Introduction Hepatocellular carcinoma (HCC) in western populations commonly occurs in the context of cirrhosis. There is therefore a rationale to offer ultrasound for surveillance for HCC with the aim of detecting malignancies at an early stage of disease for better prospects of cure. New guidelines from the National Institute of health and clinical excellence (NICE) in the UK, recommend surveillance be offered to all patients with cirrhosis.

Aims: To assess outcomes from patients enrolled in a HCC surveillance programme in a single centre in the UK, and to determine methods to identify patients most likely to benefit.

Method A single centre study of patients at the Royal Liverpool Hospital from September 2009 to October 2016. Patients were offered 6-monthly ultrasound scans. Patient characteristics collected included, age, gender, aetiology, standard biochemistry, BCLC stage, ECOG performance status, tumour size and number, and treatments offered. Additional factors included number of patients identified with curative disease, how many were offered curative treatment, and the survival at 1,2 and 5 years post diagnosis.

Results 299 patients were identified, 238 (80%), male. Median age 67 years (57-74). The disease aetiology where known was ALD=89, HBV=18, HCV=54, NASH=70, hemochromatosis=13, ALD + viral=13, others=24. BCLC stage was 0=19, A=101, B=63, C=38, D=48. Treatments offered were: resection-transplant=34, loco-regional therapy=132, sorafenib=19, palliation=92. Of 292 patients for which surveillance status was known 89 (31%), had HCC identified in surveillance. Surveillance patients were more likely to have disease detected at curable stage OR 0.56 (0.41–0.75, p<0.0001), and be offered curative therapy OR 0.65 (0.54–0.73 p<0.0001), this translated into improved patient survival at 1 year from diagnosis p=0.01, but not beyond (2 years p=0.26). 24% of surveillance patients were ECOG 0, while 18% of non-surveillance patients were ECOG 0. ECOG status was a strong predictor of death 1 year log rank=8.8, p=0.003, 2 years log rank 12.9, p<0.0001. A Duvoux score at presentation <2 was also a strong predictor of survival 1 year log rank=36.4, p<0.0001, 2 years log—rank 39.4 p<0.0001. The neutrophil to lymphocyte ration also predicted mortality at 2 years p<0.0001.

Conclusion Patients on surveillance were more likely to be offered curative therapy and be living one year from diagnosis. Survival benefit did not extend beyond this. The high proportion of patients with ECOG ≥1, suggests that many patients are inappropriately in surveillance programmes. This might explain the modest survival advantage. This could be used as a determinant to enrol patients into surveillance and should be revised on a 6-monthly basis.

Disclosure of Interest None Declared