HCC Surveillance Is Associated with Potential Harms
MetadataShow full item record
BACKGROUND: Hepatocellular carcinoma (HCC) is the 3rd leading cause of cancer death worldwide and leading cause of death in patients with cirrhosis. HCC surveillance is recommended in patients with cirrhosis to improve early detection rates. A comprehensive assessment of HCC surveillance should weigh both benefits and harms; however, no study to date has assessed potential harms. Although ultrasound and alpha fetoprotein (AFP) have minimal direct harms, there are potential downstream harms from follow-up tests that should be considered. Objective: To quantify and characterize potential harms of HCC surveillance among a large cohort of patients with cirrhosis METHODS: We conducted a retrospective cohort study among patients with cirrhosis followed at a large safety-net health system. We recorded all surveillance abdominal imaging and/or alpha fetoprotein (AFP) testing between January 2010 and December 2013. We defined a false positive surveillance test as a suspicious liver mass on ultrasound or AFP >20 ng/mL, without HCC development during follow-up. We recorded CT or MRI scans, biopsies, or any procedures performed as a direct result of surveillance testing. Predictors of harm were identified using logistic regression, with significance being defined as p<0.05. RESULTS: We identified 571 patients with cirrhosis, with median follow-up of 2.8 years. HCC surveillance was performed at least once in 551 (96%) patients. Surveillance testing led to diagnostic CT or MRI testing in 123 (21.5%) patients - 74 with one CT/MRI and 49 with multiple studies. Rates of unnecessary diagnostic testing increased from 15% if followed for ≤1 year to 25% if followed for 1-2 years to 37% if followed for ≥2 years. An additional two patients had a biopsy and one patient angiogram for false positive surveillance tests. Follow-up tests were performed due to false positive ultrasound in 47 cases, false positive AFP in 35 cases, and indeterminate ultrasound results in 41 cases. In multivariate analysis, surveillance harm was associated with viral liver disease (OR 1.60, 95%CI 1.04-2.46), receipt of hepatology subspecialty care (OR 2.32, 95%CI 1.52-3.59), and coverage by Parkland Health Plus (OR 2.21, 95%CI 1.45- 3.40). CONCLUSION: This study is the first to demonstrate HCC surveillance can be associated with potential harms. One in five patients have at least one unnecessary diagnostic test, and nearly 10% have multiple tests. Better HCC surveillance tools, with a higher positive predictive value, are urgently needed.