Introduction Hepatocellular carcinoma (HCC) causes approximately 1500 deaths per year in the UK and 95% of these patients have known established cirrhosis. Surveillance programme for patients with cirrhosis using six monthly α-fetoprotein (AFP) monitoring and ultrasound scanning (USS) is therefore recommended to ensure early identification of HCC at a stage when curative treatment is still possible. HCC identified by surveillance rather than incidental or symptomatic diagnosis results in better outcome and increased survival. Here we show the results of an audit of HCC surveillance at the Liver Transplantation Unit in Queen Elizabeth Hospital Birmingham.

Methods We identified consecutive 271 patients who had undergone HCC surveillance between January 2007 and January 2009. We divided the patients into high-risk (hepatitis B, hepatitis C, men with alcoholic liver disease or PBC and haemochromatosis) and low risk groups. We collected their demographic data, surveillance results, treatment details and outcome.

Results Mean duration of follow-up was 22 months (1–32). 156/272 (57%) patients belonged to high-risk group. Ten patients (4%) ceased surveillance as they underwent liver transplantation for reasons other than HCC. Eight patients (3%) died during surveillance, unrelated to HCC. One patient was found to have pancreatic cancer on USS done for HCC surveillance.

18/272 (6%) were found to have HCC. Of these 13 (72%) were in the high-risk group. 13 patients (72%) had single tumour whereas remaining five had multiple tumours. Two patients (11%) were detected by abnormal AFP alone with further imaging confirming the diagnosis. Both of these were unsuitable for curative treatment. Eight (44%) were identified on surveillance USS, confirmed with further imaging. Fifty-six percent of patients with HCC underwent attempted curative treatment (liver transplantation 28%, resection 6% and radiofrequency ablation 22%); 23% received controlling treatment (TACE 6%, ethanol injection 17%) and 22% were referred for palliation.

Conclusion Our results show that higher proportion (72%) of HCC were detected in high-risk group. 56% of these patients underwent treatment with curative intent. There was a significant burden of HCC in the group conventionally classified as low risk group for HCC (18%). We show that a greater proportion of HCC detected by USS were suitable for attempted curative treatment as compared to those detected by AFP monitoring. 22% of total HCC cases were not suitable for treatment other than palliation despite being on recommended surveillance. With this study we recommend that present surveillance guidelines should be followed in all patients with cirrhosis, however, there may be scope for improving early detection of HCC developing in cirrhotic patients resulting in improvement in outcome, by modification of surveillance protocol in selected patients and also by better definition of high-risk group.