Prognostic significance of differentiating necrosis from fluid collection on endoscopic ultrasound in patients with presumed isolated extrapancreatic necrosis

Introduction

Acute pancreatitis (AP) is a complex acute inflammatory disorder of the pancreas with potentially catastrophic consequences. Its clinical course is usually mild in the majority of patients, with uneventful recovery [1]. However, in up to 20% of patients the attack may be severe and may lead to a number of local and systemic complications, including organ failure, prolonged hospital course, and significant mortality [1-4]. The factors that determine the severity of AP have not been completely understood, and it is therefore often difficult to predict its clinical course [1-4]. A better understanding of the disease over the last few decades has led to the identification of clinical, investigational and radiological factors that can help in predicting the clinical course of AP [1,4].

Of these prognostic variable, pancreatic necrosis has been shown to be one of the significant determinants of poor prognosis [5-7]. Studies have shown that the majority of patients who develop organ failure, or other complications of AP, have necrotizing pancreatitis, while patients with non-necrotizing interstitial AP usually have an uneventful recovery [5-7]. Recent studies have also identified a new entity of extrapancreatic necrosis alone, which is defined as necrosis of the peripancreatic tissue only on contrast-enhanced imaging studies, with a normally enhanced pancreatic gland [8-10]. The prognosis of patients with isolated extrapancreatic necrosis is better than that of patients with pancreatic necrosis, but worse than that of patients with acute interstitial pancreatitis [8-10].

Extrapancreatic necrosis is usually diagnosed on contrast-enhanced computed tomography (CECT) as extrapancreatic changes that are more than simple fat stranding; both peripancreatic fluid collections and peripancreatic necrosis have similar appearances on CECT [8,9]. Endoscopic ultrasound (EUS) has been shown to be a better modality than CECT for differentiating fluid collections from solid-containing necrotic collections [11-13]. We have also demonstrated that EUS can detect pancreatic necrosis reliably [14]. As EUS provides high-resolution images of the pancreas and surrounding areas, we hypothesized that EUS may be able to differentiate between peripancreatic fluid collections and peripancreatic necrosis in patients with presumed isolated extrapancreatic necrosis and that these two subtypes of isolated extrapancreatic necrosis may have different clinical courses and outcomes.

Patients and methods

This study is a retrospective analysis of a prospectively collected database that includes all the patients with AP from a large tertiary care hospital. The data of all the patients with AP seen in our unit from December 2012 to November 2014 were retrieved, and patients with isolated extrapancreatic necrosis who had undergone both EUS and CECT were identified. Patients who had no contraindication for EUS and presented to us within 5 days of the onset of symptoms were enrolled. Those with coronary artery disease, pre-existing cardiac disorder or respiratory disorder, malignancy, pregnancy, underlying chronic pancreatitis, pancreatic malignancy, and patients who refused consent for EUS were excluded. Informed consent was obtained from all patients prior to the EUS procedure and the study protocol was approved by the institute’s ethics committee.

All the participants of this study were thoroughly interviewed and underwent clinical and laboratory examinations. The diagnosis of acute pancreatitis was based on suggestive clinical history (acute onset upper abdominal pain with or without radiation to the back, vomiting, constipation and obstipation), physical examination, and serum amylase elevated to greater than three times the upper limit of normal and/or positive abdominal imaging [1]. Hematological investigations, blood sugar, serum amylase, blood urea, serum creatinine, albumin, globulin, lipid profile, calcium, phosphate, arterial blood gas analysis and liver function tests were performed on the day of admission. The etiology of acute pancreatitis was based on the patient’s history of alcohol and drug use, biochemical investigations, transabdominal ultrasonography, and radiological findings. Organ failure was described as the presence of respiratory failure (PaO₂/FiO₂ <300), renal failure (serum creatinine >1.8 mg/dL), or cardiovascular failure (blood pressure <90 mmHg and non-responsive to fluids) [1]. Involvement of more than one organ system was classed as multiorgan failure, while if the organ failure persisted for >48 h it was labeled as persistent organ failure.

All patients diagnosed with acute pancreatitis underwent EUS on the day of admission, performed by either of two experienced endosonologists (SSR, DKB) using a radial echoendoscope (Pentax EG-3670 URK radial echoendoscope, Pentax Corp., Japan or GF-UE 160 radial echoendoscope, Olympus Corp., Japan) at a frequency of 7.5 MHz. EUS was performed with the patient in the left lateral decubitus position under conscious sedation with intravenous midazolam (2.5-5 mg). The uncinate process and head of the pancreas were evaluated from the duodenum, and the body and tail of the pancreas were evaluated from the stomach.

An abdominal CECT scan was performed in all patients 3 days after the onset of symptoms and the presence of pancreatic/extrapancreatic necrosis was recorded. Pancreatic necrosis was defined as focal or diffuse areas of non-enhanced pancreatic parenchyma on CECT, whereas extrapancreatic necrosis was defined as extrapancreatic morphological changes that were more than simple fat stranding [1,9].

The EUS images were retrieved for patients with isolated extrapancreatic necrosis. Isolated extrapancreatic necrosis was diagnosed on CT when there were extrapancreatic changes as defined above, with complete enhancement of the pancreatic parenchyma. On EUS, isolated extrapancreatic necrosis was characterized into two subtypes:

To differentiate between peripancreatic fluid collections and peripancreatic necrosis, the EUS images were read separately by two experienced endosonographers (SSR and DKB) who were blinded to the clinical data. In the event of disagreement between the two endosonographers, the images were reviewed jointly by both and their consensus opinion was recorded.

These two groups were compared in relation to various outcome measures: organ failure (including multiorgan failure and persistent organ failure), need for percutaneous or endoscopic or surgical interventions for local complications, and mortality.

Results

Thirty-six patients (25 male; age range 19-65 years) with isolated extrapancreatic necrosis (normally enhancing pancreatic parenchyma on CECT) were studied. The etiology of AP was alcohol in 16 (44.4%) patients, gallstone disease in 13 (36.1%), and other in 7 (19.4%) patients. Twenty-eight (77.7%) patients developed pleural effusion, 6 (16.6%) developed ascites, 20 (55.5%) developed organ failure, 13 (36.1%) developed persistent organ failure, 3 (8.3%) developed multiorgan failure, 5 (14%) needed endoscopic/percutaneous radiological intervention, 2 (5.5%) needed surgery and 2 (5.5%) patients succumbed to the illness.

On EUS, 25 (69.4%) patients had peripancreatic heterogeneously echotextured areas suggestive of peripancreatic necrosis and 11 (30.6%) patients had peripancreatic anechoic areas suggestive of peripancreatic fluid collections. None of the patients had EUS features of pancreatic necrosis, as described in our previous paper [14]. There was no significant difference between the demographic profiles of patients in these two groups (Table 1).

Table 1 Comparison of outcomes in patients with peripancreatic necrosis and peripancreatic fluid collections

Discussion

In patients with acute necrotizing pancreatitis there is usually necrosis of the pancreatic parenchyma as well as the peripancreatic fat and tissues [1]. However, some patients have necrosis of the peripancreatic tissue/fat only, with a normally enhanced pancreatic gland on contrast-enhanced imaging studies; this entity has been labeled as isolated extrapancreatic necrosis. Previous studies have demonstrated that isolated extrapancreatic necrosis is an entity that is distinct from mild interstitial non-necrotizing pancreatitis and from pancreatic necrosis, alone or combined with extrapancreatic necrosis [8-10,15]. Patients with isolated extrapancreatic necrosis have been shown to have lower rates of persistent organ failure, local complications, need for intervention and mortality, compared to patients with combined pancreatic and extrapancreatic necrosis.

We have previously shown that EUS performed on the day of admission can accurately diagnose patients with acute necrotizing pancreatitis. On EUS, patients with acute necrotizing pancreatitis had novel findings of either multiple hypoechoic or hyperechoic areas measuring more than 5 mm in diameter, whereas these findings were not present in any of the patients with acute interstitial pancreatitis [14]. We also found that the presence of peripancreatic hypoechoic areas correlated with the presence of extrapancreatic necrosis on CT; these areas were found to be soft on EUS elastography, possibly suggesting fat necrosis [14]. However, not all patients who showed extrapancreatic necrosis on CT had similar findings on EUS, with some patients having anechoic peripancreatic areas suggestive of fluid collection, rather than hypoechoic areas that suggested peripancreatic necrosis. Therefore, in the current study we compared the clinical course of these two types of extrapancreatic necrosis.

In this EUS study, we have shown that presumed isolated extrapancreatic necrosis on CT can be characterized into two subtypes: peripancreatic necrosis, which is characterized by the presence of peripancreatic heterogeneously echotextured areas, and peripancreatic fluid collection, which is characterized by the presence of peripancreatic anechoic areas. Peripancreatic necrosis and peripancreatic fluid collections are associated with different clinical courses, with patients who have peripancreatic necrosis having higher rates of organ failure (including multiorgan failure and persistent organ failure), ascites, pleural effusion, need for percutaneous, endoscopic or surgical intervention for local complications, and mortality, when compared to patients with peripancreatic fluid collection. The higher frequency of complications in the peripancreatic necrosis group is not surprising, as the presence of tissue necrosis has been shown to be associated with a higher frequency of complications and organ failure [5,16]. The causes of organ failure and complications in AP are multifactorial and may possibly be due to the release of various cytokines and vasoactive substances, leading to the activation of proinflammatory pathways, as a consequence of pancreatic and peripancreatic tissue necrosis [17,18].

Our study has a few limitations, including the small sample size and the exclusion of sick patients who had contraindications for EUS, such as those with hemodynamic or severe respiratory compromise. In addition, the interobserver variability was not assessed for either the CT or the EUS findings. Moreover, the EUS findings were not compared with the gold standard, namely surgical findings.

In conclusion, isolated extrapancreatic necrosis on CECT represents a heterogeneous group comprised of patients with peripancreatic fluid collections and/or peripancreatic necrosis. Patients who exhibit peripancreatic fluid collections on EUS have a less severe disease course than do those with peripancreatic necrosis.