Source: European Medicines Agency (EU) Revision Year: 2022 Publisher: ViiV Healthcare BV, Van Asch van Wijckstraat 55H, 3811 LP Amersfoort, Netherlands
Pharmacotherapeutic group: Antivirals for systemic use, antivirals for treatment of HIV infections, combinations
ATC Code: J05AR04
Abacavir, lamivudine and zidovudine are all NRTIs, and are potent selective inhibitors of HIV-1 and HIV-2.All three medicinal products are metabolised sequentially by intracellular kinases to the respective 5′-triphosphate (TP). Lamivudine-TP, carbovir-TP (the active triphosphate form of abacavir) and zidovudine-TP are substrates for and competitive inhibitors of HIV reverse transcriptase (RT). However, their main antiviral activity is through incorporation of the monophosphate form into the viral DNA chain, resulting in chain termination. Abacavir, lamivudine and zidovudine triphosphates show significantly less affinity for host cell DNA polymerases.
No antagonistic effects in vitro were seen with lamivudine and other antiretrovirals (tested agents: abacavir, didanosine and nevirapine). No antagonistic effects in vitro were seen with zidovudine and other antiretrovirals (tested agents: didanosine and interferon-alpha). The antiviral activity of abacavir in cell culture was not antagonized when combined with the nucleoside reverse transcriptase inhibitors (NRTIs) didanosine, emtricitabine, stavudine or tenofovir, the non-nucleoside reverse transcriptase inhibitor (NNRTI) nevirapine, or the protease inhibitor (PI) amprenavir.
HIV-1 resistance to lamivudine involves the development of a M184I or, more commonly, M184V amino acid change close to the active site of the viral RT.
Abacavir-resistant isolates of HIV-1 have been selected in vitro and are associated with specific genotypic changes in the RT codon region (codons M184V, K65R, L74V and Y115F). Viral resistance to abacavir develops relatively slowly in vitro, requiring multiple mutations for a clinically relevant increase in EC50 over wild-type virus.
The M184V or M184I variants arise in HIV-1 infected patients treated with lamivudine-containing antiretroviral therapy. Most patients experiencing virological failure with a regimen containing abacavir in a pivotal clinical trial with Combivir (fixed dose combination of lamivudine and zidovudine) showed either no NRTI-related changes from baseline (15%) or only M184V or M184I selection (78%). The overall selection frequency for M184V or M184I was high (85%), and selection of L74V, K65R and Y115F was not observed (see Table). Thymidine analogue mutations (TAMs) which are selected by zidovudine (ZDV) were also found (8%).
| Therapy | Abacavir + Combivir |
|---|---|
| Number of Subjects | 282 |
| Number of Virological Failures | 43 |
| Number of On-Therapy Genotypes | 40 (100%) |
| K65R | 0 |
| L74V | 0 |
| Y115F | 0 |
| M184V/I | 34 (85%) |
| TAMs1 | 3 (8%) |
1 Number of subjects with ≥1 TAM.
TAMs might be selected when thymidine analogs are associated with abacavir. In a meta-analysis of six clinical trials, TAMs were not selected by regimens containing abacavir without zidovudine (0/127), but were selected by regimens containing abacavir and the thymidine analogue zidovudine (22/86, 26%). In addition, the selection of L74V and K65R was reduced when co-administered with ZDV (K65R: without ZDV: 13/127, 10%; with ZDV: 1/86, 1%; L74V: without ZDV: 51/127, 40%; with ZDV: 2/86, 2%).
The M184V or M184I variants arise in HIV-1 infected patients treated with lamivudine-containing antiretroviral therapy and confers high-level resistance to lamivudine. In vitro data tend to suggest that the continuation of lamivudine in anti-retroviral regimen despite the development of M184V might provide residual anti-retroviral activity (likely through impaired viral fitness). The clinical relevance of these findings is not established. Indeed, the available clinical data are very limited and preclude any reliable conclusion in the field. In any case, initiation of susceptible NRTIs should always be preferred to maintenance of lamivudine therapy. Therefore, maintaining lamivudine therapy despite emergence of M184V mutation should only be considered in cases where no other active NRTIs are available. Similarly, the presence of TAMs gives rise to resistance to ZDV.
Clinically significant reduction of susceptibility to abacavir has been demonstrated in clinical isolates of patients with uncontrolled viral replication, who have been pre-treated with and are resistant to other nucleoside inhibitors. In a meta-analysis of five clinical trials where abacavir was added to intensify therapy, of 166 subjects, 123 (74%) had M184V/I, 50 (30%) had T215Y/F, 45 (27%) had M41L, 30 (18%) had K70R and 25 (15%) had D67N. K65R was absent and L74V and Y115F were uncommon (≤3%). Logistic regression modelling of the predictive value for genotype (adjusted for baseline plasma HIV-1 RNA [vRNA], CD4+ cell count, number and duration of prior antiretroviral therapies) showed that the presence of 3 or more NRTI resistance-associated mutations was associated with reduced response at Week 4 (p=0.015) or 4 or more mutations at median Week 24 (p≤0.012). In addition, the 69 insertion complex or the Q151M mutation, usually found in combination with A62V, V75I, F77L and F116Y, cause a high level of resistance to abacavir.
| Baseline Reverse Transcriptase Mutation | Week 4 (n=166) | ||
|---|---|---|---|
| n | Median Change vRNA (log10 c/mL) | Percent with <400 copies/mL vRNA | |
| None | 15 | -0.96 | 40% |
| M184V alone | 75 | -0.74 | 64% |
| Any one NRTI mutation | 82 | -0.72 | 65% |
| Any two NRTI- associated mutations | 22 | -0.82 | 32% |
| Any three NRTI- associated mutations | 19 | -0.30 | 5% |
| Four or more NRTI- associated mutations | 28 | -0.07 | 11% |
Phenotypic resistance to abacavir requires M184V with at least one other abacavir-selected mutation, or M184V with multiple TAMs. Phenotypic cross-resistance to other NRTIs with M184V or M184I mutation alone is limited. Zidovudine, didanosine, stavudine and tenofovir maintain their antiretroviral activities against such HIV-1 variants. The presence of M184V with K65R does give rise to crossresistance between abacavir, tenofovir, didanosine and lamivudine, and M184V with L74V gives rise to cross-resistance between abacavir, didanosine and lamivudine. The presence of M184V with Y115F gives rise to cross-resistance between abacavir and lamivudine. Appropriate use of abacavir can be guided using currently recommended resistance algorithms.
Cross-resistance between abacavir, lamivudine or zidovudine and antiretrovirals from other classes e.g. PIs or NNRTIs is unlikely.
One randomised, double blind, placebo controlled clinical study has compared the combination of abacavir, lamivudine and zidovudine to the combination of indinavir, lamivudine and zidovudine in treatment naive patients. Due to the high proportion of premature discontinuation (42% of patients discontinued randomised treatment by week 48), no definitive conclusion can be drawn regarding the equivalence between the treatment regimens at week 48. Although a similar antiviral effect was observed between the abacavir and indinavir containing regimens in terms of proportion of patients with undetectable viral load (≤400 copies/mL; intention to treat analysis (ITT), 47% versus 49%; as treated analysis (AT), 86% versus 94% for abacavir and indinavir combinations respectively), results favoured the indinavir combination, particularly in the subset of patients with high viral load (>100,000 copies/mL at baseline; ITT, 46% versus 55%; AT, 84% versus 93% for abacavir and indinavir respectively).
ACTG5095 was a randomised (1:1:1), double-blind, placebo-controlled trial performed in 1147 antiretroviral naïve HIV-1 infected adults, comparing 3 regimens: zidovudine (ZDV), lamivudine (3TC), abacavir (ABC), efavirenz (EFV) vs ZDV/3TC/EFV vs ZDV/3TC/ABC. After a median follow-up of 32 weeks, the tritherapy with the three nucleosides ZDV/3TC/ABC was shown to be virologically inferior to the two other arms regardless of baseline viral load (< or >100 000 copies/mL) with 26% of subjects on the ZDV/3TC/ABC arm, 16% on the ZDV/3TC/EFV arm and 13% on the 4 drug arm categorised as having virological failure (HIV RNA >200 copies/mL). At week 48 the proportion of subjects with HIV RNA <50 copies/mL were 63%, 80% and 86% for the ZDV/3TC/ABC, ZDV/3TC/EFV and ZDV/3TC/ABC/EFV arms, respectively. The study Data Safety Monitoring Board stopped the ZDV/3TC/ABC arm at this time based on the higher proportion of patients with virologic failure. The remaining arms were continued in a blinded fashion. After a median follow-up of 144 weeks, 25% of subjects on the ZDV/3TC/ABC/EFV arm and 26% on the ZDV/3TC/EFV arm were categorised as having virological failure. There was no significant difference in the time to first virologic failure (p=0.73, log-rank test) between the 2 arms. In this study, addition of ABC to ZDV/3TC/EFV did not significantly improve efficacy.
| ZDV/3TC/ABC | ZDV/3TC/EFV | ZDV/3TC/ABC/EFV | ||
|---|---|---|---|---|
| Virologic failure (HIV RNA >200 copies/mL) | 32 weeks | 26% | 16% | 13% |
| 144 weeks | - | 26% | 25% | |
| Virologic success (48 weeks HIV RNA <50 copies/mL) | 63% | 80% | 86% | |
In antiretroviral-naive patients treated with a combination of abacavir, lamivudine, zidovudine and efavirenz in a small, ongoing, open label pilot study, the proportion of patients with undetectable viral load (<400 copies/mL) was approximately 90% with 80% having <50 copies/mL after 24 weeks of treatment.
Currently there are no data on the use of Trizivir in heavily pre-treated patients, patients failing on other therapies or patients with advanced disease (CD4 cells <50 cells/mm³).
The degree of benefit of this nucleoside combination in heavily pre-treated patients will depend on the nature and duration of prior therapy that may have selected for HIV-1 variants with cross-resistance to abacavir, lamivudine or zidovudine.
To date there are insufficient data on the efficacy and safety of Trizivir given concomitantly with NNRTIs or PIs.
Abacavir, lamivudine and zidovudine are rapidly and well absorbed from the gastro-intestinal tract following oral administration. The absolute bioavailability of oral abacavir, lamivudine and zidovudine in adults is about 83%, 80-85% and 60-70% respectively.
In a pharmacokinetic study in HIV-1 infected patients, the steady state pharmacokinetic parameters of abacavir, lamivudine and zidovudine were similar when either Trizivir alone or the combination tablet lamivudine/zidovudine and abacavir in combination were administered, and also similar to the values obtained in the bioequivalence study of Trizivir in healthy volunteers.
A bioequivalence study compared Trizivir with abacavir 300 mg, lamivudine 150 mg and zidovudine 300 mg taken together. The effect of food on the rate and extent of absorption was also studied. Trizivir was shown to be bioequivalent to abacavir 300 mg, lamivudine 150 mg and zidovudine 300 mg given as separate tablets for AUC0-∞ and Cmax. Food decreased the rate of absorption of Trizivir (slight decrease Cmax (mean 18-32%) and increase tmax (approximately 1 hour), but not the extent of absorption (AUC0-∞). These changes are not considered clinically relevant and no food restrictions are recommended for administration of Trizivir.
At a therapeutic dose (one Trizivir tablet twice daily) in patients, the mean (CV) steady-state Cmax of abacavir, lamivudine and zidovudine in plasma are 3.49 µg/mL (45%), 1.33 µg/mL (33%) and 1.56 µg/mL (83%), respectively. Corresponding values for Cmin could not be established for abacavir and are 0.14 µg/mL (70%) for lamivudine and 0.01 µg/mL (64%) for zidovudine. The mean (CV) AUCs for abacavir, lamivudine and zidovudine over a dosing interval of 12 hours are 6.39 µg.h/mL (31%), 5.73 µg.h/mL (31%) and 1.50 µg.h/mL (47%), respectively.
A modest increase in Cmax (28%) was observed for zidovudine when administered with lamivudine, however overall exposure (AUC) was not significantly altered. Zidovudine has no effect on the pharmacokinetics of lamivudine. An effect of abacavir is observed on zidovudine (Cmax reduced with 20%) and on lamivudine (Cmax reduced with 35%).
Intravenous studies with abacavir, lamivudine and zidovudine showed that the mean apparent volume of distribution is 0.8, 1.3 and 1.6l/kg respectively. Lamivudine exhibits linear pharmacokinetics over the therapeutic dose range and displays limited binding to the major plasma protein albumin (<36% serum albumin in vitro). Zidovudine plasma protein binding is 34% to 38%. Plasma protein binding studies in vitro indicate that abacavir binds only low to moderately (~49%) to human plasma proteins at therapeutic concentrations. This indicates a low likelihood for interactions with other medicinal products through plasma protein binding displacement.
Interactions involving binding site displacement are not anticipated with Trizivir.
Data show that abacavir, lamivudine and zidovudine penetrate the central nervous system (CNS) and reach the cerebrospinal fluid (CSF). The mean ratios of CSF/serum lamivudine and zidovudine concentrations 2-4 hours after oral administration were approximately 0.12 and 0.5 respectively. The true extent of CNS penetration of lamivudine and its relationship with any clinical efficacy is unknown.
Studies with abacavir demonstrate a CSF to plasma AUC ratio of between 30 to 44%. The observed values of the peak concentrations are 9 fold greater than the IC50 of abacavir of 0.08 µg/mL or 0.26 µM when abacavir is given at 600 mg twice daily.
Metabolism of lamivudine is a minor route of elimination. Lamivudine is predominately cleared by renal excretion of unchanged lamivudine. The likelihood of metabolic drug interactions with lamivudine is low due to the small extent of hepatic metabolism (5-10%) and low plasma binding.
The 5'-glucuronide of zidovudine is the major metabolite in both plasma and urine, accounting for approximately 50-80% of the administered dose eliminated by renal excretion. 3'-amino-3'-deoxythymidine (AMT) has been identified as a metabolite of zidovudine following intravenous dosing.
Abacavir is primarily metabolised by the liver with approximately 2% of the administered dose being renally excreted, as unchanged compound. The primary pathways of metabolism in man are by alcohol dehydrogenase and by glucuronidation to produce the 5'-carboxylic acid and 5'-glucuronide which account for about 66% of the dose excreted in the urine.
The observed lamivudine half-life of elimination is 18 to 19 hours. The mean systemic clearance of lamivudine is approximately 0.32 l/h/kg, with predominantly renal clearance (>70%) via the organic cationic transport system. Studies in patients with renal impairment show lamivudine elimination is affected by renal dysfunction. Dose reduction is required for patients with creatinine clearance ≤30 mL/min (see section 4.2).
From studies with intravenous zidovudine, the mean terminal plasma half-life was 1.1 hours and the mean systemic clearance was 1.6 l/h/kg. Renal clearance of zidovudine is estimated to be 0.34 l/h/kg, indicating glomerular filtration and active tubular secretion by the kidneys. Zidovudine concentrations are increased in patients with advanced renal failure.
The mean half-life of abacavir is about 1.5 hours. Following multiple oral doses of abacavir 300 mg twice a day there is no significant accumulation of abacavir. Elimination of abacavir is via hepatic metabolism with subsequent excretion of metabolites primarily in the urine. The metabolites and unchanged abacavir account for about 83% of the administered abacavir dose in the urine the remainder is eliminated in the faeces.
Pharmacokinetic data has been obtained for abacavir, lamivudine and zidovudine separately. Limited data in patients with cirrhosis suggest that accumulation of zidovudine may occur in patients with hepatic impairment because of decreased glucuronidation. Data obtained in patients with moderate to severe hepatic impairment show that lamivudine pharmacokinetics are not significantly affected by hepatic dysfunction.
Abacavir is metabolised primarily by the liver. The pharmacokinetics of abacavir have been studied in patients with mild hepatic impairment (Child-Pugh score 5-6) receiving a single 600 mg dose; the median (range) AUC value was 24.1 (10.4 to 54.8) ug.h/mL. The results showed that there was a mean (90%CI) increase of 1.89 fold [1.32; 2.70] in the abacavir AUC, and 1.58 [1.22; 2.04] fold in the elimination half-life. No definitive recommendation on dose reduction is possible in patients with mild hepatic impairment due to substantial variability of abacavir exposure in this patient population. Based on data obtained for abacavir, Trizivir is not recommended in patients with moderate or severe hepatic impairment.
The observed lamivudine half-life of elimination is 5 to 7 hours. The mean systemic clearance of lamivudine is approximately 0.32 l/h/kg, with predominantly renal clearance (>70%) via the organic cationic transport system. Studies in patients with renal impairment show lamivudine elimination is affected by renal dysfunction.
From studies with intravenous zidovudine, the mean terminal plasma half-life was 1.1 hours and the mean systemic clearance was 1.6 l/h/kg. Renal clearance of zidovudine is estimated to be 0.34 l/h/kg, indicating glomerular filtration and active tubular secretion by the kidneys. Zidovudine concentrations are increased in patients with advanced renal failure.
Abacavir is primarily metabolised by the liver with approximately 2% of abacavir excreted unchanged in the urine. The pharmacokinetics of abacavir in patients with end-stage renal disease is similar to patients with normal renal function, and, therefore, no dose reduction is required in patients with renal impairment.
As dose adjustments of lamivudine and zidovudine may be necessary it is recommended that separate preparations of abacavir, lamivudine and zidovudine be administered to patients with severe renal impairment (creatinine clearance ≤30 mL/min). Trizivir is contraindicated in patients with end-stage renal disease (see section 4.3).
No pharmacokinetic data are available in patients over 65 years of age.
There are no data available on treatment with the combination of abacavir, lamivudine and zidovudine in animals. The clinically relevant toxicological effects of these three medicinal products are anaemia, neutropenia and leukopenia.
Neither abacavir, lamivudine nor zidovudine is mutagenic in bacterial tests, but consistent with other nucleoside analogues, they inhibit cellular DNA replication in in vitro mammalian tests such as the mouse lymphoma assay.
Lamivudine has not shown any genotoxic activity in the in vivo studies at doses that gave plasma concentrations up to 40-50 times higher than clinical plasma levels. Zidovudine showed clastogenic effects in oral repeated dose micronucleus tests in mice and rats. Peripheral blood lymphocytes from AIDS patients receiving zidovudine treatment have also been observed to contain higher numbers of chromosome breakages.
A pilot study has demonstrated that zidovudine is incorporated into leukocyte nuclear DNA of adults, including pregnant women, taking zidovudine as treatment for HIV-1 infection, or for the prevention of mother to child viral transmission. Zidovudine was also incorporated into DNA from cord blood leukocytes of infants from zidovudine-treated mothers. A transplacental genotoxicity study conducted in monkeys compared zidovudine alone with the combination of zidovudine and lamivudine at humanequivalent exposures. The study demonstrated that foetuses exposed in utero to the combination sustained a higher level of nucleoside analogue-DNA incorporation into multiple foetal organs, and showed evidence of more telomere shortening than in those exposed to zidovudine alone. The clinical significance of these findings is unknown.
Abacavir has a weak potential to cause chromosomal damage both in vitro and in vivo at high test concentrations and therefore any potential risk to man must be balanced against the expected benefits of treatment.
The carcinogenic potential of a combination of abacavir, lamivudine and zidovudine has not been tested. In long-term oral carcinogenicity studies in rats and mice, lamivudine did not show any carcinogenic potential. In oral carcinogenicity studies with zidovudine in mice and rats, late appearing vaginal epithelial tumours were observed. A subsequent intravaginal carcinogenicity study confirmed the hypothesis that the vaginal tumours were the result of long term local exposure of the rodent vaginal epithelium to high concentrations of unmetabolised zidovudine in urine. There were no other zidovudine-related tumours observed in either sex of either species.
In addition, two transplacental carcinogenicity studies have been conducted in mice. In one study, by the US National Cancer Institute, zidovudine was administered at maximum tolerated doses to pregnant mice from day 12 to 18 of gestation. One year postnatally, there was an increase in the incidence of tumours in the lung, liver and female reproductive tract of offspring exposed to the highest dose level (420 mg/kg term body weight).
In a second study, mice were administered zidovudine at doses up to 40 mg/kg for 24 months, with exposure beginning prenatally on gestation day 10. Treatment related findings were limited to lateoccurring vaginal epithelial tumours, which were seen with a similar incidence and time of onset as in the standard oral carcinogenicity study. The second study thus provided no evidence that zidovudine acts as a transplacental carcinogen.
It is concluded that as the increase in incidence of tumours in the first transplacental carcinogenicity study represents a hypothetical risk, this should be balanced against the proven therapeutic benefit. Carcinogenicity studies with orally administered abacavir in mice and rats showed an increase in the incidence of malignant and non-malignant tumours. Malignant tumours occurred in the preputial gland of males and the clitoral gland of females of both species, and in rats in the thyroid gland of males and and in the liver, urinary bladder, lymph nodes and the subcutis of females.
The majority of these tumours occurred at the highest abacavir dose of 330 mg/kg/day in mice and 600 mg/kg/day in rats. The exception was the preputial gland tumour which occurred at a dose of 110 mg/kg in mice. The systemic exposure at the no effect level in mice and rats was equivalent to 3 and 7 times the human systemic exposure during therapy.
While the clinical relevance of these findings is unknown, these data suggest that a carcinogenic risk to humans is outweighed by the potential clinical benefit.
In toxicology studies abacavir was shown to increase liver weights in rats and monkeys. The clinical relevance of this is unknown. There is no evidence from clinical studies that abacavir is hepatotoxic. Additionally, autoinduction of abacavir metabolism or induction of the metabolism of other medicinal products hepatically metabolised has not been observed in man.
Mild myocardial degeneration in the heart of mice and rats was observed following administration of abacavir for two years. The systemic exposures were equivalent to 7 to 24 times the expected systemic exposure in humans. The clinical relevance of this finding has not been determined.
Lamivudine was not teratogenic in animal studies but there were indications of an increase in early embryonic deaths in the rabbit at relatively low systemic exposures, comparable to those achieved in humans. A similar effect was not seen in rats even at very high systemic exposure.
Zidovudine had a similar effect in both species, but only at very high systemic exposures. At maternally toxic doses, zidovudine given to rats during organogenesis resulted in an increased incidence of malformations, but no evidence of foetal abnormalities was observed at lower doses.
Abacavir demonstrated toxicity to the developing embryo and foetus in rats, but not in rabbits. These findings included decreased foetal body weight, foetal oedema, and an increase in skeletal variations/malformations, early intra-uterine deaths and still births. No conclusion can be drawn with regard to the teratogenic potential of abacavir because of this embryo-foetal toxicity.
A fertility study in the rat has shown that abacavir had no effect on male or female fertility. Likewise, neither lamivudine nor zidovudine had any effect on fertility. Zidovudine has not been shown to affect the number of sperm, sperm morphology and motility in man.
© All content on this website, including data entry, data processing, decision support tools, "RxReasoner" logo and graphics, is the intellectual property of RxReasoner and is protected by copyright laws. Unauthorized reproduction or distribution of any part of this content without explicit written permission from RxReasoner is strictly prohibited. Any third-party content used on this site is acknowledged and utilized under fair use principles.
