Tenofovir disoproxil fumarate (tenofovir DF) is a bioavailable prodrug of tenofovir, a potent nucleotide analogue reverse-transcriptase inhibitor with activity against human immunodeficiency virus (HIV) and hepatitis B virus. It is administered as a single 300-mg tablet once daily. It was approved for the treatment of HIV infection on the basis of data from clinical trials demonstrating activity in treatment-experienced patients, and it was subsequently shown to be effective when used as a component of initial therapy. Tenofovir DF is active against some nucleoside-resistant strains of HIV. However, cross-resistance is associated with multiple thymidine analogue mutations that include 41L or 210W. The signature mutation is the K65R mutation, which causes variable loss in susceptibility to tenofovir DF, didanosine, and abacavir. Tenofovir DF has been well tolerated in clinical trials with durations of follow-up up to 96 weeks. It is associated with more-favorable lipid profiles than stavudine and has not been associated with the mitochondrial toxicity attributed to other nucleoside analogues.
Tenofovir disoproxil fumarate (tenofovir DF) is an orally bioavailable prodrug of tenofovir. It is the first nucleotide analogue reverse-transcriptase inhibitor (NtRTI) to have been approved by the US Food and Drug Administration (FDA) for the treatment of HIV infection. Since its approval in October 2001, this analogue of adenine 5′ monophosphate has quickly become a widely used component of antiretroviral regimens for both treatment-naive and -experienced patients on the basis of its efficacy and tolerability in several clinical trials.
Treatment-experienced patients. Tenofovir DF was approved on the basis of data from treatment-experienced patients. In Gilead study 907, a total of 550 patients for whom stable antiretroviral regimens were failing, with HIV RNA levels of 400–10,000 copies/mL, were randomized to add tenofovir DF or placebo to their existing regimens . Patients randomized to add placebo were allowed to add tenofovir DF at 24 weeks. Those randomized to add tenofovir DF experienced a decrease in the HIV RNA level of ∼0.6 log10 copies/mL, which was sustained throughout the 48-week trial. At 24 weeks, 22% of tenofovir DF recipients had HIV RNA levels of <50 copies/mL, compared with 1% of placebo recipients (P < .0001). Emergence of new resistance mutations at 24 weeks was more common in the placebo arm, which is not surprising, given that this group had more ongoing viral replication . Viral suppression to <50 copies/mL was more common among patients with lower HIV RNA levels at enrollment , which is consistent with data from other intensification studies. In the earlier Gilead 902 trial, a dose-ranging study involving patients with HIV RNA levels of 400–100,000 copies/mL, virologic response was similar to that seen in Gilead study 907 .
Treatment-naive patients. Before the availability of data from large, randomized clinical trials, several small studies hinted at the potency of tenofovir DF in treatment-naive patients. The 901 study was a randomized, placebo-controlled, dose-ranging trial involving 4 doses of tenofovir DF given as monotherapy, with 10 patients in each treatment arm (8 in the tenofovir DF arm and 2 in the placebo arm) . The median decrease in HIV RNA level was 1.2 log10 copies/mL among those taking tenofovir DF at a dosage of 300 mg per day (the median decrease was 1.6 log10 copies/mL among the 3 treatment-naive patients). In another short-term monotherapy trial, 10 treatment-naive patients received tenofovir DF for 21 days . After 21 days, the median reduction in the HIV RNA level was 1.6 log10 copies/mL, which compares favorably with the potency of nucleoside reverse-transcriptase inhibitors (NRTIs) in monotherapy trials.
Preliminary 96-week results have now been reported from Gilead 903, an ongoing, 144-week, randomized, multicenter, double-blind trial comparing tenofovir DF with stavudine, both of which are administered in combination with lamivudine and efavirenz . The study enrolled 600 treatment-naive patients from 81 sites. More than 40% of the participants had HIV RNA levels >100,000 copies/mL. By intent-to-treat, missing-equals-failure analysis, 82% of patients in the tenofovir DF arm and 78% in the stavudine arm achieved an HIV RNA level of <400 copies/mL, with no loss of response between those who had high HIV RNA levels or low CD4 cell counts at baseline. Seventy-eight percent of tenofovir DF recipients and 74% of stavudine recipients achieved an HIV RNA level of <50 copies/mL. Mean CD4 increases were 261 cells/mm3 and 266 cells/mm3, respectively.
Resistance. Although tenofovir DF belongs to a new class of agents, resistance is associated with a number of familiar NRTI resistance mutations, including the thymidine analogue mutations (TAMs) and K65R (table 1). The latter is the signature mutation and is associated with a 3- to 4-fold reduction in susceptibility in vitro. It is also selected by didanosine and abacavir, thus leading to potential cross-resistance among these drugs, and it reduces susceptibility to lamivudine . However, K65R emerges infrequently in treatment-experienced patients. In a combined analysis of the Gilead 902 and 907 studies discussed above, K65R emerged in 3.2% of patients taking tenofovir DF, and 40% developed additional TAMs . Most of the patients who developed the K65R mutation were taking other drugs (e.g., abacavir and didanosine) that could have selected for this mutation. In the 903 study that involved treatment-naive patients, 22% of the patients who experienced virologic failure while receiving tenofovir DF had the K65R mutation (along with nonnucleoside reverse-transcriptase inhibitor [NNRTI] resistance, with or without M184V) . However, because of the low rate of failure in this trial, this represented only 2.7% of the 299 tenofovir DF recipients. The K65R mutation is associated with reduced replicative capacity, and the simultaneous presence of M184V further reduces this measurement of viral fitness .
Tenofovir resistance is also affected by certain TAMs as a result of cross-resistance with the thymidine analogues, zidovudine and stavudine. In the 902 and 907 studies, the best response was seen in patients without TAMs, who experienced a 0.8-log decrease in the HIV RNA level, compared with the overall decrease of 0.6 log10 copies/mL . Patients with the weakest response (-0.2 log10 copies/mL) were those who had ⩾3 TAMs that included M41L and/or L210W (P = .013). This pattern has also been associated with a reduced response to abacavir and the thymidine analogues. HIV RNA reductions among those with ⩾3 TAMs that did not include either mutation were similar to the overall response. This is consistent with the observation that the combination of TAM mutations D67N, K70R, and 215Y is associated with minimal ATP-dependent removal of tenofovir by HIV-1 reverse transcriptase .
These 2 trials have been useful in defining the phenotypic breakpoints for tenofovir DF. In the resistance substudies from studies 902 and 907, patients who had a ⩽1.4-fold change in susceptibility had the best virologic response, whereas those with a >4-fold reduction in susceptibility had a poor response (i.e., a decrease in HIV RNA level of 0.2 log10 copies/mL). Intermediate responses were associated with a 1.5- to 4-fold decrease in susceptibility [11, 13].
Susceptibility to tenofovir DF is also reduced substantially in the presence of the uncommon T69 insertion mutation, though it remains active against virus expressing the Q151M multinucleoside resistance complex. As with zidovudine and stavudine, susceptibility is enhanced by M184V. The L74V mutation associated with didanosine resistance also results in slightly improved susceptibility to tenofovir DF. The signature K65R mutation appears to increase susceptibility to zidovudine, which may explain why emergence of K65R is unusual in patients with preexisting TAMs.
Toxicity. Tenofovir DF has been well tolerated in clinical trials to date. In the 902 and 907 trials, clinical and laboratory adverse events were no more common with tenofovir DF than with placebo. In the 903 trial, only 15% of the subjects dropped out before week 96; 1% dropped out because of adverse events. Although grade 3/4 clinical and laboratory adverse events were uncommon in both arms, stavudine was associated with significantly greater elevations in fasting triglyceride levels and total and LDL cholesterol levels . As a result, significantly more patients who received stavudine added either a statin or fibric acid derivative to their regimens, compared with those who received tenofovir DF (P < .001).
In vitro studies have suggested that tenofovir DF, along with abacavir and lamivudine, is less toxic to mitochondria than are stavudine, zalcitabine, and didanosine and probably also zidovudine, and it has limited effect on mitochondrial DNA polymerase [14, 15]. In a substudy assessing mtDNA from PBMCs, patients taking tenofovir DF had a significant increase in mtDNA levels, compared with the baseline value (P < .001), whereas those taking stavudine had a small but insignificant increase (P = .37) . Stavudine was also associated with elevated venous lactate levels: in a substudy at week 48, a total of 27% of stavudine recipients had venous lactate levels of >2.22 mmol/L, compared with 4% of tenofovir DF recipients (P < .0001). Toxicities that have been attributed to mitochondrial toxicity, including peripheral neuropathy and investigator-defined lipodystrophy, were significantly more common among patients taking stavudine (P < .001). In a substudy at week 96, patients taking tenofovir DF gained more weight (P = .002) and had significantly greater total limb fat, as determined by whole-body dual-energy x-ray absorptiometry (DEXA) scan (P < .001), than did those taking stavudine .
None of these trials has shown evidence of nephrotoxicity with tenofovir DF, a concern given the proximal renal tubular dysfunction observed in HIV-infected patients treated with adefovir dipivoxil (120 mg q.d.), another NtRTI. However, there have now been several reports of proximal renal tubular dysfunction and/or hypophosphatemia in patients taking tenofovir DF [17–19].
Tenofovir DF and intravenous tenofovir caused bone toxicity when administered in toxicology studies at high doses (6–12-times greater than the human area under the curve [AUC]) . However, clinically significant bone toxicity (e.g., as indicated by bone fractures) has not been observed in clinical trials. In study 903, through week 96, bone fractures occurred in 7 patients receiving stavudine, compared with 1 patient receiving tenofovir DF .
In reproductive studies in animals (rats and rabbits) at up to 19 times the human dose, there was no evidence of impaired fertility or harm to the fetus associated with tenofovir. The FDA has designated tenofovir DF “pregnancy category B,” indicating presumed safety, on the basis of animal studies. However, there have been no adequate and well-controlled studies involving pregnant women.
Pharmacology and drug interactions. Tenofovir DF has long serum and intracellular half-lives (∼17 h and 10–50 h, respectively), which allows once-daily dosing (appendix A). It is renally cleared and requires dose reduction in patients with significant renal impairment. Bioavailability is ∼40% higher when administered with a high-fat meal (1000 kcal and 50% fat) but is not appreciably affected with a lower-fat meal (373 kcal and 20% fat). Tenofovir DF is dosed without regard to food. Tenofovir is not a substrate, inducer, or inhibitor of the cytochrome p450 enzyme system. However, there are some interactions of clinical significance. Didanosine levels are increased when coadministered with tenofovir DF. The mechanism of this interaction is unknown, but it does not appear to be offset by taking both drugs together with food. However, when tenofovir DF is administered with a reduced dose of didanosine (250 mg of the enteric-coated formulation), the didanosine AUC is comparable to the AUCs of didanosine given at standard doses, regardless of whether the drugs are administered simultaneously or separated by 2 h, and regardless of whether they are taken with food or while fasting .
A drug interaction study has also been conducted evaluating the combination of tenofovir DF with ritonavir-boosted atazanavir (ritonavir, 100 mg q.d.; atazanavir, 300 mg q.d.) in 10 HIV-infected patients . Although most pharmacokinetic parameters did not reach statistical significance, the AUC of atazanavir was significantly reduced by 25%. However, because of the boosting effects of ritonavir, atazanavir trough concentrations remained 2–3-fold greater than those associated with atazanavir at 400 mg. Atazanavir levels are also reduced significantly when unboosted atazanavir is administered with tenofovir DF, suggesting that atazanavir should be boosted with ritonavir when combined with tenofovir DF, to maintain adequate atazanavir trough concentrations (Bristol-Myers Squibb, personal communication).
Hepatitis B virus (HBV)-HIV coinfection. Tenofovir DF is active in vivo against HBV, including virus with the lamivudine-selected YMDD mutation [23–26]. In the Gilead 903 trial, after 48 weeks, HBV-HIV-coinfected patients taking both tenofovir DF and lamivudine trended towards greater decreases in HBV DNA and alanine aminotransferase levels, were more likely to have undetectable HBV DNA levels, and were less likely to have lamivudine resistance than were those taking stavudine plus lamivudine .
The Role of Tenofovir in The Treatment of Hiv Infection
Treatment-experienced patients. Early use of tenofovir DF primarily involved treatment-experienced patients and was based on the data from the 902 and 907 studies and the needs of patients with limited treatment options. Those studies demonstrated that intensification of a failing regimen with tenofovir DF could lead to durable resuppression in a substantial proportion of patients, especially if tenofovir DF was added when the HIV RNA level was low. Both abacavir and tenofovir DF are logical choices for intensification of regimens that are not fully suppressive. However, both are subject to NRTI cross-resistance due to TAMs, which explains the lower potency observed in experienced patients. Moreover, “blind” intensification in patients with HIV RNA levels of >1000 copies/mL is usually inappropriate, because treatment decisions for such patients should be guided by the results of resistance tests. Tenofovir DF should not be considered a “salvage” agent, given that patients with extensive NRTI exposure and resistance are likely to have cross-resistance to tenofovir DF, especially if they have developed the 41L/210W TAM pattern in the course of treatment.
Treatment-naive patients and once-daily therapy. The 903 study demonstrates the potency and durability of tenofovir DF combined with efavirenz and lamivudine in treatment-naive patients, and this regimen is now included as one of the “preferred” regimens for initial therapy in the Department of Health and Human Services guidelines . Because it has a long half-life and can be administered once per day as a single tablet, tenofovir DF is attractive choice for use in initial therapy, and it is also a logical choice for use in once-daily regimens. Both of the combinations studied in the 903 study (efavirenz and lamivudine plus either tenofovir DF or the extended-release formulation of stavudine) can now be given in a once-daily dose, and many clinicians will feel comfortable extrapolating the study 903 results to once-daily dosing regimens. With appropriate dose adjustments, didanosine can also be administered once daily with tenofovir DF, though clinical data on this combination are lacking. Emtricitabine (FTC), atazanavir, and the combination of amprenavir plus ritonavir are already approved for once-daily administration, and abacavir and nevirapine have phamacokinetics that support once-daily dosing. A number of ritonavir-boosted protease inhibitor combinations may also have once-daily potential. Once-daily dosing is likely to be popular with patients, may improve adherence, and has obvious implications in settings where directly observed therapy programs are feasible, such as prisons and methadone clinics.
Toxicity and tolerability. Tenofovir DF has been well tolerated in clinical trials to date. With the possible exception of flatulence, adverse events are comparable to those seen with placebo. Long-term toxicity has not been observed thus far. The proximal renal tubular dysfunction that was observed with adefovir dipivoxil at a dose of 120 mg once per day has not been observed with tenofovir DF in clinical trials, although a small number of cases have been reported. The lack of mitochondrial toxicity is supported both by in vitro studies and clinical trial data, and lipid profiles are more favorable than those seen with stavudine. There were early concerns about loss of bone mineral density based on findings in animals treated with much higher doses. Data from DEXA scanning in the 903 study demonstrate greater bone loss in the tenofovir DF, lamivudine, and efavirenz arms, but the clinical significance remains unknown.
Resistance and sequencing. In treatment-experienced patients who add tenofovir DF to a failing regimen, resistance—specifically, the signature K65R mutation—is slow to develop. This mutation occurred in one-quarter of patients for whom a regimen of tenofovir DF, lamivudine, and efavirenz failed in the Gilead 903 trial, although, because of the low rate of failure in this trial, this represented only 2.7% of the patients treated with that regimen. K65R generally results in variably reduced susceptibility to abacavir and didanosine, as well as to tenofovir DF, and in preserved susceptibility to zidovudine and stavudine.
NRTI resistance mediated by TAMs occurs gradually and sequentially and is delayed by the presence of the M184V mutation, which always occurs before TAMs in lamivudine-containing combinations. Thus, it has been suggested that the combination of a thymidine analogue plus lamivudine is less vulnerable to resistance than is a regimen that includes a combination of lamivudine plus tenofovir DF, abacavir, or didanosine. It has also been argued that zidovudine should be included in a regimen that includes one of those drugs, because this would be expected to prevent the development of K65R.
These somewhat hypothetical resistance and sequencing concerns must be weighed against the very real potency, tolerability, and toxicity benefits of a regimen such as tenofovir DF, lamivudine, and efavirenz, which was associated with extremely low rates of virologic failure in the 903 trial. Although the routine substitution or addition of zidovudine could provide some protection against NRTI resistance for the small number of patients for whom therapy fails, it would be at the expense of increased toxicity or dosing inconvenience for all patients, including those for whom therapy is unlikely to fail. Moreover, patients who develop K65R continue to have options for NRTI therapy: their virus retains susceptibility to stavudine, is hypersusceptible to zidovudine, and has variable responses to abacavir, didanosine, and tenofovir DF.
Potential new roles for tenofovir DF. In clinical practice, tenofovir DF is already being used in ways that have not yet been studied in clinical trials. Concerns about decreased potency of the triple-nucleoside regimen, zidovudine, lamivudine, and abacavir have led many clinicians to add tenofovir DF to this combination, because it may improve potency without significantly altering tolerability or convenience. In fact, such a strategy may allow this combination to be given once per day, because 3 of the 4 drugs in the regimen have pharmacokinetics that support once-daily dosing. This combination is now being studied in clinical trials. The recent purchase of Triangle Pharmaceuticals by Gilead will lead to the development of coformulated tenofovir DF and emtricitabine, which will further expand options for simplified once-daily regimens with low pill burdens. It should be noted that the once-daily, triple-nucleoside combination of abacavir (600 mg), lamivudine (300 mg), and tenofovir DF (300 mg), which was thought to be potentially more effective than other triple-nucleoside regimens because of the potency and tolerability of the 3 agents, was associated with an unacceptably high rate of virologic failure in 2 trials [29, 30]. The reason for its poor performance is currently being investigated.
There are a number of other potential uses for tenofovir DF aside from long-term antiretroviral therapy of HIV-infected patients. Tenofovir DF is an attractive drug for use in postexposure prophylaxis regimens, given its convenience and tolerability. Data from monkey studies support the use of tenofovir DF for postexposure prophylaxis. Monkeys who were inoculated with simian immunodeficiency virus (SIV) and then given tenofovir up to 24 h after inoculation remained healthy and free of detectable SIV, whereas those who did not receive the drug died quickly of SIV infection .
Similarly, it may be a promising agent for use in pregnant women to prevent perinatal transmission. Single-dose therapy with nevirapine prevents transmission to infants, but this comes at the cost of the development of NNRTI resistance in a substantial proportion of the women who take it. Resistance to tenofovir DF is slower to develop, compared with NNRTI resistance, and it may leave more options for future therapy if resistance does occur. Tenofovir DF has also been proposed as an ideal agent for “preexposure prophylaxis,” the use of antiretroviral therapy (preferably for short periods of time) in individuals determined to be at risk of acquiring HIV infection on the basis high-risk behavior.
Finally, tenofovir DF is an important component of antiretroviral therapy, along with lamivudine or emtricitabine, in patients with HIV and HBV coinfection, given its activity against HBV and the risk of developing lamivudine-resistant HBV associated with long-term use of lamivudine monotherapy.
Fortunately, the dichotomy between potency on the one hand and convenience and tolerability on the other has become a thing of the past. It is now possible to prescribe highly potent multidrug regimens with simple once- or twice-daily dosing schedules that consist of a small number of pills. This is a welcome contrast to the complex, multiple-dose, food-restricted regimens that revolutionized the treatment of HIV disease in the late 1990s. Tenofovir DF is one of several agents that combine potency with convenience and tolerability. As with any new drug, we will need to continue to collect data on long-term toxicity before we can conclude that tenofovir DF is as benign as it appears to be on the basis of the preliminary results from clinical trials. We will also need to collect longer-term data on resistance and response to subsequent therapy in patients for whom therapy with tenofovir DF-containing regimens has failed. Such data will help to resolve unanswered questions about the optimal “sequencing” of NRTIs and NtRTIs. In the meantime, it is safe to conclude that tenofovir DF has an important role in the treatment of antiretroviral-naive and -experienced patients.
Pharmacokinetics and Dosing
• Single dose (300 mg), fasted state
• Oral bioavailability, 25%
• Time to maximum concentration (±SD), 1.0 ± 0.4 h
• Maximum concentration (Cmax; ±SD), 296 ± 90 ng/mL
• Area under the curve (AUC; ±SD), 2287 ± 685 ng × h/mL
• Multiple dose (300 mg), fed state (1000 kcal and 50% fat)
• Cmax, 326 ± 119 ng/mL
• AUC, 3324 ± 1370 ng × h/mL
• Plasma proteins, 0.7%
• Serum proteins, 7.2%
• Volume of distribution (±SD), 1.3 ± 0.6 L/kg
• Serum, 17 h
• Intracellular, 10–50 h
• Both glomerular filtration and tubular secretion, with 70%–80% eliminated unchanged in urine in 72 h
• No hepatic metabolism
Pharmacokinetic Drug-Drug Interactions
• No interaction with cytochrome P450 system
• Didanosine concentration AUC increased 44%–60%; Cmax increased 28%
• The AUC for atazanavir (300 mg q.d.) decreased 25% when administered with ritonavir (100 mg q.d.) and tenofovir DF (300 mg q.d.); the minimum concentration and AUC for atazanavir (400 mg q.d.) also decreased when administered with tenofovir DF (300 mg q.d.) without ritonavir
• Coadminstration with other drugs that are eliminated by tubular secretion, such as cidofovir, acyclovir, valacyclovir, ganciclovir, valaganciclovir, and probenecid, may increase serum concentrations of either tenofovir or the coadministered drug
Dosing in Adults
Usual. Three hundred mg per day without regard to food (equivalent to 245 mg tenofovir disoproxil).
With didanosine. Three hundred mg per day with 250 mg of enteric-coated didanosine (body weight, >60 kg). The AUC of didanosine with the 250-mg dose of enteric-coated didanosine coadministered with tenofovir is equivalent to that for a 400-mg dose of enteric-coated didanosine alone.
Renal insufficiency. Dosing interval adjustment is recommended for all patients with creatinine clearance of <50mL/min . Note that there are no safety or efficacy data with these dosing schedules. In animals, dose-limiting toxicities included nephrotoxicity and osteomalacia.
Hemodialysis. Three hundred mg once per week after completion of a total of 12 h of high-flux hemodialysis (determined on the basis of data from 4 h of dialysis 3 times per week).
Hepatic impairment. Pharmacokinetics have not been studied in patients with hepatic impairment; however, tenofovir and tenofovir DF are not metabolized by liver enzymes, so the impact of liver impairment should be limited.
The average wholesale price is $410 per month for 300 mg per day.