Impact of Lamivudine-Based Antiretroviral Treatment on Hepatitis B Viremia in HIV-Coinfected South Africans

This prospective study investigated the impact of lamivudine-containing antiretroviral therapy (ART) on HIV-positive patients in South Africa with baseline hepatitis B virus (HBV) infection. Follow-up samples from 56 HBV/HIV co-infected patients, 25 with occult HBV infection (OBI) and 31 with chronic HBV infection (CHB), were available for analysis. HBV viral loads were quantified at 6, 12, 18, and 24 months post-ART initiation by the COBAS TaqMan HBV Test 48 assay, and the HBV polymerase gene was amplified with an in-house nested polymerase chain reaction assay. During 24 months of lamivudine-based ART, 6 of 8 (75%) OBI and 4 of 6 (67%) CHB patients achieved undetectable levels of HBV DNA, while 2 patients had persistent HBV DNA levels ≥ 2 × 105 despite lamivudine-based ART for 24 months. HIV viremia was undetectable in all patients at 12 months, suggesting high adherence to ART. Several lamivudine-associated HBV resistance mutations, including L180M, A181T, M204I, and M204V, were observed. Sequence analysis also revealed a rare genotype G infection. While resource-limited settings may use lamivudine-based ART because of availability and low cost, antivirals with dual therapy against HBV and HIV (e.g., lamivudine and tenofovir) should always be recommended with the regular monitoring of HBV viremia levels.


Introduction
Human-immunodeficiency-virus (HIV) infection alters the course of the hepatitis B virus (HBV) disease by increasing the rate of chronicity, enhancing HBV replication, and accelerating liver-related morbidity and mortality [1][2][3][4][5][6][7]. In addition, HIV complicates the diagnosis of HBV infection by generating atypical serological presentations, such as occult HBV infection (OBI), defined as the detection of viral DNA in the blood in the absence of hepatitis B surface antigen (HBsAg) [2,8,9]. HBsAg prevalence in HIV-positive South African pregnant women ranges from 3.4% to 6.2% [10,11]. However, HBV prevalence in HIV-positive patients who have developed acquired immunodeficiency syndrome (AIDS) tends to be higher, with a study on South Africans initiating antiretroviral therapy (ART) reporting HBsAg, and HBV DNA prevalence of 22.9% and 40.6%, respectively [4].
South Africa is among the first African countries to formally adopt the World Health Organization's (WHO) Universal Test and Treat (UTT) guidelines on HIV treatment in support of the goals of the Joint United Nations Program on HIV/AIDS (UNAIDS) that 90% of people living with HIV are aware of their status, 90% of people with diagnosed infection receive ART, and 90% of individuals receiving ART achieve viral suppression. While expanding HIV treatment programs, significant challenges remain when managing HBV/HIV coinfected patients [7,12,13]. The current Southern African ART guidelines recommend testing for HBsAg before initiating ART to ensure that HIV/HBV-coinfected patients are placed on an appropriate ART regimen containing drugs that are also active against HBV, such as tenofovir and lamivudine (or emtricitabine) [7]. Importantly, given the high prevalence of OBI in HIV-positive patients, a large proportion of HIV/HBV coinfections are not identified by HBsAg screening alone. Lamivudine is commonly associated with HBV resistance mutations, whereas tenofovir is not [14]. In addition, patients with seronegative OBI (i.e., negative for all serological markers) may maintain persistent low-level HBV viremia during long-term lamivudine-containing ART, resulting in the reactivation of chronic HBV infection (CHB) and hepatic enzyme flares [9]. Furthermore, after initiation of lamivudine-containing ART, patients with overt CHB may progress to persistent OBI [9], which may be misinterpreted as HBV suppression if follow-up testing only consists of HBsAg. Although the cost of tenofovir is significantly declining, it remains relatively expensive compared to lamivudine, and access to tenofovir is limited in many African countries.
At the time of this study, HIV-positive patients initiating ART were placed on a first-line regimen consisting of lamivudine, stavudine, and efavirenz without being screened for HBsAg. This study investigated the impact of a lamivudine-containing ART regimen on HBV (both CHB and OBI) during management of HIV/AIDS patients in Pretoria, South Africa.

Study Design and Study Population
This prospective study included plasma samples collected between August 2004 and March 2007 (stored at −70 • C) [4] from patients initiating lamivudine-containing ART (i.e., lamivudine, stavudine, and efavirenz) as part of a state hospital HIV/AIDS management program conducted at the Tshepang HIV outpatient clinic of the Dr George Mukhari Academic Hospital in Pretoria. Of 78 participants in the parent study, current analysis was confined to HBV DNA-positive plasma samples from 56 HIV/AIDS patients, consisting of 32 females and 24 males with a mean age of 35 years (range: 20-51) who had been longitudinally followed for 6-24 months (Table 1).

Sequence Analysis
HBV PCR products were sequenced directly using the Spectru-Medix SCE 2410 Genetic Analysis System (Spectru-Medix LLC, State College, PA, USA). Evolutionary analysis was executed using Bayesian Evolutionary Analysis by Sampling Trees v1.8.4 [18] with an uncorrelated log-normal relaxed molecular clock, generalized time-reversible model, and nucleotide-site heterogeneity with gamma distribution. The analysis was run with a chain length of 100,000,000. Effective sample sizes were >1000, indicating sufficient sampling. A maximal clade credibility tree was selected after a 10% burn-in by TreeAnnotator v1.8.4. Nucleotide sequences were submitted to GenBank under accession numbers MH431627-MH431693. Sequences were also evaluated for HBV drug-resistance mutations and quality control using an online tool available in the rtHBV DB database through Stanford University.

HIV Viral Load and CD4+ Cell Counts
HIV viral loads were quantified during routine HIV management by the Virology Diagnostic Laboratory, National Health Laboratory Service, using the NucliSens EasyQ HIV-1 assay (BioMe'rieux, Boxtel, Netherlands). CD4+ cell counts were also determined as part of routine HIV treatment by an independent team from the Haematology Diagnostic Laboratory, National Health Laboratory Service using the Beckman Coulter MPL/CellMek fully automated flow-cytometer system (Beckman Coulter, Fullerton, CA, USA).

Clinical Responses during 12 Months of Lamivudine-Based ART
At the 12-month follow-up visit, 34 samples were available for HBV DNA testing. HBV DNA was undetectable in 50% (17/34), with levels in the remaining 50% ranging from < 10 to ≥ 2 × 10 5 IU/mL ( Table 2). Of these, 23.5% (4/17) (samples ZA265, ZA320, ZA131, and ZA024) had HBV DNA ≥ 20,000 IU/mL. Three of these had baseline viremia of > 2 × 10 5 IU/mL, while the fourth was undetectable but HBsAg-positive at baseline (Table 3) Table 3. Two of the three patients with HBV DNA ≥ 2 × 10 5 IU/mL were HBsAg-positive at 6 and 12 months, while the third patient was not tested for HBsAg at 12 months due to an insufficient sample, but was HBsAg positive at the six-month follow-up.

HBV Treatment Failure, Drug-Resistance Mutations, and Genotyping
Of the patients followed up at 6, 12, 18, and 24 months, a total of 6 patients had not suppressed HBV replication. At their last follow-up visit, 5 patients had HBV DNA > 2 × 10 5 IU/mL, while one patient had HBV DNA > 2 × 10 3 IU/mL. Analysis of polymerase gene sequences from baseline and follow-up samples showed the presence of resistance-associated mutations in 5 of these 6 patients, including L180M, A181T, M204I, and M204V (Table 5). Patient ZA113 had the L180M mutation and the L180M + M204V mutations at the 12-and 18-month follow-up visits, respectively, with HBV viremia increasing from 1.99 × 10 4 IU/mL at baseline to 7.08 × 10 6 at the 18-month follow-up. At baseline, patient ZA164 was HBsAg-positive with 7.67 × 10 3 IU/mL HBV DNA and the M204I variant. At the final visit at 6 months, HBV DNA level had dropped to 2.42 × 10 3 IU/mL, the M204I variant was not detectable, and the patient was still HBsAg-positive. Patient ZA131 was HBsAg-positive with undetectable HBV DNA levels at baseline, which increased to 3.39 × 10 2 IU/mL at 6 months with the M204I variant being detected. At this patient's final follow-up at 12 months, HBV viral load had increased considerably to 1.1 × 10 8 IU/mL, but there were no mutations detected, and the patient was still HBsAg-positive (Table 5).

Discussion
Monitoring HBV viremia during treatment is the standard of care for the timely detection of drug resistance [7,12,19,20].This study investigated the impact of lamivudine-based ART regimens on HBV viremia for the management of HIV/AIDS in South African patients in whom HIV and HBV prevalence rates are high. Several guidelines have been developed for the treatment of CHB in HBV/HIV coinfected patients, although the optimal timing of treatment initiation is still debated [19,20]. HBV viral load is a strong predictor for the development of hepatocellular carcinoma and is important in treatment success [13,[19][20][21].
The current study demonstrated that patients on ART suppressed HBV DNA to undetectable levels. At 6 months, 44.4% had viral suppression, and 50% at 12 months. The benefit of lamivudine-based ART regimens was also demonstrated at 18 and 24 months: 53.9% and 71.4% had undetectable HBV viremia, respectively. A study from China demonstrated that over 50% of HBV/HIV-coinfected patients with HBV DNA of < 20,000 IU/mL were able to achieve HBV suppression when treated either with lamivudine monotherapy or dual tenofovir and lamivudine therapy [22]. Regardless of baseline HBV DNA levels, the results of the current study are supported by other HBV/HIV coinfection studies demonstrating that 30-60% of patients achieved HBV DNA suppression with lamivudine monotherapy at 48 weeks [22][23][24][25][26].
In the current study, patients who experienced persistent HBV viremia with viral loads > 20,000 IU/mL, even after 18 and 24 months on a lamivudine-based HAART regimen, had baseline HBV DNA levels ≥ 20,000 IU/mL. Thus, high HBV baseline viremia may explain the lack of viral suppression in the absence of HBV drug-resistance-associated variants among these patients. A multinational HBV/HIV-coinfection study compared lamivudine-or emtricitabine-based monotherapy, and HBV dual therapy with tenofovir disoproxil fumarate and lamivudine or emtricitabine, and found that HBV/HIV-coinfected patients with baseline HBV DNA of < 20,000 IU/mL responded well regardless of treatment regimen. In addition, subjects in the monotherapy group who did not maintain a treatment response over time had HBV DNA ≥ 20,000 IU/mL [27]. Other studies from resource-limited settings demonstrated that HBV/HIV-coinfected patients with HBeAg-negative disease were likely to have low levels of HBV DNA, while patients with HBV DNA of < 20,000 IU/mL on lamivudine-based treatment could maintain viral suppression for nearly three years [22,23,25,27,28].
Mutational analysis for the current study found patients infected with lamivudine-resistant variants. Previous studies from South Africa reported the detection of HBV-resistance mutations associated with lamivudine in treatment-naïve and -experienced individuals with CHB and OBI [16,29]. The current study also detected the M204I variant at baseline in treatment-naïve patients, and other mutations such as A181T, L180M, or M204V during follow-up of individuals on a lamivudine-based regimen. All samples with resistance mutations had HBV viremia ≥ 7.67 × 10 3 IU/mL (Table 5). Previous studies reported the risk of developing resistance to be associated with baseline HBV DNA levels ≥ 20,000 IU/mL, especially in patients on lamivudine monotherapy, and a slow decline in HBV DNA level [20,27]. A recent study from China with a small number of HBV/HIV-coinfected patients demonstrated that lamivudine monotherapy was associated with higher rates of lamivudine-resistant HBV compared to tenofovir and lamivudine dual therapy, but did not stratify their analyses by baseline HBV DNA levels [26]. While phylogenetic analysis demonstrated the predominant HBV genotypes to be A and D in agreement with other reports from South Africa [30,31], genotype G was also detected for the first time in South Africa.
Limitations of the study include the modest sample size that was affected by attrition and missing data at some time points. In addition, the full profile of HBV serological markers could not be performed on every sample due to insufficient sample volumes. Adherence to therapy could not be confirmed, although HIV viremia was undetectable in all patients at 12 months suggesting high adherence. Nonetheless, the need to screen HIV-infected patients for HBV coinfection prior to ART initiation as part of the UTT strategy cannot be overemphasized. Finally, while some samples taken at 18 and 24 months had no resistance-associated mutations, routine population-based Sanger sequencing was conducted for this study that did not allow for the detection of minor variant populations. Thus, the magnitude of drug resistance may have been underestimated.
Many authors discourage using lamivudine as the only anti-HBV drug within an ART regimen among patients with high HBV DNA levels due to its low potency and the high risk for mutations that confer cross-resistance to other anti-HBV drugs such as emtricitabine, telbivudine, and entecavir [19,27]. However, lamivudine-based monotherapy may be considered in patients with low HBV DNA levels, as typically occurs during occult HBV infection when other HBV active drugs are not available. Others also called for strategic use of lamivudine-based therapies due to their low cost and availability in resource-limited settings [32,33]. In general, the use of lamivudine as the only anti-HBV drug within an ART regimen is strongly discouraged for HBV/HIV-coinfected individuals, but may be considered for individuals with low HBV DNA levels, as typically occurs during occult HBV infection and in the absence of liver cirrhosis.