Non-invasive serum fibrosis markers: A study in chronic hepatitis.

INTRODUCTION: Chronic hepatitis is specified as inflammatory disease of the liver lasting for more than six months. Role of noninvasive fibrosis markers as prognostication factors of the presence or absence of significant fibrosis on liver biopsy of patients with chronic hepatitis is the aim of this study.
METHODS: Two hundred twenty-one patients with chronic hepatitis involved in the study between 2011 and 2013. Routine biochemical indices and serum fibrosis markers such as aspartate aminotransferase (AST) to alanine aminotransferase (ALT) ratio (AAR), AST to platelet ratio index (APRI) and Fibrosis 4 score (FIB-4) were evaluated, and the histological grade and stage of the liver biopsy specimens were scored according to the Ishak scoring system. Diagnostic accuracies of these markers for prediction of significant fibrosis were assessed by Receiver Operating Characteristic (ROC) curve analysis.
RESULTS: Contemporaneous laboratory indices for imputing AAR, APRI, and FIB-4 were identified with liver biopsies. From all, 135 males (61.1%) and 86 females (38.9%), with mean age of 39.6±14.4 were studied. Significant correlation between stages of fibrosis and FIB-4, APRI and AAR were detected, with a correlation coefficient higher than that of other markers in the patients with Hepatitis B (r = 0.46), C (r = 0.58) and autoimmune hepatitis (r = 0.28). FIB-4 (AUROC = 0.84) and APRI (AUROC = 0.78) were superior to AAR at distinguishing severe fibrosis from mild-to-moderate fibrosis and gave the highest diagnostic accuracy.
CONCLUSION: Application of these markers was good at distinguishing significant fibrosis and decreased the need for staging liver biopsy specimens among patients with chronic hepatitis.

Introduction

Chronic hepatitis, regardless of etiology, is defined as a continuing disease without improvement for at least six months, although in many cases the diagnosis can be made before that time.[1] Of all causes, hepatitis B, hepatitis C and autoimmune hepatitis (AIH) are the most common causes of chronic hepatitis.[2-5] Chronic hepatitis is a common reason for the abnormal liver function tests[6] and forms the background for the cirrhosis progress[7] and hepatocellular carcinoma. Although liver biopsy is expensive, it needs hospitalization for at least 6-18 hours, is invasive and carries a risk of complications with an associated morbidity rate between 0.3% and 0.6%, and with a mortality rate of 0.05%, it remains the deterministic evaluation method for liver histology.[8-10] Moreover, sampling mistakes and inter and intra observer variations may result in under staging of cirrhosis, particularly macro nodular cirrhosis.[11-14] Considering these limitations, recent studies have focused on the progress of non-invasive markers as substitutes for information provided by the percutanous liver biopsy.[15-21] Most concentration has been on the aspartate aminotransferase (AST)–to-platelet ratio index (APRI)[22-24] and the FIB-4 index,[25-27] which is calculated from AST, alanine aminotransferase (ALT), platelet count, and patient age. Thus, based on the histology, this research was performed to explore the association between noninvasive diagnostic parameters and liver biopsy findings.

Materials and methods

We studied 221 patients who had undergone percutaneous liver biopsy due to documented hepatitis B or C infections and autoimmune hepatitis at Tabriz University of Medical Sciences clinic, Iran, from 2011 to 2013.

Histologic slides of all qualified patients were reread by one liver pathologist, who had no information about the clinical characteristics of the study patients, to avoid interobserver discrepancy. Biopsies were scored histologically using the criteria described for the Ishak system.[28] Fibrosis was determined as Ishak scores of three or more and cirrhosis as Ishak scores of five or six. None of patients had clinical, histological and biological proofs of chronic liver disease. Beside liver biopsies, a serum sample was taken from each person for further serological examinations. Serum biochemical determinations were done including total bilirubin, indirect bilirubin, ALT, AST, alkaline phosphatase (ALP), albumin, prothrombin time (PT), and platelet count. From these routine laboratory values, AAR (AST: ALT Ratio), APRI (AST: Platelet Ratio Index) and FIB-4 (Fibrosis 4 score) were calculated exactly as originally described.[22,26,29]

AST/ALT [29] = AST: ALT ratio

APRI [26] = AST level (/ULN*) / Platelet count (109/L) ×100

(*where ULN = upper limit of normal for that laboratory)

FIB-4 [22] = Age (years) × AST (U/L) / Platelet count (109/L) × [ALT (U/L)] 1/2

The results of quantitative variables are presented as Mean± SD and those of qualitative variables as numbers and percentages. Independent samples t test was used to compare quantitative variables, and differences between categorical variables were analyzed by chi-square or Fisher exact test. Defining the effect of different factors on histological findings in liver biopsy specimens was done by logistic regression analysis. Comparisons between the groups were performed using one-way ANOVA or Kruskal-Wallis test for unpaired data or regression analysis with the Spearman correlation coefficient test (r). The difference between the groups was considered to be significant when P ≤ 0.05.

In addition, the diagnostic value of each index to differentiate significant fibrosis (stage≥3) and mild-to-moderate fibrosis (stage 0-2) was measured by the area under the receiver operating characteristic curve (AUROC) and its corresponding 95% CI according to the procedure suggested by Hanley and McNeil.[30] All calculations were carried out using the SPSS software version 18.0.

Results

Of the 221 patients studied (mean age 39.6±14.4, range 13-83), there were 135 (61.1%) males (mean age 41.6, range 14-83), and 86 (38.9%) females (mean age 36.48, range 13-75). Ninety-five patients (mean age 40.6±13.9, range 13-69; 68 male and 27 female) had hepatitis B, 46 patients (mean age 45.7±11.1, range 23-70; 38 male and 8 female) had hepatitis C and 80 patients (mean age 35±15.3, range 13-83; 29 male and 51 female) had autoimmune hepatitis. Demographic and histological characteristics of subjects with chronic hepatitis are described in Table 1. We compared ALT, AST, ALP, total bilirubin, indirect bilirubin, albumin, PT, platelet count and mentioned serum fibrosis markers. As shown in Table 2, some of the laboratory markers were associated with liver biopsy findings to distinguish whether any other laboratory indices associated with the liver histology. Moreover, noninvasive significant fibrosis identification was done by performing the bivariate Spearman analysis on these three groups of patients considering all functional and biochemical data and looking for an association of parameters that would be able to identify this matter (Table 2). The fibrosis stage repartitions by the resembling scale were 18 (8.14%) F0; 99 (44.8%) F1; 33 (14.93%) F2; 38 (17.2%) F3; and 33 (14.93%) F4. The AAR, APRI and FIB-4 score in patients with different stages of fibrosis are shown in Table 3. Biomarker values were markedly associated with fibrosis stage levels (P < 0.01). Significant mean differences among biopsy fibrosis levels were indicated by mutually exclusive mean and its 95% CIs (P < 0.05). An increasing APRI and FIB-4 scores were noted with increasing stage of fibrosis in patients with hepatitis C (Table 4).

Table 1

Gender Specific Demographic and Histological Characteristics of Patients with Chronic Hepatitis

Groups	Gender	No. (%)	Age	Stage	Grade	AAR	APRI	FIB-4
	Male	68 (71.6)	42.26±13.79	1.99±1.67	4.40±2.83	0.91±0.38	1.24±2.05	1.91±1.88
Chronic HBV
	Female	27 (28.4)	36.37±13.68	2.04±1.80	4.37±3.04	0.97±0.49	0.81±0.55	1.42±1.18
	Male	38 (82.6)	45.24±11.99	2.05±1.48	4.47±2.52	0.91±0.33	1.12±0.93	1.93±1.19
Chronic HCV
	Female	8 (17.4)	48.00±4.03	1.75±1.58	5.75±4.40	1.14±0.54	0.96±1.03	2.42±2.34
	Male	29 (36.3)	35.38±19.41	3.59±1.47	5.45±2.89	0.90±037	1.41±1.35	1.63±1.24
Chronic AIH
	Female	51 (63.7)	34.75±12.67	3.20±1.64	6.24±3.10	1.05±0.43	2.44±3.42	2.08±1.45

Abbreviations: HBV, Hepatitis B virus; HCV, Hepatitis C Virus; AIH, Autoimmune Hepatitis.

All Values are mean ± SD; otherwise noted.

Table 2

Correlation of Grade (Modified Hepatic Activity Index) and Stage (Ishak Fibrosis Score) With Standard Laboratory Parameters and Simple Fibrosis Tests

		Bivariate Spearman’s Rank Correlation Coefficient
		HBV		HCV		AIH
Parameters	Mean ±SD	Grade	Stage	Grade	Stage	Grade	Stage
AST, IU/L	92.2 ±140.3	0.38†	0.27‡	0.52†	0.56†	0.34†	0.002
ALT, IU/L	106.49±145.5	0.27‡	0.15	0.38‡	0.49†	0.19‡	-0.14
Total bilirubin, mg/dL	2.33±4.36	0.15‡	0.22‡	0.04	0.34‡	0.20‡	0.12
Direct bilirubin, mg/dL	0.96±2.84	0.23	0.23‡	-0.001	0.31‡	0.19‡	0.11
ALP, IU/L	369.12±339.93	0.36†	0.28‡	0.01	0.25‡	0.23‡	0.26‡
Platelet count (109/L)	194.3±49.5	-0.52†	-0.40†	-0.30‡	-0.35‡	-0.38†	-0.31‡
PT, s	13.83±1.45	0.38†	0.27‡	0.1	0.32‡	0.17	0.20‡
Albumin, g	3.99±0.67	-0.28‡	-0.34†	-0.002	- 0.25‡	-0.09	-0.11
AAR	0.96±0.41	0.13	0.16	0.38‡	0.25	0.25‡	0.28‡
APRI	1.46±2.17	0.54†	0.41†	0.52†	0.58†	0.44†	0.07
FIB-4	1.88±1.55	0.59†	0.46†	0.51†	0.50†	0.38†	0.05

Abbreviations: ALT, Alanine aminotransferase; AST, Aspartate aminotransferase; ALP, Alkaline Phosphatase; PT, Prothrombin Time; AAR, AST/ALT Ratio; APRI, AST to Platelet Ratio Index; FIB-4, Fibrosis 4 score;HBV, Hepatitis B virus; HCV, Hepatitis C Virus; AIH, Autoimmune Hepatitis.

† P < 0.001.

‡ P < 0.05.

Table 3

Correlation of Chronic Hepatitis Stage by Invasive (Liver Biopsy Staging) and Noninvasive (ARR, APRI and FIB-4) Scores

Degree of Fibrosis (Stage) (N)	Ishak	Mean AAR (95% CI)	Mean APRI (95% CI)	Mean FIB-4 (95% CI)
No fibrosis (F0) (n = 18)	Stage 0	1.04 (0.83-1.25)	1.08 (0.21-1.95)	1.50 (0.95-2.04)
Fibrous portal expansion (F1) (n = 99)	Stage 1, 2	0.87 (0.80-0.93)	1.09 (0.70-1.47)	1.41 (1.15-1.66)
Few bridges or septa (F2) (n = 33)	Stage 3	0.93 (0.80-1.05)	1.96 (0.69-3.22)	2.07 (1.48-2.65)
Numerous bridges or septa (F3) (n = 38)	Stage 4	0.96 (0.80-1.12)	1.75 (1.36-2.13)	2.23 (1.81-2.66)
Cirrhosis (F4) (n = 33)	Stage 5, 6	1.23 (1.06-1.40)	1.94 (1.24-2.64)	2.89 (2.16-3.61)

Abbreviations: AAR, AST/ALT Ratio; APRI, AST to Platelet Ratio Index; FIB-4, Fibrosis 4 score.

Table 4

Correlation of Chronic Hepatitis Stage by Invasive (Liver Biopsy Staging) and Noninvasive (ARR, APRI and FIB-4) Scores in All Three Groups

Groups	Stage 0	Stage 1, 2	Stage 3	Stage 4	Stage 5, 6
Chronic HBV
AAR	1.05 (0.75-1.34)	0.85 (0.76-0.94)	0.77 (0.57-0.98)	1.19 (0.74-1.65)	1.05 (0.77-1.33)
APRI	0.70 (0.53-0.87)	0.99 (0.44-1.55)	0.67 (0.42-0.92)	1.26 (0.96-1.55)	2.45 (0.71-4.2)
FIB-4	1.43 (0.66-2.2)	1.31 (0.87-1.76)	1.49 (0.69-2.29)	2.39 (1.75-3.04)	3.95 (2.49- 5.4)
Chronic HCV
AAR	1.07 (0.41-1.72)	0.86 (0.74-0.98)	0.99 (0.58-1.41)	0.92 (0.61-1.23)	1.63 (0.16-3.1)
APRI	0.4 (-0.002-0.81)	0.77 (0.55-0.99)	1.21 (0.22-2.19)	2 (0.93-3.08)	2.93 (0.36-5.5)
FIB-4	1.26 (0.57-1.95)	1.46 (1.14-1.78)	1.92 (0.81-3.02)	3.52 (2.89-4.14)	5.32 (-0.56-11.2)
Chronic AIH
AAR	0.93 (-2.04-3.91)	0.92 (0.77-1.07)	0.99 (0.79-1.19)	0.85 (0.67-1.02)	1.27 (1.05-1.5)
APRI	4.77 (-35.26-44.8)	1.73 (0.49-2.98)	2.98 (0.53-5.44)	1.94 (1.3-2.58)	1.49 (0.72-2.26)
FIB-4	2.39 (-10.32-15.1)	1.58 (1.18-1.98)	2.47 (1.43-3.5)	1.79 (1.18-2.4)	1.89 (1.32-2.47)

Abbreviations: HBV: Hepatitis B virus; HCV: Hepatitis C Virus; AIH: Autoimmune Hepatitis.

All Values are Mean (95% CI); otherwise noted.

Noninvasive indexes such as FIB-4, APRI and AAR were associated markedly with the stage of fibrosis, with a correlation coefficient higher than that of other markers in the patients with hepatitis B (r= 0.46), C (r= 0.58) and autoimmune hepatitis (r=0.28). Weak to moderate correlations were found among Ishak stages of fibrosis versus APRI, FIB-4, AST, Direct bilirubin, Total bilirubin, PT and ALP in both patients with hepatitis B and C infections. In all three groups of patients there were also moderate opposite relationship between platelet count and Ishak stage of fibrosis. Hepatic fibrosis weakly correlated with PT, ALP and AAR in patients with AIH (Table 2).

Prediction of significant fibrosis was done by constructing ROC curves measuring the diagnostic precisions of AAR, APRI, FIB- 4 and platelet count ( Fig. 1). Exceptional diagnostic precision of APRI over AAR for prediction of notable fibrosis concluded by comparing AUROCs for continuous variables by the procedure proposed by Hanley and McNeil,[30] especially in patients with hepatitis B and C (Table 5). Optimal cutoff point for AAR to determine considerable fibrosis in hepatitis C was ≥ 0.7, with a sensitivity of 87% and specificity of 39%.

Fig. 1

Table 5

Performance of Simple Fibrosis Tests for Prediction of Significant Fibrosis (F3–F6) in Patients With Chronic Hepatitis

Groups	HBV			HCV			AIH
Ishak FibrosisScore vs.	Cutoff	Se/Sp	AUROC(95% CI)	Cutoff	Se/Sp	AUROC(95% CI)	Cutoff	Se/Sp	AUROC(95% CI)
AST	52.5	66/70	0.7(0.59-0.81)	42.5	87/61	0.82(0.69-0.94)	77.5	53/65	0.54(0.4-0.68)
ALT	67.5	47/65	0.6(0.48-0.72)	47	87/61	0.73(0.58-0.88)	84.5	51/61	0.49(0.36-0.63)
PT	13.35	53/75	0.63(0.51-0.76)	13.5	60/77	0.66(0.48-0.83)	13.9	56/61	0.61(0.48-0.73)
Platelet count	221	22/60	0.29(0.18-0.4)	139	80/16	0.27(0.12-0.42)	238.5	12/87	0.36(0.22-0.49)
AAR	1.37	22/94	0.55(0.43-0.68)	0.72	87/39	0.64(0.46-0.81)	1.18	33/91	0.57(0.44-0.7)
APRI	0.96	66/89	0.76(0.65-87)	1.06	73/84	0.84(0.72-0.96)	1.12	58/65	0.57(0.43-0.71)
FIB-4	1.36	81/76	0.78(0.68-0.89)	2.76	67/94	0.84(0.71-0.98)	1.82	47/74	0.55(0.41-0.68)

Abbreviations: ALT, Alanine aminotransferase; AST, Aspartate aminotransferase; PT, Prothrombin Time; AAR, AST/ALT Ratio; APRI, AST to Platelet Ratio Index; FIB-4, Fibrosis 4 score; Se, Sensitivity; Sp, Specificity.

Receiver operating characteristic (ROC) curves of four simple noninvasive tests for prediction of significant fibrosis (F3-F6) according to the Ishak system in the (A) Hepatitis B (B) Hepatitis C and (C) Autoimmune hepatitis patients. An AUC of 1.0 is characteristic of an ideal test, whereas an AUC of 0.5 or less indicates a test of no diagnostic value. AAR: AST/ALT ratio; APRI: AST to-platelet ratio index; FIB-4: Fibrosis 4 Score.

Discussion

At present, pathological examination of liver puncture tissue is the way to diagnose liver fibrosis. Usage of liver biopsy because of its invasive trait and sampling errors is still limited in clinical practice, although it is the gold standard.[31,32] Searching for noninvasive markers to diagnose liver fibrosis has demanded great attention.[33-35] Comparing pathological classification with some non-invasive markers to appraise importance of these markers in expressing pathological differences in three different groups of patients with chronic hepatitis was the main goal of this study.

In a study of long-term outcome of chronic hepatitis B based on histological grade and stage it was concluded that the serum ALT level at the time of liver biopsy was significantly correlated with the grades of lobular and porto-periportal activity. The results proved that histological grade and stage, and biochemical profile during follow-up were important prognostic factors in patients with chronic hepatitis B.[36] In an Italian multi-center study about clinical course and outcome, it was concluded that age, AST, ALT, PT, Albumin level and total bilirubin were prognostic factors,[37] although present study just suggested PT and ALP as prognostic factors in patients with AIH.

The relationship between platelet count and liver fibrosis in patients with chronic hepatitis C has been attractive topic for researchers.[18,22] However, prognostication of liver fibrosis with diagnostic value of platelet count per se have been assessed in only a few studies on these patients.[38-40] In contrast with our findings, diagnostic preciseness of prothrombin time and platelet count in patients with chronic hepatitis C was assessed by Myers et al[40] and they reported an AUROC of 0.67 for platelet count for prediction of F2-F4 fibrosis (METAVIR system).[40]

Consistent with our findings, in a large observational real-world cohort of chronic hepatitis C patients, FIB-4 and APRI were superior to AAR at distinguishing severe fibrosis from mild-to-moderate fibrosis.[41] Low diagnostic accuracy of AAR in predicting significant fibrosis in patients with hepatitis C was reported by Lackner et al[42] (AUROC 0.57).[42] In this study, we found that the optimal cutoff AAR value for diagnosing significant fibrosis in hepatitis C was ≥ 0.7, with a sensitivity of 87% and specificity of 39%. These results are in contrast with previous findings by Fouad et al;[43] who recommended an AAR value ≥ 1.2 as a cutoff value for diagnosing fibrosis.[43]

This study revealed a significant correlation between APRI and both the stage of liver fibrosis and the grade of activity. The optimal cutoff APRI value for the diagnosis of fibrosis in hepatitis C group was ≥ 1.06 that was consistent with findings by Hsieh et al;[44] who reported cutoff values of ≥1 with a sensitivity of 75.5% and specificity of 41.5%. Furthermore our study is consistent with that of Hongbo et al, [45] in which the area under the ROC of APRI was also modest (0.76) in patients with hepatitis B infection.[45] Yilmaz et al[15] reported that the APRI had an acceptable accuracy for the assessment of liver fibrosis in patients with chronic hepatitis C, but not in those with chronic hepatitis B.[15] Parsian et al[46] reported differences between severe and mild liver fibrosis by APRI with 29.0% sensitivity and 22.0% specificity.[46]

In another study Zhang et al[47] assessed the diagnostic value of FIB-4 in 212 patients with chronic hepatitis B by comparing their results with histological features.[47] The AUROC of FIB-4 for significant fibrosis was 0.733. Mahassadi et al[48] had conducted a prospective cohort study to determine the diagnostic accuracy of APRI, AAR, AP and FIB-4 index for the prediction of significant fibrosis or cirrhosis in 117 patients with chronic hepatitis B and APRI and FIB-4 index ruled out significant fibrosis with high specificity of 84.7% and 86.1%, respectively.[48] Distinguishing severe stages (F3–F4) from low or moderate stages (F0–F2) of fibrosis with high FIB-4 scores (e.g. ≥ 2.25) was studied in some researches to date.[49,50] The FIB-4 index may be of value in several respects including simple calculations and no standardization, immediate results during the patient visit and its inexpensiveness with no additional costs.[26]

Conclusion

In conclusion, the current study demonstrated that these low-cost-and-easy-to perform serum fibrosis markers, especially APRI and FIB-4, were simple methods that correlated well with the stages of fibrosis in patients with chronic hepatitis. The combination of these non-invasive markers may replace the requirement for liver biopsy. Therefore further studies with more patients are needed to evaluate these markers.

Competing interests

The authors express no opposition of profits.

Ethical issues

There is none to be declared.

What is current knowledge?

√ Hepatitis B, hepatitis C and autoimmune hepatitis (AIH) are the most common causes of chronic hepatitis.

√ Liver biopsy is the deterministic evaluation method for liver histology.

√ Usage of liver biopsy because of its invasive trait and sampling errors is still limited in clinical practice.

What is new here?

√ Low-cost-and-easy-to perform serum fibrosis markers, especially APRI and FIB-4, were simple methods that correlated well with the stages of fibrosis in patients with chronic hepatitis.

√ FIB-4 and APRI were superior to AAR at distinguishing severe fibrosis from mild-to-moderate fibrosis.

√ Significant correlation revealed between APRI and both the stage of liver fibrosis and the grade of activity.