Penetration of enfuvirtide, tenofovir, efavirenz, and protease inhibitors in the genital tract of HIV-1-infected men

One likely mechanism of virological failure is poor antiretroviral drug diffusion in sites of viral replication such as the genital tract. We measured antiretroviral drug concentrations in blood and semen in 13 HIV-infected men failing treatment. Enfuvirtide did not cross the blood-testis barrier, whereas tenofovir accumulated in semen. Unlike indinavir, semen concentrations of lopinavir, amprenavir, saquinavir and efavirenz were ineffective. These are worrying findings, because suboptimal semen drug concentrations may enhance the risk of sexually transmitted drug-resistant HIV variants.

Increasing numbers of HIV-infected patients have a history of multiple treatment failures. As HIV may evolve in a compartment-specific manner, distinct resistance patterns may arise in the male genital tract [¹]. The development of drug-resistant viral strains is possibly enhanced by suboptimal drug concentrations in semen [²]. Drug disposition into semen is influenced by drug ionization, lipophilicity, molecular weight, the degree of protein-binding, affinity for P-glycoprotein, and semen pH [³]. No data are yet available concerning the penetration of enfuvirtide and tenofovir in semen. Here we describe antiretroviral drug concentrations in seminal plasma (SP) and in blood plasma (BP) in a subset of HIV-infected men with advanced disease, failing their current treatment, and in whom HIV genotypic drug resistance was compared between the two compartments in a larger study [¹].

Thirteen HIV-infected men, failing a stable antiretroviral regimen including a median of four drugs for at least 8 weeks, with no clinical acute genital infection, were enrolled in this cross-sectional study after giving written informed consent. Their median age was 42.5 years (36-51), the median CD4 cell count at inclusion was 240 × 10⁶/l (72-460), with a median nadir of 36 × 10⁶/l (1-184). The median HIV-RNA level was 4.55 and 2.96 log₁₀ copies/ml in BP and SP, respectively, confirming therapeutic failure in both compartments. Adherence reached 100% (self-questionnaire on missed doses during the previous 4 days). Steady-state conditions were ensured. Single paired samples of blood and semen were collected the same day within one hour, and processed for viral quantification as described elsewhere [¹]. BP and SP were assessed for pharmacological measurements as previously described, using validated specific reverse-phased high-performance liquid chromatography assays coupled with spectrofluorometric or ultraviolet detection [^4-7]. No matrix effect between BP and SP was detected. The limits of quantification, the minimal expected effective BP concentration for each drug [^8-11] and the results of drug measurements in BP and SP are reported in Table 1.

Table 1

To our knowledge, this is the first study evaluating antiretroviral drug concentrations in the genital compartment of HIV-infected men with advanced disease and on complex regimens. Moreover, these are the first results of enfuvirtide and tenofovir concentrations in semen. Several points of interest deserve to be underlined.

Enfuvirtide did not cross the blood-testis barrier despite optimal concentrations achieved in BP [¹⁰], probably caused not only by the elevated molecular weight of this drug, but also by it being ionized. Because ionization increases hydrophilicity, only non-ionized drugs diffuse through biological membranes [³].

These worrying findings are questionable regarding the efficacy of enfuvirtide in other sequestered anatomical sites such as the central nervous system. Conversely, tenofovir proved to accumulate in semen with higher drug concentrations than in BP, suggesting the active transport and sequestration of this small molecule, as well as a pH partitioning effect based on its pKa. Moreover, its long half-life [¹¹] may enhance accumulation or slow elimination. These findings are consistent with previously reported results for nucleoside analogues [¹²]. Penetration of protease inhibitors (PI) in the male genital tract is hampered by their being more lipohilic and extensively bound to BP protein. Indinavir, which has more limited protein binding, is readily detectable in SP, and ritonavir boosting previously proved to enhance its penetration in SP [¹³]. We showed that, unlike indinavir, and although boosted, amprenavir, saquinavir and lopinavir, three drugs widely used within salvage regimens, achieved adequate concentrations in BP but limited concentrations in SP. Suboptimal SP concentrations of highly BP protein-bound (> 95%) lopinavir and saquinavir have already been reported [^14,15]. Amprenavir, a lipophilic PI, is approximately 90% bound to BP protein [¹⁶]. In our series, amprenavir concentrations reached in SP were approximately 10% of those in BP, which is consistent with the fact that only free drug is able to access semen. Such concentrations previously proved to suppress viral replication adequately in the genital tract of PI-naive HIV-infected men [^17,18]. Whether these concentrations would prove effective in a setting of viral resistance warrants further investigations. Our results concerning efavirenz are not concordant with previous studies [^19,20], in which concentrations in SP averaged 10% of those in BP, but were still higher than the corresponding EC₉₅ [²¹]. Our three patients on an efavirenz-containing regimen had undetectable levels in SP despite optimal concentrations in BP. The physicochemical properties that theoretically favour the attainment of detectable efavirenz concentrations in SP include lipophilicity and low P-glycoprotein affinity. However, efavirenz is highly bound to BP protein. The small number of patients and the absence of standardization of the time blood and semen were collected might be limiting, as BP and SP have different rates of accumulation and elimination, making this a dynamic process. More systemic sampling would have been preferable, but we chose not to add a sampling time burden for these HIV-infected patients with advanced disease. Nevertheless, we showed that, at a given time for a given antiretroviral drug, undetectable or suboptimal levels were observed in SP.

The blood-testis barrier preventing drug distribution to sites of viral replication remains a likely mechanism of virological failure. Therefore, with the widespread administration of combined antiretroviral regimens, it will be critical to deliver adequate drug concentrations to all compartments in which the virus can replicate in order to ensure the long-term efficacy of antiretroviral therapy and reduce the risk of sexually transmitted drug-resistant HIV strains. Moreover, new treatment strategies such as PI monotherapy [^22,23] should also take into account PI virological efficacy in sanctuary sites in order to prevent virological failure and the subsequent selection of resistant variants in those sites. Recent reports, suggesting an increasing prevalence of sexually transmitted resistant variants [²⁴], call for larger studies regarding drug-drug interactions in the genital compartment of patients with multiple therapeutic failures, and monitoring the role of local drug-selective pressure in the spread of drug-resistant viruses.

Acknowledgements

Enfuvirtide was kindly provided by Roche, Palo Alto, California, USA. The authors would also like to thank all the patients who participated in this study, and Dr Agnès Mogenet for the clinical monitoring.

Sponsorship: This work was supported by a scholarship from the Fondation pour la Recherche Médicale and by grants from Sidaction and the French National AIDS Research Agency (ANRS).