Non-invasive Fibrosis Markers for Assessment of Liver Fibrosis in Chronic Hepatitis Delta

Çağdaş Kalkan; Yusufcan Yılmaz; Beyza Doğanay Erdoğan; Berna Savaş; Esra Yurdcu; Aysun Çalışkan; Onur Keskin; Genco Gencdal; Müjdat Zeybel; Murat Törüner; A. Mithat Bozdayi; Ramazan Idilman; Cihan Yurdaydin

Disclosures

J Viral Hepat. 2023;30(5):406-416.

Abstract and Introduction

Abstract

Assessment of liver fibrosis by non-invasive means is clinically important. Studies in chronic hepatitis delta (CHD) are scarce. We evaluated the performance of eight serum fibrosis markers [fibrosis-4 score (FIB-4), aspartate aminotransferase (AST) to alanine aminotransferase (ALT) ratio (AAR), age-platelet index (API), AST-to platelet-ratio-index (APRI), Goteborg University Cirrhosis Index (GUCI), Lok index, cirrhosis discriminant score (CDS) and Hui score] in CHD and chronic hepatitis B (CHB). Liver stiffness was assessed by transient elastography (TE) in CHD. The ability of fibrosis markers to detect significant fibrosis and cirrhosis were evaluated in 202 CHB and 108 CHD patients using published and new cut-offs through receiver operating characteristics (ROC) analysis. The latter was also applied to obtain cut-offs for TE. APRI, Fib-4, API and Hui score were assessed for significant fibrosis, and APRI, GUCI, Lok index, CDS and AAR for cirrhosis determination. Fibrosis markers displayed weak performance in CHB for significant fibrosis with area under ROC (AUROC) curves between 0.62 and 0.71. They did slightly better for CHD. TE displayed an AUROC of 0.92 and performed better than serum fibrosis markers (p < 0.05 for fibrosis markers). For cirrhosis determination, CDS and Lok Index displayed an AUROC of 088 and 0.89 in CHB and GUCI, Lok index and APRI displayed AUROCs around 0.90 in CHD. TE displayed the best AUROC (0.95). Hence TE is superior to serum fibrosis markers for diagnosing significant liver fibrosis and cirrhosis. GUCI, Lok index and APRI displayed a reasonable performance in CHD, which needs further confirmation.

Introduction

Accurate assessment of liver fibrosis in patients with chronic viral hepatitis is important clinically as patients with advanced liver disease may be prioritized to treatment and/or in such patients screening strategies for hepatocellular carcinoma need to be implemented. Although liver biopsy is considered the gold standard to assess liver disease severity, due to its invasive nature and potential adverse events associated with it, it is disliked by patients.^[1] Further, sampling error may lead to false clinical impressions and it is patients with advanced liver disease where sampling errors appear to be more frequent, which could be actually harmful for patients.^[1,2] Non-invasive assessment of liver disease severity by use of serum fibrosis markers or by physical methods, such as transient elastography (TE), has been thoroughly investigated in chronic hepatitis C (CHC) and B (CHB), but little is known in chronic hepatitis delta (CHD). In two recent studies from the United States and Germany, several serum fibrosis markers were assessed in CHD.^[3,4] Both studies were in agreement that serum fibrosis markers performed less well in CHD compared to CHB and CHC. The use of TE for CHD has also been investigated recently.^[5]

Treatment of CHD consists of the use of IFNs with viral response rates obtained in roughly 25% of patients.^[6] We are at a stage where new treatments are being tested in CHD.^[7] A post-treatment beneficial viral and biochemical flare has been described with one of these new compounds^[8] which further underlines the importance of monitoring the severity of liver disease routinely to avoid decompensation.

In the current study, several serum fibrosis markers were assessed in patients with CHD and compared to chronic hepatitis B (CHB) patients. Further, TE was assessed only in patients with CHD.

1 2 3 4

Next Section

Abstract and Introduction
Patients and Methods
Results
Discussion
References

Table 1. Reference cut-offs reported in the literature used for differentiation of cirrhosis from non-cirrhosis and mild chronic hepatitis from severe chronic hepatitis.
Test	F3–F6	F5–F6
Fib-4	>3.25	–
APRI	>1.5	≥2
AAR	-	≥1
API	≥6	–
HUI Score	>0.5	–
Lok Index	–	>0.5
GUCI Score	–	>1
CDS	–	>7

Abbreviations: AAR, AST to ALT ratio; ALT, alanine aminotransferase; API, the age—platelet index; APRI, the AST platelet ratio index; AST, Aspartate aminotransferase; CDS, the cirrhosis discriminant score; FIB-4, the fibrosis 4 score; GUCI, the Goteborg University Cirrhosis Index.

Table 2. Patient characteristics
	CHB (n = 202)	CHD (n = 108)	p
Age (years), median (range)	45.48 (18–74)	52 (25–76)	<0.001
Male. n (%)	142 (70,3)	72 (66,7)	0.51
BMI (x ± SD)	27,11 ± 4,21	25,95 ± 3,51	0.065
ALT (IU/L, x ± SD)	86,88 ± 127,51	36 ± 33,36	0.001
AST (IU/L, x ± SD)	56,26 ± 54,32	33,5 ± 41,20	0.057
Platelets (×10⁹/L, x ± SD)	194,16 ± 76,36	170,50 ± 75,69	<0.001
Total bilirubin (mg/dl) (x ± SD)	0,90 ± 0,67	0,80 ± 1,53	0.007
Albumin (g/dl) (x ± SD)	4,15 ± 0,50	4,3 ± 0,61	0.53
GGT (IU/L x ± SD)	48,5 ± 69,23	30 ± 89,20	0.04
Alk. Ph. (IU/L x ± SD)	80 ± 39,39	79,50 ± 61,09	0.007
INR (x ± SD)	1,04 ± 0,11	1 ± 0,24	0.1
HBV DNA (log10 IU/ml) (x ± SD)	6,43 ± 1,80	1,32 ± 1,93	<0.001
HBeAg (%)	74 (37,9) N = 195	8 (7,8) N = 102	<0.001
Patients with cirrhosis, n (%)	35 (17.3)	38 (35.3)	0.01
HAI 8.5 ± 3.6 (n = 190) 8.0 ± 3.5 (n = 74)
0–4	26 (13,7%)	15 (20,3%)	0.19
5–8	71 (37,4%)	25 (33,8%)
9–12	64 (33,7%)	27 (36,5%)
13–18	29 (15.3%)	7 (9,5%)

Abbreviations: Alk. Phos, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; CHB, chronic hepatitis B; CHD, chronic hepatitis D; GGT, gamma glutamyl transpeptidase.

Table 3. Evaluation of published cut-offs (A) for several serum fibrosis markers for advanced fibrosis diagnosis in CHD patients and performance of these serum fibrosis markers with adjusted cut-offs and transient elastography for detecting advanced fibrosis using receiver operating characteristics curves and the Youden Index (B; Significant fibrosis: F3–6 according to Ishak et al).
A.
Tests	Cut-off	AUROC	Sensitivity	Specificity	PPV	NPV	DOR	CC
APRI	>1.5	0.769 (0.42–0.89)	0.32 (0.1–0.75)	0.91 (0.6–1.00)	0.41 (0.2–0.79)	0.88 (0.6–1.00)	4.76 (2.1–9.7)	0.52
FIB-4	>3.25	0.808 (0.51–0.99)	0.41 (0.2–0.88)	0.94 (0.43–1.00)	0.45 (0.17–0.91)	0.93 (0.7–1.00)	10.9 (3.6–19.6)	0.59
API	≥6	0.699 (0.4–0.99)	0.62 (0.33–0.97)	0.66 (0.3–1.00)	0.47 (0.2–0.8)	0.78 (0.56–1.00)	3.17 (1.4–11.2)	0.63
HUI Score	>0.5	0.733 (0.42–0.97)	0.52 (0.22–0.81)	0.86 (0.51–1.00)	0.48 (0.22–0.95)	0.88 (0.52–0.9)	6.65 (3.8–17.1)	0.63
B.
Tests	Cut-off	AUROC	Sensitivity	Specificity	PPV	NPV	DOR	CC
APRI	>0.60	0.769 (0.1–0.98)	0.68 (0.41–0.99)	0.80 (0.15–0.98)	0.56 (0.15–0.89)	0.87 (0.25–1.00)	7.21 (1.6–25.4)	0.72
FIB-4	>1.48	0.808 (0.47–0.99)	0.78 (0.14–0.96)	0.77 (0.11–0.94)	0.64 (0.12–0.89)	0.87 (0.51–1.00)	12.2 (4.3–30.6)	0.78
API	≥4.5	0.699 (0.41–0.99)	0.77 (0.1–0.95)	0.54 (0.19–0.89)	0.54 (0.13–0.95)	0.77 (0.23–0.88)	3.93 (1.5–10.6)	0.69
HUI Score	>0.54	0.733 (0.11–0.99)	0.52 (0.11–0.95)	0.94 (0.18–1.00)	0.50 (0.1–0.88)	0.95 (0.44–1.00)	18.0 (4.4–28.9)	0.66
FibroScan	8.5	0.915 (0.2–1.00)	0.87 (0.12–1.00)	0.89 (0.2–1.00)	0.78 (0.4–1.00)	0.94 (0.48–1.00)	51.7 (13.5–95.1)	0.88

Abbreviations: API, the age—platelet index; APRI, the AST platelet ratio index; AUROC, area under receiver operating characteristic; CC, Correct Classification; DOR, Diagnostic Odd's Ratio; FIB-4, the fibrosis 4 score; NPV, negative predictive value; PPV, positive predictive value.

Table 4. Evaluation of published cut-offs (A) for several serum fibrosis markers for cirrhosis diagnosis in CHD patients and performance of the same serum fibrosis markers with adjusted cut-offs and transient elastography (B) for detecting cirrhosis using receiver operating characteristics curves and the Youden Index. For FibroScan, the performance of the recently suggested cut-off of 14.0 is also provided.
A.
Tests	Cut-off	AUROC	Sensitivity	Specificity	PPV	NPV	DOR	CC
AAR	≥1	0.682 (0.15–0.98)	0.73 (0.45–0.95)	0.57 (0.22–0.89)	0.53 (0.3–0.88)	0.77 (0.5–1.00)	3.58 (1.5–19.1)	0.63
CDS	>7	0.840 (0.4–1.00)	0.41 (0.3–0.85)	0.98 (0.4–1.00)	0.72 (0.14–1.00)	0.94 (0.43–1.00)	34.1 (10.6–84.3)	0.76
Lok Index	>0.5	0.898 (0.5–1.00)	0.71 (0.6–1.00)	0.94 (0.5–1.00)	0.83 (0.47–1.00)	0.88 (0.4–1.00)	38.4 (10.8–98.1)	0.85
GUCI Score	>1	0.919 (0.7–1.00)	0.71 (0.42–1.00)	0.92 (0.4–0.96)	0.83 (0.4–0.99)	0.85 (0.52–0.95)	28.2 (10.2–88.5)	0.84
APRI	≥2	0.894 (0.6–1.00)	0.41 (0.35–0.92)	0.98 (0.22–1.00)	0.72 (0.32–0.95)	0.94 (0.42–1.00)	34.1 (10.5–69.7)	0.76
D4FS	>7.8	0.845 (0.1–0.99)	0.47 (0.12–0.88)	0.57 (0.12–0.98)	0.6 (0.20–0.95)	0.97 (0.21–0.99)	2.65 (1.6–144.68)	0.59
B.
Tests	Cut-off	AUROC	Sensitivity	Specificity	PPV	NPV	DOR	CC
AAR	≥0.97	0.682 (0.3–0.95)	0.80 (0.4–0.9)	0.56 (0.3–1.00)	0.54 (0.15–0.86)	0.81 (0.55–1.00)	5.16 (2.5–19.8)	0.65
CDS	>5.5	0.840 (0.51–0.99)	0.80 (0.3–0.9)	0.78 (0.5–1.00)	0.70 (0.4–1.00)	0.86 (0.4–1.00)	14.4 (7.1–23.8)	0.79
Lok Index	>0.37	0.898 (0.1–0.99)	0.90 (0.25–0.99)	0.76 (0.2–1.00)	0.71 (0.2–1.00)	0.92 (0.4–1.00)	29.6 (10.1–52.7)	0.82
GUCI Score	>0.64	0.919 (0.21–1.00)	0.90 (0.3–1.00)	0.81 (0.5–1.00)	0.76 (0.5–1.00)	0.93 (0.4–1.00)	39.3 (13.3–49.5)	0.85
APRI	>0.62	0.894 (0.3–1.00)	0.90 (0.3–1.00)	0.78 (0.1–0.99)	0.73 (0.5–1.00)	0.92 (0.26–1.00)	32.4 (10.6–50.7)	0.83
D4FS	>0.38	0.845 (0.16–1.00)	0.63 (0.14–0.98)	0.88 (0.12–0.99)	0.85 (0.2–0.99)	0.67 (0.11–1.00)	16.2 (1.65–54.8)	0.76
FibroScan	10.0	0.945 (0.25–1.00)	0.95 (0.3–1.00)	0.75 (0.2–0.95)	0.71 (0.35–0.95)	0.96 (0.4–1.00)	57.3 (20.9–78.3)	0.83
FibroScan	14.0	0.945 (0.28–1.00)	0.76 (0.15–0.99)	0.90 (0.4–1.00)	0.84 (0.6–1.00)	0.85 (0.4–1.00)	29.5 (10.8–39.1)	0.84

Abbreviations: AAR, AST to ALT ratio; ALT, alanine aminotransferase; API, age—platelet index; APRI, the AST platelet ratio index; AST, the Aspartate aminotransferase; AUROC, area under receiver operating characteristic; CC, correct classification; CDI, the cirrhosis discriminant score; D4FS, delta-4 fibrosis score; DOR, Diagnostic Odd's Ratio, FIB-4, the fibrosis 4 score; GUCI, Goteborg University Cirrhosis Index; NPV, negative predictive value; PPV, positive predictive value.