Determining the antiviral activity of tenofovir disoproxil fumarate in treatment-naive chronically HIV-1-infected individuals

Objective: To assess the efficacy of tenofovir disoproxil fumarate (TDF) monotherapy by following the initial rate of decline in plasma viral load, which is a measure of the efficacy of therapy in blocking viral replication.

Design: An open-label, single-site study of TDF monotherapy in 10 antiretroviral drug-naive chronically HIV-1-infected individuals.

Methods: Antiviral responses were assessed at baseline and during 21 days of monotherapy with TDF by measuring plasma HIV-1 RNA levels. The rate of change in HIV-1 RNA from baseline was determined both by linear regression and by fitting to a published model. Slopes were compared with those previously determined for ritonavir monotherapy.

Results: Over 21 days, mean plasma HIV-1 RNA levels in the TDF-treated patients fell 1.5 log₁₀ copies/ml (range, 0.7-2.0). The initial rates of decline in plasma HIV-1 RNA in the 10 TDF-treated patients and in 25 protease inhibitor-naive subjects treated with ritonavir monotherapy were nearly identical.

Conclusions: The reduction in plasma HIV-1 RNA with TDF montherapy was comparable with the decline observed in previous studies of protease inhibitor monotherapy. TDF is a potent antiretroviral agent and has comparable inherent antiviral activity with that of ritonavir, a potent protease inhibitor. These data support further study of TDF-based regimens in simplified combinations of antiviral agents as initial treatment for chronic HIV-1 infection.

Introduction

Determining the inherent potency of an antiretroviral agent in the era of highly active antiretroviral therapy (HAART) has become challenging. The rigors of drug development have been appropriately balanced with the interests and concerns of the HIV-1-infected patient, such that combination therapies incorporating new agents are routinely compared with combinations identified as 'the standard of care'. Such strategies avoid the rapid emergence of drug-resistant variants and allow for an assessment of tolerability and durability of antiviral effect [¹]. Nevertheless, for antiretroviral drugs with limited resistance-selecting potential, short courses of monotherapy may be a safe and useful method to reveal a drug's inherent potency, information critical to guiding the use of any new agent.

Tenofovir disoproxil fumarate (TDF), a prodrug of the nucleotide reverse transcriptase inhibitor (NtRTI) tenofovir (9-[2-(phosphonomethoxy)propyl]adenine), was recently approved for use in the combination therapy of HIV-1 infection, based on its activity in drug-experienced subjects [²]. The drug has been compared with stavudine as a component of combination therapy with lamivudine and efavirenz in a randomized placebo-controlled clinical trial in antiretroviral drug-naive subjects, powered to demonstrate equivalence [³]. To date, however, the inherent antiviral activity of this compound has yet to be defined. Hints at the robust potency of tenofovir were gleaned from experiments in rhesus macaques with both acute and chronic infection with SIV. Tenofovir monotherapy, when given subcutaneously, either prevented infection, even when started 24 hours after inoculation with intravenous SIVmac251 [⁴], or resulted in prolonged control of viremia with a 2-3 log₁₀ copies/ml reduction in SIV plasma viremia in three of four SIV-infected newborn rhesus macaques [⁵].

After antiretroviral therapy is initiated, a rapid and exponential decrease in the plasma viral load occurs, referred to as the 'first phase' of viral decay [^6,7]. The rate of change, or slope, of the first phase of plasma HIV-1 decay is a function of two parameters, δ, the decay constant of productively infected CD4 T cells, and ∊, the relative efficacy, defined as the ability of the drug to suppress viral replication in vivo. Assuming that δ is constant, then the slope of the first phase is directly proportional to the efficacy of the antiretroviral intervention applied. Therefore, the more negative the slope of the first phase of HIV-1 RNA decay, the greater the potency of the antiretroviral therapy being used [^8-10]. By applying these axioms to the study of the early changes in HIV-1 RNA levels, the efficacy or inherent antiviral activity of a single drug or a combination of antiviral agents can be deduced.

The present study describes a clinical experiment in 10 treatment-naive individuals who were given a 21-day course of TDF monotherapy with intensive monitoring of changes in HIV-1 RNA levels to determine the efficacy of this compound. Identical analytical methods were then used to compare the potency of TDF with a similarly studied potent protease inhibitor, ritonavir.

Methods

Ten chronically HIV-1-infected subjects were recruited for a 21-day monotherapy study of TDF. Patients were antiretroviral drug naive by history, chronically infected with HIV-1 with plasma HIV-1 RNA levels > 10 000 copies/ml and CD4 T cell counts > 50 × 10⁶ cells/l, had no acute conditions that required immediate therapy, were negative for active hepatitis B and C infection, and had not been vaccinated within 30 days of initiating TDF monotherapy. Patients were confined to the Rockefeller University Hospital Study Unit for 72 hours during which time treatment with TDF 300 mg orally daily as monotherapy was initiated and continued for 21 days. Subjects were subsequently seen as outpatients for specimen collection and safety monitoring. Plasma HIV-1 RNA levels were measured with the Amplicor UltraSensitive Cobas 1.5 Assay (Roche, Alameda, California, USA) with a lower limit of detection of 50 copies/ml. Plasma HIV-1 RNA was sampled at day 0 (baseline), every 6 h during the 72 hours of confinement, daily from days 4 through 10, and on days 12, 14, and 21.

First-phase slopes of HIV-1 RNA decay were determined for each subject by three methods in order to match the methods previously used to analyze ritonavir monotherapy trials. First, as in Ho et al. [⁶], the log₁₀ plasma viral load was plotted versus time to determine the initial slope by linear regression. Because of pharmacological and mechanistic delays (shoulder in the HIV-1 RNA decay curve), the regression curve was not required to begin falling at the start of therapy. Second, the slope of HIV-1 RNA decay was calculated using only baseline and day 7 HIV-1 RNA levels and compared with the slopes obtained by a similar analysis from 20 protease inhibitor-naïve ritonavir-treated patients [⁶]. This analysis ignored any possible initial delay in HIV-1 RNA decline. Lastly, the initial HIV-1 RNA slope was determined by fitting the viral load data to our published mathematical model of the first phase of viral decay [⁷], using a value for c, the clearance constant of plasma virus of 23/day [¹¹] and allowing both the initial delay and slope to be free variables.

To assess for the selection of drug-resistant variants during monotherapy, pre- and post-treatment (day 0 and day 21) consensus sequencing of pol from plasma-derived viruses were performed using TRUGENE HIV-1 Genotyping Kit in conjunction with the OpenGene automated DNA sequencing system (Visible Genetics, Toronto, Canada).

All participants gave written informed consent by signing a Rockefeller University Institutional Review Board reviewed and approved patient informed consent form.

Results

Study subjects had a mean baseline age of 34 years (range, 20-48), plasma viral load of 4.3 log₁₀ copies/ml (range 3.7-5.1), and CD4 count of 645 × 10⁶ cells/l (range, 340-1260) (Table 1). TDF therapy was extremely well tolerated and was not associated with any moderate or severe adverse events and all 10 subjects completed the 21-day course of therapy.

Table 1

The mean decrease in plasma HIV-1 RNA observed after 21 days of TDF monotherapy was 1.5 log copies/ml (range, 0.7-2.0; median, 1.6) (Fig. 1). The average individual first phase viral decay slopes, as determined by linear regression, was 0.39/day (range, 0.24-0.59) (Fig. 2). This is comparable with the average decay slope, 0.34/day, determined by linear regression using similar data obtained in a cohort of 20 protease inhibitor-naive chronically infected subjects treated with ritonavir monotherapy [⁶]. Analysis of the first-phase slopes using day 0 and day 7 viral load data for the TDF study and that previously published in 20 ritonavir-treated protease inhibitor-naive subjects [⁶] gave decay slopes of 0.32/day and 0.34/day, respectively.

Fig. 1

Fig. 2

A best-fit analysis for mean initial HIV-1 RNA decay slope was carried out by fitting the viral load data to our published mathematical model of the first phase of viral decay [⁷], which incorporates an initial delay or shoulder in the HIV-1 RNA curve. The estimated decay slope was found to be 0.48 ± 0.15/day, not statistically different from that obtained by linear regression. The same model was used to analyze data collected from five protease inhibitor-naive subjects treated with 1200 mg ritonavir daily and gave a nearly identical decay slope value of 0.47 ± 0.16/day [⁷]. The shoulder or delay in decay in HIV-1 RNA was longer and more variable in the TDF-treated cohort (1.39 ± 1.06/day) than in the ritonavir-treated patients (0.80 ± 0.39/day); however, this difference was not statistically significant (P = 0.26). Note that the slopes estimates obtained from the model are higher, though not significantly so, than the ones obtained by simple linear regression in both groups of patients. This difference may arise because the model more accurately accounted for the initial delay and hence showed a steeper slope.

Though the emergence of resistance to TDF is a slow process both in vitro [¹²] and in vivo [¹³], pre- and post-treatment genotypic analyses of patient virus were performed to determine whether drug resistance emerged during the 21 days of monotherapy. Two subjects had resistance-conferring mutations in reverse transcriptase (RT) before therapy. Plasma HIV-1 harbored M184V (a lamivudine resistance-conferring mutation) in one subject (patient 201) and K103N in another (patient 204). Both patients insisted that they had no previous exposure to antiviral therapy. As viruses with the M184V amino acid substitution are more susceptible to the effects of TDF in vitro [¹⁴], it is important to note that the degree of HIV-1 RNA reduction (1.8 log copies/ml) in the patient harboring this mutation was not substantially different from the mean, 1.5 log copies/ml. Furthermore the first-phase slope (-0.41/day) was also almost identical to the mean value, -0.39/day. The subject harboring the K103N amino acid substitution had a 0.7 log copies/ml reduction in HIV-1 RNA and a -0.24/day first-phase slope, lower than average but comparable with results obtained in subject 203, who harbored wild-type virus. There was sufficient plasma HIV-1 RNA at day 21 in 8 of 10 subjects to allow for successful amplification and consensus sequencing of HIV-1 RT. HIV-1 RNA values were below detection and 134 copies/ml plasma, respectively, in the two subjects in whom amplification was not successful. No changes were observed in RT in all eight subjects (data not shown) following the 21 days of monotherapy.

Discussion

TDF is a prodrug of tenofovir, a NtRTI, recently approved in North America, Australia, and Europe for the treatment of HIV-1 infection in combination with other antiretroviral agents. When added to failing HAART regimens in heavily pretreated individuals, the mean reduction in HIV-1 RNA was approximately 0.6 log copies/ml [²], slightly greater than that observed for zidovudine (0.5 log copies/ml) [¹⁵] and stavudine (0.5 log copies/ml) [¹⁶] monotherapy in drug-naive subjects. More recently, a randomized placebo-controlled clinical trial of TDF versus stavudine in combination with lamivudine and efavirenz in treatment-naive patients found the two regimens to be comparable in efficacy and durability after 48 weeks of therapy, with 82% of subjects achieving plasma HIV-1 RNA levels below detection (50 copies/ml plasma) [³]. These studies clearly support the use of TDF in the treatment of both naive and experienced subjects. However, as the armamentarium of antiviral agents increases in size and diversity, it becomes more challenging to define clearly a particular role for a new compound. We believe that defining the inherent antiviral activity of a new drug serves ultimately to guide the development of clinical trials for assessing optimal use.

Here we have documented that TDF monotherapy is able to effect an approximate 1.5 log copies/ml reduction in HIV-1 RNA in drug-naive patients, an antiviral effect similar to that seen in selected studies of protease inhibitor monotherapy [^17-19] and slightly greater than that seen for lamivudine [¹⁵] and abacavir [²⁰] monotherapy. We believe that these data confirm the robust potency of TDF, as was initially suggested in studies of tenofovir (PMPA) in the SIV/macaque model [^4,5].

Furthermore, using three methods to analyze the first-phase decay of HIV-1 RNA during TDF monotherapy allowed comparison with similarly analyzed data from studies of ritonavir monotherapy. There were nearly identical rates of plasma HIV-1 RNA decay with TDF and ritonavir, consistent with comparable relative efficacy. Taken together, these results suggest that TDF may provide adequate antiviral activity around which a simple once-daily regimen can be designed.

Though early changes in HIV-1 RNA levels have correlated with short-term control of viral suppression in vivo [^9,10], caution must be exercised in the interpretation of these results. Initial changes in HIV-1 RNA levels may not necessarily translate into longer-term suppression of viral replication in combination therapies, as issues such as adherence, host factors and other pharmacological variables may override the issue of inherent antiviral activity. Nevertheless, further studies of TDF in well-designed large randomized controlled clinical trials are clearly indicated.

Concerns regarding the use of monotherapy in drug-naive subjects are clearly warranted. It would be ill-advised to test novel agents with a low genetic threshold for high-level resistance in a trial of this design. As earlier studies showed no emergence of resistant HIV-1 variants after 4 weeks of TDF monotherapy [²¹], we believed that the study design employed here would not have a deleterious effect on our study participants. Indeed, we have shown that 21 days of monotherapy did not select for HIV-1 variants with any additional resistance-conferring amino acid substitutions in RT. That two subjects harbored resistance-conferring changes in RT at study entry was surprising and merits comment. Given the wide access to antiretroviral drugs as well as increasing rates of transmission of drug-resistant variants [^22,23], it seems prudent to screen patients for evidence of previous drug exposure or the presence of transmitted resistant virus prior to study entry, particularly in small intensive trials such as this. We believe, however, that the data and conclusions generated remain valid in this particular study, as the patients infected with HIV-1 variants with resistance-conferring changes in RT had a response to TDF that does not appear substantially different from the eight subjects harboring wild-type virus.

Current guidelines for the use of antiviral therapies advise initial therapy to include a combination of two nucleoside RT inhibitors and either one protease inhibitor or one non-nucleoside RT inhibitor [²⁴]. However, alternatives, such as triple nucleoside RT inhibitor therapy with zidovudine, lamivudine, and abacavir, formulated in a fixed combination pill, may be considered, given demonstrated antiviral activity and ease of administration. It is conceivable, given the potency, once-daily dosing, tolerability [²], resistance profile [^12,13], and pharmacokinetic profile [²¹] of TDF, that TDF-based regimens may well serve as yet another alternative to now standard initial HAART combinations. We believe our data justifies the design and execution of prospective trials to test this hypothesis.

Acknowledgements

We are grateful to David D. Ho for ongoing support and scientific input, to Amiel Balagtas for assistance with protocol development, to Wen Chen for assistance in the preparation of the manuscript, and to the Rockefeller University Hospital Nursing staff for their assistance

Sponsorship: This work was supported by Gilead Sciences, the Rockefeller University General Clinical Research Grant M01-RR00102, NIH grants AI47033, AI28433 and RR0655 and the Columbia-Rockefeller-Aaron Diamond AIDS Research Center CFAR Grant P30-AI42848. We thank Roche Molecular Systems for providing supplies for HIV-1 RNA testing; Visible Genetics for providing materials for genotype testing.