Unexpected increase of Hepatitis B Surface Antigen (HBsAg) levels during treatment with nucleos(t)ide analogous 12–24 months before virological breakthrough and resistance

Introduction: HBsAg levels differ in the course of chronic hepatitis B virus infection and may predict the response to treatment with interferon alpha. Recently, we have shown that significant HBsAg decline during treatment with nucleos(t)ide analogues (NA) is associated with later HBsAg loss (Jaroszewicz et al., AVT 2011; GASL 2011). So far there is little data if resistance to NA affects HBsAg levels.

Methods: In this retrospective study, HBsAg was quantified longitudinally in 30 patients with virological breakthrough (VB) due to NA resistance (VB-R: 27 LAM, 3 ADV). We have selected a control group of 30 patients on NA therapy (4 LAM, 3 ADV, 6 LAM&ADV de novo, 13 ETV, 4 TdF) without VB who have been treated for at least 29 months (median time until resistance in group VB-R). In addition, 5 patients were analyzed who had VB due to non-compliance. HBsAg was quantified in all patients using Roche Elecsys II, some samples were additionally tested by Abbott Architect.

Results: HBsAg level increased significantly in patients with later VB and resistance. Surprisingly, the increase of HBsAg peaked 12–24 months before VB, despite virological response at that time. 20/30 patients showed an HBsAg increase to >120% 12–24 months before VB. A second smaller peak was detected at the time of VB. In contrast, HBsAg slowly declined in the control group and only 2 patients without VB showed an HBsAg increase to >120%. Also, patients with VB due to non-compliance demonstrated a decrease in HBsAg until VB.

Conclusion: HBsAg quantification may be a useful marker to monitor the outcome of NA therapy. A slow HBsAg decline can be observed in most patients with maintained on-treatment virological response and a significant decline may even indicate later HBsAg loss. However, an HBsAg increase may predict resistance to NA treatment more than 12 months before virological breakthrough. The surprisingly early HBsAg increase 12–24 months before VB is unexpected and demands further research.

Table: HBsAg levels during NA treatment in patients with virological breakthrough due to resistance (VB-R) and in controls without VB and non-compliance (NC).

	Baseline (%)	24 mo before VB (%)	12–24 mo before VB (%)	6–12 mo before VB (%)	3–6 mo before VB (%)	VB (med 29 mo after baseline) (%)	End of FU (%)
n=30 VB-R	100	133±26.8	157±25.4*	115±20.2**	83.2±18.2*	116±17.2**	63.4±10.3
n=30 Control	100	78.6±6.95	55.0±6.33	43.7±5.68	44.7±7.10	35.5±4.74	26.8±12.1
n=5 NC	100	n.a.	57.7±16.6	53.3±14.6	16.8±8.50	59.7±19.7	n.a.

TS: Baseline HBsAg (before treatment)=100%, HBsAg at other time-points (mo=months before VB) are shown as Mean±SEM% from baseline. VB (virological breakthrough) or equivalent time (29 months); *p<0,05;**p<0,001 (VB-R vs. control)

Hepatitis B - Nucleos(t)ide Analogues - Quantitative HBsAg - Resistance