AODTH-009 Quantitative elastography is useful in the diagnosis of chronic pancreatitis

    loading  Checking for direct PDF access through Ovid

Abstract

Introduction

Endoscopic ultrasound (EUS) is widely used in the assessment of chronic pancreatitis (CP), however the Rosemont criteria that was initially developed has been criticised for being subjective and susceptible to intra-observer variability. Quantitative tissue stiffness measurement of the pancreas with EUS-elastography (EUS-E) is a potential non-subjective alternative to diagnose CP. It has been described as a tool for diagnosing CP in a single centre study with EUS-E readings/strain ratio of >2.25 found to have a diagnostic accuracy of >90%; but has not been validated. We aimed to assess if EUS elastography could accurately identify CP compared to EUS Rosemont scores.

Method

Patients referred for pancreatic EUS with suspicion of CP and those referred for EUS for unexplained upper abdominal pain were recruited prospectively. All patients had pancreatic computerised tomography (CT) or magnetic resonance cholangiopancreatography (MRCP) prior to EUS. Patients were also asked to return a stool sample for faecal elastase-1 (FEL-1) to assess for exocrine dysfunction. Two endoscopists conducted standard pancreatico-biliary EUS examinations using linear EUS endoscopes (UCT-260). Rosemont grades were recorded. EUS-E was performed to obtain strain ratios from the head, body and tail of pancreas, the mean of three ratios was included in analysis. Linear regression was used to determine if there was an association between strain ratios and number of features of CP. Student’s t test was used to determine if there was a difference in strain ratio in patients with low and normal FEL-1 and when comparing the strain ratios of the CP group and abdominal pain controls.

Results

25 patients were recruited (9 CP group median age 51.2, range 24–82, 3 male and 16 abdominal pain group median age 53.7, range 36–70, 5 male). There was a statistically significant difference in average strain ratios when comparing the two groups (11.8 versus 2.1 p=0.017). Using a cut off of 2.25 we report sensitivity of 100% and specificity of 60% for diagnosis of CP (PPV 57.1%, NPV 100%). Higher strain ratios predicted finding a greater number of parenchymal features of CP on EUS suggestive of CP (R2=0.34 p=0.0046). The strain ratios of those with normal (>200 µg/g) and low (<200 µg/g) FEL-1 results were significantly different (2.11 versus 5.14 p=0.029).

Conclusion

EUS-E is a useful adjunct to existing tests to exclude a diagnosis of CP with unexplained abdominal symptoms. Further recruitment to this study will improve the reliability to enable routine use in clinical diagnostic practice.

Disclosure of Interest

None Declared

Related Topics

    loading  Loading Related Articles