icon star paper   HIV Articles  
Back grey_arrow_rt.gif
The Role of Nucleoside and Nucleotide Reverse Transcriptase Inhibitor Backbones in Antiretroviral Therapy
  Part III
JAIDS Journal of Acquired Immune Deficiency Syndromes: Volume 37 Supplement 1 1 September 2004
Young, Benjamin MD, PhD
From the Rose Medical Center, Denver, CO
Several dual nucleoside reverse transcriptase inhibitor (NRTI) combinations provide efficacy when combined with a 3rd agent. However, there are a number of issues with current NRTI and nucleotide reverse transcriptase inhibitor (NtRTI) combinations that often lead to treatment failure and limited treatment options. These issues include suboptimal potency, drug interactions, toxicities, tolerability issues, and selection of resistance mutations that confer cross-resistance. Options for simplified NRTI backbones include fixed-dose combinations and agents that allow once-daily dosing; however, once-daily treatment choices are currently limited because of a lack of data on potential combinations. This article provides an historical perspective on the use of NRTI backbones in the treatment of HIV infection and outlines the advantages and disadvantages of currently available backbone combinations. In addition, it provides a brief introduction to backbone combinations under investigation as potential options for initial therapy. In an environment where several NRTI/NtRTI backbones offer comparable efficacy, treatment decisions will increasingly be made based on toxicity, resistance, and convenience considerations.
Early nucleoside reverse transcriptase inhibitor (NRTI) backbone combinations improved upon the results seen with nucleoside monotherapy and included lamivudine (3TC) + zidovudine (ZDV), ZDV + didanosine (ddI), 3TC + stavudine (d4T), and ddI + d4T. More recently, clinical trials examining these dual-NRTI combinations with a protease inhibitor (PI), nonnucleoside reverse transcriptase inhibitor (NNRTI), or 3rd NRTI documented sustained reductions in viral load and a normalization of CD4 cell counts. Data on these combinations suggested that while powerful NRTI backbones work well in a variety of settings, the 3rd drug is the most important determinant of the efficacy of the regimen.
Several recent clinical studies highlight the potency of specific nucleoside/nucleotide backbones in combination antiretroviral therapy (ART). ACTG 384 provided additional evidence of differences in efficacy and tolerability among NRTI backbones. In this study, time to first regimen failure (P = 0.002) and time to first virologic failure (P = 0.0006) favored fixed-dose ZDV/3TC combined with efavirenz (EFV) compared with other 3-drug regimens. In a post hoc analysis, the median time to severe toxicity was shorter for ZDV/3TC than for d4T + ddI—sounds like misprint, should say time to severe toxicity was shorter for d4T/ddI-- (104 weeks for ZDV/3TC and 52 weeks for d4T + ddI). Based on this study and others, ZDV/3TC + EFV was recommended as a preferred regimen in the ART guidelines published by the US Department of Health and Human Services (DHHS). Gilead study 903 showed that the nucleotide reverse transcriptase inhibitor (NtRTI) tenofovir DF (TDF), when combined with 3TC and EFV, showed similar efficacy compared with d4T + 3TC + EFV but had fewer NRTI-associated toxicities and led to more favorable lipid profiles. These data have prompted inclusion of the TDF + 3TC + EFV regimen in the list of preferred initial regimens recommended by the DHHS.
A number of toxicity issues associated with current NRTI combinations may influence treatment options, adherence, durability of effect, and ultimate treatment outcome. Some NRTIs, particularly the dideoxynucleosides, have been associated with peripheral neuropathy, pancreatitis, lipoatrophy, lactic acidemia and acidosis, and hepatic steatosis. d4T, in particular, has been associated with lipoatrophy. HIV-associated neuromuscular weakness syndrome has also been associated with d4T to a greater extent than with other NRTIs. Tenofovir DF may be associated with renal toxicity, especially in patients with preexisting renal dysfunction. ZDV has been associated with myopathy, anemia, and neutropenia, although the incidence of these side effects was greater earlier in the era of monotherapy, when the doses of ZDV commonly prescribed were higher than those recommended today. Emtricitabine (FTC) has been associated with hyperpigmentation of the palms and soles. Abacavir (ABC) is associated with a hypersensitivity reaction that occurs in approximately 5% of treated patients.
Overlapping toxicity profiles among NRTIs can lead to additive toxicities in some cases. For example, the incidence of peripheral neuropathy varied among a number of major clinical trials using different NRTI backbones and was higher with d4T-containing regimens, particularly with the backbone ddI + d4T. As noted above, a post hoc analysis of ACTG 384 found that the median time to severe toxicity was 104 weeks for ZDV + 3TC and 52 weeks for d4T + ddI.6 Note that, as shown in the Kaplan-Meier curves, differences in toxicity were not apparent for the first 6-9 months of treatment. The fact that treatment-limiting toxicities may emerge after several months of treatment underscores the importance of studying toxicity profiles and making treatment decisions that are informed by a data-driven understanding of the toxicity profiles of the individual drugs and the combination regimens under consideration. In view of these data, it is now recommended that the combination of d4T and ddI be avoided because of the high rate of mitochondrial DNA dysfunction and associated increased risk for toxicity.
Gilead study 903, a study conducted to evaluate the safety and efficacy of TDF, has provided additional insight into the toxicity profile of the comparator agent, d4T. This study compared TDF with d4T, both combined with 3TC + EFV, in ART-naive patients. An interim 96-week analysis showed a significantly higher (P< 0.001) rate of certain adverse events in the d4T arm compared with the TDF arm, including peripheral neuritis/neuropathy (TDF, 3%; d4T, 10%) and investigator-defined lipodystrophy (TDF, 1%; d4T, 12%). The incidence of lactic acidosis was also higher in the former arm (TDF, 0%; d4T, 1%), but the difference was not statistically significant. Examination of the change in lipids in this study revealed statistically significant differences in all parameters, favoring the TDF arm. Increases in fasting triglycerides, total cholesterol, low-density lipoprotein (LDL) cholesterol, and high-density lipoprotein (HDL) cholesterol were 4 mg/dL, 30 mg/dL, 11 mg/dL, and 9 mg/dL, respectively, in the TDF arm and 104 mg/dL, 52 mg/dL, 20 mg/dL, and 7 mg/dL, respectively, in the d4T arm.
The toxicity profile of particular drugs is not the only risk factor for the emergence of treatment-related toxicity. Host factors may also play an important role. For example, Lichtenstein et al. studied the effect of CD4 count nadir as a risk factor for the development of lipoatrophy in the HIV Outpatient Study (HOPS) cohort. The study prospectively evaluated patients for clinical signs of lipoatrophy at 2 visits, 21 months apart. Of 337 patients with no lipoatrophy at survey 1, 44 (13.1%) developed moderate or severe lipoatrophy 21 months later. In multivariate analyses, significant risk factors for the emergence of treatment-related toxicity included white race, CD4 count at survey 2 <100 cells/mm3, and body mass index <24 kg/m2. When controlling for severity of HIV disease, use of ART or time on ART was not a significant risk factor. CD4 count and other factors associated with previous or current severity of HIV infection had the strongest association with incidence of lipoatrophy. These data provide an important counterpoint to the suggestion that ART should be delayed as long as possible to defer the potential for treatment-emergent toxicities. Since CD4 cell count almost invariably declines in HIV-infected individuals over time, and a low CD4 cell count is associated with the emergence of treatment-related side effects, it is important to weigh the potential risks of delaying treatment when evaluating whether to initiate ART in a given patient. As indicated by these data, these risks include not only those associated with the progression of HIV disease itself but also the variable response to therapy associated with delayed initiation of therapy.
As indicated above, overlapping drug toxicity profiles can influence the emergence of treatment-related side effects. In some cases, the increased incidence of treatment-emergent side effects may be due to simple additive toxicity. In other cases, drug-drug interactions may have important implications for treatment-emergent toxicity. One such interaction that has been studied recently is that between TDF and ddI. When ddI is coadministered with TDF, the area under the curve (AUC) increases by 46% for ddI taken without food and by 60% for ddI taken with food. In contrast, ddI has no effect on the AUC of TDF. Recent data suggest that the 250-mg dose of ddI EC yields similar plasma levels to the 400-mg dose without TDF, and it is currently recommended that dosage of ddI EC be reduced to 250 mg daily in patients weighing >60 kg who are also taking TDF. There are no data on the effect of a ddI dosage adjustment for patients weighing <60 kg. ddI is associated with peripheral neuropathy and pancreatitis; TDF alone is not associated with these side effects. However, through cohort studies and clinical experience, it has been observed that coadministration of ddI with TDF appears to potentiate these ddI-related toxicities. Young et al. performed a systematic search of all patients in the HOPS cohort who received ddI + TDF for >14 days prior to July 2003. A total of 207 patients were identified. Of these, 68% received high-dose ddI, 32% received low-dose ddI (250 mg once daily), and 26% were switched from high- to low-dose ddI. The investigators found that peripheral neuropathy occurred in 11.4% of those on high-dose and 2.5% of those on low-dose ddI. Pancreatitis occurred in 4.4 and 0% of patients, respectively. Although the individual differences in pancreatitis and peripheral neuropathy did not reach statistical significance, the aggregate endpoint of any discontinuation was statistically higher for the high-dose ddI patients.
Tenofovir DF is a fairly recent addition to the class and the first NtRTI to receive Food and Drug Administration (FDA) approval. Emerging data about its pharmacokinetics are of particular interest. The pharmacokinetics of TDF are altered in patients with renal impairment. The effects of lopinavir/ritonavir (LPV/r) coadministration on TDF: Tenofovir AUC increases by 32% with this combination; tenofovir minimum concentration (Cmin) increases by 51%; and tenofovir maximum concentration increases by 15%. Plasma concentrations of LPV/r were unaffected by the combination. A recent presentation suggested no apparent excessive risk of renal toxicity among persons who received TDF compared with those who received other antiretroviral therapies. Nevertheless, because of drug interactions between TDF, ddI, and LPV/r, prospective study is required to determine if there is an increased risk of TDF-related renal toxicities or ddI-related pancreatitis and hyperlactatemia. In studies in which TDF was combined with an RTI and either LPV/r or atazanavir (ATV), no renal toxicity was observed.
Pharmacokinetic interactions with the combination of TDF, ddI EC, and the PI ATV have also been evaluated.19 With this combination, the AUC of ATV decreased by 26% and the AUC of TDF increased by 25% compared with the AUC of the respective single agents, whereas ddI exposure did not change. The Cmin of TDF increased 25%, the Cmin of ddI increased 89%, and the Cmin of ATV decreased 39%. The mechanism of the 2-way interaction between ATV and TDF is not known. This combination has shown efficacy and safety in treatment-experienced patients when RTV is included.
There are a number of resistance issues with current NRTI/NtRTI combinations that may influence treatment options, durability of effect, and ultimate outcomes. These include the resistance profiles associated with failure of individual NRTIs/NtRTIs and of particular combinations, and the effect of the interplay between agents within a regimen. An important early resistance analysis of NRTI therapy was conducted by Kuritzkes et al.20 as part of the NUCA 3001 study.21 In this study, despite the rapid emergence of 3TC resistance, combination therapy with 3TC + ZDV resulted in greater reduction in plasma viral load over 24 weeks than ZDV monotherapy. Emergence of the K70R mutation, a thymidine analogue mutation (TAM) associated with ZDV resistance in previous studies, was significantly reduced in patients who received ZDV/3TC compared with patients receiving ZDV alone.20 These data provided the first suggestion that coadministration of 3TC might delay or prevent the emergence of ZDV resistance and contribute to the sustained antiretroviral activity of the 3TC/ZDV combination.20
Resistance patterns vary as a consequence of not only the NRTI/NtRTI backbone but also the other drug(s) in the regimen. The term nucleoside-associated mutations (NAMs) includes both TAMs and other mutations that appear to be associated with exposure to multiple nucleoside compounds and whose presence in various combinations may confer cross-resistance to NRTIs/NtRTIs. A number of studies have evaluated the emergence of NAMs. Regimens containing a ZDV/3TC backbone in combination with various 3rd agents have been studied to determine the genotypic and phenotypic resistance profiles of these combinations. In these studies, the M184V mutation emerged frequently, but TAMs emerged considerably less frequently overall. The presence of the M184V mutation alone in a patient with limited virologic rebound may indicate an adherence issue. Indeed, in one study the presence of M184V did not correlate with increased rates of treatment failure. Continued treatment during an extended period with detectable viremia will eventually lead to the selection of TAMs, but close monitoring usually allows time for treatment intervention, if necessary, before TAMs appear.
As indicated above, the NUCA 3001 study suggested that the presence of the M184V mutation delays or prevents the emergence of TAMs. Other data suggest that the emergence of NAMs occurs less frequently in 3TC-containing regimens than in regimens that do not contain 3TC. One such study is the Agence Nationale de Recherches sur le SIDA (ANRS) 070 (ALBI) study by Picard et al.,29 a randomized controlled trial in which ART-naive patients received dual-NRTI therapy with either d4T + ddI or ZDV/3TC. Resistance testing was performed at baseline and again after virologic failure on isolates obtained from 21 patients in each treatment group. The T215Y mutation, a TAM, emerged in 61.9% of patients receiving d4T + ddI and in 9.5% of patients receiving ZDV/3TC (P < 0.0001). In addition, the Q151M mutation, which is associated with resistance to multiple NRTIs, was selected in 14% of patients receiving d4T + ddI but in none of those receiving ZDV/3TC. Among patients who experienced virologic failure on ZDV/3TC, 100% of isolates carried the M184V mutation.29 These data suggest that the combination of d4T + ddI is associated more frequently with the emergence of NAMs, including TAMs and multidrug resistance mutations, than is the combination of ZDV/3TC. It should be noted, however, that this study was conducted in the setting of dual-NRTI therapy. While this study provides valuable information about differential resistance patterns among NRTI backbones, triple combination therapies have been shown to provide superior potency and a lower incidence of resistance.
Data from Gilead Study 903 shed some light on the emergence of NRTI/NtRTI and NNRTI resistance with either 3TC + d4T + EFV or 3TC + TDF + EFV.30 In this study, EFV resistance was the most frequent type of resistance to emerge at 144 weeks (the difference in the rate of EFV resistance between the 2 arms was not statistically significant). The M184V mutation occurred at the same rate in both arms. The K65R mutation, which is associated with TDF resistance, was present at virologic failure in 17% of those experiencing virologic failure on TDF, representing 2.7% of those randomly assigned to the TDF arm. The K65R mutation did not occur alone; it occurred only in combination with EFV resistance mutations with or without the M184V mutation.
The ZODIAC (Ziagen Once Daily in Antiretroviral Combination Therapy) study provides some insight into the emergence of resistance among patients receiving regimens containing a backbone of ABC + 3TC.31 Patients in this study received ABC once or twice daily combined with 3TC and EFV (n = 770). Few patients experienced virologic failure (once daily, 9.9%; twice daily, 8.3%). Among those with virologic failure, a large proportion had low viral loads, which allowed investigators to perform genotypic testing and phenotypic testing on only 56 and 57%, respectively. The numbers of emergent mutations were similar in each group. The most common resistance mutations were M184V/I (48%), K103N (45%), and L74V (26%). There was 1 incident case of the K65R mutation. Importantly, phenotypic susceptibility to TDF, ZDV, and d4T was retained in all instances.
Options for simplified NRTI backbones include fixed-dose combinations. The ZDV/3TC fixed-dose combination tablet (Combivir; GlaxoSmithKline) is dosed as 1 pill twice daily with or without food and has no drug interactions with NNRTIs or PIs. The ABC/3TC/ZDV fixed-dose combination tablet (Trizivir; GlaxoSmithKline) is dosed as 1 pill twice daily with or without food and has no drug interactions with NNRTIs or PIs. The availability of these fixed-dose combinations has made it possible to construct compact and convenient twice-daily regimens. Currently in development are fixed-dose combinations of TDF/FTC and ABC/3TC, both of which are expected to be dosed once daily. Currently, ABC is dosed twice daily; however, pharmacokinetic data suggest that ABC may be dosed once per day (see article by Piliero on PK in this issue).
The availability of fixed-dose combination ABC/3TC/ZDV offered the option of a compact, convenient, well-tolerated triple-nucleoside regimen with the potential to achieve levels of viral suppression comparable to those of PI-containing regimens. In a study designed prior to FDA approval of the fixed-dose triple combination, Vibhagool et al.32 showed that ABC + ZDV/3TC was comparable to indinavir (IDV) + ZDV/3TC in a 48-week open-label study in ART-naive adults. In this study, ABC + ZDV/3TC was comparable to IDV + ZDV/3TC in all viral load strata, including among patients with HIV RNA >100,000 copies/mL at baseline. The triple-nucleoside arm performed better in patients with baseline viral loads of <100,000 copies/mL than in those with higher baseline viral loads. An adherence substudy found that significantly more patients (P < 0.001) reported >95% adherence in the ABC + ZDV/3TC arm (72%) than in the IDV + ZDV/3TC arm (45%); in addition, median adherence was predictive of virologic outcome (P < 0.01). However, a subsequent study, ACTG 5095, raised concerns about the efficacy of the triple-nucleoside regimen.34 ACTG 5095 was a double-blind study of ABC/3TC/ZDV vs. EFV + ZDV/ 3TC vs. EFV + ABC/3TC/ZDV (n = 1147). After an average of 32 weeks, a total of 167 study volunteers experienced virologic failure: 21% in the group receiving ABC/3TC/ZDV alone vs. 10% in the other 2 groups combined. Virologic failure occurred sooner and more often in those receiving ABC/3TC/ZDV alone, regardless of baseline viral load. Although the response rates of patients on the triple-NRTI regimen in this study were similar to those seen in previous clinical trials of other standard HAART regimens, the National Institute of Allergy and Infectious Diseases Data and Safety Monitoring Board recommended that the ABC/3TC/ZDV-alone arm be stopped because of the superior response of the pooled EFV arms.
Despite the results of ACTG 5095, interest remains in the potential for using triple-nucleoside combinations in appropriate clinical situations, e.g., in patients with lower baseline viral loads or as maintenance therapy in patients whose viral loads were initially suppressed on a PI- or NNRTI-containing regimen. Moyle and Gazzard conducted a study of ABC/3TC/ZDV maintenance therapy in patients with high baseline viral loads. This was a retrospective clinical cohort analysis of 32 patients with pretreatment HIV RNA level of >100,000 copies/mL. Patients whose HIV RNA dropped to <50 copies/mL for a median of 28 months were switched to ABC/3TC/ZDV. At a median of 6.5 months, ABC/3TC/ZDV maintained effective virologic suppression regardless of high baseline viral load or low CD4 cell counts.
The hypersensitivity reaction to ABC, which is seen in approximately 5% of patients, requires that the drug be immediately and permanently discontinued. Symptoms generally resolve within 24 hours after discontinuation. After discontinuation, rechallenge should never be attempted, as this can result in serious, even life-threatening reactions.
In addition to fixed-dose combinations, options for simplified NRTI backbones include agents that allow once-daily dosing. The TDF + 3TC combination is the only once-daily backbone currently listed as a preferred dual-NRTI combination when combined with EFV, based on the results of the Gilead 903 trial. TDF + 3TC is not currently recommended for use as a backbone in PI-containing regimens, although data are accumulating demonstrating the safety and efficacy of TDF + 3TC and TDF/FTC in combination with PIs (e.g., LPV/r and ATV/r). Fixed-dose combination NRTI backbones in development include ABC/3TC and TDF/FTC. Studies have demonstrated the efficacy of ABC/3TC dosed once daily with EFV. In addition, there are studies showing the efficacy of ABC/3TC in initial therapy dosed twice daily with NRTIs, PIs, and NNRTIs. Gathe et al. reported 48-week results of a study evaluating treatment initiation with TDF + FTC + either LPV/r twice daily (n = 115) or LPV/r once daily (n = 75). Baseline median viral load among all patients was approximately 65,000 copies/mL; baseline median CD4 count was 214 cells/uL. In an intent-to-treat analysis at week 48, 70 and 64% of patients in the once-daily and twice-daily arms, respectively, achieved viral loads <50 copies/mL. Total cholesterol and triglyceride levels increased by <2% in each group; HDL cholesterol rose by >25%; and there was no change in the LDL/HDL ratio in either group. Moderate-to-severe diarrhea occurred in 16 vs. 5% of those in the once-daily and twice-daily arms, respectively (P = 0.04).17 One patient in each arm developed acute renal failure with a rise in creatinine to >3.0 mg/dL. The first patient, an elderly man, had a baseline creatinine clearance of 40 mL/min, which now requires a TDF dose reduction, not the standard at the time. The second patient, a 54-year-old man, experienced renal failure at week 38. Both patients showed improvement upon discontinuation of the study drug. A head-to-head study is currently under way to evaluate the safety and efficacy of a TDF/FTC fixed-dose combination compared with the fixed-dose combination of ZDV/3TC.
A number of studies have demonstrated good results with a once-daily NRTI backbone of ddI + 3TC; however, most of these studies have been uncontrolled. In study FTC-301, the combination of ddI and FTC was shown to be superior to ddI and d4T when combined with an NNRTI, though this comparator regimen is now recognized to be less potent and more toxic than standard backbones and is no longer recommended for use.
Limited data are available on other once-daily NRTI combinations, including TDF + ddI, TDF + ABC, and ddI + ABC, and questions remain about the efficacy, safety and tolerability, and resistance profiles of these combinations. Recent studies demonstrating high rates of virologic failure with the combinations of ABC + 3TC + TDF and ddI + 3TC + TDF raise concerns about combining drugs that select for the K65R mutation, especially in treatment-naive patients.
Several dual-NRTI combinations offer durable efficacy as well as good safety, tolerability, and convenience when combined with a 3rd agent. The current (March 2004) DHHS guidelines recommend ZDV/3TC or d4T + 3TC for use with LPV/r and recommend ZDV/3TC, d4T + 3TC, or TDF + 3TC as NRTI backbones of choice for use with EFV in initial therapy. FTC is a recommended alternative to 3TC in all the preferred PI- or NNRTI-containing regimens listed here. Other alternative NNRTI-containing regimens include 3TC or FTC in combination with ddI or ABC and nevirapine (NVP) or EFV. ABC is a recommended alternative to ZDV or d4T in combination with 3TC or FTC and LPV/r. Other PI-containing alternative regimens include 3TC or FTC with ZDV, d4T, or ABC and ATV, fosamprenavir (FPV), FPV/r, IDV/r, NFV, or saquinavir/ritonavir (SQV/r). Triple-NRTI regimens, ABC + 3TC + ZDV or d4T are listed as alternatives to preferred or alternative NNRTI and PI regimens.
Long-term toxicity risks, convenience issues, or problematic resistance profiles have been identified with certain NRTI backbones. For example, d4T and ddI should not be combined, as this combination appears to cause a high rate of mitochondrial DNA dysfunction and associated toxicity risk. Lipoatrophy has been associated with NRTI therapy, particularly with regimens containing d4T, and, to a lesser degree, with ZDV and ddI. The combination of ZDV and ddI has a problematic resistance profile and may be inconvenient because ddI is taken on an empty stomach and ZDV is better tolerated with food.
Promising new options are on the horizon. Once-daily combinations of existing NRTIs may increase convenience and facilitate greater adherence in some patients. New fixed-dose combinations may also improve convenience and promote increased adherence. The NRTI backbone is likely to remain the cornerstone of antiretroviral regimens, as both well-established and novel backbones are combined with potent, tolerable, and convenient PIs, NNRTIs, and other 3rd agents, and as new QD and fixed-dose NRTI backbone options become available.
  icon paper stack View Older Articles   Back to Top   www.natap.org