Conference Coverage

Pronounced racial differences in HBsAg loss after stopping nucleos(t)ide



Loss of the hepatitis B surface antigen (HBsAg), a marker for functional cure of hepatitis B infection, is nearly six times more common among White patients than Asian patients following cessation of therapy with a nucleotide or nucleoside analogue, investigators in the RETRACT-B study group report.

Among 1,541 patients in a global retrospective cohort, the cumulative rate of HBsAg loss 4 years after cessation of therapy with entecavir (ETV), tenofovir disoproxil fumarate (TDF), or other nucleoside/nucleotide analogue (“nuc” or NA) was 11% in Asian patients, compared with 41% in Whites, which translated in multivariate analysis into a hazard ratio (HR) of 5.8 (P < .001), said Grishma Hirode, a clinical research associate and PhD candidate at the Toronto Centre for Liver Disease.

“On univariate Cox regression, the rate of S [antigen] loss was significantly higher among older patients, among [Whites], and among tenofovir-treated patients prior to stopping,” she said during the virtual annual meeting of the American Association for the Study of Liver Diseases.

Although NAs are effective at suppressing hepatitis B viral activity, functional cure as indicated by HBsAg loss is uncommon, Ms. Hirode noted.

“Finite use of antiviral therapy has been proposed as an alternative to long-term therapy, and the rationale for stopping nuc therapy is to induce a durable virologic remission in the form of an inactive carrier state, and ideally a functional cure,” she said.

The RETRACT-B (Response after End of Treatment with Antivirals in Chronic Hepatitis B) study group, comprising liver treatment centers in Canada, Europe, Hong Kong, and Taiwan, studies outcomes following cessation of nucleos(t)ide analogue therapy.

The investigators looked at data on 1,541 patients, including those with both hepatitis B e-antigen (HBeAg) positive and HBeAg-negative disease at the start of therapy, all of whom were HBeAg negative at the time of antiviral cessation and had undetectable serum HBV DNA. Patients with hepatitis C, hepatitis D and/or HIV co-infection were excluded, as were patients who had received interferon treatment less than 12 months before stopping.

The mean age at baseline was 53 years. Men comprised 73% of the sample. In all, 88% of patients were Asian, 10% White, and 2% other.

In patients for whom genotype data was known, 0.5% had type A, 43% type B, 11% type C, and 2% type D.

Nearly two-thirds of patients (60%) were on ETV at the time of drug cessation, 29% were on TDF, and 11% were on other agents.

In all, 5% of patients had cirrhosis at the time of nucleos(t)ide cessation, the mean HBsAg was 2.6 log10 IU/mL, and the mean alanine aminotransferase (ALT) level was 0.6 times the upper limit of normal.

The median duration of NA therapy was 3 years.

The cumulative rates of HBsAg loss over time among all patients was 3% at 1 year, 8% at 2 years. 12% at 3 years, and 14% at 4 years. Cumulative rates of antigen loss at year 4 were significantly greater for patients 50 and older vs. those younger than 50 (18% vs. 9%, respectively, P = .01), Whites vs. Asians (41% vs. 11%, P < .001), and in those who had been on TDF vs. ETV (17% vs. 12%, P = .001). There was no significant difference in cumulative HBsAg loss between patients who were HBeAg positive or negative at the start of NA therapy.

Cumulative rates of retreatment were 30% at 1 year, 43% at 2 years, 50% at 3 years, and 56% at 4 years. The only significant predictor for retreatment was age, with patients 50 and older being significantly more likely to be retreated by year 4 (63% vs. 45%, respectively, P < .001).

In a univariate model for HBsAg loss, the HR for age 50 and older was 1.7 (P = .01), the HR for White vs. Asian patients was 5.5 (P < .001), and the HR for TDF vs. ETV was 2.0 (P = .001).

A univariate model for retreatment showed an HR of 1.6 for patients 50 and older; all other parameters (sex, race, NA type, and HBeAg status at start of therapy) were not significantly different.

In multivariate models, only race/ethnicity remained significant as a predictor for HBsAg loss, with a HR of 5.8 for Whites vs. Asians (P < .001), and only age 50 and older remained significant as a predictor for retreatment, with a HR of 1.6 (P < .001).

The 4-year cumulative rate of virologic relapse, defined as an HBV DNA of 2000 IU/mL or higher) was 74%, the rate of combined DNA plus ALT relapse (ALT 2 or more times the upper limit of normal) was 56%, and the rate of ALT flares (5 or more times the upper limit of normal) was 33%.

In all, 15 patients (1%) experienced hepatic decompensation, and 12 (0.96%) died, with 9 of the deaths reported as liver-related.


Next Article: