2 The procedure
2.1.1 Pancreatic necrosis (also called necrotising pancreatitis) is a serious complication of acute pancreatitis that can occur in some patients. It is associated with significant morbidity, requiring prolonged hospitalisation, and high mortality.
2.1.2 Traditionally pancreatic necrosis has been treated by open necrosectomy via laparotomy, but image-guided drainage or laparoscopic drainage may also be used.
2.2.1 Percutaneous retroperitoneal endoscopic necrosectomy aims to remove necrotic tissue under direct vision. The procedure is less invasive and may improve prognosis compared with traditional open surgery. Percutaneous drainage may be attempted as part of the management prior to the procedure.
2.2.2 With the patient under general anaesthesia, an endoscope (which may be rigid or flexible) is inserted via a posterolateral approach into the retroperitoneal space to visualise the area of necrosis. Dead tissue is removed, for example using suction, lavage or forceps, and debrided where necessary using forceps. Drains may be placed for irrigation in the postoperative period. The procedure may be repeated if required.
Sections 2.3 and 2.4 describe efficacy and safety outcomes from the published literature that the Committee considered as part of the evidence about this procedure. For more detailed information on the evidence, see the overview.
2.3.1 A randomised controlled trial (RCT) of 88 patients treated by a step-up protocol involving drainage followed-up as required by percutaneous retroperitoneal endoscopic necrosectomy versus primary open necrosectomy, reported mortality rates of 19% (8/43) and 16% (7/45) respectively (p = 0.70) (patients in this study were followed-up for up to 6 months from hospital discharge). In the group randomised to drainage followed as required by percutanenous retroperitoneal endoscopic necrosectomy 60% (26/43) of patients underwent the procedure, 35% (15/43) of patients required drainage alone and 5% (2/43) of patients with multiple organ failure were too unstable for the procedure and underwent endoscopic transgastric drainage.
2.3.2 A non-randomised controlled study of 189 patients treated by the procedure or open pancreatic necrosectomy reported mortality rates of 19% (26/137) and 38% (20/52) respectively (p = 0.009) (follow-up not stated).
2.3.3 A non-randomised controlled study of 30 patients treated by the procedure or open necrosectomy reported in-hospital mortality rates of 7% (1/15) and 40% (6/15) respectively (p = 0.08).
2.3.4 The non-randomised controlled study of 30 patients treated by the procedure or open necrosectomy reported postoperative multiple organ failure in 13% (2/15) and 67% (10/15) of patients respectively (p = 0.008).
2.3.5 The RCT of 88 patients comparing drainage followed as required by percutanenous retroperitoneal endoscopic necrosectomy with primary open necrosectomy, reported a composite rate of major complication or death in 40% (17/43) and 69% (31/45) of patients in either group respectively (p = 0.006) (follow-up of up to 3 months from hospital discharge).
2.3.6 The Specialist Advisers listed key efficacy outcomes as a reduction in mortality and morbidity, reduction of requirement for postoperative critical care, number of interventions required and length of hospital stay.
2.4.1 The RCT of 88 patients comparing percutanenous retroperitoneal endoscopic necrosectomy with primary open necrosectomy, reported fistula formation or perforation requiring intervention in 33% (14/43) and 22% (10/45) of patients respectively (p = 0.32) (patients in this study were followed-up for up to 6 months from hospital discharge).
2.4.2 Bowel perforation occurred in 7% (1/15) of patients treated by the procedure and in 13% (2/15) of patients treated by open necrosectomy in the non-randomised controlled trial of 30 patients (p = not significant). In the same study, pancreatic fistula developed in 13% (2/15) of patients and 0% (0/15) of patients respectively (p = not significant; follow-up not stated).
2.4.3 The RCT of 88 patients reported that bleeding requiring intervention occurred in 16% (7/43) of patients treated in the percutanenous retroperitoneal endoscopic necrosectomy group and in 22% (10/45) of patients treated by open necrosectomy (p = 0.48).
2.4.4 The Specialist Advisers stated that adverse events reported in the literature include incomplete drainage and/or sepsis control, iatrogenic injury to the kidney or spleen, colonic necrosis, pseudocyst formation, venous thrombosis and death.