Mac-2 binding protein glycosylation isomer as a potential biomarker of hepatocellular carcinoma in HCV-cured patients

Abstract

Objective

Mac-2 binding protein glycosylation isomer (M2BPGi) is produced in the extracellular matrix and serves as an indicator of hepatic stellate cell activation. Assessing M2BPGi levels could aid in predicting hepatocellular carcinoma (HCC) in individuals with hepatitis C virus (HCV). The objective of this study was to evaluate the usefulness of M2BPGi as a biomarker for HCC in HCV patients and its association with disease severity and progression.

Methods

This study included patients who were cured of chronic hepatitis C virus. The patients were divided into three subgroups: HCV without cirrhosis, HCV with cirrhosis, and HCV with HCC. These subgroups were then compared to a subgroup of healthy volunteers. In addition to routine laboratory investigations, M2BPGi levels were measured in all the enrolled subjects.

Results

The level of serum M2BPGi was significantly greater in the HCV with cirrhosis and HCC groups than in the control group (P < 0.001). Additionally, it was significantly greater in multifocal HCC than in those with unifocal HCC (P < 0.001), and it was directly proportional to the size of the focal lesion of HCC (P = 0.001). The cutoff for serum M2BPGi in diagnosing HCC was ≥ 0.869 (C.O.I), with an AUC of 0.762, a sensitivity of 78.6%, and a specificity of 61.9% (P = 0.004). Furthermore, the cutoff for predicting multifocality was > 0.93 (C.O.I), with an AUC of 0.73, sensitivity of 66.7%, and specificity of 63.8% (P = 0.03). Although the AFP level was still superior in predicting cirrhosis and HCC, the M2BPGi level was better at predicting the size and diagnostic value of HCC when the AFP level was normal. The cutoff for M2BPGi in this case was 0.903(C.O.I), with a sensitivity of 80%, specificity of 75%, and an accuracy of 76.25%. M2BPGi was independently associated with the CRP level (β = 0.484, P = 0.001) and the size of the HCC focal lesion (β = 1.422, P = 0.001).

Conclusion

M2BPGi can be used as an effective marker to assess the biological behavior and aggressiveness of HCC. Further studies are warranted on a large scale of patients to confirm our findings.

Introduction

Viral hepatitis caused by the hepatitis C virus (HCV) may pose a considerable public health risk because of the devastating complications it can induce, most notably liver cirrhosis and hepatocellular carcinoma (HCC). Every year, approximately 71 million people worldwide develop HCV infection, resulting in nearly 700,000 deaths. By 2025, one million people are expected to be diagnosed with HCC each year. HCC is the fifth most common type of cancer and the second leading cause of cancer-related death [1].

HCC is caused by the amplification of the HCV protein in infected liver cells, which results in their mutation and malignant transformation. The primary cause of malignant transformation is assumed to be repetitive inflammation, damage, and regeneration. Other risk factors for HCC with HCV include male sex, smoking, obesity, diabetes, and HBV or human immunodeficiency virus co-infection [2].

Mac-2 binding protein (M2BP) is a glycoprotein that is produced in the extracellular matrix. It is also known as Wisteria floribunda agglutinin positive Mac-2 binding protein (WFA + -M2BP). The sugar chain structure of M2BP evolves in step with the worsening of hepatic fibrosis. It is extensively glycosylated and interacts with galectin-3 which controls fundamental cellular processes such as growth, proliferation, differentiation, and inflammation as well as cell–matrix interactions and other extracellular proteins such as collagens IV-, V-, and VI and fibronectin. It triggers the production of interleukin-1, 6, and other cytokines, and its source in liver tissues is hepatic stellate cells (HSCs) thus it reflects the activation of HSCs during the progression of liver fibrosis [3, 4].

The Mac-2 binding protein glycosylation isomer is being studied to potentially detect fibrosis progression in various liver diseases such as primary biliary cirrhosis, nonalcoholic fatty liver disease, and chronic hepatitis B virus infection. It has also shown promise as a biomarker for HCC under these conditions [5, 6]. Recent cross-sectional research has revealed a correlation between the serum level of M2BPGi and the progression of liver fibrosis during chronic HCV infection and hepatocarcinogenesis [7].

The current study was designed to assess M2BPGi levels as a potential biomarker for HCC in a cohort of HCV patients and to correlate these levels with the severity of disease progression.

Patients and methods

Patients with chronic hepatitis C virus were enrolled and assessed in this case–control study which was conducted between January 2023 and June 2023 at the Internal Medicine Department, Faculty of Medicine, Zagazig University.

The study included both males and females over the age of 18 who had been diagnosed with chronic HCV. All participants achieved a sustained virological response after treatment with direct-acting antiviral agents, primarily sofosbuvir combined with daclatasvir and ribavirin.

The exclusion criteria for the study were pregnancy; acute or chronic renal failure; congestive heart failure; thyroid dysfunction; acute infections; mixed or additional causes of liver disease; excessive alcohol consumption; the presence of hepatitis A, hepatitis B, cytomegalovirus or HIV; autoimmune or drug-induced liver disease; and the use of immunosuppressive medications.

Normal healthy volunteers (group A, n = 14, Fibroscan 3.8 ± 1.12 kPa) were compared with a cohort of cured HCV patients. The patients were stratified into group B consisting of HCV patients without cirrhosis (n = 14, 9.14 ± 1.17 kPa); group C consisting of HCV patients with cirrhosis (n = 14, 15.6 ± 1.8 kPa), and group D consisting of HCV patients with HCC (n = 14, 20.9 ± 3.1 kPa). All the subjects underwent a thorough history and comprehensive clinical examination.

Routine laboratory investigations included a complete blood count (hemoglobin, white blood cells (WBCs), and platelets), kidney function tests (blood urea nitrogen (BUN) and creatinine), liver function tests (alanine aminotransferase (ALT), aspartate aminotransferase (AST), AST/ALT ratio, total and direct bilirubin and albumin, alkaline phosphatase), prothrombin time (PT), international normalized ratio (INR), serum sodium, C-reactive protein (CRP), coagulation profile, fasting blood sugar, glycosylated hemoglobin (HbA1c), serology (HCV Ab), quantitative serum HCV RNA levels by real-time quantitative PCR (COBAS Ampliprep/TaqMan HCV monitor, with a detection limit of 15 IU/ml; Roche Diagnostic Systems), alpha-fetoprotein (AFP) (chemiluminescence immunoassay using a Cobas e 411 Analyzer, Roche Diagnostics, USA), and calculation of FIB-4 [Age in years × AST [U/L])/platelet count [10⁹/L]) × √ALT (U/L)] [8], The Child-Turcotte-Pugh (CTP) score evaluates the severity of liver dysfunction based on total bilirubin levels, serum albumin levels, prothrombin time, presence of ascites, and encephalopathy. The total score can range from 5 to 15, with Class A representing score of 5–6, Class B 7–9, and Class C 10–15 [9], the model for end-stage liver disease (MELD-Na) is calculated by first determining the MELD score using the formula [9.57 × log (creatinine) + 3.78 × log (total bilirubin) + 11.2 × log (INR) + 6.43]. The MELD-Na score is then calculated by using the formula [MELD − Na − [0.025 × MELD × (140 − Na)] + 140], with Na < 125 being calculated as 125 and Na > 140 being calculated as 140 [10].

Specific investigations

M2BPGI was measured via a double-antibody sandwich (ELISA) to assay the level of M2BPGI in the samples (Shanghai Sunred Biological Technology Co., Ltd.). M2BPGI was added to the monoclonal antibody enzyme well, which was pre-coated with M2BPGI monoclonal antibody. M2BPGI levels were indexed by the Cut-off Index (C.O.I.) which was calculated by dividing the M2BPGi count of the serum sample by the positive control and the negative control [11].

Abdominal ultrasonography

Abdominal ultrasonography provided detailed information about the liver’s texture, the presence of cirrhosis, and HCC. It also included comments on the number and size of focal lesions and the presence of portal vein thrombosis.

Transient elastography (TE)

Liver stiffness was measured by TE using the FibroScan device (EchoSens, Paris, France). Patients were examined in the supine position with their right upper extremity lifted. The detection site was fixed at 1.0–2.0 cm beneath the right liver capsule, away from the intrahepatic vessels and the gallbladder. The mean of 10 consecutive measurements during a single examination session was used for statistical analysis. Reliable TE measurements were defined as median values of 10 valid LS measurements, with an interquartile range < 30% and a success rate (SR) ≥ 60%. The cutoff values used were correlated with the METAVIR fibrosis scoring system as follows: F0–F1: 2–7 kPa. F2: 7–9.5 kPa. F3: 9.5–12.5 kPa, and F4 > 12.5 kPa (presence of cirrhosis) [12].

The diagnosis of hepatocellular carcinoma (HCC) was made through clinical evaluation, which included increased alpha-fetoprotein levels (with a threshold of > 200 ng/mL; > 400 ng/mL indicates a poor prognosis with specificity above 95%). Additionally, ultrasound imaging criteria and confirmation via triphasic computed tomography (CT) of the abdomen (which revealed a hypervascular pattern with arterial enhancement and rapid washout during the portal venous phase) were used [13, 14].

Statistical analysis

SPSS version 26 (Statistical Package for the Social Sciences) was used to analyze the data. Quantitative variables are described as means and standard deviations or medians and interquartile ranges, according to the type of data. Chi-square, Kruskal‒Wallis, one-way ANOVA tests with pairwise comparisons, and Fisher’s LSD comparisons were used as appropriate. The ROC curve was used to determine the best cutoff of M2BPGi for the diagnosis of cirrhosis and HCC multifocality and the size of HCC, and it was compared with that of AFP. To assess the strength and direction of the correlation between two continuous variables, the Spearman rank correlation coefficient (for non-normally distributed data) was used. Linear regression analysis was performed to detect parameters independently associated with serum M2BPGi levels. The level of statistical significance was set at P < 0.05. The sample size was calculated after assuming that the mean M2BPGI was 0.74 ± 0.5 vs. 1.38 ± 1.1 for F1 vs. F2. At 80% power and 95% CI, the estimated sample will be 56 cases, with 14 cases in each group (4) by Open Epi.

Results

Forty-two patients with chronic hepatitis C were candidates for antiviral therapy and were treated with sofosbuvir-based treatment regimens according to the treatment recommendation of the National Committee for Control of Viral Hepatitis in Egypt. All patients achieved a sustained virological response and were compared to a healthy control group (group A, n = 14) that was matched for age and sex.

A total of 167 patients were selected. One hundred twenty-five patients were excluded due to pregnancy in 3 patients, associated HBV in 10 patients, drug-induced liver disease in 20 patients, uncontrolled diabetes in 37 patients, sepsis in 12 patients, lymphoma in 3 patients, and missed follow-up or refused participation in 40 patients. The patients were randomly assigned in a sequential manner into three groups according to the inclusion criteria and results of FibroScan, as shown in Table 1: HCV patients without liver cirrhosis (n = 14), those with cirrhosis (n = 14), and those with documented HCC and liver cirrhosis (n = 14). Smoking and diabetes were significantly prevalent in group D, which included patients with HCC (P < 0.001, and 0.008, respectively). There was no significant difference between the studied groups in terms of age and sex (Table 1).

Table 1. Comparison between the studied groups regarding demographic and laboratory data

	Group A	Group B	Group C	Group D	χ2	P
	N = 14	N = 14	N = 14	N = 14
Gender:
Male	10 (71.4%)	5 (35.7%)	9 (64.3%)	11 (78.6%)	MC	0.089
Female	4 (28.6%)	9 (64.3%)	5 (35.7%)	3 (21.4%)
Smoking	1 (7.1%)	2 (14.3%)	6 (42.9%)	10 (71.4%)	MC	< 0.001**
Diabetes (n)	0 (0%)	3 (21.4%)	5 (35.7%)	8 (57.1%)		0.008
Age (year)	61.9 ± 4.9	58.5 ± 10.04	66.21 ± 8.94	62.71 ± 4.91	4.3	0.682
Hemoglobin (g/dl)	14.18 ± 1.06	11.54 ± 2.14	9.46 ± 1.14	7.32 ± 2.17	20.856	< 0.001**
WBCs (103/μl)	6.92 ± 1.94	8.69 ± 2.5	8.25 ± 5.28	8.51 ± 2.81	0.79	0.505
Platelet (103/μl)	328.43 ± 66.19	231.57 ± 82.1	154.21 ± 70.41	155.0 ± 45.78	20.967	< 0.001**
LSD	P1 < 0.001**	P2 0.004*	P3 0.976	P4 < 0.001**	P5 < 0.001**	P6 0.004*
Fasting blood sugar (mg/dl)	95.86 ± 5.56	107.93 ± 15.52	126.86 ± 26.54	166.21 ± 37.94	13.2	< 0.001**
HbA1c	5.28 ± 0.64	5.84 ± 0.9	6.29 ± 0.84	7.44 ± 1.52	11.111	< 0.001*
LSD	P1 = 0.069	P2 = 0.183	P3 = 0.020*	P4 = 0.001*	P5 < 0.001*	P6 = 0.022*
	Median(IQR)	Median(IQR)	Median(IQR)	Median(IQR)	KW	p
CRP (mg/dl)	2(1–2.7)	10.9(6.03–20.75)	38.5(23.25–60.5)	42(31.75–77)	38.229	< 0.001**
Pairwise	P1 0.006*	P2 0.029*	P3 0.498	P4 < 0.001**	P5 < 0.001**	P6 0.004*
Serum protein (gm/dl)	6.57 ± 0.37	6.25 ± 0.78	6.12 ± 0.58	6.36 ± 0.82	1.143	0.34
Serum albumin (gm/dl)	4.01 ± 0.45	3.23 ± 0.82	2.77 ± 0.79	2.32 ± 0.56	16.041	< 0.001**
LSD	P1 0.003*	P2 0.077	P3 0.086	P4 < 0.001**	P5 < 0.001**	P6 0.001**
	Median (IQR)	Median (IQR)	Median (IQR)	Median (IQR)	KW	p
Total bilirubin (mg/dl)	0.8 (0.6–0.9)	1.05 (0.77–1.33)	1.08 (0.7–1.43)	2.8 (1.36–5.6)	17.112	0.001**
Pairwise	P1 0.206	P2 0.614	P3 0.023*	P4 0.077	P5 < 0.001**	P6 0.005*
ALT(IU/L)	13.5 (10–19.5)	20 (12.9–39.75)	21.25 (17–34.68)	43 (15.74–52.2)	10.854	0.013*
Pairwise	P1 0.108	P2 0.676	P3 0.217	P4 0.043*	P5 < 0.001**	P6 0.099
AST(IU/L)	13(10.5–16.25)	19.4(13.58–38.5)	34(18.08–51.6)	73(42–122)	25.906	< 0.001**
Pairwise	P1 0.028*	P2 0.357	P3 0.06	P4 0.002*	P5 < 0.001**	P6 0.005*
AST/ALT ratio	0.85(0.75–0.98)	1.21(0.59–1.87)	1.79(0.79–2.37)	1.94(0.87–4.42)	8.939	0.03*
Pairwise	P1 0.391	P2 0.19	P3 0.61	P4 0.03*	P5 0.007*	P6 0.069
FIB-4	–	1.13 (0.69–2.9)	3.36 (2.77–4.59)	5.52 (2.19–10.98)	31.3	< 0.001*
Pairwise	–	P2 = 0.012*	P3 = 0.041*	P4 < 0.001*
PT	12.99 ± 0.78	15.3 ± 0.82	15.54 ± 1.35	17.51 ± 2.43	21.3	< 0.001**
LSD	P1 < 0.001**	P2 0.67	P3 < 0.001**	P4 < 0.001**	P5 < 0.001**	P6 < 0.001**
INR	0.91 ± 0.2	1.2 ± 0.09	1.22 ± 0.1	1.44 ± 0.19	28.129	< 0.001**
LSD	P1 < 0.001**	P2 0.685	P3 < 0.001**	P4 < 0.001**	P5 < 0.001**	P6 0.135
	Median (IQR)	Median (IQR)	Median (IQR)	Median (IQR)	KW	p
Creatinine (mg/dl)	0.7(0.7–0.8)	0.8(0.68–1.03)	0.9(0.78–1.2)	1.1(0.68–1.03)	8.848	0.031*
Pairwise	P1 0.213	P2 0.293	P3 0.67	P4 0.022*	P5 0.006*	P6 0.139
BUN (mg/dl)	22(18.75–24.5)	25(18.25–36)	26(23.75–33.25)	41(2.5–55.75)	18.134	< 0.001**
MELD-Na score	–	5.79 ± 1.1	11.93 ± 3.41	20.36 ± 6.36	44.274	< 0.001*
LSD	–	P2 < 0.001*	P3 < 0.001*	P4 < 0.001*
Pairwise	P1 0.13	P2 0.42	P3 0.064	P4 0.02*	P5 < 0.001**	P6 0.008*
Serum M2BPGi (C.O.I)	0.75 (0.6–0.79)	0.86 (0.63–1.03)	0.97 (0.78–1.09)	0.99 (0.87–1.25)	18.177	< 0.001**
Pairwise	P1 0.073	P2 0.19	P3 0.363	P4 0.002*	P5 < 0.001**	P6 0.026*
AFP (ng/ml)	9.7 ± 4	16.1 ± 8.7	24.4 ± 8.95	221.9 ± 325(SE 87)***		0.002

LSD least significant difference, CRP C-reactive protein, FIB-4 Fibrosis Index Based on 4 Factors, PT prothrombin time, INR international normalized ratio, BUN blood urea nitrogen, COI cutoff index, M2BPGi Mac-2 binding protein glycosylation isomer, AFP Alpha-fetoprotein

χ2 chi-square test, MC Monte Carlo test, p1 difference between group A and group B, p2 difference between group B and C, p3 difference between group C and D, p4 difference between group A and C, p5 difference between group A and group D, p6 difference between group B and D. LSD Fisher least significance difference test

^*p p < 0.05 is statistically significant

^**p p < 0.001 is statistically highly significant

^***SE standard error is selected if SD exceeds the mean value

There was a significant difference between the studied groups in terms of hemoglobin and platelet levels. These levels were lower in the HCC group than in the other groups, which can be attributed to portal hypertension and previous episodes of upper gastrointestinal bleeding. In terms of CRP (which reflects the intense inflammatory state), FBS, and HbA1c, their mean values were significantly greater in both the HCC subgroup and the HCV-related cirrhosis subgroup. This can be attributed to the higher prevalence of diabetes in these groups (Table 1).

Serum transaminases, bilirubin, PT, INR, creatinine, and BUN were significantly greater in the HCC group than in the other groups, with lower serum albumin in the HCC subgroup than in the other subgroups, reflecting deranged liver function with an increased catabolic state (Table 1). FIB-4 and MELD-Na scores were significantly higher in the HCC group than in the other groups, implying a worse prognosis in these patients (Table 1). The AFP level was significantly greater in the HCC patients than in the healthy controls (P = 0.001), patients with HCV without cirrhosis (P = 0.002), and patients with HCV with cirrhosis (P = 0.002).

The disease duration was significantly longer in the HCC group, with a median of 16.5 years (range 15–20) (P < 0.001). CTP class B was prevalent in group C, whereas class C was prevalent in the HCC group. Manifestations of hepatic decompensation and hematemesis were more common in the HCC group (P < 0.001) (Table 2).

Table 2. Comparison between the studied patient groups regarding presenting symptoms and ultrasonographic data

		Group B	Group C	Group D	χ2	P
		N = 14 (%)	N = 14 (%)	N = 14 (%)
Clinical	Child class:				21.6
	B	0 (0%)	13 (92.9%)	3 (21.4%)		< 0.001*
	C	0 (0%)	1 (7.1%)	11 (78.6%)
Disease duration (years)		Median(IQR)	Median(IQR)	Median(IQR)	KW	p
		8(6.75–12.25)	14.5(11.75–17.25)	16.5(15–20)	22.508	< 0.001**
Pairwise comparison		P1 0.003*	P2 0.095	P3 < 0.001**
Presenting symptom	Jaundice	0 (0%)	1 (7.1%)	8 (57.1%)	MC	< 0.001*
	P (χ2)	P1 > 0.999	P2 0.012*	P3 < 0.001**
	Pallor	2 (14.3%)	6 (42.9%)	13 (92.9%)	27.9	< 0.001*
	P (χ2)	P1 < 0.001**	P2 < 0.001**	P3 < 0.001**
	LL edema	0 (0%)	14 (100%)	14 (100%)	19.8	< 0.001*
	Ascites	0 (0%)	10 (71.4%)	14 (100%)	16.3	< 0.001*
	Hematemesis	0 (0%)	6 (42.9%)	10 (71.4%)	18.1	< 0.001*
	P (χ2)	P1 = 0.016*	P2 = 0.252	P3 < 0.001*
Ultrasound	Liver ultrasound:
	Average	9 (64.3%)	2 (14.3%)	0 (0%)
	Shrunken	0 (0%)	5 (35.7%)	10 (71.4%)	3.549	0.06
	Enlarged	5 (35.7%)	7 (50%)	4 (28.6%)
	Cirrhosis	0 (0%)	14 (100%)	14 (100%)	19.8	< 0.001*
	P (χ2)	P1 < 0.001*	–-	P3 < 0.001*
	Ascites
	Moderate	0 (0%)	11 (78.6%)	3 (21.4%)	26.3	< 0.001*
	Severe	0 (0%)	3 (21.4%)	11 (78.6%)
	P (χ2)	P1 < 0.001*	P2 = 0.007*	P3 < 0.001*

χ² chi-square test, MC Monte Carlo test

^*p < 0.05 is statistically significant

^**p ≤ 0.001 is statistically highly significant, p1 difference between groups B and C, p2 difference between groups C and D, p3 difference between groups B and D

Diabetes was prevalent in 16 patients, with 3 patients (21.4%) in group B, 5 patients (35.7%) in group C, and 8 patients (57.1%) in group D (P = 0.008). The mean HbA1c of these patients was 7.41 ± 1.2. Compared with non-diabetic patients, diabetic patients had a non-significantly greater M2BPGi level (0.93 ± 0.2 vs. 0.87 ± 0.24 (C.O.I), P = 0.404).

Evaluation of serum M2BPGi levels in the studied groups

Mac-2 binding protein glycosylation isomer was significantly different between the study groups, being higher in the HCV with cirrhosis and HCC subgroups than in the control group (P < 0.001) (Table 1), and being significantly greater in the multifocal HCC subgroup than in the unifocal HCC subgroup (P < 0.001), and its level was directly proportional to the size of the HCC lesion (P = 0.013) (Table 3).

Table 3. Serum M2BPGi level variation according to size and number of HCC lesions

	Median (IQR)	Serum M2BPGi (C.O.I)	P
Number of HCC nodules	2(1–2)
Unifocal	5 (35.7%)	0.98 (0.87–1.1)	< 0.001
Multifocal	9 (64.3%)	1.2 (1.1–1.25)
Size of largest HCC nodule (cm)	3.8 ± 1.01 (2.6–6.1)
< 3 cm	3(21.4%)	0.86 ± 0.02
3–5 cm	9(64.3%)	1.1 ± 0.2	0.013
> 5 cm	2(14.3%)	1.2 ± 0.19

IQR interquartile range, COI cutoff index

Correlation of the serum M2BPGi concentration with the AFP concentration in the studied groups

There was no significant correlation between AFP and M2BPGi levels (r = 0.057, P = 0.676). AFP correlated with the presence of HCC (r = 0.492, P = 0.000), number (r = 0.443, P = 0.001), and size of the lesions (r = 0.354, P = 0.007).

Four patients (28.6%) in the HCC group had an AFP level > 200 ng/ml (649.3 ± 341.8 ng/ml) and their corresponding M2BPGi level (0.96 ± 0.2 C.O.I). Ten HCC patients (71.4%) had an AFP concentration < 200 ng/ml (51 ± 23 ng/ml) with M2BPGi (1.06 ± 0.198 C.O.I) (P = 0.41), indicating a possible diagnostic rule of M2BPGi in HCC with a low level of AFP, as shown in Table 5.

ROC curve analysis revealed that the area under the curve (AUC), with corresponding cutoff values of serum AFP, was significantly greater than that of serum M2BPGi with respect to the diagnosis of cirrhosis, HCC, and multifocality. This was accompanied by higher corresponding values for sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy. However, the serum M2BPGi level was superior for the prediction of large HCC lesions, as depicted in Table 4 and Fig. 1. The cutoff value of M2BPGi for diagnosing HCC in patients with a low level of AFP (< 200 ng/ml) was determined to be 0.903 C.O.I, with a sensitivity of 80%, specificity of 75%, and accuracy of 76.25% (Table 5).

Table 4. Performance of serum M2BPGi in the diagnosis of cirrhosis, HCC, and multifocality of HCC among studied participants

		Cutoff	AUC	Sensitivity	Specificity	PPV	NPV	Accuracy	P
Cirrhosis	M2BPGi (C.O.I)	≥ 0.785	0.744	78.6%	60.7%	50%	85%	66.7%	0.002
	AFP (ng/ml)	≥ 13	0.859	87.9%	73.9%				0.000
HCC	M2BPGi (C.O.I)	≥ 0.869	0.762	78.6%	61.9%	40.7%	89.7%	66.1%	0.00400.000
	AFP (ng/ml)	> 29	0.957	85.7%	83.3%
Multifocal HCC	M2BPGi (C.O.I)	≥ 0.93	0.73	66.7%	63.8%	26.1%	90.9%	64.3%	0.03
	AFP (ng/ml)	27	0.888	88.9%	74.5%	54.1%	95.7%	78.38%	0.000

p < 0.05 is statistically significant, AUC area under curve, COI cutoff index, PP positive predictive value, NPV negative predictive value

Fig. 1 [Images not available. See PDF.]

ROC curve showing the comparative performance of both AFP and serum M2BPGi levels in the diagnosis of cirrhosis (A), HCC (B), and multifocal HCC (C) among the studied participants

Table 5. Comparative approach between serum AFP (ng/ml) and serum M2BPGi (C.O.I) in the studied groups

AFP1	Group		N	Minimum	Maximum	Mean	Std. deviation
< 200 ng/ml	Healthy (G1)	MAC2BP	14	.54	.84	0.71	0.106
	HCV non-cirrhotic (G2)	MAC2BP	14	.54	1.14	0.842	0.21
	HCV with cirrhosis (G3)	MAC2BP	14	.59	1.58	0.964	0.251
	HCV with HCC (G4a)	MAC2BP	10	.72	1.31	1.06	0.198
> 200 ng/ml	HCV HCC(G4b)	MAC2BP	4	.84	1.26	0.96	0.2

Table 6 shows the parameters that were significantly correlated with serum M2BPGi levels. All of these variables were significantly positively correlated; however, hemoglobin, platelet count, serum albumin concentration, and serum sodium concentration showed a significantly negative correlation.

Table 6. The correlation between serum M2BPGi and the studied parameters

	r	P
Disease duration	0.465	0.000
Hemoglobin	− 0.385	0.003
Platelet count	− 0.411	0.002
Serum albumin	− 0.402	0.002
PT	0.392	0.002
INR	0.437	0.001
BUN	0.361	0.006
Sodium	− 0.396	0.003
CRP	0.545	0.000
Child score	0.472	0.000
MELD-Na score	0.449	0.001
Cirrhosis	0.401	0.002
Ascites	0.452	0.002
HCC	0.371	0.005
Largest size of HCC nodule	0.418	0.001
Number of HCC nodules	0.362	0.006

PT prothrombin time, INR international normalized ratio, BUN blood urea nitrogen, CRP creative protein, MELD-Na model for end-stage liver disease-sodium, HCC hepatocellular carcinoma

Linear regression with backward selection analysis of parameters independently associated with serum M2BPGi among the studied participants revealed that CRP (β = 0.485, P = 0.001), the presence of HCC (β =  − 1.446, P = 0.001), and a larger HCC size (β = 1.42, P = 0.001) were independently associated with this parameter (Table 7). The R square value (0.49) indicated that this prediction model can explain 49% of the cases (F = 7.86, P = 0.000), with a regression constant of 0.836.

Table 7. Linear stepwise regression analysis of parameters independently associated with serum M2BPGi among studied participants

Model	Unstandardized coefficients		Standardized coefficients	t	Sig	95.0% confidence interval for B
	B	Std. error	Beta			Lower bound	Upper bound
(Constant)	.836	.126		6.629	0.000	.582	1.089
Fibroscan value	.018	.010	.538	1.808	0.077	− .002	.038
DURATION	− .014	.008	− .456	− 1.775	0.082	− .031	.002
PLT	− .001	.000	− .244	− 1.697	0.096	− .001	.000
CRP	.004	.001	.485	3.530	0.001	.002	.006
HCC	− .758	.228	− 1.446	− 3.327	0.002	− 1.215	− .300
Largest HCC	.189	.053	1.422	3.546	0.001	.082	.296

HCC hepatocellular carcinoma, PLT platelets, PT prothrombin time, CRP C-reactive protein

Discussion

The risk of developing cirrhosis-related consequences and HCC is increased in individuals with severe fibrosis. Consequently, it is important to predict patients at increased risk. The pathophysiological function of M2BPGi is not well understood. M2BPGi is produced by HSCs and causes Kupffer cells to express Mac-2, which in turn stimulates HSCs and elevates alpha-smooth muscle actin. Therefore, M2BPGi is related to the stage of liver fibrosis and plays a significant role in the development of hepatic fibrosis [5].

In the present study, smoking and diabetes were more prevalent in HCC patients than in the other groups (P < 0.001, and 0.008, respectively). Smoking appears to increase ALT levels only in HCV infection, probably by increasing oxidative stress, impairing immunological function, and creating insulin resistance, all of which are linked to HCV-related HCC [15]. Additionally, earlier hospital-based cross-sectional investigations revealed a correlation between higher serum M2BPGi levels and high fasting plasma glucose and HbA1c levels, as well as a greater proportion of those with impaired glucose metabolism [16, 17]. Some claim that the addition of blood M2BPGi levels to conventional risk variables for type 2 diabetes enhances the capacity to predict the future risk of the disease [18]. Liver cirrhosis increases glucose intolerance and diabetes through a number of mechanisms, including insulin resistance and decreased insulin secretion, with a confirmed link between HCC and diabetes mellitus [19, 20].

Hemoglobin levels and platelet counts (P = 0.007, and < 0.001, respectively) were significantly lower in the HCC group than in the other groups and were negatively correlated with M2BPGi levels, reflecting the indirect associations of M2BPGi levels with the severity of fibrosis and portal hypertension. Previous studies agreed with our study and revealed that anemia is a common event in patients with advanced cirrhosis and HCC [21–23]. Previous research has revealed that serum M2BPGi levels are positively correlated with the INR and negatively correlated with the platelet count [24].

C-reactive protein was greater in the HCC group than in the other groups and was positively correlated with M2BPGi levels (P = 0.000). It was also identified as one of the independently associated variables (P = 0.001). In support of our findings, Pieri et al. [25] reported that high initial CRP levels were associated with the progression of cirrhosis and the development of HCC. A meta-analysis conducted on 1885 patients with HCC from 11 cohorts (representing 10 studies, 7 of which were conducted in Asia) demonstrated a substantial correlation between elevated blood CRP levels and poor overall survival as well as recurrence-free survival (HR 2.15, 95% CI 1.76–2.63) (HR 2.66, 95% CI 1.54–4.58) in patients with HCC [26]. CRP has been linked to systemic inflammation, insulin resistance, and HCC in cirrhotic patients.

Compared with those of the other groups, FIB-4, CTP, and MELD scores were significantly greater in patients with liver cirrhosis and HCC. However, the CTP and MELD score were positively correlated with the M2BPGi level (P = 0.000, 0.001). This finding is consistent with previous studies that have shown that patients with severe fibrosis or cirrhosis have higher FIB-4 levels. Performing FIB-4 followed by M2BPGi could help decrease unnecessary referrals and/or liver biopsies [27]. Additionally, M2BPGi values differ significantly between patients with high and low CTP scores [28].

Mac-2 binding protein glycosylation was significantly different between the studied groups. The difference between the cirrhosis and HCC groups, compared with the control and HCV non-cirrhotic groups was substantial. According to research by McGlynn et al. [29], blood M2BPGi levels in patients with HCV infection are associated with fibrosis stage, dynamic fibrosis evolution, antiviral responses, the presence of esophagogastric varices, the risk of HCC development or recurrence, extrahepatic malignancy, and survival. A study revealed that M2BPGi is a marker for the presence of esophagogastric varices in cirrhotic patients after HCV eradication with DAAs [30]. In addition, we found that the M2BPGi level was directly proportional to the number and size of HCC lesions (P < 0.001, and P = 0.013, respectively).

The best cutoff of serum M2BPGi for the diagnosis of cirrhosis was 0.785 C.O.I, with an AUC of 0.741, a sensitivity of 78.6%, and a specificity of 60.7% (P = 0.002). A study showed that ROC curve analysis for the prediction of hepatic cirrhosis had a cutoff value of 0.78 C.O.I for M2BPGi, with a sensitivity of 90% and specificity of 58% [31]. Wei et al. [32] reported that M2BPGi has a high AUC value for detecting cirrhosis (AUC = 0.811, 95% CI 0.735–0.860). Additionally, Heo et al. [33] reported an AUC of 0.704 for cirrhosis.

In the present study, the optimal blood M2BPGi threshold for the diagnosis of HCC was ≥ 0.8685 C.O.I, with an AUC of 0.762, a sensitivity of 78.6%, a specificity of 61.9%, a positive predictive value of 40.7%, a negative predictive value of 89.7%, and an overall accuracy of 66.1% (P = 0.004). This finding is in agreement with a study that reported that M2BPGi could be a significant marker of hepatocellular carcinoma recurrence after achieving SVR in patients with HCV, with an AUC value of 0.75 (sensitivity 62.5%, specificity 82.5%; P = 0.0015) [34]. According to earlier research, individuals with chronic hepatitis B, chronic hepatitis C, or non-alcoholic fatty liver disease can use M2BPGi levels as a serum biomarker of HCC [35]. However, the cutoff value of M2BPGi for the prediction of HCC with HBV seems to be greater than that with HCV, as shown in a previous study that revealed a cutoff value at 48 weeks > 1.5 with a hazard ratio of 34.9 (95% confidence interval, 4.3–284.9 [36].

AFP still appears to be superior to M2BPGi, as denoted by a larger AUC in predicting cirrhosis, HCC, and multifocality. However, M2BPGi levels were found to have greater diagnostic value for large HCC lesions and HCC lesions with low AFP levels.

Among the variables that were significantly correlated with M2BPGi levels, CRP levels, HCC, and the largest HCC nodule size were independently associated with M2BPGi levels. Saleh et al. [24] reported that pre-treatment serum M2BPGi levels were negatively correlated with platelet count, but positively correlated with bilirubin, INR, and AST levels. Yamasaki et al. [37] reported a significant positive correlation between serum M2BPGi and bilirubin levels (r = 0.091, P = 0.001) and a significant negative correlation with platelet count (r =  − 0.147, P < 0.001).

The limitations of the current study are the relatively small sample size, the single-center nature of the study, and the lack of follow-up. However, one important result of this study is the diagnostic value of serum M2BPGi levels in HCC patients with low AFP levels. Therefore, future studies are needed to evaluate serum M2BPGi levels as an early marker of HCC, especially in patients with low levels of AFP and advanced liver cirrhosis.

In conclusion, M2BPGi can be used as an effective marker that could be related to the biological behavior of the tumor or its aggressiveness; further studies are warranted on a large scale of patients to confirm our findings.

Acknowledgements

Not applicable.

Authors’ contributions

Conceptualization: AS Hanafy. Data curation: KAM Abdelaziz, FN Mohammad, AS Ibrahim. Formal analysis: all authors. Methodology: AS Hanafy. Writing—original draft: KAM Abdelaziz, AS Hanafy. Writing—review and editing: all authors. Approval of final manuscript: all authors.

Funding

No funding agent.

Data availability

The data can be provided upon request.

Declarations

Ethics approval and consent to participate

Informed consent was signed by patients who provided their permission to participate in the study. The study methods were approved by the research ethics board of the Zagazig Faculty of Medicine, Egypt (ZU-IRB-9718/2022). The research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki and its later amendments.

Consent for publication

Patients gave their consent for their data to be used for publications and research without revealing their identities.

Competing interests

The authors declare that they have no competing interests.

Abbreviations

AFP

Alpha fetoprotein

AST

Aspartate aminotransferase

ALT

Alanine aminotransferase

AUC

Area under curve

BUN

Blood urea nitrogen

COI

Cutoff index

CT

Computed tomography

CTP

Child-Turcotte-Pugh score

FIB-4

Fibrosis Index Based on 4 Factors

HCC

Hepatocellular carcinoma

HSCs

Hepatic stellate cells

INR

International normalized ratio

kPa

Kilopascals

M2BPGi

Mac-2 binding protein glycosylation isomer

PLT

Platelets

PT

Prothrombin time

MELD

Model for end-stage liver disease

ROC

Receiver operating characteristic

SVR

Sustained virological response

TE

Transient elastography

USG

Abdominal ultrasonography

WBCs

White cell count

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Llovet, JM; Kelley, RK; Villanueva, A; Singal, AG; Pikarsky, E; Roayaie, S et al. Hepatocellular carcinoma. Nat Rev Dis Primers; 2021; 7, 1 6. [DOI: https://dx.doi.org/10.1038/s41572-020-00240-3] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33479224]

2. Shen, C; Jiang, X; Li, M; Luo, Y. Hepatitis virus and hepatocellular carcinoma: recent advances. Cancers (Basel); 2023; 15, 2 533.[COI: 1:CAS:528:DC%2BB3sXivVeht7o%3D] [DOI: https://dx.doi.org/10.3390/cancers15020533] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36672482]

3. Nagata, H; Nakagawa, M; Nishimura-Sakurai, Y; Asano, Y; Tsunoda, T; Miyoshi, M et al. Serial measurement of Wisteria floribunda agglutinin positive Mac-2-binding protein is useful for predicting liver fibrosis and the development of hepatocellular carcinoma in chronic hepatitis C patients treated with IFN-based and IFN-free therapy. Hepatol Int; 2016; 10, 6 pp. 956-964. [DOI: https://dx.doi.org/10.1007/s12072-016-9754-1] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27435935]

4. Mak, LY; To, WP; Wong, DK; Fung, J; Liu, F; Seto, WK; Lai, CL; Yuen, MF. Serum Mac-2 binding protein glycosylation isomer level predicts hepatocellular carcinoma development in E-negative chronic hepatitis B patients. World J Gastroenterol; 2019; 25, 11 pp. 1398-1408.[COI: 1:CAS:528:DC%2BC1MXhsFCntrfE] [DOI: https://dx.doi.org/10.3748/wjg.v25.i11.1398] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30918432][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6429346]

5. Tamaki, N; Kurosaki, M; Loomba, R; Izumi, N. Clinical utility of Mac-2 binding protein glycosylation isomer in chronic liver diseases. Ann Lab Med; 2021; 41, 1 pp. 16-24.[COI: 1:CAS:528:DC%2BB38XitVWlu74%3D] [DOI: https://dx.doi.org/10.3343/alm.2021.41.1.16] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32829576][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7443525]

6. Jun, T; Hsu, YC; Ogawa, S; Huang, YT; Yeh, ML; Tseng, CH et al. Mac-2 binding protein glycosylation isomer as a hepatocellular carcinoma marker in patients with chronic hepatitis B or C infection. Hepatol Commun; 2019; 3, 4 pp. 493-503.[COI: 1:CAS:528:DC%2BC1MXmvVKnsr8%3D] [DOI: https://dx.doi.org/10.1002/hep4.1321] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30976740][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6442699]

7. Inoue, T; Tsuzuki, Y; Iio, E; Shinkai, N; Matsunami, K; Fujiwara, K et al. Clinical evaluation of hepatocarcinogenesis and outcome using a novel glycobiomarker Wisteria floribunda agglutinin-positive Mac-2 binding protein (WFA+-M2BP) in chronic hepatitis C with advanced fibrosis. Jpn J Infect Dis; 2018; 71, pp. 177-183.[COI: 1:CAS:528:DC%2BC1MXlvVKrsbk%3D] [DOI: https://dx.doi.org/10.7883/yoken.JJID.2017.459] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29491234]

8. Sterling, RK; Lissen, E; Clumeck, N; Sola, R; Correa, MC; Montaner, J et al. Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection. Hepatology; 2006; 43, 6 pp. 1317-1325.[COI: 1:CAS:528:DC%2BD28XmsF2nur0%3D] [DOI: https://dx.doi.org/10.1002/hep.21178] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16729309]

9. Pugh, RN; Murray-Lyon, IM; Dawson, JL; Pietroni, MC; Williams, R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg; 1973; 60, 8 pp. 646-649.[COI: 1:STN:280:DyaE3s3jsVWqsw%3D%3D] [DOI: https://dx.doi.org/10.1002/bjs.1800600817] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/4541913]

10. Nagai, S; Chau, LC; Schilke, RE; Safwan, M; Rizzari, M; Collins, K et al. Effects of allocating livers for transplantation based on model for end-stage liver disease-sodium scores on patient outcomes. Gastroenterology; 2018; 155, 5 pp. 1451-1462.e3. [DOI: https://dx.doi.org/10.1053/j.gastro.2018.07.025] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30056096]

11. Bui, HH; Nguyen, ST; Phan, ST; Nguyen, KM; Nguyen, CD. Evaluating M2BPGi as a marker for liver fibrosis in patients with chronic hepatitis B. Dig Dis Sci; 2023; 68, 12 pp. 4407-4417.[COI: 1:CAS:528:DC%2BB3sXitF2lurjF] [DOI: https://dx.doi.org/10.1007/s10620-023-08143-5] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37861877][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10635958]

12. Castera, L; Forns, X; Alberti, A. Non-invasive evaluation of liver fibrosis using transient elastography. J Hepatol; 2008; 48, 5 pp. 835-847. [DOI: https://dx.doi.org/10.1016/j.jhep.2008.02.008] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18334275]

13. Galle, PR; Foerster, F; Kudo, M; Chan, SL; Llovet, JM; Qin, S et al. Biology and significance of alpha-fetoprotein in hepatocellular carcinoma. Liver Int; 2019; 39, 12 pp. 2214-2229. [DOI: https://dx.doi.org/10.1111/liv.14223] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31436873]

14. Shah, S; Shukla, A; Paunipagar, B. Radiological features of hepatocellular carcinoma. J Clin Exp Hepatol; 2014; 4, Suppl 3 pp. S63-S66. [DOI: https://dx.doi.org/10.1016/j.jceh.2014.06.009] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25755613][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4284210]

15. Chuang, S-C; Lee, Y-CA; Hashibe, M; Dai, M; Zheng, T; Boffetta, P. Interaction between cigarette smoking and hepatitis B and C virus infection on the risk of liver cancer: a meta-analysis. Cancer Epidemiol Biomark Prev; 2010; 19, pp. 1261-1268.[COI: 1:CAS:528:DC%2BC3cXls1Clt7g%3D] [DOI: https://dx.doi.org/10.1158/1055-9965.EPI-09-1297]

16. Sugiura, T; Dohi, Y; Takase, H; Yamashita, S; Murai, S; Tsuzuki, Y et al. Serum levels of Mac-2 binding protein increase with cardiovascular risk and reflect silent atherosclerosis. Atherosclerosis; 2016; 251, pp. 192-196.[COI: 1:CAS:528:DC%2BC28XhtVKntL%2FJ] [DOI: https://dx.doi.org/10.1016/j.atherosclerosis.2016.06.027] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27344370]

17. Park, H; Jun, DW; Park, HK; Kim, M. Clinical implications of serum Mac-2-binding protein glycan isomer as a novel biomarker of advanced hepatic fibrosis in diabetes. Ann Transl Med; 2020; 8, 23 1583.[COI: 1:CAS:528:DC%2BB3MXjvVyls7o%3D] [DOI: https://dx.doi.org/10.21037/atm-20-5216] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33437782][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7791238]

18. Higashioka, M; Hirakawa, Y; Hata, J; Honda, T; Sakata, S; Shibata, M et al. Serum Mac-2 binding protein glycosylation isomer concentrations are associated with incidence of type 2 diabetes. J Clin Endocrinol Metab; 2023; 108, 7 pp. e425-e433. [DOI: https://dx.doi.org/10.1210/clinem/dgad011] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36638007][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10271221]

19. Nadelson, J; Satapathy, SK; Nair, S. Glycated hemoglobin levels in patients with decompensated cirrhosis. Int J Endocrinol; 2016; 2016, 8390210. [DOI: https://dx.doi.org/10.1155/2016/8390210] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27882051][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5110874]

20. Lai, SW; Chen, PC; Liao, KF; Muo, CH; Lin, CC; Sung, FC. Risk of hepatocellular carcinoma in diabetic patients and risk reduction associated with anti-diabetic therapy: a population-based cohort study. Am J Gastroenterol; 2012; 107, pp. 46-52.[COI: 1:CAS:528:DC%2BC38XkslOnug%3D%3D] [DOI: https://dx.doi.org/10.1038/ajg.2011.384] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22085817]

21. Finkelmeier, F; Bettinger, D; Köberle, V; Schultheiß, M; Zeuzem, S; Kronenberger, B; Piiper, A; Waidmann, O. Single measurement of hemoglobin predicts outcome of HCC patients. Med Oncol; 2014; 31, 1 806. [DOI: https://dx.doi.org/10.1007/s12032-013-0806-2] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24326985]

22. Yang, YY; Lin, HC; Lee, WC; Huang, YT; Hou, MC; Lee, FY et al. Plasma erythropoietin level in patients with cirrhosis and its relationship to the severity of cirrhosis and renal function. J Gastroenterol Hepatol; 2003; 18, pp. 1156-1161.[COI: 1:CAS:528:DC%2BD3sXpsVOgtr8%3D] [DOI: https://dx.doi.org/10.1046/j.1440-1746.2003.03144.x] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/12974902]

23. Bruno, CM; Neri, S; Sciacca, C; Bertino, G; Di Prima, P; Cilio, D et al. Plasma erythropoietin levels in anaemic and non-anaemic patients with chronic liver diseases. World J Gastroenterol; 2004; 10, pp. 1353-1365.[COI: 1:CAS:528:DC%2BD2cXnsVCntrs%3D] [DOI: https://dx.doi.org/10.3748/wjg.v10.i9.1353] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15112358][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4622782]

24. Saleh, SA; Salama, MM; Alhusseini, MM; Mohamed, GA. M2BPGi for assessing liver fibrosis in patients with hepatitis C treated with direct-acting antivirals. World J Gastroenterol; 2020; 26, 21 pp. 2864-2876.[COI: 1:CAS:528:DC%2BB3cXitVSnu7rI] [DOI: https://dx.doi.org/10.3748/wjg.v26.i21.2864] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32550761][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7284180]

25. Pieri, G; Agarwal, B; Burroughs, AK. C-reactive protein and bacterial infection in cirrhosis. Ann Gastroenterol; 2014; 27, 2 pp. 113-120. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24733601][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3982625]

26. Zhang, Y; Wang, R; Yang, X. FIB-4 index serves as a noninvasive prognostic biomarker in patients with hepatocellular carcinoma: a meta-analysis. Medicine (Baltimore); 2018; 97, 51[COI: 1:CAS:528:DC%2BC1cXisFCks7jM] [DOI: https://dx.doi.org/10.1097/MD.0000000000013696] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30572498]

27. Kim, M; Jun, DW; Park, H; Kang, BK; Sumida, Y. Sequential combination of FIB-4 followed by M2BPGi enhanced diagnostic performance for advanced hepatic fibrosis in an average risk population. J Clin Med; 2020; 9, 4 1119. [DOI: https://dx.doi.org/10.3390/jcm9041119] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32295166][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7230806]

28. Moon, HW; Park, M; Hur, M; Kim, H; Choe, WH; Yun, YM. Usefulness of enhanced liver fibrosis, glycosylation isomer of Mac-2 binding protein, Galectin-3, and soluble suppression of tumorigenicity 2 for assessing liver fibrosis in chronic liver diseases. Ann Lab Med; 2018; 38, 4 pp. 331-337.[COI: 1:CAS:528:DC%2BC1MXhvVejt7c%3D] [DOI: https://dx.doi.org/10.3343/alm.2018.38.4.331] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29611383][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5895862]

29. McGlynn, KA; Petrick, JL; El-Serag, HB. Epidemiology of hepatocellular carcinoma. Hepatology.; 2021; 73 Suppl 1, Supp1 1 pp. 4-13. [DOI: https://dx.doi.org/10.1002/hep.31288] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32319693]

30. Kikukawa, K; Uchida-Kobayashi, S; Tamori, A; Yoshida, K; Kotani, K; Motoyama, H et al. Serum Mac-2-binding protein glycosylation isomer predicts esophagogastric varices in cirrhotic patients with chronic hepatitis C virus infection treated with IFN-free direct-acting antiviral agent: M2BPGi levels predict varices in SVR patients. Ann Hepatol; 2020; 19, 4 pp. 367-372.[COI: 1:CAS:528:DC%2BB3cXisFKiurjO] [DOI: https://dx.doi.org/10.1016/j.aohep.2020.04.002] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32444247]

31. Hiyoshi, M; Yano, K; Nanashima, A; Ikenoue, M; Imamura, N; Fujii, Y et al. Usefulness of serum Mac-2 binding protein glycosylation isomer in patients undergoing hepatectomy: A case controlled study. Ann Med Surg (Lond); 2019; 17, 48 pp. 17-22. [DOI: https://dx.doi.org/10.1016/j.amsu.2019.10.014]

32. Wei, B; Feng, S; Chen, E; Li, D; Wang, T; Gou, Y et al. M2BPGi as a potential diagnostic tool of cirrhosis in Chinese patients with Hepatitis B virus infection. J Clin Lab Anal; 2018; 32, 2 e22261. [DOI: https://dx.doi.org/10.1002/jcla.22261] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28544156]

33. Heo, JY; Kim, SU; Kim, BK; Park, JY; Kim, DY; Ahn, SH et al. Use of wisteria floribunda agglutinin-positive human Mac-2 binding protein in assessing risk of hepatocellular carcinoma due to hepatitis B virus. Medicine (Baltimore); 2016; 95, 14 e3328.[COI: 1:CAS:528:DC%2BC28XmtFeqs7o%3D] [DOI: https://dx.doi.org/10.1097/MD.0000000000003328] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27057911]

34. Yugawa, K; Maeda, T; Nagata, S; Sakai, A; Edagawa, M; Omine, T et al. Mac-2-binding protein glycosylation isomer as a novel predictor of hepatocellular carcinoma recurrence in patients with hepatitis C virus eradication. Ann Surg Oncol; 2022; 29, 4 pp. 2711-2719. [DOI: https://dx.doi.org/10.1245/s10434-021-11011-z] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34729653]

35. Jang, SY; Tak, WY; Park, SY; Kweon, YO; Lee, YR; Kim, G et al. Diagnostic efficacy of serum Mac-2 binding protein glycosylation isomer and other markers for liver fibrosis in non-alcoholic fatty liver diseases. Ann Lab Med; 2021; 41, 3 pp. 302-309.[COI: 1:CAS:528:DC%2BB38XitVWksbY%3D] [DOI: https://dx.doi.org/10.3343/alm.2021.41.3.302] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33303715][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7748098]

36. Murata, A; Amano, N; Sato, S; Tsuzura, H; Tomishima, K; Sato, S et al. On-treatment serum Mac-2 binding protein glycosylation isomer (M2BPGi) level and risk of hepatocellular carcinoma development in patients with chronic hepatitis B during nucleot(s)ide analogue therapy. Int J Mol Sci; 2020; 21, 6 2051.[COI: 1:CAS:528:DC%2BB3cXhvVejsLjO] [DOI: https://dx.doi.org/10.3390/ijms21062051] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32192084][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7139452]

37. Yamasaki, K; Tateyama, M; Abiru, S; Komori, A; Nagaoka, S; Saeki, A; Hashimoto, S et al. Elevated serum levels of Wisteria floribunda agglutinin-positive human Mac-2 binding protein predict the development of hepatocellular carcinoma in hepatitis C patients. Hepatology; 2014; 60, 5 pp. 1563-1570.[COI: 1:CAS:528:DC%2BC2cXhvVansbvF] [DOI: https://dx.doi.org/10.1002/hep.27305] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25042054]