The Diagnosis and Treatment of Local Complications of Acute Necrotizing Pancreatitis in China: A National Survey

Background The optimal management strategy in acute necrotizing pancreatitis (ANP) is debated, and compliance with current guidelines in China is not known. In this study, we performed a national survey on this topic in China. Methods An online questionnaire about the diagnosis and treatment of local complications of ANP was distributed through a national collaborative network. The local and systemic complications were defined according to the Revised Atlanta Classification. Results There were 321 survey respondents from the 394 who opened the link (response rate 81%) from 208 hospitals located in 30/34 provinces across China. There was a lack of consensus in terms of early diagnosis of infected pancreatic necrosis (IPN) as the respondents chose to depend on clinical symptoms (70/321, 22%), organ failure (82/321, 26%), imaging changes (84/321, 26%), and fine needle aspiration (51/321, 16%), respectively. A “step-up” approach has been widely adopted in patients with IPN (294/321, 92%). The decision for initial intervention (without confirmed/suspected infection) was based on clinical condition, CT imaging, or laboratory indicators for most respondents (229/321, 71%). Conclusion While the “step-up” approach has been widely adopted, there is still significant variation in regard to the diagnosis of infection, the best timing for drainage, and the indications for early intervention.


Introduction
Acute pancreatitis (AP) is one of the most common gastrointestinal disorders and a major cause of acute hospital admission in China [1]. Pancreatic necrosis occurs in 5-10% of all AP episodes [2]. The most important determinant of severity in patients with AP is persistent organ failure (POF) [3] which can account for more than half of the deaths from AP [4,5]. The local complications of AP are responsible for considerable morbidity and have been redefined in the Revised Atlanta Classification (RAC) [2,6], and there has been a notable evolution in their treatment over the last a few years [7].
The early treatment of severe AP is mostly supportive with the primary aim of correcting intravascular volume depletion and organ support (e.g., vasopressors, mechanical ventilation, and renal replacement). The treatment of acute local complications of AP, like acute pancreatic fluid collections (APFC) and acute necrotic collections (ANC), is rarely indicated in the acute setting [8,9]. The local complications can get infected in approximately a third of patients with ANP and therefore become infected pancreatic necrosis (IPN). IPN is a challenging and potentially fatal complication, with a mortality of 15-30% [10][11][12]. Several criteria for the diagnosis of IPN have been reported, but the optimal diagnostic strategy remains uncertain [8,[13][14][15][16]. The approaches include the development of clinical markers of infection, laboratory markers (including procalcitonin), the detection of gas within the collection on contrast-enhanced computed tomography (CECT), and positive fine needle aspiration (FNA) culture results.
The current recommended treatment of confirmed IPN is antibiotics and drainage (percutaneous or endoscopic). This has replaced the "surgery-first" approach with necrosectomy and drainage [17]. The "step-up" approach of drainage first (followed by endoscopic or percutaneous debridement of necrosis) has gradually become the standard of treatment in recent years, although uncertainty remains around the optimal timing, drainage, and debridement techniques [8,11,12,18]. This means that there is a significant variation in the management strategies of AP between different centers or even within centers. Further, it is not known to what extent the current evidence for the "step-up" approach has been implemented in China. Therefore, the aim of this study was to determine the current diagnostic and treatment strategies for local complications of ANP in China.

Materials and Methods
An online 11-question survey ( Table 1) was distributed to Chinese doctors using a commercial platform tool through the Acute Pancreatitis Network (APnet) which is a webbased platform designed to promote education and research for health professionals involved in the care of patients with AP [19]. The subscribers are based in four different medical specialties, intensive care medicine, emergency medicine, gastroenterology, and general surgery. Both the local complications like ANC, APFC, walled-off necrosis (WON), and IPN and systemic complications like organ failure and persistent organ failure were defined according to the RAC terminology [2].
The survey was conducted over a 4-week period (July 9 th , 2018~Aug 5 th , 2018), with the invitation to participate issued via email, a notice on the APnet, and social media (WeChat, Tencent, Shenzhen China) with weekly reminders. The physicians conducting the survey were excluded from this survey. This was a "closed" survey, and participants were required to log in first to prevent duplicate entries. It was also a voluntary survey without any incentives offered. There were 11 questions in the survey including 2 adaptive question. Question 1.1 was displayed based on responses to question 1, and question 1.1.1 based on question 1.1. Completeness would be checked by the website before the questionnaire was submitted. Participants were able to review and change their answers through a Back button before submitting.
Data were collected anonymously and analyzed using IBM SPSS 24.0 software. Only completed questionnaires would be analyzed. The data were categorical variables and described as counts and percentages. All methods were carried out under the relevant guideline [20]. Approval was obtained from the institutional review board of the Jinling Hospital, and informed consent was incorporated into the questionnaire and signed prior to answering the questions.

Treatment
Strategy for Sterile ANC. The decision to drain patients without signs of infection was based on clinical condition, CT imaging, or laboratory indicators for 71% (N = 229) respondents, while 18% (N = 57) indicated they would never consider drainage, and 11% (N = 35) said they would offer immediate drainage whenever possible (Table 1). Of the 229 respondents who would drain "based on clinical condition, CT imaging, or laboratory indicators," most (N = 126, 55%) were mainly concerned about organ failure (OF), while for 17% (N = 39), it was persistent abdominal pain or abdominal distension, 15% (N = 34) CT evidence of increasing extent of pancreatic/peripancreatic necrosis, and 8% (N = 18) laboratory indicators. Among the 126 doctors who would offer drainage because of OF, 45% (N = 57) would do so for POF, 42% (N = 53) for deteriorating original OF, 33% (N = 41) for new-onset OF, 17% (N = 21) for OF at least 1 week, and 7% (N = 9) for OF of 2-week duration or more.
Interventions for patients without signs of infection included percutaneous catheter drainage (PCD) in 47% (n = 153), percutaneous or endoscopic drainage depending on the location of the ANC in 42% (N = 135), endoscopic drainage in 4% (N = 12), surgical drainage in 4% (N = 12), and other drainage technique in 3% (N = 9). 57% of the respondents (N = 33) from gastroenterology would choose percutaneous or endoscopic drainage depending on the location, while respondents from other departments preferred PCD (Supplement Table 1).

Treatment Strategy for Demarcated Pancreatic Necrosis.
For the WON with symptoms but no suspicion of infection, 82% (N = 262) of the respondents would take a step-up approach starting with drainage (percutaneous or endoscopic), while 14% (N = 45) would opt for open surgical debridement/drainage and 4% (N = 14) would choose  Step-up approach with PCD as the initial treatment 162 50 Step-up approach with endoscopic catheter drainage as the initial treatment 14 4 Step-up approach with PCD or endoscopic catheter drainage as the initial treatment based on the location of ANC 118 37 Direct laparotomy debridement and drainage 15 5 Others 12 4 (9) For walled-off necrosis (WON) with symptoms such as persistent abdominal pain, stomach outlet obstruction, and emaciation, what is your preferred treatment?
Observation without drainage 6 2 Step-up approach with PCD as the initial treatment 97 30 Step-up approach with endoscopic catheter drainage as the initial treatment 24 7 Step-up approach with PCD or endoscopic catheter drainage as the initial treatment based on the location of WON 141 44 Direct laparotomy debridement and drainage 45 14 Others 8 3  Gastroenterology Research and Practice other treatments, including observation without drainage (Table 1).  (Table 1).

Treatment Strategy for Infected Complications of AP.
Patients with confirmed infection would be offered drainage if the collection was well demarcated or walled off for 35% (N = 112) of the respondents, whereas 25% (N = 79) would base the decision on the lack of satisfactory response to antibiotic treatment, 22% (N = 72) would undertake immediate drainage for confirmed infection and would not consider other factors, and 14% (N = 45) for occurrence or evolution of OF, and 4% (N = 13) for other reasons, including a positive blood culture (Table 1). For patients with confirmed IPN but an immature (not well demarcated or encapsulated) collection, 50% (N = 162) of respondents would administer antibiotics and drain (percutaneous or endoscopic) with a view to performing a delayed necrosectomy after 4 weeks, while 25% (N = 79) would administer antibiotics and wait for demarcation, and 16% (N = 79) would administer antibiotics and only consider drainage if antibiotics failed. Immediate debridement (endoscopic or surgical) was favored by only 5% (N = 16) respondents (Table 1).
For patients with confirmed IPN and a mature (well demarcated, encapsulated, or walled off) collection, 78% (N = 250) of respondents would administer antibiotics and arrange immediate drainage (percutaneous or endoscopic), 8% (N = 25) would not consider drainage until antibiotics had proven futile, and 8% (N = 25) would administer antibiotics and arrange immediate minimally invasive debridement ( Table 1).
The preferred drainage strategy for confirmed IPN was the step-up approach for 92% (N = 294) of the respondents. This would be with either percutaneous or endoscopic drainage based on the collection locations and available expertise. Only 5% (N = 15) would consider open surgical debridement and drainage (Table 1).

Discussion
This is the first national survey in China to document the approach to the diagnosis and treatment of the local complications of ANP. The results highlight significant variation in practice across China while also demonstrating the significant uptake of the "step-up" approach [21], which has been shown in other studies to be associated with a significantly reduced requirement for open surgical treatment and improved outcomes [22,23].
The survey was distributed widely to doctors in China using the web-based platform of the Acute Pancreatitis Network, which encompasses 4 specialties (general surgery, intensive care medicine, emergency medicine, gastroenterology) and 3511 centers. Responses were obtained from hospitals located in 30 of the 34 provincial regions (except Tibet, Hongkong SAR, Taiwan, China, and Macau SAR). 394 subscribers logged onto the survey, of whom 321 responded (response rate 81%).
The indication and timing of intervention for sterile local complications of AP were a marked lack of consensus. The reasons for intervention were largely based on organ failure, and this included persistent, worsening, and new onset of organ failure. It was notable that the majority of respondents supported early intervention in this setting even in the absence of any recommendations for this in current guidelines [11,22]. The International Pancreas Association and  5 Gastroenterology Research and Practice the American Pancreas Association jointly published guidelines recommending that when ANC has explicitly excluded infections, if organ failure persists for several weeks after onset, intervention therapy may be considered, but preferably not until ANC evolves to WON [11]. In the latest consensus published by the European Endoscopy Association, it was suggested that invasive intervention was indicated for patients with POF or "failure to thrive" over several weeks, but the concepts of "several weeks" and "failure to thrive" are still ambiguous [22]. More recently, a small study showed some benefit of early aggressive drainage of sterile ANC in reducing SIRS and POF and for the early diagnosis of IPN [23].
The optimal approach to the diagnosis of infected local complications of AP is important for deciding about intervention, but this remains controversial, and there was no consensus [24]. Only 26% (N = 84) of the respondents thought that gas within the local complication/collection was the most important diagnostic criteria. Collection with gas had high specificity but poor sensitivity. Previous study showed that gas were seen in only 42% of patients with infected necrosis [25]. Image-guided FNA-bacterial culture was previously widely used to confirm the suspicion of infection, but most experts rarely or never use it [24,25] in part because of the 12-25% risk of false-negative result [26,27] and the risk of introducing infection. Clinical and laboratory markers can be abnormal for other reasons, including a sterile systemic inflammatory response and extrapancreatic infections [28]. After the first 14 days, clinical signs of infected necrosis are much more reliable. It is possible to attain high levels of accuracy in the identification of infected necrosis based on clinical criteria [29]. Without an early and accurate approach to the definitive diagnosis of IPN, intervention is based on the probability of infection using clinical, radiological, and biochemical methods, highlighting a significant knowledge gap. The results of this survey indicate that 71% (229/321) respondents would drain patients without proven infection based on concerns about clinical condition, CT imaging, or laboratory indicators, and this would be based on organ failure (persistent, new-onset, and deteriorating) in 55%,126/229 respondents. This is despite the lack of evidence that draining a sterile ANC is of benefit and acceptable risk. And it is not known what proportion of these patients would prove to have infection and how many would have infection introduced by the procedure. One of the advantages of a more liberal approach to drainage in a patient that is deteriorating, whether it is on the basis of cytokine or infection mediated organ failure, is that definitive bacteriology will be obtained from the ANC/IPN by drainage (in the same way FNA was used). Further research is required to determine whether, in the absence of an early and accurate method to diagnose IPN, organ failure itself can be used as a reliable indicator for drainage, with or without proven infection.
The timing of intervention in patients with complicated AP is one of the most difficult issues for clinicians treating these patients. The widespread adoption of the step-up approach in China is noteworthy, but the optimal timing remains unclear for antibiotics, drainage, and debridement.
Intravenous antibiotic treatment is commonly used in this period as a bridge between conservative and invasive treatment and can sometimes itself obviate the need of any additional intervention in selected cases [10,30]. However, the long-term use of antibiotics may result in an increased incidence of Candida infections and antibiotic resistance [30]. The recommendation for drainage and debridement (IAP/APA) has been to wait for 4 weeks to allow demarcation because the risks are lower if the necrosis is walled off [18]. An international survey of expert pancreatologists revealed a lack of agreement on the need to postpone drainage for 4 weeks [24]. The Dutch group failed to find any relationship between the timing of catheter drainage and the outcome and concluded that waiting 4 weeks was not always necessary [16]. They have gone on to show that clinically relevant WON (largely or fully encapsulated) occurs in 43% patients within the first 3 weeks [31]. Several observational studies have also suggested that a collection does not have to be "walled off" for safe and successful catheter drainage of IPN [30,32]. The results of this survey are in line with that trend, with only a third of respondents indicating that encapsulation was a major factor in whether to drain or not. For the timing of debridement, less than 10% of the respondents would choose an aggressive approach like immediate debridement (before or after 4 weeks) when intervention was considered necessary.
There are some important limitations to this study. While there are responses from almost all (30/34) of the provincial regions in China and from the 4 relevant medical specialties, it is not possible to prove that the results of the survey are representative of China as a whole. Those receiving the survey request were those that expressed an interest in acute pancreatitis, but not necessarily those who made clinical or policy decisions regarding its management. In the absence of any previous national data on the management of AP, it is also not possible to say to what extent the responses reflect a change in the approach to the diagnosis and treatment of local complications of ANP. The survey has highlighted the need to adopt a standardized approach to the definition of complications and interventions in this field. The "VRP" approach to describing interventions for the local complications as we used in the questionnaire, where the visualization method, route, and purpose of intervention are tightly defined, should be considered for future studies [7,33].

Conclusion
In conclusion, there is evidence that the "step-up" approach to the management of IPN has been widely adopted in most parts of China, with PCD or endoscopic catheter drainage as the initial treatment. There is also evidence of significant variation in practice, and this serves to highlight areas warranting further research. These include the optimal way to diagnose infection; the optimal timings for antibiotics, drainage, and debridement; and the indications for intervention when infection is not suspected in relation to ANC. Whether organ failure, as a systemic complication, should be considered as an indication for local intervention in this setting warrants further investigation. 6 Gastroenterology Research and Practice

Data Availability
All relevant data analyzed during this study are presented in tabular form in this published article. The original datasets used during the current study are available from the corresponding author on reasonable request.

Conflicts of Interest
The authors declare that they have no competing interests.