Walled-off pancreatic necrosis: endoscopy vs surgery

Research-informed explainer — last updated April 2026

If your doctor has told you that you have walled-off pancreatic necrosis, the most important thing to know is this: in most cases, an endoscopic procedure can drain it without surgery. A 2018 randomized trial published in Gastroenterology found that patients treated endoscopically had a 12% rate of major complications versus 41% for those treated surgically — and their hospital bills averaged $75,000 versus $117,000 [2]. Surgery is still necessary in some situations, but it is no longer the default.

This explainer draws on peer-reviewed research from four gastroenterologists and advanced endoscopists listed in the Convene directory: Mark Topazian, M.D., at Mayo Clinic, who co-authored the landmark U.S. multicenter series on direct endoscopic necrosectomy for this condition; Muhammad Hasan, MD, at AdventHealth Orlando, who published both the major randomized trial comparing endoscopy to surgery and a randomized trial on stent selection for drainage; Irving Waxman, MD, at Rush University Medical Center, whose work on lumen-apposing metal stents helped establish the endoscopic drainage technique; and Santhi Vege, M.D., also at Mayo Clinic, who authored the American Pancreatic Association consensus statement and the ACG guideline on managing acute pancreatitis and its complications.

What walled-off pancreatic necrosis actually is

Walled-off pancreatic necrosis, or WON, is a late complication of severe acute pancreatitis. During a bad episode of pancreatitis, part of the pancreas — or the tissue surrounding it — can die (necrose). Over the following weeks, the body walls off that dead tissue with a thick fibrous capsule. The result is a collection of dead tissue and fluid, encased in this wall, usually located near the pancreas or behind the stomach.

The 2012 revised Atlanta classification — the international consensus document that standardized how pancreatologists define these complications worldwide — distinguishes WON from simpler pancreatic pseudocysts [7]. Pseudocysts contain only fluid; WON contains solid necrotic debris. That debris is what makes drainage harder and what drove decades of surgical management.

WON typically matures around four weeks after the initial pancreatitis attack. Trying to intervene before that wall forms — before the collection is well-encapsulated — increases risk substantially, which is why both the ACG guideline and expert consensus recommend waiting when it is safe to do so [6] [8].

When treatment is actually needed

Not every walled-off collection requires drainage. Many WON cases resolve on their own. Intervention is indicated primarily when the necrosis becomes infected, and sometimes when a large sterile collection is causing persistent symptoms — pain, nausea, inability to eat — that do not improve over weeks.

Infected WON is the clearest indication. CT imaging showing gas bubbles inside the collection, or aspiration culture growing bacteria, confirms infection. Patients with infected WON typically look sick: persistent fever, elevated inflammatory markers, and failing to improve despite supportive care. Without drainage, infected WON carries a high mortality rate.

Symptomatic sterile WON is a more judgment-dependent call. The 2012 APA consensus statement on interventions for necrotizing pancreatitis, co-developed by Santhi Vege and colleagues, recommends intervention for symptomatic sterile necrosis only when conservative management has clearly failed — not as a first step [6].

At a glance

What the key trials actually showed

The most cited randomized trial on this question enrolled 66 patients with necrotizing pancreatitis at multiple U.S. centers and compared an endoscopic transluminal approach to minimally invasive surgery [2]. The results were not close. The endoscopy group had a composite major complication or death rate of 11.8% at six months. The surgery group reached 40.6%. Pancreatic fistulas — a common and debilitating surgical complication — occurred in 28% of the surgery patients and zero percent of the endoscopy patients. Average total costs were $75,830 for endoscopy versus $117,492 for surgery.

These results led to a significant shift in practice guidelines. Endoscopic drainage is now the preferred first approach at centers with the relevant expertise.

For patients who need active removal of solid necrotic debris — a procedure called direct endoscopic necrosectomy — the evidence also comes from systematic study. A multicenter U.S. series involving six tertiary centers and 104 patients found that endoscopic necrosectomy achieved resolution in 91% of cases, with a median of three procedures and an acceptable safety profile [1]. Mark Topazian at Mayo Clinic was among the co-authors of that work.

How the endoscopic procedure works

Endoscopic drainage typically starts with EUS — endoscopic ultrasound — where a specialist uses an endoscope with an ultrasound probe on the tip to see the WON collection through the wall of the stomach or small bowel. Once the collection is identified and confirmed to be in a safe location for puncture, a stent is placed directly through the stomach wall into the collection. This creates a channel through which the fluid and debris can drain into the gut.

Two types of stents are used: plastic pigtail stents and lumen-apposing metal stents (LAMS), which have flanged ends that hold them in position across the stomach wall. A 2014 prospective multicenter study by Irving Waxman and colleagues confirmed the safety and efficacy of the LAMS approach for pancreatic fluid collections, documenting favorable outcomes in a real-world patient population [5].

A subsequent randomized trial directly compared LAMS to plastic stents in 60 patients with WON [3]. The headline finding was that LAMS did not show superiority over plastic stents for overall treatment success — both approaches achieved comparable resolution rates. However, the trial found that LAMS were associated with a higher rate of stent-related adverse events (32% versus 7%) when left in place beyond three weeks. The practical takeaway from that trial: if LAMS are used, removing or exchanging the stent by the three-week mark substantially reduces complications.

When fluid drainage alone is not enough — when solid necrotic debris fills much of the collection — the endoscopist can pass instruments through the stent opening to physically break up and remove debris. This is direct endoscopic necrosectomy. It usually takes multiple sessions but avoids abdominal incisions entirely.

The disconnected duct problem

One complication that changes the management plan is disconnected pancreatic duct syndrome (DPDS). In DPDS, the pancreatitis has permanently disrupted a segment of the main pancreatic duct, leaving a portion of the pancreas disconnected and continuously leaking pancreatic juice. When this happens, WON tends to recur after stent removal because the underlying leak never heals.

A study published in Annals of Surgery, with Muhammad Hasan among the authors, followed 361 patients undergoing endoscopic drainage across more than a decade and found that DPDS was present in 46% of them [4]. Patients with DPDS were far more likely to need additional procedures, hybrid treatments combining endoscopy with percutaneous drainage, or rescue surgery. Crucially, the study found that leaving a permanent transmural stent in place dramatically reduced recurrence in DPDS patients — 1.7% recurrence with a permanent stent versus 17.4% without one.

If your imaging or clinical course suggests DPDS, ask your gastroenterologist whether permanent stenting is the right plan for you.

When surgery is still the answer

Endoscopic drainage requires two things: a mature, well-formed wall around the collection, and a collection that is located close enough to the stomach or duodenum to allow safe stent placement. When either condition is absent, endoscopy may not be feasible.

Surgery also becomes necessary when endoscopic treatment fails — the collection does not resolve, infection persists, or bleeding occurs. Hemorrhage into a necrotic collection is a serious complication that may require angioembolization or surgery. Patients with WON extending far from the stomach, or with collections in locations where a stent cannot be safely placed, may need percutaneous drainage (a CT-guided catheter placed through the skin) or surgery as the primary approach.

The APA consensus statement is explicit that the choice of approach should depend on local expertise [6]. A center without experienced endoscopists trained in necrosectomy should not attempt endoscopic drainage of complex WON — the technique has a real learning curve, and outcomes differ substantially between high-volume and low-volume centers.

Questions to ask your gastroenterologist

• Has my collection matured enough for drainage, or should we wait longer?
• Does your center perform direct endoscopic necrosectomy, and how many cases per year?
• Is there evidence on my imaging of infected necrosis, or is this symptomatic sterile WON?
• Do I have a disconnected pancreatic duct, and if so, would permanent stenting change the plan?
• What are the backup options if endoscopic drainage does not fully resolve the collection?
• Should I be managed at a specialized pancreatic center given the complexity of my case?

The bottom line

For most patients with walled-off pancreatic necrosis who need treatment, endoscopic drainage is the starting point. The randomized trial data is clear: fewer major complications, no pancreatic fistulas, and significantly lower cost compared to surgery. The technique requires waiting until the collection is mature and having access to a center with advanced endoscopy expertise. Surgery remains the right choice when anatomy prevents endoscopic access, when endoscopy fails, or when a hemorrhagic complication requires it. If you have been told you have WON and are being considered for an intervention, asking whether endoscopic drainage is an option at your center is the right first question.