Alpha-fetoprotein in hepatocellular carcinoma surveillance: Wake not the dead


  • Potential conflict of interest: Nothing to report.

To the Editor:

We read with interest the letter by Marrero and El-Serag that calls for the inclusion of alpha-fetoprotein (AFP) in the American Association for the Study of Liver Diseases (AASLD) updated guidelines for the management of hepatocellular carcinoma (HCC).1, 2 However, we disagree with their conclusions and feel that the AASLD recommendation to perform HCC surveillance with ultrasonography (US) alone is supported by solid evidence.1, 2

The evidence supporting surveillance programs for HCC with liver US with or without AFP testing stems from the results of a randomized controlled trial and from cohort studies showing that surveillance improves both detection rate of early HCCs and patient survival.3-5 However, it is clear that the authors of the AASLD guidelines took into account the numerous limitations of AFP testing, and therefore it is no surprise that they did not include this serological marker in their HCC surveillance recommendations.2 In fact, although we may agree with Marrero and El-Serag that the Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis (HALT-C) trial is a suboptimal setting to assess the role of AFP for the early detection of HCC, this study had the precious gifts of providing prospectively collected data and to include a large population of patients who were mainly at risk of developing HCC.6 Furthermore, data were available both at HCC diagnosis and 1 year before, thus being as close as possible to everyday clinical practice and therefore providing the best evidence currently available.2, 6 In this study, the sensitivity of AFP at a cutoff of 20 ng/mL was low (i.e., 61%) at the time of HCC diagnosis, yet at 22% it was unacceptably low 12 months before, when HCC was likely present in the majority of patients.6 These results are strikingly similar to those obtained in a case-control study carried out in a completely different population, where a 20 ng/mL AFP cutoff showed 60% sensitivity for the diagnosis of HCC,7 although sensitivity was unacceptably low (i.e., approximately 50%) for single nodules and/or lesions smaller than 3 cm.8

We feel that the extensors of the updated AASLD guidelines did not ignore the “highest level of evidence for the efficacy of US combined with AFP in research studies”2 as affirmed by Marrero and El-Serag,1 but evaluated both efficacy and cost-effectiveness. Indeed, the combination of AFP and US leads to a mere 6%-8% increase in sensitivity for the detection of early HCC as compared to US alone, with a doubling in the rate of false-positives and at an unbearable increase (by 84%) in surveillance-related costs.9, 10 Therefore, AFP provides no additional benefit to US, as recently concluded even in the meta-analysis by the Marrero group,10 with a significant worsening of the cost-effectiveness of surveillance.9 To conclude, we feel that the use of AFP as a surveillance test for HCC should be regarded as a memory, and any effort to increase the awareness and application of the currently proposed surveillance guidelines among physicians in clinical practice should be embraced.

Edoardo G. Giannini M.D., Ph.D.*, Fabio Farinati M.D.†, Franco Trevisani M.D.‡, * Department of Internal Medicine, Gastroenterology Unit, University of Genova, Genova, Italy, † Department of Surgical and Gastroenterological Science, Gastroenterology Unit, University of Padova, Padova, Italy, ‡ Department of Clinical Medicine, Unità di Semeiotica Medica, Alma Mater Studiorum–University of Bologna, Bologna, Italy.