Introduction Liver stiffness measurement (LSM) using Transient Elastography (TE) is a powerful non-invasive tool for staging hepatic fibrosis. Our aim was to establish the ability of LSM in predict ing liver-related complications and mortality.

Method We retrospectively reviewed consecutive patients who underwent outpatient TE between April 2013 & April 2014 (with follow up until April 2016) at a teaching hospital in SE England. Cox regression analysis was applied to determine the relationship between LSM and liver-related events or mortality. Liver-related events were defined as one or more of the following: development of portal hypertension, hepatic decompensation, hepatocellular cancer and or need for transplantation.

Results A total of 403 patients fulfilled study criteria of whom 155 (38.5%) had LSM > 8 kPa indicating clinically significant hepatic fibrosis. Overall 94 (23%) had cirrhosis (LSM > 13 kPa). The mean age of the cohort was 50±13.4 years and 271 (67%) being males. Indication for performing TE was viral hepatitis (n=225, 56%) followed by alcohol (47, 12%) and non-alcoholic fatty liver disease (n=41, 10%). Of those with cirrhosis (n=94), the mean age of this subgroup was 56±10 years and 68 (72%) were males and median LSM was 22.1 kPa (13.1–75 kPa). Of those who had bloods results available at baseline (88/94), the majority had Child-Pugh score (CPS) A (n=73, 83%) while all of the remaining were in CPS B (n=15, 17%). The mean MELD score was 9.8±4.6. Almost half of the patients (n=44/94, 47%) with cirrhosis developed one or more liver-related complications. The majority (n=40/94) had the complications at baseline (i.e within 6 months following index LSM.). The median LSM score at baseline for those with liver-related complications was significantly higher than that of the patients without these events (35.3 vs 20.9 kPa, p=0.003). Moreover, of the 55 patients who underwent gastroscopy, those with oesophageal varices at baseline (20, 36.3%) had significantly higher median LSM than those without varices (48 vs 21.5 kPa, p=0.0003). Age, gender, aetiology of liver disease and presence of comorbidities did not independently influence the development of varices or liver-related events. Twelve patients (13%) with cirrhosis (n=94) died within a median time to death of 8 months (1–24 months). The median stiffness score of those who died was not different from that of the surviving group (24.3 vs 21.7 kPa).

Conclusion This retrospective cohort study suggests that TE could potentially be a useful non-invasive tool in predicting portal hypertension and or development of hepatic decompensation in individuals with cirrhosis.

Disclosure of Interest None Declared