In Vitro Resistance Profile of the Candidate HIV-1 Microbicide Drug Dapivirine

Abstract

Antiretroviral-based microbicides may offer a means to reduce the sexual transmission of HIV-1. Suboptimal use of a microbicide may, however, lead to the development of drug resistance in users that are already, or become, infected with HIV-1. In such cases, the efficacy of treatments may be compromised since the same (or similar) antiretrovirals used in treatments are being developed as microbicides. To help predict which drug resistance mutations may develop in the context of suboptimal use, HIV-1 primary isolates of different subtypes and different baseline resistance profiles were used to infect primary cells in vitro in the presence of increasing suboptimal concentrations of the two candidate microbicide antiretrovirals dapivirine (DAP) and tenofovir (TFV) alone or in combination. Infections were ongoing for 25 weeks, after which reverse transcriptase genotypes were determined and scrutinized for the presence of any clinically recognized reverse transcriptase drug resistance mutations. Results indicated that suboptimal concentrations of DAP alone facilitated the emergence of common nonnucleoside reverse transcriptase inhibitor resistance mutations, while suboptimal concentrations of DAP plus TFV gave rise to fewer mutations. Suboptimal concentrations of TFV alone did not frequently result in the development of resistance mutations. Sensitivity evaluations for stavudine (d4T), nevirapine (NVP), and lamivudine (3TC) revealed that the selection of resistance as a consequence of suboptimal concentrations of DAP may compromise the potential for NVP to be used in treatment, a finding of potential relevance in developing countries.

INTRODUCTION

In lieu of an effective vaccine, antiretroviral (ARV)-based microbicides may offer a means to reduce the sexual transmission of HIV-1. Indeed, results released in July 2010 from a double-blind randomized phase IIb clinical trial (CAPRISA 004) demonstrated that a 1% tenofovir (TFV) vaginal gel reduced HIV-1 acquisition by at least 39% and, moreover, by 54% in women who highly adhered (>80%) to the dosing strategy (1). Thus, ARVs traditionally used for treatment might help quell the sexual transmission of HIV-1. However, in cases where the TFV-based microbicide did not protect against infection, these data strongly suggest that efficacy is dependent on user adherence. This may indicate that (i) some users became infected when no drug was present at the site of transmission or (ii) some users became infected when a suboptimal concentration of the drug was present at the site of transmission. Taking this into consideration, it is important that the current treatment options remain effective for users that may become infected while using an ARV-based HIV-1 microbicide.

HIV-1 drug resistance remains a salient obstacle in treatment efficacy. The extraordinary capacity of HIV-1 to mutate (largely due to error-prone reverse transcriptase [RT]) leads to the development of drug resistance within HIV-1-infected individuals and, further, to the transmission of drug-resistant (DR) HIV-1 among individuals. Although microbicides are intended to completely thwart the initial infection foothold within an individual, poor adherence to dosing strategy may result in breakthrough and/or development of DR HIV-1. In the case of undiagnosed HIV-1 infection, further use of an ARV-based microbicide could facilitate the development of drug resistance at the individual level and potentially increase the incidence of DR HIV-1 transmission among individuals. This is because the same (or similar) drugs that are currently being used in treatment regimens are also being developed as microbicides. Thus, it is necessary to address the possibility that particular drug treatments may be compromised due to the selection and/or development of DR HIV-1 in the presence of suboptimal doses of candidate microbicide ARVs that are formulated as a microbicide.

Dapivirine (DAP) and tenofovir (TFV) are strong candidate microbicide ARVs. These drugs are a nonnucleoside reverse transcriptase inhibitor (NNRTI) and a nucleotide reverse transcriptase inhibitor (NtRTI), respectively. A wealth of evidence supports the continued development of DAP, both alone and in combination with TFV, as an effective microbicide. In culture, DAP showed potent activity against HIV-1_IIIb and HIV-1_Ba-L in the context of cell-free and cell-associated infection (32). Furthermore, cell-associated HIV-1 vaginal infection of SCID mice was prevented with a DAP gel (12) (2.25 mm), and DAP has been shown to maintain potent in vitro activity against a spectrum of wild-type (WT) and NNRTI-resistant clinical HIV-1 isolates in MT4 cells (14). More recently, DAP alone and in combination with tenofovir was shown to be active against clinical isolates cultured in peripheral blood mononuclear cells (PBMCs) (16). The combined effect of DAP and TFV was also evaluated in the context of WT and DR HIV-1, and the combination was shown to be effective against HIV-1 harboring the NNRTI resistance mutation Y181C (27). This notwithstanding, and although the resistance profile of tenofovir has largely been elucidated both in vitro and in vivo (6, 8, 11, 15, 20, 23, 29–31, 36–38), resistance mutations that might be selected as a result of DAP with TFV remain undetermined. Furthermore, DAP could conceivably select for resistance mutations that might impact the efficacy of other NNRTIs used to treat HIV-1 infection. Therefore, we evaluated the resistance profile of DAP in culture.

HIV-1 primary isolates of different subtypes and resistance profiles were used to infect primary cells in vitro in the presence of increasing, suboptimal concentrations of DAP and TFV, alone and in combination. The infections were ongoing for 25 weeks, after which reverse transcriptase (RT) resistance genotypes were determined. The results indicated that suboptimal concentrations of DAP alone facilitated the development/emergence of common NNRTI resistance mutations, while DAP in combination with TFV gave rise to fewer mutations. Suboptimal doses of TFV alone resulted in the development/emergence of only one nucleoside reverse transcriptase inhibitor (NRTI) mutation, K65R, in one of the nine different samples tested. Notably, mutations at amino acid position 138 were more frequent in the DAP plus TFV selection experiments, arising in all samples except when either NNRTI mutation Y181C or E138A was present at baseline. Since a common first-line therapeutic regimen used among ARV recipients in many low-income countries is coformulated stavudine (d4T), nevirapine (NVP), and lamivudine (3TC) (Triamune) (1), we also investigated whether the DR viruses selected in this study remained susceptible to these drugs. Pre- and postselection HIV-1 sensitivity evaluations for d4T, NVP, and 3TC indicated that the potential use of NVP-based treatments may be the most affected by selections with suboptimal concentrations of DAP plus TFV, since the resultant viruses were ∼15- and 8-fold more resistant to NVP than preselection viruses in terms of 50% effective concentration (EC₅₀) and EC₇₅, respectively. Thus, suboptimal long-term use of a DAP- or DAP plus TFV-based microbicide may facilitate the development of DR virus in HIV-1-positive individuals, and this may affect treatment outcome in resource-limited countries.

MATERIALS AND METHODS

Primary HIV-1 isolates.

The primary HIV-1 specimens 5326, 8116, 10583, 10680, 4742, and 4743 were collected with informed consent from different ARV-naïve individuals. Samples 10583 and 10680 were collected from the same patient, with informed consent, at different times of infection. Each sample was amplified using cord blood mononuclear cells (CBMCs) stimulated for 72 h with phytohemagglutinin (PHA) in complete RPMI 1640 medium (Invitrogen) supplemented with 10% fetal bovine serum (FBS; Gibco), penicillin-streptomycin (P-S; Invitrogen), l-glutamine (Invitrogen), and interleukin-2 (IL-2; Roche).

Generation of replication-competent genetically homogeneous HIV-1.

The NL4-3 wild-type (WT) molecular clone was obtained from the NIH AIDS Research and Reference Reagent Program (http://www.aidsreagent.org/) and amplified using kits (Qiagen) according to the manufacturer's instructions. To introduce the K103N and the Y181C RT mutations, site-directed mutagenesis was performed on the WT clone using XL-Gold 2 SDM kits (Stratagene). Sequencing of the RT region confirmed the presence of the mutations in both plasmids and the absence of any deviant mutations. To generate genetically homogenous/clonal HIV-1, pNL4-3, pNL4-3-K103N, and pNL4-3-Y181C were transfected separately into 293T cells using Lipofectamine 2000 (Invitrogen). Cells were subsequently incubated in antibiotic-free Dulbecco's modified eagle medium (DMEM) (2% FBS) for 6 h prior to washing with phosphate-buffered saline (PBS; Gibco) and replacement with fresh growth medium. At 24 h postwash, culture supernatants were collected and stored at −150°C for future use. HIV-1 strains produced from pNL4-3, pNL4-3-K103N, and pNL4-3-Y181C were designated HIV-1_NL4-3-WT, HIV-1_NL4-3-K103N, and HIV-1_NL4-3-Y181C, respectively.

Antiviral compounds.

TFV and DAP were kindly provided by the International Partnership for Microbicides (IPM), Silver Spring, MD. Nevirapine (NVP) was obtained from Boehringer Ingelheim, Inc. Stavudine (d4T) and lamivudine (3TC) were obtained from Sigma, Inc., and the NIH AIDS Research and Reference Reagent Program, respectively. Stock solutions were prepared from powders using dimethyl sulfoxide (DMSO) as solvents for DAP and 3TC, while H₂O was used for d4T and NVP. Dilutions of compound stocks were completed in DMEM immediately prior to experimentation.

Selection of resistance mutations to DAP, TFV, and DAP plus TFV in CBMCs.

Phytohemagglutinin (PHA)-stimulated CBMCs were infected with viruses (multiplicity of infection [MOI] of 0.1 to 1.0) for 2 h, incubated at 37°C, and subsequently washed with RPMI 1640 medium (Invitrogen) plus human IL-2 (hIL-2) (20 U/ml; ABI Inc.) and 10% FBS and seeded into a 24-well plate at a density of 2 × 10⁶ to 4 × 10⁶ cells per well. Selection for resistance was performed with increasing concentrations of drugs at starting concentrations below the 50% effective concentration (EC₅₀) of the drugs. As controls, all viruses were simultaneously passaged without drugs. RT assays were performed weekly to monitor viral replication as described previously (25). Based on the ratio of the RT values in cell culture supernatants from control wells to those from test wells at the previous round of replication, drug concentrations were increased at subsequent passages. Selection at a particular drug concentration was considered to be complete when repeated passage revealed that RT levels in culture fluids were relatively equal to those in the respective control wells tested at the same time. Virus-containing cell culture supernatants were harvested and kept at −80°C for subsequent genotypic analysis whenever drug concentrations were increased. Selections for drug resistance were performed over 25 weeks.

Genotyping.

Viral RNA was extracted from the culture supernatants at 0 and 25 weeks using a Qiagen QIAmp viral extraction kit (Mississauga, Ontario, Canada). Viral RNA was amplified via RT-PCR, and nested PCR was employed to amplify the HIV-1 RT gene. The resulting DNA was purified using a QIAquick PCR purification kit (Qiagen), as specified by the manufacturer. The presence of the expected 1.7-kb PCR product was confirmed by running 5 μl of each product on a 1% agarose gel. The samples were directly sequenced with subtype-specific PR primers using a BigDye Terminator cycle sequencing kit (version 1.1; Applied Biosystems). The sequences were run on an ABI Prism 3130xl genetic analyzer (Applied Biosystems). The data were analyzed using SeqScape software (version 2.5), and the PCR sequences were aligned using Bioedit software, version 7.0.

Phenotypic drug susceptibility.

The phenotypes of HIV-1₅₃₂₆ viruses against two NRTIs, d4T and 3TC, and one NNRTI, NVP, were determined before and after selection with DAP plus TFV. RTIs, at concentrations ranging from 1.5 nM to 100 μM, were added to TZM-bl cells (10⁴ cells/well) in 96-well plates grown in 100 μl DMEM (Gibco) supplemented with 10% fetal bovine serum (Gibco), 1% penicillin-streptomycin, and 1% l-glutamine (Invitrogen). Immediately after drug addition, cells were infected with each pre- and postselected virus. At 24 to 48 h postinfection, cells were rinsed with 100 μl PBS and lysed with 50 μl/well 1× cell lysis reagent (Promega). Cell lysates were then transferred to a white, opaque 96-well plate (Corning). Luciferase assay reagent (Promega) was then added to each well, and relative luminescence units (RLU)/well were measured by a luminometer (Microbeta²; Perkin Elmer). Infection standardization was accomplished by adding virus to always result in 100 50% tissue culture infective doses (TCID₅₀), as calculated in TZM-bl cells. The inhibitory effect of each drug alone and in combination was calculated as percent inhibition of infection relative to results obtained in the absence of drug.

RESULTS

DAP-, TFV-, and DAP plus TFV-selected drug resistance mutations in the RT gene associated with clinically relevant genotypic profiles.

WT and NNRTI-resistant HIV-1 primary patient isolates of subtypes B and C were challenged with increasing, yet always suboptimal, concentrations of DAP, TFV, or DAP plus TFV over 25 weeks (∼6 months). Clonal subtype B HIV-1 (derived from the molecular clone pNL4-3) was challenged likewise in parallel to control for initial input viral variation and to help differentiate between the emergence of preexisting (initially undetected) minority viral populations characteristic of patients' quasispecies and the occurrence of a spontaneous mutation(s) subsequently selected during the 25-week period. Interestingly, mutations present in the infecting inocula were not deselected, i.e., mutations present at the start of the selections were also present at the end of the selection process. The clinically recognized RT mutations (18) that were observed among the selected viruses included K65R, V90I, L100I, K101E, V106I, V108I, E138K, E138G, Y181C, and Y188L. All amino acid residue changes in RT are summarized inTable 1. Initial and final drug concentrations for each selection are listed in Table 2.

Table 1.

Reverse transcriptase drug resistance mutations selected with suboptimal concentrations of DAP, TFV, and DAP plus TFV at week 25 (∼6 months)

^aDrug-sensitive (WT) amino acid.

^bPosition of amino acid according to WT HXB2 RT reference sequence.

^cHIV-1 primary isolate or strain designation.

^dGenotype of RT at start of selection.

^eAmino acid after 25 weeks of selection with DAP (black), TFV (green), or DAP plus TFV (red).

^fMix of amino acids detected in the selected HIV-1 sample after 25 weeks.

^gParentheses indicate amino acid present at beginning of experiment and after 25 weeks of selection.

^h*, Missense mutation indicating that K65R might have emerged if selection continued further.

Table 2.

Summary of initial (day 0) and final (week 25) drug concentrations for each virus tested

Isolate or strain	DAP		TFV		DAP + TFV
	Initial concn (nM)	Final concn (nM)	Initial concn (nM)	Final concn (nM)	Initial concn (nM)	Final concn (nM)
Subtype B
5326	0.1	250	10	1,000	0.5 + 10	1,000 + 5,000
8116	0.1	5	10	2,500	1 + 25	10 + 100
HIV-1NL4-3-WT	0.1	500	10	5,000	0.5 + 10	50 + 500
HIV-1NL4-3-K103N	1	2,500	10	10,000	5 + 10	500 + 1,000
HIV-1NL4-3-Y181C	1	250	10	2,500	5 + 10	1,000 + 1,000
Subtype C
10583	1	5,000	10	2,500	5 + 10	500 + 1,000
10680	0.1	2,500	10	2,500	0.5 + 10	25 + 250
4742	0.1	1,000	10	5,000	0.5 + 10	250 + 500
4743	0.1	500	10	2,500	0.5 + 10	10 + 100

Mutations selected with DAP alone.

The most common NNRTI resistance-associated mutation selected in the presence of DAP alone was located at position 138. Typically, a glutamic acid (E) is located at position 138 in WT HIV-1, but DAP selections utilizing WT HIV-1 resulted in a change to lysine (K), while selections utilizing HIV-1_NL4-3-K103N resulted in a mix of K, G, R, and E. Two DAP-only selections with subtype B HIV-1 resulted in an I or E at RT amino acid position 100 instead of L, and an I was observed in HIV-1₅₃₂₆ and clonal HIV-1_NL4-3-K103N selections. HIV-1_NL4-3-WT, HIV-1_NL4-3-Y181C, and HIV-1₈₁₁₆ did not deviate from the WT phenotype at position 100 under DAP pressure. DAP selections also resulted in mutations at positions 101, 106, 138, 179, and 181 in some but not all subtype B HIV-1 viruses that were studied.

Subtype C HIV-1 selections resulted in a different resistance profile, with the most common mutations selected by DAP occurring at positions 138 and 181, i.e., E138K and Y181C, respectively (two of four selections). Mutations at positions L100I and E138K were observed in DAP selections involving HIV-1₁₀₆₈₀, a subtype C virus that possessed a WT genotype at the beginning of the selection. A mix of amino acid I and the wild type at positions 90 and 106 was also apparent in postselected HIV-1₁₀₆₈₀. Mutations L100I and V179I were observed for HIV-1₁₀₅₈₃, in addition to the Y181C mutation, which was present at the beginning of the selection. A subtype C virus containing the A98S and E138A substitutions at baseline maintained these mutations while also developing Y181C and Y188H. HIV-1₄₇₄₃, a subtype C primary isolate originally containing the G190A mutation, developed K103M and Y181C without losing the G190A substitution.

Mutations selected with TFV alone.

Two different TFV selections produced single mutations, with the most common, as expected, being K65R in the subtype C primary isolate HIV-1₁₀₅₈₃. A mix of V179V/I was observed with the subtype B primary isolate HIV-1₅₃₂₆.

Mutations selected with DAP plus TFV.

Selections utilizing DAP in combination with TFV yielded mutation patterns similar to those observed for DAP alone, with some exceptions. Notably, mutations at position 138 were more frequent in the DAP plus TFV selections and were observed for all viruses tested except those that had a Y181C mutation prior to the selections with both DAP plus TFV and DAP alone. Under DAP plus TFV pressure, but not under DAP pressure alone, viruses containing the G190A mutation prior to selection commonly developed an additional E138K mutation, although E138G and a mix of E138E/K were observed for HIV-1_NL4-3-K103N and HIV-1₄₇₄₃, respectively. For both subtypes tested, the DAP plus TFV selections produced a mutation at position 100 for only one virus, i.e., HIV-1₅₃₂₆. The mutations K101E/K and K101T/K in subtype C HIV-1₁₀₅₈₃ and HIV-1₄₇₄₃, respectively, were unique to the DAP plus TFV selections.

DAP plus TFV-selected HIV-1 remains susceptible to 3TC and d4T but not NVP.

Pre- and postselected HIV-1₅₃₂₆ was tested for the capacity to infect TZM-bl cells in the presence of 3TC, d4T, and NVP. The results showed that the preselected viruses were susceptible to all three drugs. However, after 6 months of pressure with DAP plus TFV, the selected viruses were susceptible to d4T and 3TC but less so to NVP (Fig. 1), compared to the preselected counterpart virus. The impact on NVP appeared to be directly proportional to the EC evaluated, i.e., resistance was greater at the EC₅₀ (∼15-fold more resistant than that for preselected virus) than that at the EC₇₅ (∼8-fold more) (Table 3). For DAP plus TFV-selected HIV-1₅₃₂₆, the changes in susceptibility to d4T and 3TC were ∼3- and ∼2-fold more resistant, respectively, at the EC₅₀. However, DAP plus TFV-selected HIV-1₅₃₂₆ showed only modest changes in susceptibility to d4T and 3TC at the EC₇₅ (1.63- and 0.6-fold more resistant, respectively).

Fig 1.

DAP plus TFV affects NVP efficacy in an infectivity assay. Sensitivity to d4T (A), NVP (B), and 3TC (C) of HIV-1 before (●) and after (■) selection with DAP plus TFV over 25 weeks in vitro. x axes indicate drug concentrations (μM) tested. Data points depict the means and standard deviations from two independent experiments performed in triplicate.

Table 3.

EC₅₀s and EC₇₅s of HIV-1₅₃₂₆ before and after selection with TFV plus DAP

Drug	Selection	EC50 (nM)a	P valueb	EC75 (nM)a	P valueb	Fold resistancea
						EC50	EC75
d4T	Preselection	148 ± 14.1	0.0137	595 ± 8.49	0.1069
	Postselection	541.5 ± 64.3		930.5 ± 169		3.367 ± 0.322	1.363 ± 0.019
NVP	Preselection	67.34 ± 0.067	0.0002	189.35 ± 23.55	0.0007
	Postselection	997.7 ± 18.9		1,488 ± 41.7		14.89 ± 1.62	7.931 ± 1.21
3TC	Preselection	327 ± 31.7	0.0094	1,863 ± 202	0.0549
	Postselection	775 ± 53.2		1,178 ± 123		2.390 ± 0.393	0.632 ± 0.002

^aMean (n = 2) ± SD. Fold resistance values indicate increase in resistance compared to that seen with preselected counterpart virus.

^bP values calculated using unpaired t tests.

DISCUSSION

ARV-based microbicides are intended to prevent the initial HIV-1 infection foothold. Poor adherence to dosing strategy may result in the breakthrough of transmission-fit WT or DR HIV-1 and/or the development of DR HIV-1 in individuals who are HIV-1 positive but unaware of their infection status. In the case of undiagnosed HIV-1 infection, further use of an ARV-based microbicide could potentially increase the incidence of DR HIV-1 transmission among individuals. In this study, various WT and DR HIV-1 strains of subtypes B and C were grown in tissue culture under suboptimal concentrations of the candidate microbicide inhibitors DAP and/or TFV for a duration of 6 months, following which the genotype of each tested virus was evaluated for clinically relevant RT drug resistance mutations. Interestingly, suboptimal concentrations of DAP and DAP plus TFV permitted emergence of more RT mutations than suboptimal concentrations of TFV alone. Of all nine HIV-1 isolates tested, only HIV-1₁₀₆₈₀ developed the K65R mutation in response to suboptimal TFV pressure, and HIV-1₅₃₂₆ developed a mixture of WT and mutated viruses at amino acid RT position 179. Thus, suboptimal concentrations of TFV might be less detrimental to HIV-1 RT inhibitor-based treatment regimens than suboptimal concentrations of DAP, presumably due to the higher barrier for resistance to TFV. The observation that DAP plus TFV selections produced fewer drug resistance mutations than selections performed with DAP alone further supports this notion.

Although the activity of DAP against laboratory-adapted and drug-resistant primary isolates of HIV-1 has previously been evaluated and the in vitro development of DAP HIV-1 resistance mutations was previously reported (27a), our results confirm L100I and E138K as DAP-associated resistance mutations. However, this is the first study to demonstrate that suboptimal use of a TFV-based microbicide may potentially be less harmful (in terms of generating NNRTI resistance mutations) than that of a microbicide which includes both TFV and DAP, a finding of potential importance with regard to treatment options in developing countries with limited access to ARVs.

The fact that DAP can select for the E138K mutation in RT is of concern, as this establishes that cross-resistance between DAP and both initial and subsequently developed inhibitors of the NNRTI family of drugs is likely. The selection of L100I in RT under suboptimal DAP pressure occurred most frequently for both subtype B HIV-1₅₃₂₆ and HIV-1_NL4-3-K103N and subtype C HIV-1₁₀₅₈₃ and HIV-1₁₀₆₈₀ over 25 weeks. If DAP resistance is similar to that of etravirine (ETV), L100I should reduce but not preclude the clinical utility of DAP (18, 21). E138K confers resistance to ETV, DAP, and the most recently approved NNRTI, rilpivirine (RPV) (2, 3, 7, 9, 22).

As ETV and RPV are closely related to DAP, both in structure and in efficacy, it is not surprising that E138K emerged in our DAP and DAP plus TFV selection experiments. Importantly, E138K was the most common mutation in all selections, i.e., 4 of 7 with DAP alone and 5 of 7 with DAP plus TFV. The possibility that Y181C may preclude the development of E138K, possibly due to a cost to viral fitness, should be investigated, since Y181C-containing viruses do not appear to be fitness impaired (33).

The development of Y181C under suboptimal use of a DAP-containing microbicide is also important, given that this mutation can confer high levels of resistance to NVP (18).

G190A can also occasionally develop during failure of an NVP- or efavirenz (EFV)-containing regimen (4, 5, 17, 24). Since both types of regimen are available in developing countries, the possibility that G190A may be transmitted sexually by someone receiving NVP or EFV exists (19). In this context, it may be important that G190A HIV-1 remains susceptible to DAP in biochemical assays (39). However, G190A confers a fitness cost to HIV-1, presumably due to a reduction in RNase H activity, decreased RNA-dependent DNA synthesis from the tRNA^{Lys, 3} primer (34), and an increase in RT-RT heterodimer stability (13). In contrast, RNase H activity and primer binding are increased upon exposure to EFV (26, 35), suggesting that G190A may develop under NNRTI pressure to compensate for the increase in RNase H activity and RNA-dependent DNA synthesis of strong-stop DNA (10).

A common first-line triple-therapy-based ARV regimen among patients in low-income countries is coformulated NVP, 3TC, and d4T (Triamune) (28). Since a microbicide is likely comprised of a combination of ARVs, we tested the susceptibility of DAP plus TFV-selected HIV-1 to NVP, 3TC, and d4T (Fig. 1) and showed that DAP plus TFV-selected HIV-1 would remain susceptible to 3TC and d4T, an unsurprising result since the selected virus did not develop NRTI resistance-associated mutations. However, the DAP plus TFV-selected virus did demonstrate resistance to NVP, a matter of concern. Since poor adherence to the use of microbicides may select for drug-resistant viruses, it is important to be able to compare concentrations of drugs attained in clinical studies versus tissue culture evaluations, but a comprehensive statement on this topic is not yet possible due to a paucity of data. Once sufficient data are available, we hope to address this topic in a future manuscript.