Introduction

Duodenal perforation (DP) is a rare severe condition characterized by high morbidity (ranging from 38 to 75%) and mortality (ranging from 7 to 40%) [1,2,3].

In 2019, the World Society of Emergency Surgery and American Association for the Surgery of Trauma (WSES-AAST) provided a classification of duodenal injuries into four grades according with the AAST-OIS (Organ Injury Scale) classification and the hemodynamic status [4].

Iatrogenic DP represents a wide category involving potentially life-threatening injuries mostly related to surgical, endoscopic, or radiologic procedures complications or trauma.

The adequate management strategy must be carefully identified and tailored on patient clinical features (progressive signs of intra-abdominal sepsis), duodenal fistula out-put (low < 100 ml/die, medium > 100 ml/die < 500 ml/die or high > 500 ml/die output) [5], and mostly on his hemodynamic stability. The main objectives of DP treatment are resuscitation, source control, nutritional support, and prompt rehabilitation of intestinal continuity.

The available DP treatments can be divided in non-interventional, non-surgical, and surgical interventional procedures [6].

Non-interventional Management

Non-interventional treatment consists of fasting, total parenteral or enteral nutrition (below the DP site), antibiotics, intravenous proton pump inhibitors, and nasogastric tube (NGT) placement. The use of somatostatin analogs is still under debate. This type of treatment could be administered in selected stable patients without septic shock signs and with low/medium fistula output (< 500 ml/die).

Non-surgical Interventional Management

Diversion Technique

The main diversion method is represented by the placement of a curved covered endoscopic stent (duodenal stent) with the aim to cover the duodenal hole [7, 8]. Literature data reports further diversion techniques represented by radiologic or endoscopic duodenostomy that could be performed by transhepatic approach or percutaneous drainage location [9, 10].

Repair Technique

The non-surgical interventional repair technique consists of different endoscopic procedures of DP closure by direct positioning of tissue adhesive agents, direct suture with dedicated devices, or clip placement [11,12,13].

Surgical Interventional Management

Many surgical procedures were proposed. Peritoneal cavity washing and abdominal drainage placement, positioning of a duodenostomy tube [14, 15], direct suture of the duodenal defect, duodenal repair with a rectus abdominis flap [16], fistula closure by Roux-en-Y duodenojejunostomy [5, 17], biliogastric diversion, laparostomy, and pancreatoduodenectomy [18] were reported in literature, although final result of surgery were often unsuccessful [10, 19].

The best treatment of DP in terms of safety and efficacy still remains unclear, and a proper management pathway flowchart was never described. Unlike endoscopic iatrogenic DPs, for which many papers debated the adequate treatment schedule, to date there is low- or no-evidence in literature concerning the other iatrogenic DPs entities, mostly surgery- related.

This systematic review aims to identify the best treatment pathway for iatrogenic DPs related to any surgical procedure.

Patients and Methods

Survey Strategy

The present systematic review was registered to the International prospective register of systematic review (PROSPERO) with CRD42022307958 ID number. We conducted both the literature search and the systematic review according with the Cochrane Collaboration guidance to seek bias minimizing [20]. This study was conducted in agreement with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [21].

Literature Research

Our literature investigation was performed to identify all papers regarding surgical-related iatrogenic DPs, without any limitations of language, publication date, and status (Epub Ahead of Print, In-Process, In-Data-Review and Other Non-Indexed Citations). On 28th of February 2022, with the support of the Federate Library of Medicine of Turin, the authors searched the following databases: Ovid Medline, Embase, and Cochrane CENTRAL Register of Controlled Trials. Table 1 suppl shows each database research method.

Table 1 Details of the studies included in the review
Full size table

The primary outcome was the type of DP’s treatment. The secondary endpoints were the efficacy (in term of rate of DP healing, time to fistula resolution, rate of mortality due to DP fistula) of DP’s treatment, the length of hospital stay, and the costs of the procedure.

Only Italian or English articles were selected at the end. We included only article with proven diagnosis of surgical-related DP, human studies, studies enrolling patients with age \(\ge\) 18 years, and studies including at least one outcome of interest among their results.

Exclusion criteria were other causes of perforation not related to surgery (peptic ulcer disease, drug- and/or trauma- induced, radiologic-, endoscopic- and vascular- procedure related, etc.), metanalyses, systematic reviews, reviews, letters, oral communications, videos, abstracts, papers conducted by the same institution or reporting overlapping patients (in these circumstances, the studies with smallest sample size or published earlier were excluded).

Five authors (H.A.R.A., E. B., G. M., F. S. and M. C.) selected the suitable articles that met inclusion criteria firstly evaluating their titles and subsequently their abstracts. Furthermore, the resulted papers were checked by a sixth reviewer (R.R.). Hence, all authors meticulously analyzed the whole texts of each selected study to verify its pertinence. A repeated study examination with a team discussion was necessary to solve the cases with any eventual disagreement.

Results

Studies identification and selection method was summarized in Fig. 1 according to PRISMA guidelines [21]. A total of 11,097 papers were identified from the main electronic databases research (Ovid MEDLINE 4391, Embase 6498, Cochrane CENTRAL 208) from November 1903 to February 2022 with the support of “Biblioteca Federata di Medicina, Università degli studi di Torino”. After the deduplication process, seven thousand five hundred and sixty-eight articles resulted; thereafter, 5 authors proceeded to screen all articles by title and subsequently by abstract. At the end of the process, thirty residual studies were meticulously reviewed, and 18 of them were excluded later. In Fig. 1 the reasons for the studies exclusion were summarized. Finally, a total of 12 articles were included in the present review [7, 13, 22,23,24,25,26,27,28,29,30,31].

Fig. 1
figure 1

PRISMA 2020 flow diagram for new systematic reviews which included searches of databases, registers, and other sources. *Consider, if feasible to do so, reporting the number of records identified from each database or register searched (rather than the total number across all databases/registers). **If automation tools were used, indicated how many records were excluded by a human and how many were excluded by automation tools

Full size image

Study and Patient Characteristics

The included articles were case reports and case series involving a few patients with iatrogenic duodenal perforation due to a prior surgical procedure. A total of 22 patients were included on accrual; historical patient records were analyzed. The mean age of included patients was 54.5 years and was equivalent between sexes (Table1).

Duodenal Perforation Details

All duodenal perforations’ characteristics were showed in Table 1. The most common reported iatrogenic cause of DP was prior laparoscopic cholecystectomy (72.7%). Other causes were right nephrectomy, right hemicolectomy together with small bowel resection, choledochoduodenostomy, cholecystectomy with common bile duct exploration, colonic fistula closure, and benign biliary stricture repair. The median day of symptoms appearance was the second postoperative day; similarly, the median time to diagnosis was the second postoperative day. However, Mendoza et al. [13] reported a late DP with symptoms appearance 210 days after surgery and the diagnosis 22 days later.

More than a half of DPs was located in the first portion of the duodenum (57.3%), 33.4% in the second portion, 9.1% in the third portion, and none in the last portion.

Despite most of the studies did not describe the size of the perforation, the mean defect’s dimension reported was 9 mm.

We retrospectively proceeded to classify all the DPs with the WSES-AAST duodenum injuries scale [4]. According to this classification, most of them were severe injuries (type III and type IV, 58.4%), while the remnant were moderate (type II, 41.6%).

Furthermore, the duodenal fistula output was mentioned only in three articles [22, 29, 31] and was higher than 500 ml/24 h.

Type of Approach

Most of the injuries were related to minimally invasive surgery (72.7%). The type of approach is detailed in Table 1.

Type of DP treatment was not related to the type of approach.

Treatments and Outcomes

The different types of treatment were detailed in Table 2. Only one author [25] reported a non-interventional, conservative management that failed to heal the defect. Different types of interventional management both non-surgical and surgical were reported.

Table 2 Duodenal perforation treatments and outcomes
Full size table

Non-surgical Interventional Management

Four different types of endoscopic management were mentioned by authors. Half of these conservative treatments were unsuccessful. One paper [7] reported a failing placement of metallic clips. One author [29] described the placement of a long-coated metal stent prosthesis that unfortunately migrated. Therefore, the stent was relocated and fixed with clips, and two double pigtail drainages were positioned across the duodenal hole.

The two successful endoscopic managements were reported by one article [31], who described the placement of a Vicryl mesh plug fixed to the borders of the perforation with Endo clips, while one author [13] employed the over-the-scope-clip (OTSC) OVESCO.

Finally, one paper [7] reported the heal of the perforation with the innovative use of a percutaneous Amplatzer Septal Occluder.

Surgical Interventional Management

Direct surgical suture of the DP was proposed by 3 authors [7, 27, 30]. Unfortunately, this type of treatment always failed.

Three studies reported a direct suture of DP with the simultaneous T-tube duodenostomy (TTD) positioning [24,25,26]. This type of management resulted in the duodenal defect healing in 2 of these 3 patients [25]. Two authors reported the use of an abdominal drainage as the unique treatment; this procedure was characterized by a complete recovery of the injury [23, 30]. Another author [25], after 2 successive failures of DP treatment, successfully added the pyloric exclusion together with a jejunostomy placement to the sole suture of the perforation over a DTT. Similarly, one paper reported a treatment pathway characterized by gastric resection, closure of the duodenal stump, and direct repair of the DP by suture [24]. Besides, more extensive surgeries have been described by other authors. In one [22] article, DP was managed with Whipple’s pancreatoduodenectomy leading to a successful patient outcome. One article [28] reported the adoption of a serosal patch procedure using a jejunal Roux-en-Y loop to repair the duodenal injury, with a successful rate of 75%. Nonetheless, one of four patients included in this article underwent redo-surgery for secondary biliodigestive anastomosis and duodenal patch revision. Otherwise, another author [27] described an innovative technique of pancreas-sparing, ampulla-preserving duodenectomy. The authors concluded that this type of surgery is feasible and associated with satisfactory outcomes. In one [24] article, a duodenojejunostomy and a Whipple’s pancreatoduodenectomy were performed in two cases, both with successful healing of duodenal hole.

Nine of the twelve papers included in this review mentioned the median time to DP resolution (20 days) and only three of them reported the median time to oral intake recovery (7 days).

The mortality rate was 13.6% with a median hospital stay of 38.5 days.

No article cited the costs of the procedures and of the hospital stay.

Discussion

Perforation of the duodenum consequent to prior surgical procedures is a rare and severe complication carrying serious consequences if not adequately and promptly managed. The best treatment is still under debate, and several strategies were proposed over the years without a documented superiority of one of them in term of safety, efficacy, and costs.

Many DP-related complications have been reported, such as intra-abdominal abscesses, wound infections, diffuse peritonitis, sepsis, malnutrition, pancreatitis, abdominal bleeding, and pneumonia [32]. This complication may lead to longer hospitalization times and increased patient mortality. In the present study, the mean hospital stay was 59.75 days, and the postoperative mortality was 13.6%.

Although medical therapy is generally associated with better outcomes when successful, surgery is still mandatory in case of severe abdominal sepsis, bleeding not otherwise manageable, and failure of conservative management. In the present review, only one patient (4.5%) was treated with conservative approach. Most of patients (95.5%) were treated with interventional approach (radiological, endoscopic, and surgical).

However, reoperation is often ineffective owing to postoperative edema and inflammation, and the prognosis of patients undergoing surgery for DP remains very poor. To improve the outcome of these patients, many surgical procedures were proposed from washing the peritoneal cavity together with placement of an abdominal drainage to positioning of a tube duodenostomy [14, 15], closure of the fistula, fistula repair with a rectus abdominis flap [16], fistula closure by Roux-en-Y duodenojejunostomy [17, 33], biliogastric diversion, laparostomy, and pancreatoduodenectomy [18]. Nevertheless, to date, surgeons are often unsure about the best treatment to employ, and the results of these treatments are often unsuccessful [34].

The available DP treatments can be divided in two major groups: conservative (non-interventional and interventional) and surgical procedures. The best management should be patient-tailored, taking into account the cause of perforation, both the size and the site of the defect, the time to diagnosis, and above all the clinical setting of the patient. The most important issues to be addressed to achieve a successful healing of the perforation and to reduce the related mortality consist in an early diagnosis and a prompt patient handling. Prior to determine the best DP management, unstable patients must be treated with intravenous fluid resuscitation and with broad spectrum antibiotics. A standard management must initially provide the best nutritional support and furthermore the drainage of the abdominal cavity to avoid intra-abdominal collections. The nutritional strategy is one of the main debated issues of DP treatment. Several types of nutritional management have been proposed, from fasting and total parenteral nutrition to the placement of nasojejunal or jejunostomy feeding tubes.

In a clinically stable patient with a well contained perforation, three main types of treatments can be proposed. The first type is a non-interventional conservative management characterized by sepsis control with antibiotics/antifungals, aggressive fluid resuscitation, and ipernutrition support, as described in one article [25]. Unfortunately, our literature overview documented that this treatment has low efficacy. The second type is the interventional conservative management, and it is most often proposed in WSES-AAST grade I–III patients, showing high successful rates, ranging between 50 and 90% as reported in literature [7, 13, 29, 31, 35]. Hence, many of these patients can be safely managed without surgery. The main challenge is to define when a patient can be treated conservatively and when he becomes candidate for salvage surgery due to previous conservative treatment failure.

Nevertheless, surgery plays a leading role in unstable patients (WSES-AAST grade IV) or in type < IV injuries, including failure of conservative treatment, persistent hole, large free or retroperitoneal collection, evidence of peritonitis, or surgical site infection. The main purposes of surgical treatments are both the sepsis control and the repair of the defect, with or without duodenal diversion [22, 24,25,26,27,28, 34].

The DP can be sutured primarily following debridement of necrotic tissue. Minor perforations promptly diagnosed are successful managed with a direct suture [26]. However, patients with delayed diagnosis or with larger duodenal injury should be managed with duodenal diversion to decrease the risk of suture dehiscence. The duodenal exclusion can be achieved with T-tube decompression placement (DTT), duodenal diverticulation, and gastric resection [35]. The DTT is performed together with direct sutures and can lead to decompression of the gastrointestinal tract. Nonetheless, its use is controversial because the advantage of the decompression is associated with the creation of a new duodenal hole; unless, it is introduced in the existing duodenal defect as described by two authors [25, 26, 36]. The duodenal diverticulation consists in a Billroth II gastric resection, direct suture of the DP, and DTT placement; this technique can be performed together with truncal vagotomy and biliary diversion [37]. Gastric exclusion consists of DP suture, pylorus exclusion ( both with manual or mechanical suture), and the creation of a gastrojejunostomy [38,39,40]. After few weeks, the suture can be break down with the aim to rehabilitate the gastro-intestinal tract. Management of DPs, including conservative non-interventional, percutaneous or endoscopic (conservative interventional), and surgical strategies, is reported in Fig. 2.

Fig. 2
figure 2

Duodenal perforation management flowchart. WSES-AAST, World Society of Emergency Surgery-American Association for Surgery of Trauma; DTR, duodenal T-tube; OTSC, over the scope clip

Full size image

To our best knowledge, this systematic review represents the most updated comprehensive overview of available treatments for duodenal perforation related to prior surgical interventions. The interpretation of our results must take into account the high risk of bias related to the miscellaneous data, different clinical contest, and above all the limited scientific value of papers with small sample size retrospective descriptions, case series, or even case reports. Hence, each successful treatment of DP must be balanced with the limited scientific evidence of a small cohort study.

Conclusion

The present systematic review shows a wide spectrum of managements that can be proposed to patients carrying a duodenal perforation related to prior surgery. The decision on which treatment should be  adopted must take into account patient’s clinical setting and many duodenal injury-related factors (size, site, and time to diagnosis) (Fig. 2). This flow chart represents the actual way to face a duodenal leak following surgery.