HBsAg quantification in clinical practice

Pr Albert TRAN, Hôpital l’Archet 2, Nice, FranceDr Denis OUZAN, Institut Arnault Tzanck, Saint

Laurent du Var, France

Disclosure (A. Tran)

• BMS• Gilead• Merck• Abbvie• Janssen

HBV lifecycle

Chan H et al. J Hepatol 2011;55:1121

Correlation between qHBsAg, cccDNA and HBV DNA

HBsAg and genotype

The correlation between HBsAg and HBV DNA inserum wasnoticed also within thegenotypes(Fig. 3A–D),but this correlation was not significant in genotype Apatients. HBsAg levels were higher in patients carryingHBV genotype C, but this difference mainly reflected thehigher rate of HBeAg positivity. If the patients weregrouped according to HBeAg status, as shown inFigure 3E, the levels of HBsAg did not differ significantlybetween the different genotypes, neither in the HBeAg-positive nor in the HBeAg-negative group. The HBVDNA/HBsAg ratio (log1 0 value difference) was lower inHBeAg-negative patients infected with genotype A (0.63)as compared with B (1.24), C (1.23) or D (1.38,P = 0.004). Accordingly, HBsAg levels in the IC groupwere lower in genotype D than in genotype A (2.62 vs.3.38 log1 0 IU/ml, P = 0.01).

Discussion

Recent reports indicate that HBsAgquantification mightbeauseful complement to HBV DNA quantification forclinical assessment and treatment monitoring inpatients with chronic HBV infection (3, 17, 18). Therationale for this is the fact that subviral particles(HBsAg) and virions (HBV DNA) are synthesized viadifferent pathways, and therefore may reflect liver viralload and viral suppression differently.

This study confirms the previously described associa-tion between HBsAg levels and clinical stage, and alsoshowsan association with liver inflammation assessed byhistological scoring and ALT. Given that an associationbetween HBV DNA and liver damage iswell known thiswas not surprising, but some previous studies have notfound an association between HBsAg and ALT (8), andstudiesof HBsAg levelsand histological scoresarerare.

HB

V D

NA

(lo

g c

opie

s/m

l) –

HB

sAg

(log

IU/m

l)

e+AC e-AC IC

–1

0

1

2

3

4

5

6

–2

HB

V D

NA

(lo

g co

pie

s/m

l)/H

BsA

g (

log

IU/m

l)e+AC e-AC IC

5

4

3

2

1

0

1110

98765

HBeAg–HBeAg+

4321

HB

V D

NA

(lo

g co

pies

/ml)

–6cccDNA/cEq

–5

HBV DNA BL = 8.318 + 1.42 * cccDNA/cell; Rˆ2 = .609

–4 –3 –2 –1 0 1

HBeAg–HBeAg+

HB

sAg

(lo

g IU

/ml)

cccDNA/cEqHBsAg BL = 4.84 + .542 * cccDNA/cell; Rˆ2 = .411

0

1

2

3

4

5

6

–6 –5 –4 –3 –2 –1 0 1

HBeAg–HBeAg+

HB

sAg

(log

IU/m

l)

HBV DNA (log copies/ml)HBsAg BL = 2.221 + .279 * HBV DNA BL; Rˆ2 = .387

0

1

2

3

4

5

6

1 2 3 4 5 6 7 8 9 10 11

HB

sAg

(log

IU/m

l)

Age (years)HBsAg BL = 4.516 - .024 * Age; Rˆ2 = .068

0

1

2

3

4

5

6

10 20 30 40 50 60 70

(A) (D)

(B) (E)

(C) (F)

Fig. 1. Correlation between cccDNA, HBsAg and HBV DNA (A-C). Impact of clinical stage on the ratio between HBV DNA and HBsAg(D shows log ratio, Eshows ratio of log values). Fshows correlation between HBsAg and age.

Liver International (2013)© 2013 John Wiley & SonsA/S. Published by John Wiley & SonsLtd4

HBsAg, liver damage and clinical staging of chronic HBV infect ion Larsson et al.

HBsAg and genotype

The correlation between HBsAg and HBV DNA inserum wasnoticed also within thegenotypes(Fig. 3A–D),but this correlation was not significant in genotype Apatients. HBsAg levels were higher in patients carryingHBV genotype C, but this difference mainly reflected thehigher rate of HBeAg positivity. If the patients weregrouped according to HBeAg status, as shown inFigure 3E, the levels of HBsAg did not differ significantlybetween the different genotypes, neither in the HBeAg-positive nor in the HBeAg-negative group. The HBVDNA/HBsAg ratio (log1 0 value difference) was lower inHBeAg-negative patients infected with genotype A (0.63)as compared with B (1.24), C (1.23) or D (1.38,P = 0.004). Accordingly, HBsAg levels in the IC groupwere lower in genotype D than in genotype A (2.62 vs.3.38 log1 0 IU/ml, P = 0.01).

Discussion

Recent reports indicate that HBsAg quantification mightbeauseful complement to HBV DNA quantification forclinical assessment and treatment monitoring inpatients with chronic HBV infection (3, 17, 18). Therationale for this is the fact that subviral particles(HBsAg) and virions (HBV DNA) are synthesized viadifferent pathways, and therefore may reflect liver viralload and viral suppression differently.

This study confirms the previously described associa-tion between HBsAg levels and clinical stage, and alsoshowsan association with liver inflammation assessed byhistological scoring and ALT. Given that an associationbetween HBV DNA and liver damage iswell known thiswas not surprising, but some previous studies have notfound an association between HBsAg and ALT (8), andstudiesof HBsAg levelsand histological scoresarerare.

HB

V D

NA

(lo

g co

pies

/ml)

– H

BsA

g(lo

g IU

/ml)

e+AC e-AC IC

–1

0

1

2

3

4

5

6

–2

HB

V D

NA

(lo

g co

pies

/ml)/

HB

sAg

(log

IU/m

l)

e+AC e-AC IC

5

4

3

2

1

0

1110

98765

HBeAg–HBeAg+

4321

HB

V D

NA

(lo

g co

pies

/ml)

–6cccDNA/cEq

–5

HBV DNA BL = 8.318 + 1.42 * cccDNA/cell; Rˆ2 = .609

–4 –3 –2 –1 0 1

HBeAg–HBeAg+

HB

sAg

(log

IU/m

l)

cccDNA/cEqHBsAg BL = 4.84 + .542 * cccDNA/cell; Rˆ2 = .411

0

1

2

3

4

5

6

–6 –5 –4 –3 –2 –1 0 1

HBeAg–HBeAg+

HB

sAg

(log

IU/m

l)HBV DNA (log copies/ml)

HBsAg BL = 2.221 + .279 * HBV DNA BL; Rˆ2 = .387

0

1

2

3

4

5

6

1 2 3 4 5 6 7 8 9 10 11

HB

sAg

(log

IU/m

l)

Age (years)HBsAg BL = 4.516 - .024 * Age; Rˆ2 = .068

0

1

2

3

4

5

6

10 20 30 40 50 60 70

(A) (D)

(B) (E)

(C) (F)

Fig. 1. Correlation between cccDNA, HBsAg and HBV DNA (A-C). Impact of clinical stage on the ratio between HBV DNA and HBsAg(D shows log ratio, Eshows ratio of log values). Fshowscorrelation between HBsAg and age.

Liver International (2013)© 2013 John Wiley & SonsA/S. Published by John Wiley & SonsLtd4

HBsAg, liver damage and clinical staging of chronic HBV infect ion Larsson et al.

Larsson S et al. Liver Int 2013 (in press)

qHBsAg and genotypes A and D

Jaroszewicz et al. J Hepatol 2010;52:514

qHBsAg and HBV DNA in acute hepatitis B

Jaroszewicz et al. J Hepatol 2010;52:514

qHBsAg in clinical practice

• Combined with HBV DNA• Chronic Hepatitis B

– Natural history (A. TRAN)– Treatment (D. Ouzan)

Questions

• Do you measure qHBsAg in complement of HBV DNA in routine (follow up)?

– Yes– No

Case report• 32 yrs old• Female• Caucasian• No familial history of liver cancer• Liver US exam: normal• ALT 24 UI/L, AST 22 UI/L• Platelet 250 000/mm3• HBV DNA < 20 UI/mL• e negative• qHBsAg 45 UI/mL

The adding value of qHBsAg in natural history

• Prediction of inactive carrier • Prediction of HBs loss• Prediction of liver fibrosis and cellular

hepatocarcinoma

Inactive carrier and CAH HBe-

Martinot-Peignoux M et al. Ann Biol Clin 2013

qHBsAg according to diferent replicative phases of CHB

Martinot-Peignoux M et al. Ann Biol Clin 2013

Brunetto MR et al Gastroenterology 2010;139:483

qHBsAg and genotype D inactive carriers

Prediction of Inactive carrier

qHBsAgHBV-DNA

<1000 IU/mL <2000 IU/mL

HBs Carriers (IC) 209 (56)

Diagnostic accuracy 91.1%Sensibility 91.1%

Specificity 95.4%

PPV 87.9%

NPV 96.7%

Inactive carriers: qAgHBs < 1000 UI/mL

and HBV DNA < 2000 UI/mL

qHBsAg and inactive carriers

HBsAg (log IU/ml)

Immunotolerance (7)

HBeAg+ (25)

e seroconversion (17)

CHB HBeAg- (46)

Inactive carrier (22)

Chan HL, et al. Hepatology 2010

117 Patients FU 99±16 mois

Long term follow up of qHBsAg

Avec l’aimable autorisation de Denis Ouzan

qHBsAG can predict the loss of HBsAG

Tseng et al, Gastroenterology 2011

1 3 5 7 9 11 13 15 ans

020

4060

80 %

Incidence of spontaneous loss of HBsAg

HBsAg < 100 UI/ml

HBsAg100-999 UI/ml

HBsAg > 1000 UI/ml

HBsAg seroclearance stratified by HBV DNA levels in Hbe- patients infected by genotype B and C

qHBsAg (IU/mL) HBV DNA*≥ 2000 UI/mL

(n=837)

HBV DNA*Detectable-1999 UI/mL (n=935)

HBV DNA*Undetectable

(n=719)

≥ 1000 1 1 1

100-999 1.84 (1.04-3.25)

4.26 (2.49-7.28)

10.22 (3.62-28.86)

<100 6.04 (3.23-11.31)

8.88 (5.30-14.89)

38.93 (14.47-104.73)

*adjusted rate ratio (95% CI)

Liu J et al. J Hepatol 2013;58:853

Changes in qHBsAg prior to HBsAg seroclearance

Undetectable HBV DNA Detectable HBV DNA

Liu J et al. J Hepatol 2013;58:853

qHBsAG according to the fibrosis severity

Martinot-Peignoux M et al. J Hepatol 2013;58:1089Seto WK et al. Plos One 2012;7:e43087

qHBsAG 3.85 log IU (gen B or C) : NPV 91%, ≤ F1 - > F1

Ratio GGT to qHBsAg to predict significant fibrosis among e positive patients

Xun YH et al. J Gastroenterol Hepatol 2013;28:1746

HBV DNA an independent risk factor for HCC and cirrhosis : REVEAL study

• HBV DNA level was a strong predictor of HCC1 and cirrhosis2, independent of HBeAg status and serum ALT

1. Chen CJ, et al. JAMA. 2006; 295:65-73. 2. Iloeje UH, et al. Gastroenterology 2006; 130:678-686.

Year of follow-up

Cu

mu

lati

ve

inc

ide

nc

e o

f H

CC

(%

) 15%

1.3–1.4%

4%

12%

Entire cohort N=3653

Patient age distribution: 30−39 years: 33%; 40−49 years: 28%; 50−59 years: 29%; � 60 years: 10%

Incidence of HCC and qHBsAg in HBe- patients with low viral load (< 2000 IU/mL)

Tseng TC et al. Gastroenterology 2012;142:1140

Adj. HR 13.7 (95%CI 4.8-39.3)