Abstract and Introduction
Abstract
Background Anecdotally, liver size is important in determining prognosis in patients with end-stage liver disease (ESLD).
Aims To assess if a ratio of liver area and abdominal area on cross-sectional imaging could accurately predict mortality in ESLD.
Methods A retrospective–prospective cohort study was performed on patients with ESLD in a training set. The censor point used was date of patient death or liver transplant (LT). The liver to abdominal area ratio (LAAR) was calculated using the formula {LAAR = [liver area (cm)/abdominal area (cm)] × 100}. A validation set was collected from a different institution.
Results Three hundred and sixteen patients were identified. Complete imaging and survival data were available in 158 subjects, 100 male (63%). The LAAR score detected progression to death/LT in our cohort (P < 0.003). Its prognostic accuracy at 90, 360 and 720 days, using the optimal cut-off (32.1), from baseline CT date to death/LT using the log-rank test was P = 0.28, P = 0.06 (OR 1.347, 95% CI 0.94–1.94) and P < 0.0001 (OR 1.89, 95% CI 1.25–2.85) respectively. On multivariate analysis, LAAR (P = 0.008), MELD (P = 0.004) and MELD-Na (P = 0.03) were independently associated with the primary study outcome measurement at 720 days. The validation set of 52 patients confirmed the utility of the LAAR to determine risk of death or need for LT, AUROC 0.89 (0.78–0.97), and P < 0.0001.
Conclusions The liver to abdominal area ratio (LAAR) score offers a new paradigm in disease modelling in end-stage liver disease (ESLD) and offers prognostic accuracy at 2 years from computer tomography (CT) imaging.
Introduction
Cirrhosis occurs as the consequence of chronic and sustained liver injury and marks the clinical and histological manifestation of end-stage liver disease (ESLD). It consists of a compensated phase in which the patient is often clinically well and a decompensated phase in which the patient develops the classical symptoms and signs of cirrhosis including ascites, hepatic encephalopathy (HE), jaundice, spontaneous bacterial peritonitis (SBP) and variceal bleeding. The onset of decompensated cirrhosis is an ominous landmark in the natural history of chronic liver disease and is associated with a grave prognosis. Liver transplantation (LT) is the only curative option for these patients, but this significant advance in the management of ESLD has been hampered by donor organ shortages and prolonged waiting times on the LT wait list.
One of the skills required by liver clinicians is the ability to identify cirrhotic patients with poor prognostic features most likely to benefit from LT. When LT is not a viable option, accurate prognostication allows a change in emphasis from active treatment to end-of-life care. The timing of referral for LT assessment is vital. If the referral is too late, the patient may become too unfit for general anaesthesia or surgery. Clinical parameters including ascites, nutrition and HE have been considered important to identify patients with poor prognostic features. Furthermore, biochemical parameters such as serum sodium, serum creatinine, serum bilirubin, and international normalised ratio (INR) have been found to be independent factors predictive of death in patients with ESLD. Disease modelling has led to the development of prognostic scoring systems including the Model for End-stage Liver Disease (MELD), the United Kingdom Model for End-stage Liver Disease (UKELD), the integrated MELD (iMELD) and the MELD-Sodium (MELD-Na) being utilised for liver transplant recipient selection and prioritisation.
There are anecdotal data that liver size is important in determining patient survival in ESLD. Computer Tomography Derived Liver Volume (CTLV) has been reported as a prognostic factor in acute liver failure, but the use of liver size as a determinant of survival has not been previously evaluated in the context of ESLD. CTLV requires software that is not routinely available in all radiology units. The present 'proof of concept' study addressed the hypothesis that a simple geometric method could be used to determine the risk of death or LT in patients with ESLD.