Comparison of clinical outcomes between mesh-reinforced pancreatojejunostomy and pancreatogastrostomy following pancreaticoduodenectomy: a cohort study

Pancreaticoduodenectomy (PD) is considered to be one of the most difficult and complex surgery since Whipple and colleagues published the first report of patients undergoing PD in 1935 [1]. In addition, PD has been regarded to be a standard surgical procedure in pancreatic cancer and other lesions which are located in periampullary region, while it is always with high postoperative complications rates [2‐5]. Generally, postoperative pancreatic fistula (POPF), biliary fistula, intra-abdominal fluid collection, intra-abdominal hemorrhage, and delayed gastric emptying are common post-pancreaticoduodenectomy complications even there are advances in surgical techniques in recent years, and it leads to prolonged hospital stays and increased medical costs [6, 7]. Among all the postoperative complications, POPF is regarded as Achilles heel of PD. To prevent complications, various strategies have been tried including pharmacologic prophylactic approaches and surgical techniques [8, 9]. These include octreotide, occluding the main duct with rubber or fibrin glue, pancreatic duct stenting, suture ligation of the pancreatic duct, pancreaticoenterostomy with the jejunum or stomach, modification of suturing techniques, and others [10‐12]. Although the rate of mortality after PD has decreased to less than 5%, the morbidity rate is still higher, ranging from 40 to 60% [13]. As the most harmful complications, the incidence of POPF has been reported, with highly variable rates depending on various definitions, ranging from 10 to 29% [14]. Postoperative complications at or around the anastomosis following PD were the most frequent and dangerous, occurring almost 25 times with five deaths [15]. Therefore, improvements in perioperative management and surgical technique remain crucial to decrease complications following PD.

For almost 10 years, we have used the new mesh-reinforced technique described by Peng and his colleagues to minimize postoperative complications after PD [16‐18]. This safe and simple technique is characterized by a strip of polypropylene mesh, and it is wrapped around the pancreatic stump. Wang et al. [17] reported that no leakage occurred in 10 patients undergoing mesh-reinforced pancreaticojejunostomy (PJ) in PD. Only one patient developed grade A pancreatic leakage after mesh-reinforced pancreaticogastrostomy (PG) in a series of 13 initial cases [14, 18]. However, a large sample analysis is needed to evaluate whether the mesh-reinforced technique is safe and effective following PD. Some studies suggested that PG reduced biliary fistula and delayed gastric emptying and postoperative collections in patients following PD more so than did PJ [19, 20]. Therefore, whether mesh-reinforced PG is better than mesh-reinforced PJ in postoperative complications following PD remains unknown.

In this study, we evaluated the outcomes of the mesh-reinforced technique in 110 patients undergoing PD. Either PG or PJ was performed depending on the surgeon’s preference. Postoperative complications were assessed for both types of reconstruction.

In this study, the safety and efficacy of mesh-reinforced pancreatojejunostomy and pancreatogastrostomy in patients following pancreatoduodenectomy were compared. We found that patients undergoing pancreatoduodenectomy with the mesh-reinforced technique suffered fewer postoperative complications, especially lower rates of pancreatic leakage. This retrospective study also indicated that PG with mesh-reinforced pancreatoduodenectomy resulted in fewer complications and improved the quality and safety of the pancreatic anastomosis more so than did the PJ with mesh-reinforced following PD.

The reported frequency of pancreatic fistula after pancreaticoduodenectomy varied from 10 to 30%, and pancreatic fistula was considered a serious event that even might be life-threatening in almost half of patients [15, 26, 27]. The reasons for pancreatic fistula were primarily its soft parenchyma, making it more likely to develop parenchymal lacerations from shear forces applied during tying of sutures [17]. In addition, the anastomotic technique, the diameter of pancreatic duct, the general condition of the patients and the various definitions of pancreatic fistula could also affect the incidence of postoperative pancreatic fistula [14, 28‐30]. It was reported that soft pancreatic parenchyma and the diameter of main pancreatic duct less than 3 mm were risk factors for pancreatic leakage [31, 32]. To prevent these complications, various methods during the preoperative, intraoperative, and postoperative periods have been proposed, especially surgical techniques [33, 34]. For example, it is reported that the application of the novel embeddedness-like pancreaticojejunostomy anastomosis technique in PD was effective and could reduce the incidence of pancreatic fistula [35]. In our hospital, we also introduce a different method of mesh-reinforced technique in PD to reduce postoperative complications.

The rate of pancreatic fistula of 10% in our study was lower than that of other studies. The technique used in our procedure to reduce morbidity and mortality was modified according to the binding pancreaticojejunostomy described by Peng et al. [16]. There exist some advantages of this technique as previously described [36]. First, the safe anchor site was provided by the mesh for sutures, and this resulted in the ability to avoid leakage and bleeding after surgery for soft and fragile pancreatic tissue. Second, it was more convenient to wrap the pancreatic stump with the bowel loop by changing the shape of the pancreas with mesh reinforcement. In addition, the left edge of the mesh was sutured tightening to the posterior wall of the bowel loop, and it facilitated the pancreas sliding into the bowel loop. Fourth, the likelihood of pancreatic leakage and bleeding was diminished by the mesh compression of pancreatic tissue. Finally, the mesh stimulated the growth of fibroblasts and promoted the healing process between the pancreatic capsule and the bowel mucosa [36‐38]. In our study, we proposed various mesh-reinforced surgical techniques to prevent complications, and there were lower rates of several surgical complications.

There are two methods for the restoration of pancreatic drainage into the gastrointestinal tract in PD, including PG and PJ [19, 39, 40]. Not all previous studies yielded similar results. Yeo et al. [41] reported that the rate of pancreatic fistula was almost the same for the PG (12.3%) and PJ (11.1%), suggesting that PG was not safer than PJ. However, Bassi et al. [19] found that biliary fistula, postoperative collections and delayed gastric emptying rates were significantly lower in patients with PG than in PJ. Schlitt et al. [20] also suggested that PG was significantly safer than PJ with respect to the incidence of pancreatic fistula. Therefore, there remained a need to compare the two methods following PD. Mesh-reinforced PG and mesh-reinforced PJ were compared with each other in our study. The mesh-reinforced PG showed advantages in PD in that its comprehensive complication index was lower than that of the mesh-reinforced PJ group. We also showed a significant benefit of mesh-reinforced PG reconstruction compared with mesh-reinforced PJ with respect to intra-abdominal fluid collection and delayed gastric emptying. Pancreatic and biliary fistulas also showed lower trends of incidence in mesh-reinforced PG.

There are additional factors supporting our conclusion that mesh-reinforced PG was superior to PJ. First, the blood supply to the stomach is rich, promoting healing between the stomach and the pancreatic stump [42]. Second, in the acidic gastric environment, pancreatic enzymes will not be activated, preventing deleterious tissue digestion around the anastomosis [43]. Third, a nasogastric tube can be used to decompress, and an endoscope can easily access the anastomosis when needed [43, 44]. However, the learning curve of the surgeon and the experience of the center also influence the method we chose. As a result, we suggest that mesh-reinforced PG can be performed in patients following PD to prevent postoperative complications, especially pancreatic fistulas.

There are several limitations in current study. First of all, as this study was a retrospective observational analysis, it could not exclude the impact of residual confounding factors completely, including the fact that patient distribution between groups depended on surgeon’s preference. Second, as it is just a single-center study; multicenter studies are needed for further confirmation. Third, the results lacked a control group without a mesh-reinforced technique; a randomized controlled trial would be necessary to conduct to avoid such bias in the future.

The current study confirmed that the mesh-reinforced technique may be beneficial in patients following PD. Although mesh-reinforced PG did not show significant differences in pancreatic fistulas and biliary fistulas, the CCI, intra-abdominal fluid collection, and delayed gastric emptying were significantly lower in patients with the mesh-reinforced PG than in mesh-reinforced PJ. In summary, the mesh-reinforced technique is preferred during pancreaticoduodenectomy, especially mesh-reinforced pancreatogastrostomy.