Tumor-associated autoantibodies in combination with alpha-fetoprotein for detection of early stage hepatocellular carcinoma

• Published in: PloS one 2020-05, Vol.15 (5), p.e0232247-e0232247
• Main Authors: Welberry, Christopher, Macdonald, Isabel, McElveen, Jane, Parsy-Kowalska, Celine, Allen, Jared, Healey, Graham, Irving, William, Murray, Andrea, Chapman, Caroline
• Format: Article
• Language: English
• Subjects: Alpha fetoproteins; Antibodies; Antigen (tumor-associated); Antigens; Autoantibodies; Autoimmunity; Biological markers; Biology and Life Sciences; Biomarkers; Biomedical research; Cancer; Cancer staging; Carcinoma; Diagnosis; Disease control; Diseases; Enzyme-linked immunosorbent assay; Enzymes; Ethics; Glycoproteins; Health risks; Health screening; Health surveillance; Hepatitis; Hepatocellular carcinoma; Liver; Liver cancer; Liver diseases; Medical research; Medical screening; Medical tests; Medicine and Health Sciences; Methods; Mortality; Oncology; Patients; Proteins; Reclassification; Research and Analysis Methods; Studies; Testing; Tumors; α-Fetoprotein
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0232247

Hepatocellular carcinoma (HCC) continues to be a leading challenge in modern oncology. Early detection via blood-based screening tests has the potential to cause a stage-shift at diagnosis and improve clinical outcomes. Tumor associated autoantibodies (TA-AAbs) have previously shown the ability to distinguish HCC from patients with high-risk liver disease. This research aimed to further show the utility of TA-AAbs as biomarkers of HCC and assess their use in combination with Alpha-fetoprotein (AFP) for detection of HCC across multiple tumor stages. Levels of circulating G class antibodies to 44 recombinant tumor associated antigens and circulating AFP were measured in the serum of patients with HCC, non-cancerous chronic liver disease (NCCLD) and healthy controls via enzyme-linked immunosorbent assay (ELISA). TA-AAb cut-offs were set at the highest Youden's J statistic at a specificity ≥95.00%. Panels of TA-AAbs were formed using net reclassification improvement. AFP was assessed at a cut-off of 200 ng/ml. Sensitivities ranged from 1.01% to 12.24% at specificities of 95.96% to 100.00% for single TA-AAbs. An ELISA test measuring a panel of 10 of these TA-AAbs achieved a combined sensitivity of 36.73% at a specificity of 89.89% when distinguishing HCC from NCCLD controls. At a cut-off of 200 ng/ml, AFP achieved a sensitivity of 31.63% at a specificity of 100.00% in the same cohort. Combination of the TA-AAb panel with AFP significantly increased the sensitivity for stage one (40.00%) and two (55.00%) HCC over the TA-AAb panel or AFP alone. A panel of TA-AAbs in combination with AFP could be clinically relevant as a replacement for measuring levels of AFP alone in surveillance and diagnosis strategies. The increased early stage sensitivity could lead to a stage shift with positive prognostic outcomes.