Bilirubin is a superior biomarker in hepatocellular carcinoma diagnosis and differential diagnosis from benign liver disease

Abstract

We analyzed the difference in serum level of bilirubin between hepatocellular carcinoma (HCC) patients and benign liver disease (BLD) patients. We established a novel diagnostic model combining bilirubin, protein induced by vitamin K absence or antagonist-II (PIVKA-II), and alpha-fetoprotein (AFP) in diagnosing HCC. When the bilirubin level increased, the positive-predictive value (PPV) of bilirubin in the diagnosis of HCC decreased, while the negative predictive value (NPV) increased. The area under the receiver operating characteristic curves (AUCs) of the novel diagnostic model were 0.935 and 0.862 in the diagnosis of HCC and HCC < 3.0 cm (sensitivity 82.45% and 71.58%, specificity 95.77% and 93.45%), which was significantly higher than PIVKA-II, AFP, or both in the diagnosis of HCC and HCC < 3.0 cm (P < 0.001), and the AUC value of the novel diagnostic model further increased in the diagnosis of HCC and HCC < 3.0 cm with serum level of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL(0.965 vs. 0.935 and 0.910 vs. 0.862 ). In conclusion, bilirubin was a superior biomarker in the diagnosis of HCC. The combination of bilirubin, PIVKA-II, and AFP has superior diagnostic value for HCC and early-stage HCC.

Introduction

Hepatocellular carcinoma (HCC) is the third most common cause of cancer-related death, and the number of deaths due to HCC is about 700,000 every year globally. In China, HCC is the fourth most common cancer and the third most common cause of cancer death^1,2. Of new HCC cases worldwide, 50% occur in China annually ³. Chronic hepatitis B (CHB) is associated with an increased risk of HCC, and more than 50% of global HCC cases and 60–80% of HCC cases in Asia and Africa are related to CHB infection ^4,5.

The 5-year survival rate of patients with HCC is closely related to the stage at which HCC is diagnosed, with the rate exceeding 70% for early-stage HCC and being lower than 10% for advanced-stage HCC, and the median survival time is only 1‒2 years ^6–8. Because of the difficulties in diagnosing early-stage HCC, most cases are diagnosed at an advanced stage, which is the main reason for the poor prognosis of patients with HCC ^9,10.

Alpha-fetoprotein (AFP) and protein induced by vitamin K absence or antagonist-Ⅱ (PIVKA-Ⅱ) are widely used in HCC diagnosis ^11–13. These markers have been included in the guidelines for HCC diagnosis issued by many national hepatology societies ^14–16. However, in clinical application, about 50% of HCC cases and 60‒80% of early-stage HCC cases do not express AFP, which is the main reason for the poor sensitivity of AFP in the diagnosis of HCC overall as well as diagnosis of early-stage HCC ^17,18. Moreover, AFP is also expressed in cases with benign liver diseases (BLDs), such as cirrhosis and chronic hepatitis B (CHB) and C (CHC), which complicates the difficulty in the differential diagnosis of HCC vs. BLD ^19,20. As a newly discovered tumor marker in HCC diagnosis, PIVKA-Ⅱ has a higher diagnostic value than AFP, with a sensitivity for CHB-related early-stage HCC exceeding 50% and a specificity of about 90% ²¹. Li et al. analyzed the data of 49 articles, including 14,118 cases of HCC and 1,544 cases of early HCC. They demonstrated that the sensitivity of AFP and PIVKA-Ⅱ was only 59% and 63%. The specificity was only 86% and 91%, respectively, for HCC diagnosis, while the sensitivity was only 48% and 45%, and the specificity was only 89% and 95%, respectively, for early-stage HCC diagnosis ²².

Bilirubin is the final decomposition product of hemoglobin, a diagnostic marker of liver disease and hematological disease. Because of its antioxidant effect, bilirubin is widely believed to be a protective factor for inflammation, diabetes, cardiovascular disease, metabolic syndrome, and chronic liver disease. Mild unconjugated hyperbilirubinemia may play a role in preventing cardiovascular diseases and cancer ^23–25. Several studies have shown that the serum level of bilirubin in HCC patients is far lower than that in BLD patients ^26,27.

Accordingly, we hypothesized that bilirubin could be used as a biomarker in the differential diagnosis of HCC from BLD. We investigated whether a mathematical model incorporating PIVKA-Ⅱ, AFP, and bilirubin could improve the efficiency of HCC diagnosis and differential diagnosis of HCC from BLD.

Materials And Methods

Study setting and patients.  A total of 3,481 patients, including 718 HCC patients and 2,763 BLD patients, were retrospectively enrolled at the Affiliated Hospital of Northern Sichuan Medical College from May 2016 to November 2020. The diagnostic criteria of HCC were in accordance with the guidelines for the diagnosis and treatment of primary HCC issued by the Chinese Society of Clinical Oncology (2018.V1) ¹⁴. The 2,763 BLD patients that were enrolled included 986 patients with CHB, 3 patients with CHC, 541 patients with hepatitis B or C-related cirrhosis, 16 patients with alcoholic hepatitis, 90 patients with alcoholic cirrhosis, 80 patients with drug-induced hepatitis, 43 patients with drug-induced cirrhosis, 245 patients with calculous cholangitis, 411 patients with calculous cholecystitis, 39 patients with gallbladder polyps, 6 patients with liver tuberculosis, 67 patients with liver cyst, 38 patients with liver abscess, 2 patients with liver leiomyosarcoma, 135 patients with liver haemangioma, 14 patients with fatty liver, 34 cases of primary biliary cirrhosis, 8 patients with autoimmune hepatitis and 5 patients with primary biliary cholangitis. All BLD patients were followed up for at least 6 months, and none developed HCC during the follow-up period. None of the HCC or BLD patients received any clinical treatment related to HCC and BLD before collecting blood samples.

Detection of serum levels of PIVKA-Ⅱ, AFP, and bilirubin.   Serum levels of PIVKA-Ⅱ were detected by chemiluminescent microparticle immunoassay (Architect i1000, Abbott Laboratories, Chicago, IL, USA), and serum levels of AFP were detected by electrochemiluminescence immunoassay (Cobas E602, Roche, Inc., Mannheim, Germany). Serum levels of bilirubin were detected by biochemical rate assay (AU5800, Beckman Coulter, Inc., Brea, CA, USA).

Data processing mode of combined application of PIVKA-Ⅱ and AFP.  The cut-off value of PIVKA-Ⅱ and AFP in the diagnosis of HCC was determined using receiver operating characteristic (ROC) curves. The multiple serum levels of PIVKA-Ⅱ and AFP relative to their corresponding cut-off value were expressed by M_cut-off. This study evaluated the performance of PIVKA-Ⅱ combined with AFP (PA combination) in the diagnosis of HCC by analyzing the sum of M_cut-off of PIVKA-Ⅱ and AFP ^28,29.

Statistical analysis.  Data are expressed by median (interquartile range) or number (%). The mathematical model of PIVKA-Ⅱ² × AFP / Bilirubin³ was used as the combined diagnostic model, incorporating PIVKA-Ⅱ, AFP, and bilirubin (PAB combination), in the diagnosis of HCC. Pearson’s chi-square test was used to compare enumeration data between two groups or multiple groups. The Mann‒Whitney U test was used to compare measurement data between two groups. Pearson’s correlation analysis was used for two-factor correlation analysis. ROC curves were used to determine the cut-off values, the area under the ROC curves (AUC_S), sensitivity, and specificity of different diagnostic models in the diagnosis of HCC. Statistical analyses were performed using SPSS version 19.0 (IBM SPSS Inc., Armonk, NY, USA) and Medcalc version 12.3 (MedCalc Software bvba, Ostend, Belgium). The statistical significance of all tests was defined as P < 0.05, based on two-tailed tests.

Ethics approval and consent to participate.  The conduct of this study was approved by the Ethics Committee of Affiliated Hospital of North Sichuan Medical College (2022ER184-1). All methods were performed in accordance with the relevant guidelines and regulations. The ethics committee waived the need for informed consent due to the retrospective character. 

Results

Patient characteristics.  HCC patients and BLD patients were predominantly middle-aged or older and male. The median age of HCC patients was 58 (49‒67) years, which was significantly higher than that of BLD patients (52 [44‒62] years) (P < 0.001). The serum levels of PIVKA-Ⅱ, AFP, and the PAB combination in HCC patients were higher than those in BLD patients (P < 0.001), while the serum levels of bilirubin in HCC patients were lower than that in BLD patients (P < 0.001). The proportion of HCC patients with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL was higher than that in BLD patients (P < 0.001). The serum levels of PIVKA-II, AFP and PAB combination were positively correlated with tumour size (r = 0.456, 0.315, and 0.150; P < 0.001) (Fig. 1). The correlation between the PAB combination and tumor size was significantly weaker than that of PIVKA-Ⅱ and AFP (P < 0.001). The clinical characteristics of the 3,481 patients are shown in Table 1.

Expression of bilirubin in HCC patients and BLD patients.  Table 2 shows that the proportion of cases with serum levels of bilirubin ≥ 20.0 μmol/L, ≥ 50.0 μmol/L, ≥ 100.0 μmol/L, and ≥ 200.0 μmol/L cases in the BLD and HCC groups, and the corresponding positive predictive value (PPV) and negative predictive value (NPV). The proportions of cases with these different serum bilirubin levels were higher in the BLD group than in the HCC group (all P < 0.05). The proportions of cases in the BLD group with these different serum levels of bilirubin who also had serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL were significantly higher than those in the total BLD group for the corresponding serum bilirubin levels (all P < 0.001). The proportions of cases in the HCC group with serum levels of bilirubin ≥ 20.0 μmol/L, ≥ 100.0 μmol/L, and ≥ 200.0 μmol/L who also had serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL were not significantly different from that of the total HCC group for the corresponding serum bilirubin levels (all P > 0.05). However, the proportion of HCC cases with serum levels of bilirubin ≥ 50.00 μmol/L cases who also had serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL was different from that in the total HCC group for this level of serum bilirubin (P = 0.004). The proportions of BLD and HCC cases decreased with the increasing serum bilirubin levels (P < 0.001). As the serum bilirubin levels increased, the negative predictive value gradually increased, while the positive predictive value gradually decreased (both P < 0.001).

Serum levels of PIVKA-Ⅱ and AFP correlated positively with serum levels of bilirubin in BLD patients (r = 0.107, P < 0.001; r = 0.167, P < 0.001); their correlation coefficients were slightly higher than those in HCC patients (r = 0.084, P = 0.025; r = 0.120, P = 0.001) (Fig. 2), although these differences were not statistically significant (P > 0.05).

Performance of the PAB combination in the diagnosis of HCC.   BLD cases as the control group, the AUC of PA combination in the diagnosis of HCC was 0.883 (95%CI: 0.867‒0.899), which was significantly higher than that of AFP (0.815 [95%CI: 0.795‒0.834]) (P < 0.001), but not significantly different from that of PIVKA-Ⅱ (0.883 vs. 0.884) (P = 0.934) (Fig. 3A). Using the PAB combination in the diagnosis of HCC, the AUC increased to 0.935 (95%CI: 0.923‒0.947) (Fig. 3A), which was significantly higher than that of the PA combination (P < 0.001). The sensitivity, specificity, PPV, and NPV of the PAB combination for HCC diagnosis were 82.45%, 95.77%, 83.51%, and 95.45%, respectively, which were better than those of the PIVKA-Ⅱ, AFP, and PA combination (Table 3).

There were 906 cases of BLD and 646 cases of HCC with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL. BLD cases with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL as the control group, the AUCs of PIVKA-Ⅱ, AFP, and PA combination in the diagnosis of HCC with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL were 0.870 (95%CI: 0.851‒0.888), 0.738 (95%CI: 0.712‒0.764), and 0.879 (95%CI: 0.861‒0.897) (Fig. 3B), which were slightly decreased as compared to the values of HCC. The AUC of PAB combination in the diagnosis of HCC with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL was further increased to 0.965 (95%CI: 0.956‒0.973) (Fig. 3B), which was significantly higher than that of PA combination (P < 0.001). The sensitivity, specificity, PPV, and NPV of PAB combination in the diagnosis of HCC with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL were 86.38%, 94.92%, 92.38%, and 90.72%, respectively, which were better than those of PIVKA-Ⅱ alone, AFP alone, and PA combination (Table 3).

Performance value of PAB combination in the diagnosis of HCC< 3.0 cm.  There were 95 cases of HCC with tumour size < 3.0 cm (HCC < 3.0 cm). 2,763 BLD cases as control group, the AUC of PA combination in the diagnosis of HCC< 3.0 cm was 0.808 (95% confidence interval (CI): 0.763‒0.853), which was slightly higher than that of PIVKA-Ⅱ (0.756 [95%CI: 0.696‒0.817]) and of AFP (0.749 [95%CI: 0.697‒0.801]), but without statistically significant difference (P = 0.178 and P = 0.096) (Fig. 4A). The AUC of the PAB combination in the diagnosis of HCC< 3.0 cm was 0.862 (95%CI: 0.815‒0.910), which was slightly higher than that of the PA combination, but without statistically significant difference (P = 0.104) (Fig. 4A). The sensitivity, specificity, PPV, and NPV of PAB combination in the diagnosis of HCC< 3.0 cm were 71.58%, 93.45%, 27.31%, and 98.97%, respectively. Except that the sensitivity of the PAB combination was lower than that of the PA combination (81.05%), the other diagnostic performance parameter values were better for the PAB combination than for the PIVKA-Ⅱ, AFP, and PA combination (Table 4).

There were 80 cases of HCC < 3.0 cm with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL. 906 BLD cases with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL as control group, the AUCs of PIVKA-Ⅱ, AFP, and PA combination in the diagnosis of HCC< 3.0 cm were 0.695 (95%CI: 0.630‒0.759), 0.636 (95%CI: 0.570‒0.703), and 0.732 (95%CI: 0.673‒0.790) (Fig. 4B). The AUC of PAB combination in the diagnosis of HCC< 3.0 cm with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL was 0.910 (95%CI: 0.878‒0.943), which was higher than that of PA combination (P < 0.001) (Fig. 4B). The sensitivity, specificity, PPV, and NPV of PAB combination in the diagnosis of HCC< 3.0 cm with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL were 86.25%, 85.10%, 33.82%, and 98.59% respectively. Except that the specificity of the PAB combination was lower than that of AFP (91.06%), the other diagnostic performance parameter values of the PAB combination were better than those of the PIVKA-Ⅱ, AFP, and PA combination (Table 4).

Discussion

This study established a combined mathematical diagnostic model incorporating levels of PIVKA-II and AFP combined with bilirubin levels for HCC diagnosis. ROC curve analysis showed that the performance of PAB combination in the diagnosis of HCC and early-stage HCC was better than that of PIVKA-II, AFP, or PA combination, with sensitivity exceeding 70% and specificity exceeding 90%. Moreover, the performance of PAB combination in the diagnosis of HCC and early-stage HCC with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL was further improved. Those results indicated that bilirubin was a superior biomarker in diagnosing HCC and its differential diagnosis from BLD. Bilirubin combined with PIVKA-II and AFP in the diagnosis of HCC can improve the probability of detecting HCC at an early stage and reduce the occurrence of missed diagnosis and misdiagnosis to improve the survival time and quality of life of HCC patients.

Bilirubin is considered a protective factor for BLD and plays a role in preventing tumorigenesis ^23-25. The data in this study showed that the serum levels of bilirubin in BLD cases were significantly higher than those in HCC cases. The proportion of elevated serum levels of bilirubin in BLD cases was significantly higher than that in HCC cases, which was consistent with existing research reports ^26,27. Interestingly, the correlations between serum levels of PIVKA-Ⅱ, AFP, and bilirubin in BLD cases were slightly stronger than that in HCC cases. At the same time, among BLD and HCC patients with serum levels of PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL, the proportions of BLD cases with high levels of bilirubin were significantly higher than that of HCC cases. It is unclear whether this reflects the initial liver compensation mechanism in BLD patients and a hepato-protective measure, which is worth further study. PIVKA-Ⅱ and AFP are highly expressed in some BLD cases, which is essential for the difficulty in the differential diagnosis of HCC from BLD ^19,20,30. As shown in Table 2, with the increase in serum bilirubin levels, the PPV of bilirubin in the diagnosis of HCC decreased gradually, and the NPV increased gradually, indicating that bilirubin can be used as a biomarker in the differential diagnosis of HCC from BLD: a higher serum bilirubin level indicates a lower probability of HCC and a higher probability of BLD. However, because of the poor performance of bilirubin in the diagnosis of HCC, based on the characteristics of the difference in bilirubin serum levels in BLD patients and HCC patients, we established a mathematical model in which bilirubin is combined with PIVKA-Ⅱ and AFP for HCC diagnosis and differential diagnosis of HCC from BLD. The sensitivity and specificity of the PAB combination in the diagnosis of HCC and HCC < 3.0 cm were significantly improved, indicating that the PAB combination could facilitate HCC diagnosis and differential diagnosis of HCC from BLD. As shown in Figure 1, the correlation between PAB combination and tumor size was markedly weaker than that of PIVKA-Ⅱ and AFP, indicating that PAB combination was more suitable than PIVKA-Ⅱ and AFP in the diagnosis of early-stage HCC.

PIVKA-II ≥ 40 mAU/mL and AFP ≥ 20 ng/mL are often used as cut-off values in HCC diagnosis ^31-33. Generally, before the appearance of clinical symptoms, abnormal PIVKA-II or AFP levels should attract the attention of clinicians and patients. Moreover, HCC and HCC < 3 cm cases with such PIVKA-Ⅱ and/or AFP levels account for more than 80% of all HCC and of all HCC < 3.0 cm cases, respectively. Therefore, we evaluated the performance of the PAB combination in diagnosing HCC and HCC < 3.0 cm cases in those with serum PIVKA-II ≥ 40 mAU/mL and/or AFP ≥ 20 ng/mL. In such HCC and HCC < 3.0 cm cases, the performance parameter values of PIVKA-Ⅱ, AFP, and PA combination decreased, while the corresponding values of PAB combination improved. The results indicated that PIVKA-Ⅱ, AFP, and PA combinations were less effective in differentiating between BLD and HCC, particularly early-stage HCC, in which the serum levels of PIVKA-Ⅱ and AFP are far less than those in the middle and late stages of HCC ²⁸, which complicates the differential diagnosis. PAB combination will thus be more suitable in diagnosing HCC with abnormal PIVKA-Ⅱ and AFP results.

The study had some limitations. The PPV of PAB combination for the diagnosis of HCC < 3.0 cm was only about 30%, which may be related to the small number of HCC < 3.0 cm cases (accounting for only 13.23% of all HCC cases in the same period). 

Given the difficulty of diagnosing HCC early, we derived a model by which HCC can be diagnosed earlier and distinguished from BLD. Follow-up monitoring of PIVKA-Ⅱ, AFP, and bilirubin should be strengthened in high-risk populations, such as cases with CHB infection or liver cirrhosis ^4,5, to improve diagnosis and treatment of HCC at an earlier stage, and thereby improve the survival rate, prognosis, and quality of life of HCC patients.

Declarations

Availability of Data and Materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Acknowledgments

This work was supported by the Science and Technology Project of Nanchong (20SXQT0337), and the Nanchong City’s S&T strategic cooperation projects between the city and the university (19SXHZ0274).

Author Contributions Statement

L.Y. Data curation, Writing-original draft. L.X.L. Data curation, Writing-original draft. J.Y. Statistical analysis, Figures, Tables. T.W. Conceptualization, Methodology, Project administration. W.Q. Conceptualization, Methodology, Project administration. All authors reviewed the manuscript.

Competing interests

The authors declare no competing interests.

References

1. Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5): E359-E386. 
2. Chen W, Zheng R, Baade PD, et al. Cancer statistics in China, 2015. CA Cancer J Clin. 2016;66(2):115-132. 
3. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. CA Cancer J Clin. 2015;65(2):87-108. 
4. Lavanchy D. Hepatitis B virus epidemiology, disease burden, treatment, and current and emerging prevention and control measures. J Viral Hepat. 2004;11(2):97-107.
5. de Martel C, Maucort-Boulch D, Plummer M, Franceschi S. World-wide relative contribution of hepatitis B and C viruses in hepatocellular carcinoma. Hepatology. 2015;62(4):1190-1200.
6. El-Serag HB. Hepatocellular carcinoma. N Engl J Med. 365(12): 1118-1127. 
7. Everhart JE, Ruhl CE. Burden of digestive diseases in the United States Part III: Liver, biliary tract, and pancreas. Gastroenterology. 2009;136(4):1134-1144. 
8. Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med. 2008;359(4):378-390. 
9. Llovet JM, Burroughs A, Bruix J. Hepatocellular carcinoma. Lancet. 2003;362(9399):1907-1917. 
10. Bruix J, Reig M, Sherman M. Evidence-Based Diagnosis, Staging, and Treatment of Patients With Hepatocellular Carcinoma. Gastroenterology. 2016;150(4):835-853. 
11. Yu R, Tan Z, Xiang X, Dan Y, Deng G. Effectiveness of PIVKA-II in the detection of hepatocellular carcinoma based on real-world clinical data. BMC Cancer. 2017;17(1):608. 
12. Grandhi MS, Kim AK, Ronnekleiv-Kelly SM, Kamel IR, Ghasebeh MA, Pawlik TM. Hepatocellular carcinoma: From diagnosis to treatment. Surg Oncol. 2016;25(2):74-85. 
13. Kim SH, Moon DB, Kim WJ, et al. Preoperative prognostic values of α-fetoprotein (AFP) and protein induced by vitamin K absence or antagonist-II (PIVKA-II) in patients with hepatocellular carcinoma for living donor liver transplantation. Hepatobiliary Surg Nutr. 2016;5(6):461-469.
14. Chinese Society of Clinical Oncology, Guidelines of Chinese Society of Clinical Oncology (CSCO) Hepatocellular Carcinoma (in Chinese). Beijing: People's Medical Publishing House,2018.
15. Kudo M, Izumi N, Kokudo N, et al. Management of hepatocellular carcinoma in Japan: Consensus-Based Clinical Practice Guidelines proposed by the Japan Society of Hepatology (JSH) 2010 updated version. Dig Dis. 2011;29(3):339-364.
16. Izumi N. Diagnostic and treatment algorithm of the Japanese society of hepatology: a consensus-based practice guideline. Oncology. 2010;78 Suppl 1:78-86. 
17. Ertle JM, Heider D, Wichert M, et al. A combination of α-fetoprotein and des-γ-carboxy prothrombin is superior in detection of hepatocellular carcinoma. Digestion. 2013;87(2):121-131. 
18. Reichl P, Mikulits W. Accuracy of novel diagnostic biomarkers for hepatocellular carcinoma: An update for clinicians (Review). Oncol Rep. 2016;36(2):613-625.
19. Park SJ, Jang JY, Jeong SW, et al. Usefulness of AFP, AFP-L3, and PIVKA-II, and their combinations in diagnosing hepatocellular carcinoma. Medicine (Baltimore). 2017;96(11): e5811. 
20. Forner A, Bruix J. Biomarkers for early diagnosis of hepatocellular carcinoma. Lancet Oncol. 2012;13(8):750-751. 
21. Gao J, Song P. Combination of triple biomarkers AFP, AFP-L3, and PIVAKII for early detection of hepatocellular carcinoma in China: Expectation. Drug Discov Ther. 2017;11(3):168-169. 
22. Li C, Zhang Z, Zhang P, Liu J. Diagnostic accuracy of des-gamma-carboxy prothrombin versus α-fetoprotein for hepatocellular carcinoma: A systematic review. Hepatol Res. 2014;44(10): E11-E25.
23. Fevery J. Bilirubin in clinical practice: a review. Liver Int. 2008;28(5):592-605. 
24. Seyed Khoei N, Jenab M, Murphy N, et al. Circulating bilirubin levels and risk of colorectal cancer: serological and Mendelian randomization analyses. BMC Med. 2020;18(1):229. 
25. Weaver L, Hamoud AR, Stec DE, Hinds TD Jr. Biliverdin reductase and bilirubin in hepatic disease. Am J Physiol Gastrointest Liver Physiol. 314 (6) (2018) G668-G676.
26. Wang T, Zhang KH, Hu PP, et al. Combination of dual serum fluorescence, AFP and hepatic function tests is valuable to identify HCC in AFP-elevated liver diseases. Oncotarget. 2017;8(58):97758-97768. 
27. Park SJ, Jang JY, Jeong SW, et al. Usefulness of AFP, AFP-L3, and PIVKA-II, and their combinations in diagnosing hepatocellular carcinoma. Medicine (Baltimore). 2017;96(11): e5811. 
28. Wang Q, Chen Q, Zhang X, et al. Diagnostic value of gamma-glutamyltransferase/aspartate aminotransferase ratio, protein induced by vitamin K absence or antagonist II, and alpha-fetoprotein in hepatitis B virus-related hepatocellular carcinoma. World J Gastroenterol. 2019;25(36):5515-5529.
29. Wang G, Lu X, Du Q, et al. Diagnostic value of the γ-glutamyltransferase and alanine transaminase ratio, alpha-fetoprotein, and protein induced by vitamin K absence or antagonist II in hepatitis B virus-related hepatocellular carcinoma. Sci Rep. 2020;10(1):13519. 
30. Feng H, Li B, Li Z, Wei Q, Ren L. PIVKA-II serves as a potential biomarker that complements AFP for the diagnosis of hepatocellular carcinoma. BMC Cancer. 2021;21(1):401. Published 2021 Apr 13. 
31. Park H, Park JY. Clinical significance of AFP and PIVKA-II responses for monitoring treatment outcomes and predicting prognosis in patients with hepatocellular carcinoma. Biomed Res Int. 2013;2013: 310427. 
32. Cui R, Wang B, Ding H, Shen H, Li Y, Chen X. Usefulness of determining a protein induced by vitamin K absence in detection of hepatocellular carcinoma. Chin Med J (Engl). 2002;115(1):42-45.
33. Kim KH, Kim JY, Yoo JS. Mass spectrometry analysis of glycoprotein biomarkers in human blood of hepatocellular carcinoma. Expert Rev Proteomics. 2019;16(7):553-568. 

Tables

Table 1. Clinical characteristics of the 3,481 patients. Data are expressed as median (interquartile range) or number (%); NA: Not applicable. PIVKA-II, protein induced by vitamin K absence or antagonist-II; AFP, alpha-fetoprotein; PAB combination, PIVKA-II and AFP combined with bilirubin; HCC, hepatocellular carcinoma; BLD, benign liver disease. 

Table 2. Expression characteristics of bilirubin in HCC patients and BLD patients. Data are expressed as numbers (%). PIVKA-II, protein induced by vitamin K absence or antagonist-II; AFP, alpha-fetoprotein; HCC, hepatocellular carcinoma; BLD, benign liver disease; PPV, positive predictive value; NPV, negative predictive value.

Table 3. Performance value of different diagnostic models in the diagnosis of HCC. PIVKA-II, protein induced by vitamin K absence or antagonist-II; AFP, alpha-fetoprotein; PA combination, PIVKA-II combined with AFP; PAB combination, PIVKA-II and AFP combined with bilirubin; AUC, the area under the receiver operating characteristic curve; CI, confidence interval; HCC, hepatocellular carcinoma; PPV, positive predictive value; NPV, negative predictive value.

Table 4. Performance value of different diagnostic models in the diagnosis of HCC< 3.0 cm. PIVKA-II, protein induced by vitamin K absence or antagonist-II; AFP, alpha-fetoprotein; PA combination, PIVKA-II combined with AFP; PAB combination, PIVKA-II and AFP combined with Bilirubin; AUC, the area under the receiver operating characteristic curve; CI, confidence interval; HCC, hepatocellular carcinoma; PPV, positive predictive value; NPV, negative predictive value.