The safety of tenofovir-emtricitabine for HIV pre-exposure prophylaxis (PrEP) in individuals with active hepatitis B

Abstract

BACKGROUND

Pre-exposure prophylaxis (PrEP) with daily oral FTC/TDF prevents HIV infection. The safety and feasibility of HIV PrEP in the setting of HBV infection was evaluated.

METHODS

The iPrEx study randomized 2499 HIV-negative men and transgender women who have sex with men to once-daily oral FTC/TDF versus placebo. Hepatitis serologies and transaminases were obtained at screening and at the time PrEP was discontinued. HBV DNA was assessed by PCR and drug resistance was assessed by population sequencing. Vaccination was offered to individuals susceptible to HBV infection.

RESULTS

Of 2499 participants, 12 (0.5%; including 6 randomized to FTC/TDF) had chronic HBV infection. After stopping FTC/TDF, 5 of 6 in the active arm had LFTs performed at follow-up. LFTs remained within normal limits at post-stop visits except for a Grade 1 elevation in 1 participant at post-stop week 12 (ALT=90, AST=61). There was no evidence of hepatic flares. PCR of stored samples showed that 2 participants in the active arm had evidence of acute HBV infection at enrollment. Both had evidence of grade 4 transaminase elevations with subsequent resolution. Overall, there was no evidence of TDF or FTC resistance among tested genotypes. Of 1633 eligible for vaccination, 1587 (97.2%) received at least 1 vaccine; 1383 (84.7%) completed the series.

CONCLUSIONS

PrEP can be safely provided to individuals with HBV infection if there is no evidence of cirrhosis or substantial transaminase elevation. HBV vaccination rates at screening were low globally, despite recommendations for its use, yet uptake and efficacy were high when offered.

Introduction

Antiretroviral pre-exposure prophylaxis (PrEP) with daily oral emtricitabine and tenofovir disoproxil fumarate (FTC/TDF) prevents acquisition of HIV infection in adults^1–4 and is recommended by the Centers for Disease Control and Prevention (CDC) in the United States as part of a comprehensive package of preventative measures for individuals at substantial risk of HIV acquisition⁵. FTC/TDF PrEP has an excellent safety profile and is well tolerated with subclinical effects on renal function⁶ and bone mineral density^7,8.

FTC/TDF is also active against hepatitis B, however, and concern has been raised that providing PrEP to individuals infected with hepatitis B could lead to hepatitis flares and injury, especially in the setting of suboptimal PrEP adherence or PrEP discontinuation⁹. This concern stems from data on HIV-HBV co-infected individuals with chronic hepatitis B infections and cirrhosis who have high rates of hepatitis flares, sometimes leading to hepatic failure, if they discontinue treatment for HIV with agents that are dually active against both HIV and HBV^10,11.

The effect of PrEP use on hepatitis B infection is not well understood, and most PrEP trials excluded participants with circulating HBV surface antigen at baseline^2,12. There are no reported cases of flares in HIV-uninfected persons with chronic hepatitis B infection who have discontinued FTC/TDF PrEP, although clinical experience with PrEP remains limited. Due to the complexity of hepatitis B management, CDC PrEP guidelines recommend that individuals positive for hepatitis B surface antigen be referred to a clinician who specializes in the treatment of hepatitis B prior to initiation of PrEP⁵.

Furthermore, individuals susceptible to hepatitis B infection should be vaccinated⁵, but recent data from the National Health and Wellness Survey suggest that only 26.3% of adults and 34.2% of men who have sex with men (MSM) in the US have ever received a hepatitis B vaccine¹³. Although sparse, data on Latin American MSM cohorts also reveal low vaccination rates: only 7% in a cohort in Argentina reported ever receiving a hepatitis B vaccine¹⁴. Even when HIV-infected individuals were found to be susceptible for hepatitis B, only 25% were vaccinated in a multicenter study of HIV-positive MSM in the US¹⁵. PrEP delivery could offer an additional opportunity to evaluate hepatitis B serostatus and increase vaccination rates.

We sought to provide information about hepatitis B global epidemiology, vaccine uptake, and the safety of PrEP use among hepatitis B infected men and transgender women who have sex with men and transgender women (MSM/TGM) in the iPrEx study.

Methods

The iPrEx study randomized 2,499 MSM to evaluate the safety and efficacy of once-daily oral FTC/TDF PrEP for HIV prevention². Study visits were scheduled every 4 weeks after enrollment². A secondary endpoint of the iPrEx study was the proportion of hepatic flares among subjects with active hepatitis B infection.

Hepatitis B surface antibody (HBsAb), hepatitis B core antibody (HBcAb), and hepatitis B surface antigen (HBsAg), and hepatitis C antibody were obtained at screening for all participants. In addition, hepatitis serologies were repeated at the time PrEP was stopped if HBV infection status was ambiguous at baseline or if study participants were susceptible at baseline. Participants at screening who were positive for HBcAb and negative for HBsAb were also tested for hepatitis B core antibody IgM (HBcAb IgM) to distinguish acute from chronic infection. Individuals were excluded from the trial if the HBcAb IgM was positive, if they had indications for HBV treatment based on local practice standards, or had elevations in liver function tests (LFTs) including alanine aminotransferase (ALT), aspartate aminotransferase (AST), or total bilirubin more than 2 fold the upper limit of normal (ULN)) within 28 days of enrollment, or signs of hepatic cirrhosis. All participants with evidence of active hepatitis B infection were excluded from participation at sites in Brazil due to local concerns about safety and treatment monitoring requirements.

Hepatitis B e antigen (HBeAg) and hepatitis B e antibody (HBeAb) were tested for enrolled participants with a positive HBsAg and when the activity of HBV remained ambiguous after initial screening (i.e., when HBcAb was positive, HBsAb was negative, and HBcAb IgM was negative). Participants with a positive HBsAg were followed with a medical history and LFTs at weeks 4, 8, 12, 16, 20, 24 weeks after stopping study drug to assess for flares. Participants were also tested for elevated LFTs if they had signs or symptoms of hepatitis during the study and when deemed clinically appropriate.

For enrolled individuals with a positive HBsAg at any time during the trial, HBV DNA was measured via polymerase chain reaction. A HBV genotype was obtained for all participants with measurable HBV DNA.

Narratives were constructed for participants who were randomized to the FTC/TDF arm and were found to have acute hepatitis B infection.

Participants found to be HBV susceptible at screening were offered vaccination starting at their enrollment visit or at any time thereafter during study participation. When warranted by local medical practice standards, sites tested for HBsAb after vaccination and could provide an additional vaccine dose if HBsAb titers were insufficient. Flares were to be managed by specialists in internal medicine at the study site or referred to local specialists. Nucleoside and nucleotide analogue reverse transcriptase inhibitors were available for treatment of hepatitis B flares at study sites.

In a subset of participants, blood plasma was tested for the presence of FTC and tenofovir (TFV), and peripheral blood mononuclear cells (PBMC) were tested for FTC triphosphate (FTC-TP) and TFV diphosphate (TFV-DP)¹⁶.

Definitions

Active hepatitis B was defined as either acute hepatitis B or chronic hepatitis B infection.

Acute hepatitis B was defined by HBcAb, HBcAb IgM and HBsAg positivity, and HBsAb negativity.

Chronic hepatitis B was defined by HBsAg and HBcAb positivity with HBsAb and HBcAb IgM negativity. Other hepatitis B designations are described in Table 1. A hepatic flare was defined as an increase in AST or ALT to > five times ULN at any visit or > 2.5 times ULN for three months, within 24 weeks of permanently discontinuing the study drug.

Table 1.

Interpretation of the Hepatitis B Panel at baseline

Tests	Results	Interpretation and management
HBsAg anti-HBc anti-HBs	negative negative negative	Susceptible. Offer vaccination series. If accepted, flag chart as "HBV vaccine series." if vaccination is refused, then test the entire panel when stopping study drug and assign HBV flare risk accordingly.
HBsAg anti-HBc anti-HBs	negative positive positive	Immune due to natural infection. Flag chart as "HBV Immune." No further testing.
HBsAg anti-HBc anti-HBs	negative negative positive	Immune due to hepatitis B vaccination. Flag chart as " HBV Immune." No further testing.
HBsAg anti-HBc anti-HBs anti-HBc IgM	positive positive negative positive	Acutely infected, Flag chart as " At Risk for HBV Flare." Not eligible for enrollment
HBsAg anti-HBc anti-HBs anti-HBc IgM	positive positive negative negative	Chronically infected Flag chart as “At Risk for HBV Flare” Test for HBeAg and anti-HBe at enrollment
HBsAg anti-HBc anti-HBs	negative positive negative	Four interpretations possible. Might be recovering from acute HBV infection, Might be distantly immune and anti-HBs test is not sensitive enough to detect low level of anti-HBs; Might be susceptible with a false positive anti-HBc; Might be undetectable level of HBsAg in the serum and the person is actually chronically infected or undetectable HBsAg due to presence of mutant HBV.
HBsAg anti-HBc anti-HBs	Other result pattern notlisted above	Possible assay error

Results

Screening

4,905 individuals were screened for iPrEx. Baseline hepatitis B serologies were obtained from 4,459 individuals (Table 2).

TABLE 2.

Hepatitis B Serologies among Individuals who Screened for iPrEx

	N	%
Individuals Screened for iPrEx	4,459
Susceptible	2,828	63.4%
Immune due to natural infection	794	17.8%
Immune due to hepatitis B vaccination	529	11.9%
Infected with Hepatitis B (acute & chronic)	61	1.4%
Acutely Infected	22	0.5%
Chronically Infected	39	0.9%
Four interpretations possible	246	5.5%

According to self-report, only 23.4% (1038/4431) had ever received a hepatitis B vaccine of which less than half 45.7% (447/977) had received all three recommended vaccines. Vaccination rates varied widely by site (range: 20.6 to 90.7%) with US and Brazilian sites reporting the highest rates (83 to 91%) and the Andean sites reporting the lowest (21 to 27%).

Enrollment

Chronic HBV cases

Among 2,499 enrolled participants, twelve (0.48%) (6 cases in each arm, mean age=31 (range 20–54)) had chronic HBV infection, all with positive HBsAg and HBcAb, and negative HBsAb and HBcAb IgM. Five were from the Andes (4 Peru, 1 Ecuador), 4 from Thailand, 1 from South Africa, and 2 from the US. All tested negative for hepatitis C antibodies.

Chronic HBV cases randomized to FTC/TDF (Figures 1,2a)

Figure 1.

Hepatitis B virus and LFTs over time for the 6 participants with chronic hepatitis B randomized to FTC/TDF

Figure 2.

a. Cases of chronic hepatitis B infection randomized to FTC/TDF

b. Cases of acute hepatitis B infection randomized to FTC/TDF

Of the 6 HBsAg+ cases in the active arm, 5 were HBeAg negative and HBeAb positive, and 1 was HBeAg positive and HBeAb negative. The HBV DNA at baseline ranged from 0 to 8,849 copies/ml in HBeAg negative participants and was 336,000,000 copies/ml in the HBeAg positive participant at baseline. Of the 5 individuals with HBV genotypes obtained, 3 had genotypes obtained after some exposure to FTC/TDF PrEP and none had evidence of antiviral drug resistance. (One participant randomized to the placebo arm had the 80P mutation with possible resistance to lamivudine and telbivudine). After drug discontinuation, 5 of 6 participants had LFTs performed at one or more post-stop follow-up visits (range 6–7 visits). LFTs remained within normal limits at all post-stop visits for 5 of 6 participants, with no evidence of hepatic flares. In one participant randomized to FTC/TDF, grade 1 LFT elevations were seen at post-stop week 12 visit, with AST 61 (ULN 41) and ALT 90 (ULN 38); at the post-stop week 24 visit, AST was 37 and ALT 60. These elevations did not meet protocol criteria for hepatic flares¹⁷. Post-stop testing was not available in 1 participant who moved away. All 6 participants had evidence of detected drug at one or more visits. One participant with HBV viral load suppressed on FTC/TDF later developed a HBV viral rebound to 24,703 copies at a time when no drug was detected prior to permanent study drug discontinuation. The transaminases from the same date were ALT=44 and AST=51. Overall, while on randomized treatment HBV viral loads ranged from 0 to 6,205 at time points with detected drug and 0 to > 3.41 × 10⁹ at time points with no detected drug.

Narratives for 2 acute HBV cases randomized to receive FTC/TDF (Figure 2b)

A 25 year-old man was randomized to FTC/TDF and dispensed study drug at enrollment. This participant was hepatitis B susceptible at screening (HBsAg, HBcAb, and HBsAb all negative) with normal LFTs. At week 4, the participant had elevated LFTs (AST=205, ALT=669), and serologies showed HBsAg, HBsAb, HBeAg negative; HBcAb, HBcAb IgM, HBeAb positive at that visit. LFTs normalized two weeks later and by week 28, he developed positive HBsAb and positive total HBcAb suggesting resolution of infection with protective immunity. Enrollment specimens were retrospectively tested and found to have an HBV viral load of 30,684, no evidence of resistance, HBsAg positive, HBcAb negative, HBsAb negative and normal LFTs.

A 35 year-old participant was randomized to FTC/TDF and dispensed study drug on the day of enrollment. At screening, the participant had normal AST/ALT (35/35), with serologies consistent with chronic HBV infection: positive HBsAg, HBcAb, and HBeAg and negative HBsAb, HBcAb IgM, HBeAb. Testing at enrollment (15 days after his screening visit), indicated that his AST/ALT had increased to 214/304 (grade 3) prior to PrEP initiation. Eight days after enrollment, repeat LFTs at an interim visit indicated grade 4 elevations (ALT 1061, AST 1473) and PrEP was stopped. At that time there was evidence of an acute HBV infection with interval appearance of detected HBcAb IgM and continuing positive HBsAg, HBcAb, positive HBeAg, and negative HBsAb and HBeAb. Eleven weeks later, his AST/ALT returned to normal, and he was restarted on study drug 3 weeks later. Retrospective analysis at enrollment revealed a HBV viral load > 3.41 × 10⁹ copies that was undetectable by week 24 and remained undetectable for the duration of participant’s time in the study. At visits at study weeks 28 and 72, HBV serology showed interval loss of HBsAg and HBeAg and interval appearance of HBeAb, yet still no detectable HBsAb. The patient self-terminated his participation in the study early at week 80 citing time constraints.

Vaccine Uptake

Of the 2,499 enrolled participants, 1,633 (65.4%) were eligible for hepatitis B immunization. 1,587 (97.2%) received at least one hepatitis B immunization, and 1,383 (84.7% of total eligible) received the complete series of three injections.

Vaccine Response

Of the 1,383 participants who received the complete hepatitis B vaccination series with three doses and post-vaccination serologic testing, 86.9% (1,021/1,175) developed immunity as measured by a positive HBsAb. 44.4% (12/27) and 74.5% (38/51) of individuals who received one or two doses respectively developed immunity. Two individuals who did not attain immunity were administered a fourth dose. Neither individual developed subsequent immunity.

Discussion

In this large multinational randomized placebo-controlled clinical trial of oral FTC/TDF for HIV PrEP, hepatitis B exposure was highly prevalent (> 25% of participants screened with evidence of prior or active infection). While over 60% of the participants were HBV susceptible at screening, fewer than 25% had reported ever receiving an HBV vaccine. Among those who had received a vaccine, fewer than half had received all 3 doses recommended by the CDC¹⁸; this deficiency varied widely among the sites and was especially high in the Andes.

In contrast to low vaccination rates prior to the trial, vaccine uptake during the trial was excellent with over 97% of eligible participants receiving at least one dose and 85% receiving the complete series; the vaccine response rate (as reflected by HBsAb seroconversion) for those completing the series was 87%, consistent with CDC data regarding vaccine immunogenicity¹⁹. The acute HBV infections that occurred during the study highlight how prevention through vaccination is imperative, and the low rates of global vaccination provide a clear opportunity for disease prevention. Engaging adult MSM/TGW in PrEP programs may offer a unique opportunity to increase vaccination rates. Interestingly, there is also some evidence from a recent retrospective study that use of hepatitis B-active anti-retroviral therapy for treatment of HIV (lamivudine, emtricitabine, and tenofovir) led to a lower incidence of hepatitis B acquisition than the use of antiretrovirals without hepatitis B activity²⁰. This effect was especially apparent when tenofovir was used, which provides another protective benefit of FTC/TDF use and is consistent with our data showing that no acute HBV infections occurred after 4 weeks on FTC/TDF; both cases of acute HBV infection that occurred in the placebo arm occurred later. These findings suggest that FTC/TDF might serve to prevent hepatitis B-susceptible people from acquiring hepatitis B before completion of the vaccine series.

Among the 12 participants enrolled in the study with chronic hepatitis B, almost all of those tested had no evidence of antiviral drug resistance (8/9). One isolate from a participant randomized to the placebo arm showed possible resistance to lamivudine and telbivudine. No participant had virus associated with resistance mutations against tenofovir, and no participants experienced a flare either during the trial or after drug discontinuation, although one participant whose LFTs were normal while on study drug developed a low-grade, asymptomatic, and self-limited elevation of the transaminases after PrEP discontinuation. The majority with chronic hepatitis B randomized to FTC/TDF (5 of 6) had visits when drug was detected followed by visits when no drug was detected, indicating inconsistent adherence, and yet, none experienced a hepatitis B flare. It is possible that flares were absent because individuals with evidence of liver dysfunction were excluded. The risk of flares with FTC discontinuation appears to be limited to people with advanced liver disease²¹.

Interestingly, tenofovir use in chronic hepatitis B in the setting of cirrhosis has led to regression of fibrosis²². Regarding the 2 cases of acute HBV infection within the FTC/TDF arm, 1 resolved with development of immunity and the other resolved clinically with loss of circulating antigen but remained HBsAb negative. It is not known whether this person may have a recurrence of HBV antigenemia after stopping PrEP.

The principal limitation of this study is the small number of enrolled participants with active HBV infection, which limits the ability to provide definitive conclusions regarding the safety of PrEP use in the context of HBV infection in MSM/TGW or conclusive guidance related to best practices. Our results are reassuring, however, in that none of the participants in iPrEx with chronic HBV infection developed a flare during intermittent use of PrEP and even after drug discontinuation.

We agree with the CDC that screening for HBsAg should occur prior to PrEP initiation; while most providers prescribing PrEP are compliant with most PrEP guidelines, only approximately 60% of providers are screening for HBV infection²³. Obtaining HBsAg alone followed by HBV DNA (when available) for individuals with a positive HBsAg may be the only required testing prior to initiating PrEP, especially in resource-limited settings in which obtaining additional serologies may be more expensive than vaccination.

We did not find that testing for HBsAb, HBcAb, HBcAb IgM, HBeAb, or HBeAg added value for the purposes of assessing PrEP eligibility. Importantly, persons with isolated HBcAb (and negative HBsAg and HBsAb) were all HBV DNA negative with normal LFTs, indicating that this common pattern of serologies was consistent immunity from prior infection rather than active infection with a HBV mutant. Individuals with HBV infection who are interested in PrEP can be safely started on FTC/TDF, which will also provide excellent activity against HBV²⁴; consistent adherence may be required to prevent development of resistance.

This experience, combined with prior evidence showing no hepatitis flares among 23 women during and after oral TDF PrEP¹², demonstrates that PrEP can often be safely started and stopped in persons with HBV infection who do not have cirrhosis or transaminase elevations at PrEP initiation. As information is limited and treatment for HBV infection is complex, referral to a specialist is appropriate when available, especially for PrEP discontinuation. At that time, a separate decision regarding continuing FTC/TDF or switching management strategies for HBV infection must be reached⁵. Such a decision would be based on further testing to evaluate cirrhosis to determine the need for ongoing hepatitis B treatment^25,26.

Individuals infected with HBV should have monitoring of LFTs (and possibly HBV viral load where available) every 6–12 months while on PrEP to ensure that the emergence of HBV resistance to FTC/TDF has not occurred, a theoretical concern with suboptimal adherence⁵. Although TDF presents a high barrier to developing HBV drug resistance especially when given as combination therapy²⁷, inconsistent adherence due to intermittent dosing may lead to drug resistance²⁸, yet no cases have been reported. We did not observe any TDF resistance during the iPrEx study during which adherence to the daily regimen occurred in a minority of participants².