Society: AGA
Background: Around 15% of acute pancreatitis patients develop necrosis of the pancreatic/peri-pancreatic tissue resulting in necrotic debris surrounded by a wall of granulation tissue called “walled-off pancreas necrosis” (WON). Patients with WON have a prolonged clinical course, often associated with infection within the necrotic debris. Direct endoscopic necrosectomy has been developed to remove the necrotic debris but is technically challenging due to the gluey debris and limited endoscopic accessories, requiring a median of 3-6 necrosectomy sessions for complete removal. Here, we aim to develop a chemical dissolution product to facilitate the breaking down and dissolution of the solid necrosome, allowing immediate aspiration in a single session.
Methods: Human pancreas necrosis specimens from endoscopic necrosectomies were collected and characterized using near-infrared spectroscopy to help identify candidate dissolution chemicals. The liquid dissolution chemicals were converted into a foam formulation to improve spreading within the WON cavity and contact with the debris. A novel endoscopic foam generation cap device was developed to deliver the dissolution foam through cyst gastrostomy (Fig.1). Ex-vivo high-throughput dissolution studies were performed on the porcine-pancreas necrosis model to optimize the foam formulation, requiring the lowest concentration of dissolution chemical and ensuring the stability of the foam (Fig.2). In-vivo porcine WON model was developed through implantation of the donor pancreas into the abdominal cavity of the recipient pig, to evaluate efficacy of dissolution foam and foam generation cap device. Safety studies were performed evaluating the effect of dissolution foam on granulation tissue wall, blood vessels, and gastric mucosa.
Results: Human pancreas necrosis debris characterization on spectrometry showed the presence of complex fats and proteins, nearly identical to the debris from the porcine pancreas necrosis model (p < 0.01). Hypochlorite (3.5%) based foam formulation with surfactants and stabilizing polymers showed >90% dissolution of necrotic debris at 10 mins, allowing complete suction through an endoscope (Fig.2). Stability testing of the foam formulation showed >95% intact bubbles at up to 18 hours. The in-vivo porcine study demonstrated adequate foam delivery through cyst gastrostomy opening (Fig. 2D). Safety studies showed no significant damage on gross and histopathological examination.
Conclusion: This novel foam formulation can chemically dissolve the WON necrotic debris in less than 10 minutes without damaging the granulation tissue wall or blood vessels and can be effectively delivered using our endoscopic foam generation cap device. This technology will potentially allow a single procedure complete necrosectomy avoiding the need for repeat interventions.

Figure 1: A – Foam generating cap that attaches to endoscope and converts injected solution into foam. B – In-vitro foam generation set-up to evaluate foaming from the dissolution solution, along with the cap attachment device. Peristaltic pump and gas source used to mimic endoscope.
Figure 2: A – Ex-vivo dissolution study demonstrating complete dissolution of necrotic debris in < 10 mins and can be suctioned with an endoscope. B – Illustration of dissolution foam delivery into the necrotic cavity using cap attachment, through the cystgastrostomy. C – Custom 3D printed foam generation cap attachment prototype. D – In-vivo porcine study demonstrating the feasibility of foam delivery using the cap attachment, through cyst gastrostomy lumen-apposing metal stent.
Background: Post-ERCP pancreatitis (PEP) is a common complication, that occurs in 5-10% of patients who first received Endoscopic retrograde cholangiopancreatography (ERCP). Recently, the efficacy of early refeeding for acute pancreatitis (AP) has been reported, however, the effect of early refeeding has not been reported in patients with PEP. Therefore, we investigated the safety and efficacy of early oral refeeding (ERF) compared to delayed refeeding (DRF) in patients with mild PEP through a prospective randomized controlled multicenter study.
Methods: Eligible patients were randomized to the ERF and DRF groups with a 1:1 ratio. In the ERF group, an oral diet was started 24 hours after PEP was diagnosed. In the DRF group, an oral diet was started after confirmation of restoring normal bowel sound, and abdominal pain decreasing below visual analog scale 2. The oral diet was started with sips of water and built up sequentially in the order of clear liquid diet - soft diet, considering patient tolerability. The oral refeeding was interrupted when the pain scale of the patient increased to VAS 5 points or more, or the patient refused diet due to abdominal pain or other reasons. After the interruption of refeeding, the diet was restarted after the amylase/lipase level decreased below the upper normal limit, the abdominal pain disappeared, and the bowel movement was restored. If the patient could tolerate more than 24 hours after completion of the diet build-up, it was determined that patient met the discharge criteria. The primary outcome was the hospitalization period for PEP, and secondary outcomes were the progression to severe AP, readmission rate (<30 days), mortality, and morbidity rate related to PEP.
Results: Between Feb. 2021 and Sep. 2022, 78 patients were enrolled in 9 referral centers and randomized to ERF (n=40) and DRF (n=38) groups (Figure 1). There was no significant difference between groups in baseline characteristics (age, sex, past history of acute pancreatitis, ERCP indications, smoking, alcohol, and other comorbidities) and procedural parameters (cannulation method, total procedure time, rate of difficult cannulation, pancreatic duct cannulation, and biliary stent insertion). The baseline severity of PEP was not different between groups. During refeeding, 4 patients in the ERF group and 3 patients in the DRF group were undergone interruption of refeeding (10% vs. 8%, p=0.745). The hospitalization period for PEP was significantly decreased in the ERF group compared to the DRF group (2.9 ± 1.6 vs. 3.9 ± 2.0 days, p=0.023, Figure 2). Severe AP, readmission (<30 days), mortality, and morbidity rate related to PEP did not occur in both groups.
Conclusion: Early oral refeeding significantly reduced hospital stays in patients with post-ERCP pancreatitis compared to delayed oral refeeding and did not increase PEP-related safety issues.

Figure 1. Study flow chart
Figure 2. The length of hospital days