Introduction

Management of resectable periampullary tumors is currently standardized regarding surgical methodology and perioperative therapy. However, specific postoperative complications in proximal pancreatectomy are still the leading cause of unsatisfactory immediate results. Failure of pancreticodigestive anastomoses followed by postoperative pancreatic fistula prevails in the structure of specific complications [1, 2]. Preoperative acute pancreatitis (AP) naturally transforms pancreatic parenchyma, often results intra- and parapancreatic necrosis and significantly changes topographic and anatomical relationships of the pancreatoduodenal zone [3]. These features determine the need for a non-standard approach to extensive pancreatic resections and surgical methodology in these patients. Rare combination of diseases (previous AP + resectable periampullary tumor) excludes stereotypical approaches and development of evidence-based guidelines and substantiate advisability of this study.

The purpose of the study was to evaluate the early outcomes of radical surgical treatment of patients with resectable periampullary tumors and previous AP.

Material and methods

A retrospective study is based on collective experience of 654 extensive pancreatectomies (pancreaticoduodenectomy, total pancreatectomy) for the period from 2009 to 2018. All procedures were carried out in 4 hospitals (Bakhrushin Brothers Clinical Hospital, Medical and Rehabilitation Center, Pirogov Clinical Hospital, Blokhin National Cancer Research Center). A retrospective analysis of the immediate results included 9 patients with resectable periampullary tumors and preoperative AP. Analysis of clinical, intra- and postoperative characteristics of patients was based on the databases of surgical centers (Table 1).

Preoperative AP was confirmed by clinical, laboratory and instrumental data in accordance with the international guidelines [4]. All patients received complex conservative treatment that was successful in 8 cases. One patient underwent 2 repeated laparotomies due to widespread intra- and parapancreatic necrosis followed by parapancreatic abscesses. There were no other surgical interventions for AP. Preoperative examination before resections was standard and included contrast-enhanced CT of abdominal cavity to assess resectability, pancreatic parenchyma and parapancreatic tissues. It should be noted that percutaneous biopsy of pancreatic head resectable tumor was applied in 7 patients outside the international recommendations in other hospitals. This event resulted AP.

All patients were divided into 2 groups (Table 1): group 1 — pancreaticoduodenectomy (n = 5); group 2 — total pancreatectomy (n = 4). Surgeries were performed by 6 surgeons. Stomach resection implied antrumectomy; lymph node dissection was standard in accordance with ISGPS classification. Pancreaticoduodenectomy was accompanied by invaginational pancreaticoenterostomy (MUST technique) in 3 patients and pancreaticojejunostomy by Cattel-Warren in 2 patients.

Incidence and severity of postoperative complications and mortality were analyzed in early postoperative period. Severity of complications was graded in accordance with Clavien-Dindo classification. Complications grade IIIb and over were classified as severe ones. Specific postoperative complications included postoperative pancreatic fistula, delayed gastric emptying and arrosive bleeding. Severity of these complications was stratified according to ISGPS classification [5]. Statistical analysis was not performed due to small sample size in both groups.

Results and discussion

Our data correlate with previous researches (Table 2). We would like to emphasize the rarity of combination of these diseases once again that completely excludes stereotypical intervention and also dictates the need for personalized clinical approach to pre- and postoperative planning of treatment. In the available national and foreign literature over the past 10 years, we found only 3 studies devoted to elective pancreatic surgery after previous AP. All studies are retrospective [6, 7, 9].

In our study, preoperative AP occurred in 1.37% of patients with resectable periampullary tumors. The authors from one of the most authoritative European centers of surgical pancreatology (Heidelberg University Surgical Hospital, Germany) came to similar conclusions. Over 200 extensive pancreatectomies are annually performed in this center. M. Erkan et al. [6] reported only 9 patients with preoperative AP for the period 2002-2006. This event significantly influenced surgical strategy and perioperative therapy. S. Asari et al. [7] performed 248 pancreaticoduodenectomies for 6 years (2006-2012). Six patients had preoperative AP.

AP is a polyetiologic disease. In general population, excessive alcohol intake and biliary lithiasis are the most common triggering factors. However, according to our study and literature data, the main etiological factors of AP in patients with periampullary tumors are different from ones in general population. In our study, AP was caused by percutaneous biopsy of resectable pancreatic head tumor in 7 out of 9 patients (77.7%), and only 2 patients had alcoholic AP. We would like to emphasize this etiological structure. According to the international guidelines and consensus, preoperative biopsy is not indicated in an operable patient with typical diagnostic data on resectable solid tumor of the pancreatic head [8]. In our opinion, preoperative morphological verification does not affect treatment strategy in this situation. Moreover, this approach is also dangerous due to possible post-manipulation AP (or other complications), significant impairment of intra- and postoperative period and often delay of specialized care. Meanwhile, clinical studies convincingly demonstrate that early specialized treatment is a significant predictor of long-term outcomes in patients with extremely aggressive pancreatoduodenal malignancies [8].

S. Asari et al. [7] reported curative and diagnostic ERCP as a cause of AP in 5 out of 6 patients. M. Erkan et al. (n = 10) [6] reported AP in all patients after endoscopic manipulations within major duodenal papilla, bile and pancreatic ducts. In our study, ERCP-associated AP was not identified that is primarily due to another approach to preoperative biliary decompression (antegrade drainage).

Pancreaticodigestive anastomosis is an "Achilles' heel" of proximal pancreatectomies. This complications determines the vast majority of severe postoperative complications and mortality [1, 2]. Obviously, pancreatic inflammation, focal destruction of parenchyma or parapancreatic tissue and previous infected necrotic inflammation change pancreatic parenchyma. These processes can make pancreaticodigestive anastomosis either unsafe or impossible [7]. It is reasonable to assume that resection of the pancreas compromised by previous destructive inflammation will be accompanied by higher morbidity. Indeed, M. Erkan et al. [7] reported higher incidence of postoperative complications up to 60% (33% in patients without preoperative AP) and prolonged hospital-stay (19.5 vs. 14.5 days, respectively). Y-H. Chen et al. [9] summarized an experience of 179 pancreaticoduodenectomies (34 patients after previous AP). Considering multivariate analysis data, the authors concluded that preoperative AP is an independent risk factor of pancreatic fistula (OR 2.91; p = 0.032) and severe complications (OR 4.7; p = 0.009).

In our sample, severe pancreatic fistula complicated 3 out of 5 pancreaticoduodenectomies. In all 3 patients with persistent postoperative pancreatic fistula, destruction of major pancreatoduodenal arteries required redo urgent laparotomy. Emergency angiography was impossible due to severe hemorrhage and hemodynamic instability. One patient suffered 6 episodes of arrosive bleeding followed by 6 (!)reconstructions of hepatic artery and its bifurcation. Postoperative period after pancreaticoduodenectomy was uneventful in 2 patients. There were no postoperative complications after TP.

Our own results and data of the above-mentioned studies allow us to recognize TP as a safer intervention in patients with previous AP. In our opinion, higher "surgical reliability" of TP is determined by 2 factors. First, there is no need for pancreaticodigestive anastomosis with a compromised pancreas. Secondly, total debridement of focal para- and intrapancreatic necrosis reliably prevents further persistence of infectious process in postoperative period. M. Erkan et al. [7] performed 4 pancreaticoduodenectomies and 6 TP in patients with periampullary tumors and previous AP and came to similar conclusions. In patients with preoperative AP, the authors recommend TP in the slightest doubt about safe pancreatoenterostomy or in case of massive parapancreatic necrotic foci. Importantly, this study was performed at the Heidelberg University Hospital. The team of this hospital was one of the first who postulated primary (so-called preventive, “technical”) TP if safe pancreatic anastomosis was impossible [10]. Rare combination of preoperative AP and subsequent pancreatectomy dictates the need to make non-standard tactical and technical decisions by an experienced surgeon. We would also like to emphasize that the risk of severe metabolic disorders following TP is currently overestimated. Correct insulin therapy and modern enzyme drugs ensure adequate metabolic control and good quality of life that was demonstrated in several studies [10].

Limitations of this study include retrospective design, as well as certain technical intraoperative nuances used in various hospitals (in particular, methodology of pancreaticodigestive anastomosis)

Conclusion

Primary TP may be a safer surgical strategy compared to pancreaticoduodenectomy in patients with resectable periampullary tumors and previous AP. Randomized trials are required for evidence-based recommendations. However, these trials seem difficult due to rare combination of these diseases and differences in incidence and course of AP. Preoperative percutaneous biopsy in patients with resectable pancreatic head tumors and typical X-ray signs is a redundant diagnostic method that can be accompanied by such severe complications as AP and/or focal pancreatic necrosis. Preoperative AP is followed by forced delay of specialized care (first of all, surgical management) that can impair long-term outcomes.

The authors declare no conflicts of interest.