Diagnostic Performance of AFP, AFP-L3, or PIVKA-II for Hepatitis C Virus-Associated Hepatocellular Carcinoma: A Multicenter Analysis

Abstract

Background and Aim: Alpha-fetoprotein (AFP), a lens culinaris agglutinin-reactive fraction of AFP (AFP-L3), is a protein that is induced by vitamin K deficiency or antagonist-II (PIVKA-II) that has been clinically used as a serum biomarker for early detection and diagnosis of hepatocellular carcinoma (HCC). Diagnostic performance of each serum biomarker alone, or their combinations for the detection of hepatitis C virus (HCV)-associated HCC were compared. Methods: Serum AFP, AFP-L3, and PIVKA-II levels were evaluated in patients with HCV-associated HCC, and those with chronic HCV infection without HCC (HCV-controls). The areas under the curve (AUC), sensitivity, and specificity were compared to identify the diagnostic performance of each serum HCC biomarker alone or in combination. Results: Overall, 172 HCV controls and 105 patients with HCV-associated HCC were enrolled. The AFP, AFP-L3, and PIVKA-II levels were significantly increased among patients with HCV-associated HCC when compared with HCV patients without HCC (p < 0.001). When these biomarkers were analyzed individually, PIVKA-II revealed the best predictive performance (AUC: PIVKA-II 0.90 vs. AFP 0.80 vs. AFP-L3 0.69, p < 0.001). In evaluating the combinations of any two biomarkers, the best predictive performance was found in PIVKA-II + AFP (0.93 vs. AFP + AFP-L3 0.78, p = 0.001; and PIVKA-II + AFP-L3 0.89, p < 0.001), which had no difference compared to the predictive performance of the combination of all three serum biomarkers (AFP + AFP-L3 + PIVKA-II 0.93, p = 0.277). Similar results were identified in the subgroups of patients with HCV-induced cirrhosis, and among patients with early-stage HCC defined by BCLC and TNM staging. Conclusions: The addition of the PIVKA-II test to routine AFP test maybe provide a more suitable biomarker approach to detect HCV-induced HCC in patients with HCV infection undergoing HCC surveillance.

Diagnostic accuracy of serum protein induced by vitamin K absence (PIVKA‐II), serum a‐fetoprotein and their combination for hepatocellular carcinoma among Caucasian cirrhotic patients with diagnostic or non‐diagnostic serum a‐fetoprotein levels

“…Studies suggest that AFP and PIVKA-II are complementary biomarkers as their production occurs through different pathways. [15][16][17][18] In patients with very early/early HCC the diagnostic accuracy of each biomarker was almost equal (AUROC 65.8% for AFP and 65% for PIVKA-II) and did not improve with the combination of them (AUROC 64.2%) whereas there was an additive, complementary effect of PIVKA-II to AFP in the HCC diagnosis in intermediate and advanced stages, according to the results of our study. It seems that the combination of the two biomarkers is not the most useful tool available to predict the presence of very early or early HCC in advanced cirrhotic patients, irrespective of the imaging results.…”

“…PIVKA‐II revealed the best predictive performance compared to AFP and AFP‐L3, in studies which evaluated HCC patients and patients with chronic hepatitis C 15 or patients with benign liver lesions, liver metastases, and other gastrointestinal malignancies 16 . It is important to note, that in these studies the control group was either non‐cirrhotic patients 16 or mainly patients with compensated Child‐Pugh A cirrhosis as presented in the study by Liu S et al., 15 in contrast to our study in which approximately half of the study population exhibited advanced decompensated liver cirrhosis. Non‐cirrhotic control groups may overestimate the diagnostic accuracy of each individual biomarker as well as the combination of them in the detection of HCC, as it has been observed that liver cirrhosis significantly impacts serum PIVKA‐II and AFP levels 13,14 …”

“…Moreover, the combination of biomarkers seems to improve the diagnostic accuracy for HCC detection in the intermediate (BCLC‐B) and advanced (BCLC‐C) stage, but not in very early/early stages (BCLC‐0/A), a finding that could possibly suggest the limited role of the two evaluated biomarkers alone or in combination in the prediction of HCC among cirrhotic patients and their usefulness in surveillance programs without concomitant imaging. PIVKA‐II revealed the best predictive performance compared to AFP and AFP‐L3, in studies which evaluated HCC patients and patients with chronic hepatitis C 15 or patients with benign liver lesions, liver metastases, and other gastrointestinal malignancies 16 . It is important to note, that in these studies the control group was either non‐cirrhotic patients 16 or mainly patients with compensated Child‐Pugh A cirrhosis as presented in the study by Liu S et al., 15 in contrast to our study in which approximately half of the study population exhibited advanced decompensated liver cirrhosis.…”

“…13,14 Moreover, the addition of serum PIVKA-II levels to routine AFP test might provide a more suitable biomarker approach to detect HCV-induced HCC in patients with chronic HCV infection. 15 The clinical value of the combination of the two biomarkers has already been studied in small regional studies and seems to be superior compared to each test alone for the diagnosis of HCC, according to a recent study from China. 16 On the other hand, data from prospective Phase III studies concerning the utility of serum PIVKA-II levels in the diagnosis of HCC among Caucasian patients with liver cirrhosis of any etiology, which is mainly the target population of HCC surveillance, are scarce 17,18 as well as information regarding the diagnostic accuracy of each biomarker separately and their combination in HCC patients according to the stage of the malignant disease.…”

“…Serum PIVKA‐II levels in HCC patients are significantly elevated compared to healthy controls, which suggests that PIVKA‐II could be a useful complement to AFP in HCC diagnosis 13,14 . Moreover, the addition of serum PIVKA‐II levels to routine AFP test might provide a more suitable biomarker approach to detect HCV‐induced HCC in patients with chronic HCV infection 15 . The clinical value of the combination of the two biomarkers has already been studied in small regional studies and seems to be superior compared to each test alone for the diagnosis of HCC, according to a recent study from China 16 .…”