Infected pancreatic necrosis: drain first, but do it better

Several trends are evolving that will shape the future management of infected pancreatic necrosis (IPN) and its complications. First, open necrosectomy is no longer considered the standard of care in many leading centres and it is no longer considered an absolute requirement for IPN. This has occurred during the proliferation of minimally invasive treatments for IPN¹ and the emerging evidence that they result in improved outcomes.² When the Liverpool group compared contemporaneous patients with similar demographics and pre-operative parameters who had minimal access retroperitoneal necrosectomy or open necrosectomy, the former had a lower morbidity, mortality and need for intensive care, reflected in a lower rise in post-operative severity markers and onset of new organ failure.³ We now recognize that the metabolic impact of open necrosectomy can exceed the physiologic reserve of patients with multiple organ dysfunction syndrome.

The second trend has been due to the realization that there should be a delay before intervention, whenever clinically possible, and for at least 3 weeks from the onset of the disease.⁴‘Staying the surgeon's hand’ allows the maturation of infected and necrotic lesion(s) and allows more time for intensivists to stabilize and support patients. The placement of an external drain into these lesions can ‘take the heat out of the fire’ by relieving pus under pressure, stemming the systemic inflammatory response and sometimes allowing reversal of organ dysfunction. This ‘buys time’ before definitive intervention on a more mature lesion in an improved patient.

The third trend was highlighted by the first randomized controlled trial involving minimally invasive necrosectomy, the PANTER trial, which compared open necrosectomy with a ‘step-up approach’ where drainage, by percutaneous or endoscopic means, was followed by videoscope-assisted retroperitoneal debridement.⁵ The patients who had the step-up approach had a significant decrease in the composite endpoint (major complications and/or death) and the onset of new multiple organ failure, with fewer admissions to the intensive care unit. This study provides compelling evidence for adopting primary percutaneous drainage (PCD) followed by minimally invasive necrosectomy in patients with suspected IPN. There are some settings where intervention by percutaneous or endoscopic drainage is not appropriate in necrotizing pancreatitis, for example, retroperitoneal haemorrhage, intestinal necrosis and duodenal/biliary obstruction, and there are occasional situations where PCD is not considered technically feasible.⁶

Finally, it is recognized that PCD can be the sole treatment in some patients with IPN, avoiding an unnecessary necrosectomy altogether. There has been extensive experience with PCD as a secondary treatment for residual collections after open necrosectomy, and an increasing experience of it to delay necrosectomy, but there is much less experience with it as sole treatment. We do not know whether PCD is best performed when infection is suspected, when it is confirmed or when it can be delayed. Some interventional radiologists have advocated primary PCD for some time but widespread adoption has not occurred.⁷ However, recent data suggest that this might soon change. A comprehensive systematic review found that 56% of patients, which included those with both sterile and IPN, did not require a surgical necrosectomy after PCD.⁸ There will remain debate as to the proportion of patients who can be treated solely by PCD. The 56% from the systematic review⁸ and 35% from the PANTER trial⁵ are likely to be an overestimate, because of selection bias and other design limitations, especially with the source studies. The role of sole PCD in treating IPN needs further evaluation, but it will not go away. The challenge is how to do it better.

One clue to improving the outcome from sole PCD comes from the striking variation in technique revealed in the systematic review.⁸ Only 5 of the 11 studies used the Seldinger technique, narrow calibre drains were used frequently and only one study undertook routine stepwise dilation for upsizing. The average number of drains used for each patient was only two and the number used at any one time was impossible to determine. In spite of wide variation in technique there is considerable benefit from PCD and standardization should bring further benefit. This would likely require a more vigorous drainage approach in which drains are exchanged routinely to avoid occlusion and upsized routinely to facilitate egress. Larger and multiple drains would also enable more effective irrigation, including continuous and even pulsed methods. Questions remain as to which type of fluid should be used for irrigation and whether it can be used actively to accelerate liquefaction. The development of such an ‘active or augmented’ approach to drainage may well be the way forward. Establishing larger diameter access ports should also allow repeat procedures to be performed without the need for general anaesthesia,⁹ and even as an outpatient. If hospital stay can be reduced there will be notable economic benefits from PCD.¹⁰

The trail has been blazed, traffic is increasing, but the road has yet to be paved. Further work is required, not only in regards to standardizing PCD techniques, but in determining ways to do it better. The downside to PCD in this setting is that it cannot be applied to all lesions in all patients and there is a substantial failure rate. However, it is not the failure rate that is important but that PCD will avoid the need for further intervention in a proportion of patients. The challenge is to drive the proportion even higher with more efficient and effective PCD techniques, resulting in a decrease in hospital stay and mortality.

Conflicts of interest

None declared.