic head or the periampullary region. Mortality in patients undergoing PD is recorded to be below 5% for general advance in surgical technique; however, postoperative morbidity remains high at 30–50% [1, 2, 3]. The main factor is postoperative pancreatic fistula (POPF), which can lead to severe secondary complications such as postoperative hemorrhages and intra-abdominal abscesses [4, 5]. Therefore, prevention and adequate treatment of POPF has always been of high priority [6]. Considerable techniques including pancreaticojejunostomy with duct to mucosa anastomosis or intussusceptions, main duct stenting and pancreaticogastrostomy have been described for safe surgical management of pancreatic remnants; however, no single method has made evident to the scientific community its superiority [7, 8, 9, 10, 11, 12, 13]. Since August 2005, our institute has attempted to reduce the frequency of pancreatic fistula (PF) by using new method termed mesh-reinforced anastomosis [14]. We have previously reported in previous studies that this technique appears to be safe, simple, and quick [14, 15]. The purpose of this retrospective study is to compare perioperative outcomes of mesh-reinforced pancreaticojejunostomy (PJ) with the conventional surgical procedure of pancreaticojejunostomy (PJ). This study was conducted by the same pancreatic team of the same institute.

Despite the remarkable progress in surgical technique and perioperative care during the last decades, the pancreatic-enteric anastomosis remains the “Achilles heel” of PD. Although more than 80 different methods have been described for safe surgical management of the pancreatic-enteric anastomosis, none have been proven to be superior techniques over others, and consequently became widely accepted [20]. Because most of these surgical methods include stitches that penetrate through the pancreatic parenchyma and the soft pancreatic tissue, it makes the pancreas vulnerable to the formation of PF [21, 22]. Peng and coworkers [23] performed a comprehensive three-layer invagination anastomosis, called binding anastomosis, to protect the pancreatic anastomosis from PF. Remarkably, a 0% rate of PF has been reported using this technique. However, this procedure includes complex and troublesome maneuvers. To effectively prevent the PF, we have designed a new technique, namely “mesh-reinforced pancreaticojejunostomy.” This technique using single-layer continuous suturing is far less complex when compared with the binding pancreaticojejunostomy. In the present study, there is no significant difference between mesh-reinforced PJ and conventional PJ in regard to operative time (p > 0.05). This new technique of mesh-reinforced pancreaticojejunostomy is favored for patients with a soft pancreatic remnant. Mesh around the pancreatic remnant provides a safe anchor site for the suture, which is particularly suitable for soft pancreatic parenchyma to avoid anastomotic dehiscence. In the present study, the rate of overall PF was recorded to be 15% in the mesh-reinforced PJ group and 34% in the conventional PJ group with a significant difference (Table 3).

Several previous studies have evaluated risk factors of pancreatic fistula after pancreatic-enteric anastomosis. These risk factors include age, prolonged jaundice, and intraoperative blood loss, all of which have been associated with an increased risk of PF [24, 25]. In the current study (Table 5), demographic factors, such as age, gender, and prolonged jaundice were not statistically associated with pancreatic fistula. Intraoperative parameters, such as soft pancreatic texture and pathology diagnoses, were found to positively correlate with the risk of PF on univariate analysis (p < 0.05). However, pathological diagnosis failed to maintain its statistical significance in the multivariate model. In the stepwise multivariate logistic regression analysis (Table 5), independent factors influencing PF rates were the texture of the organ and type of conventional pancreaticojejunostomy (p < 0.05). Thus, these results were similar with a number of studies [26, 27, 28], which reported that a soft pancreatic remnant is more likely to develop PF.

To identify new surgical techniques that can substantially lead to decrease mortality rates is a challenging task due to the fact that operative mortality in patients undergoing PD is already low. Therefore, the length of postoperative hospital stay comes to be an important representation of patients’ condition and surgical outcome. A shorter length of postoperative hospital stay is considered a predictor of less-invasive surgical procedures and less cost of medical expense. Our data demonstrates that length of hospital stay was significantly reduced in mesh-reinforced PJ group (18 ± 9 vs 23 ± 12 days; p = 0.016). This shows mesh-reinforced pancreaticojejunostomy developed a fast-track postoperative course. The possible explanation is that surgical complications and POPF rate were significantly less common in the mPJ group (p < 0.05 Table 3). The occurrences of POPF and surgical complications contribute to an increased length of postoperative hospital stay. The cost of mesh increased hospital cost, but there was no significant difference in total hospital cost between the two groups (p > 0.05 Table 3). We found that the length of postoperative hospital stay, surgical complications, and POPF rates were significantly less common in the mPJ group (p < 0.05 Table 3) which reduced the total hospital cost.

The advantages of mesh-reinforced PJ [14, 15] are as follows: Firstly, mesh provided a safe anchor site for the suture to avoid laceration of anastomotic and postoperative bleeding; secondly, mesh compression to pancreatic tissue minimized the chance of pancreatic leakage and bleeding; thirdly, mesh was thought to stimulate growth of fibroblast and enhance the anastomotic healing process. We considered that these advantages of mesh decreased pancreatic fistula rate, hospital stay, and the complication rate.

Theoretically, the use of mesh has potential disadvantages. As an implanted foreign body, mesh may increase the risk of intra-abdominal infection. However, our data showed there was no statistical difference between the two groups concerning occurrence of abdominal infections (p > 0.05). We think it may be because of the fact that the mesh was completely wrapped by the jejunal loop during the procedure. On the other hand, the mesh had contractility, which may result in pancreatic atrophy or pancreatic duct dilation. Prolene-Mesh had a contractility of around 20% [29]. We used polypropylene-mesh reinforcement in 10 pigs in an animal experiment, which showed that the pancreatic duct was dilated, and the mesh was rejected 3 months after mesh-reinforced pancreatojejunostomy in all experimented pig subjects. However, the pancreatic duct of the pig was too fine to have a stent tube placed inside. Our data showed there was no statistical difference between the two groups concerning dilated pancreatic duct at the 1-year follow-up, demonstrating that the stent tube withstood compression by the mesh strip. Unfortunately, the anastomotic patency and postoperative pancreatic function was not examined in our study.

In conclusion, this retrospective single-center study showed that mesh-reinforced pancreaticojejunostomy appears to be a simple and safe technique for pancreaticojejunostomy. It provided a safe anchor site for the suture, which was especially suitable for the soft and fragile pancreatic texture to avoid laceration of anastomotic sutures and prevent pancreatic leakage. It should be applicable to all types of pancreatic remnant. Our study was a retrospective analysis of medical records. It should be mentioned that it has all the disadvantages that are associated with any retrospective series. A prospective randomized control study is necessary to confirm the efficacy of this procedure in future.