IS ABDOMINAL ULTRASONOGRAPHY A RELIABLE MODALITY TO DIAGNOSE COMPENSATED ADVANCED CHRONIC LIVER DISEASE?

IntroductionCompensated advanced chronic liver disease (cACLD) refers to a cohort of asymptomatic patients with advanced fibrosis who do not yet exhibit significant biochemical evidence of synthetic dysfunction or clinical and radiological manifestations of portal hypertension. Ultrasonography (US) is a first-line radiological investigation for patients with suspected chronic liver disease but its role in diagnosing cACLD is not well-established. The aim of this study is to assess the diagnostic accuracy of US in identifying patients with cACLD using categorical cut offs of Transient Elastography (TE) based on Baveno VI consensus.MethodsA retrospective cross-sectional observational study was conducted at the Royal Wolverhampton Hospital NHS Trust for patients with suspected chronic liver disease who underwent US, TE and screening blood tests within 12 months of each other as part of their investigations. Patients were included if; TE recorded a valid reading and the US report mentioned a radiological description suggestive of cirrhosis as per the nomenclature agreed with the local specialist sonographers (coarse echotexture, irregular liver margin, nodular or heterogeneous parenchyma, or cirrhosis). Patients were excluded if; US showed signs of portal hypertension, there was a past history of hepatic decompensation, liver transplant or heart failure, or if they were currently being evaluated for acute hepatitis. Sensitivity, specificity, positive and negative predictive value (PPV & NPV) of US to detect cACLD were calculated using categorical cut offs of TE based on Baveno VI consensus (TE >15 kPa as positive and <10 kPa negative for cACLD respectively).Results1528 of 3357 screened patients were included; mean age 51 years, 64% men and 71% of white ethnicity. Aetiology of liver disease included [%, n]; ALD (20, 312) NAFLD (29, 452), Hepatitis C (11.4, 175), Hepatitis B (11, 168), PBC (4, 63), AIH (2.7, 42), Haemochromatosis (2.7, 41), unspecified (18, 275). The overall sensitivity and specificity of US for diagnosis of cACLD was 64.6% (p<0.0001) and 78% (p<0.0001) respectively, with PPV of 40% (p<0.0001) and NPV of 91% (p<0.0001). A similar trend was seen across all aetiologies. We also found that a platelet count (PLT) of <150 x 10^9/L and serum albumin concentration (ALB) of <35 g/L improved the PPV of US for detecting cACLD to 84% (95% CI 73—95%, p<0.0086), while a normal PLT and ALB reduced the PPV to 21% (95% CI 15.4—26.2%, p<0.0001). Similarly, the NPV of US for excluding cACLD was 93% (95% CI 91—94.7%, p<0.0001) when both PLT and ALB were in normal range.ConclusionsUS alone is not a reliable imaging modality for diagnosing cACLD. In suspected cases, TE should be carried out to determine the severity of hepatic fibrosis before a diagnosis of cACLD is made.