Sign Out
Pharmacodynamics: Exposure-Response Relationship in Treatment-Experienced Adult Subjects: The relationship between maraviroc, modeled plasma trough concentration (Cmin) (1 to 9 samples per subject taken on up to 7 visits), and virologic response was evaluated in 973 treatment-experienced HIV-1 infected subjects with varied optimized background antiretroviral regimens in Trials A4001027 and A4001028. The Cmin, baseline viral load, baseline CD4+ cell count, and overall sensitivity score (OSS) were found to be important predictors of virologic success (defined as viral load less than 400 copies per mL at 24 weeks). Table 1 illustrates the proportions of subjects with virologic success (%) within each Cmin quartile for 150-mg twice-daily and 300-mg twice-daily groups. (See Table 1.)
Click on icon to see table/diagram/imageExposure-Response Relationship in Treatment-Naive Adult Subjects: The relationship between maraviroc, modeled plasma trough concentration (Cmin) (1 to 12 samples per subject taken on up to 8 visits), and virologic response was evaluated in 294 treatment-naive HIV-1-infected subjects receiving maraviroc 300 mg twice daily in combination with lamivudine/zidovudine in Trial A4001026. Table 2 illustrates the proportion (%) of subjects with virologic success less than 50 copies per mL at 48 weeks within each Cmin quartile for the 300-mg twice-daily dose. (See Table 2.)
Click on icon to see table/diagram/imageEighteen of 75 (24%) subjects in Q1 had no measurable maraviroc concentration on at least one occasion versus 1 of 73 and 1 of 74 in Q3 and Q4, respectively.
Effects on Electrocardiogram: A placebo-controlled, randomized, crossover trial to evaluate the effect on the QT interval of healthy male and female volunteers was conducted with 3 single oral doses of maraviroc and moxifloxacin. The placebo-adjusted mean maximum (upper 1-sided 95% CI) increases in QTc from baseline after 100, 300, and 900 mg of maraviroc were -2 (0), -1 (1), and 1 (3) msec, respectively, and 13 (15) msec for moxifloxacin 400 mg. No subject in any group had an increase in QTc of greater than or equal to 60 msec from baseline. No subject experienced an interval exceeding the potentially clinically relevant threshold of 500 msec.
Clinical Studies: Clinical Studies in Adult Subjects: The clinical efficacy and safety of CELSENTRI are derived from analyses of data from 3 trials in adult subjects infected with CCR5-tropic HIV-1: Trials A4001027 and A4001028 in antiretroviral treatment-experienced adult subjects and Trial A4001026 in treatment-naive subjects. These trials were supported by a 48-week trial in antiretroviral treatment-experienced adult subjects infected with dual/mixed-tropic HIV-1, Trial A4001029.
Trials in CCR5-Tropic, Treatment-Experienced Subjects: Trials A4001027 and A4001028 were double-blind, randomized, placebo-controlled, multicenter trials in subjects infected with CCR5-tropic HIV-1. Subjects were required to have an HIV-1 RNA greater than 5,000 copies per mL despite at least 6 months of prior therapy with at least 1 agent from 3 of the 4 antiretroviral drug classes (greater than or equal to 1 NRTI, greater than or equal to 1 NNRTI, greater than or equal to 2 PIs, and/or enfuvirtide) or documented resistance to at least 1 member of each class. All subjects received an optimized background regimen consisting of 3 to 6 antiretroviral agents (excluding low-dose ritonavir) selected on the basis of the subject's prior treatment history and baseline genotypic and phenotypic viral resistance measurements. In addition to the optimized background regimen, subjects were then randomized in a 2:2:1 ratio to CELSENTRI 300 mg once daily, CELSENTRI 300 mg twice daily, or placebo. Doses were adjusted based on background therapy as described in Dosage & Administration, Table 10.
In the pooled analysis for Trials A4001027 and A4001028, the demographics and baseline characteristics of the treatment groups were comparable (see Table 3). Of the 1,043 subjects with a CCR5-tropism result at screening, 7.6% had a dual/mixed-tropism result at the baseline visit 4 to 6 weeks later. This illustrates the background change from CCR5- to dual/mixed-tropism result over time in this treatment-experienced population, prior to a change in antiretroviral regimen or administration of a CCR5 co-receptor antagonist. (See Table 3.)
Click on icon to see table/diagram/imageThe Week 48 results for the pooled Trials A4001027 and A4001028 are shown in Table 4. (See Table 4.)
Click on icon to see table/diagram/imageAfter 48 weeks of therapy, the proportions of subjects with HIV-1 RNA less than 400 copies per mL receiving CELSENTRI compared with placebo were 56% and 22%, respectively. The mean changes in plasma HIV-1 RNA from baseline to Week 48 were -1.84 log10 copies per mL for subjects receiving CELSENTRI + OBT compared with -0.78 log10 copies per mL for subjects receiving OBT only. The mean increase in CD4+ cell count was higher on CELSENTRI twice daily + OBT (124 cells per mm3) than on placebo + OBT (60 cells per mm3).
Trial in Dual/Mixed-Tropic, Treatment-Experienced Subjects: Trial A4001029 was an exploratory, randomized, double-blind, multicenter trial to determine the safety and efficacy of CELSENTRI in subjects infected with dual/mixed co-receptor tropic HIV-1. The inclusion/exclusion criteria were similar to those for Trials A4001027 and A4001028 as previously mentioned and the subjects were randomized in a 1:1:1 ratio to CELSENTRI once daily, CELSENTRI twice daily, or placebo. No increased risk of infection or HIV-1 disease progression was observed in the subjects who received CELSENTRI. Use of CELSENTRI was not associated with a significant decrease in HIV-1 RNA compared with placebo in these subjects and no adverse effect on CD4+ cell count was noted.
Trial in Treatment-Naive Subjects: Trial A4001026 was a randomized, double-blind, multicenter trial in subjects infected with CCR5-tropic HIV-1 classified by the original TROFILE tropism assay. Subjects were required to have plasma HIV-1 RNA greater than or equal to 2,000 copies per mL and could not have: 1) previously received any antiretroviral therapy for greater than 14 days, 2) an active or recent opportunistic infection or a suspected primary HIV-1 infection, or 3) phenotypic or genotypic resistance to zidovudine, lamivudine, or efavirenz. Subjects were randomized in a 1:1:1 ratio to CELSENTRI 300 mg once daily, CELSENTRI 300 mg twice daily, or efavirenz 600 mg once daily, each in combination with lamivudine/zidovudine. The efficacy and safety of CELSENTRI are based on the comparison of CELSENTRI twice daily versus efavirenz. In a pre-planned interim analysis at 16 weeks, CELSENTRI 300 mg once daily failed to meet the pre-specified criteria for demonstrating non-inferiority and was discontinued.
The demographic and baseline characteristics of the maraviroc and efavirenz treatment groups were comparable (Table 5). Subjects were stratified by screening HIV-1 RNA levels and by geographic region. The median CD4+ cell counts and mean HIV-1 RNA at baseline were similar for both treatment groups. (See Table 5.)
Click on icon to see table/diagram/imageThe treatment outcomes at 96 weeks for Trial A4001026 are shown in Table 6. Treatment outcomes are based on reanalysis of the screening samples using a more sensitive tropism assay, enhanced sensitivity TROFILE HIV tropism assay, which became available after the Week 48 analysis; approximately 15% of the subjects identified as CCR5-tropic in the original analysis had dual/mixed- or CXCR4-tropic virus. Screening with enhanced sensitivity version of the TROFILE tropism assay reduced the number of maraviroc virologic failures with CXCR4- or dual/mixed-tropic virus at failure to 12 compared with 24 when screening with the original TROFILE HIV tropism assay. (See Table 6.)
Click on icon to see table/diagram/imageThe median increase from baseline in CD4+ cell counts at Week 96 was 184 cells per mm3 for the arm receiving CELSENTRI compared with 155 cells per mm3 for the efavirenz arm.
Pharmacokinetics: See Table 7.
Click on icon to see table/diagram/imageAbsorption: Peak maraviroc plasma concentrations are attained 0.5 to 4 hours following single oral doses of 1 to 1,200 mg administered to uninfected volunteers. The pharmacokinetics of oral maraviroc are not dose proportional over the dose range.
The absolute bioavailability of a 100-mg dose is 23% and is predicted to be 33% at 300 mg. Maraviroc is a substrate for the efflux transporter P-gp.
Effect of Food on Oral Absorption: Coadministration of a 300-mg tablet with a high-fat breakfast reduced maraviroc Cmax and AUC by 33% in healthy adult volunteers. Studies with the tablet formulation demonstrated a reduced food effect at higher doses.
There were no food restrictions in the adult trials (using the tablet formulation) that demonstrated the efficacy/antiviral activity and safety of maraviroc [see Pharmacodynamics: Clinical Studies as previously mentioned].
Distribution: Maraviroc is bound (approximately 76%) to human plasma proteins, and shows moderate affinity for albumin and alpha-1 acid glycoprotein. The volume of distribution of maraviroc is approximately 194 L.
Elimination: Metabolism: Trials in humans and in vitro studies using human liver microsomes and expressed enzymes have demonstrated that maraviroc is principally metabolized by the cytochrome P450 system to metabolites that are essentially inactive against HIV-1. In vitro studies indicate that CYP3A is the major enzyme responsible for maraviroc metabolism. In vitro studies also indicate that polymorphic enzymes CYP2C9, CYP2D6, and CYP2C19 do not contribute significantly to the metabolism of maraviroc.
Maraviroc is the major circulating component (~42% drug-related radioactivity) following a single oral dose of 300 mg [14C]-maraviroc. The most significant circulating metabolite in humans is a secondary amine (~22% radioactivity) formed by N-dealkylation. This polar metabolite has no significant pharmacological activity. Other metabolites are products of mono-oxidation and are only minor components of plasma drug-related radioactivity.
Excretion: The terminal half-life of maraviroc following oral dosing to steady state in healthy subjects was 14 to 18 hours. A mass balance/excretion trial was conducted using a single 300-mg dose of 14C-labeled maraviroc. Approximately 20% of the radiolabel was recovered in the urine and 76% was recovered in the feces over 168 hours. Maraviroc was the major component present in urine (mean of 8% dose) and feces (mean of 25% dose). The remainder was excreted as metabolites.
Specific Populations: Patients with Hepatic Impairment: Maraviroc is primarily metabolized and eliminated by the liver. A trial compared the pharmacokinetics of a single 300-mg dose of CELSENTRI in subjects with mild (Child-Pugh Class A, n = 8) and moderate (Child-Pugh Class B, n = 8) hepatic impairment with pharmacokinetics in healthy subjects (n = 8). The mean Cmax and AUC were 11% and 25% higher, respectively, for subjects with mild hepatic impairment, and 32% and 46% higher, respectively, for subjects with moderate hepatic impairment compared with subjects with normal hepatic function. These changes do not warrant a dose adjustment. Maraviroc concentrations are higher when CELSENTRI 150 mg is administered with a potent CYP3A inhibitor compared with following administration of 300 mg without a CYP3A inhibitor, so patients with moderate hepatic impairment who receive CELSENTRI 150 mg with a potent CYP3A inhibitor should be monitored closely for maraviroc-associated adverse events. The pharmacokinetics of maraviroc have not been studied in subjects with severe hepatic impairment [see Hepatotoxicity under Precautions].
Patients with Renal Impairment: A trial compared the pharmacokinetics of a single 300-mg dose of CELSENTRI in adult subjects with severe renal impairment (CrCl less than 30 mL per minute, n = 6) and ESRD (n = 6) with healthy volunteers (n = 6). Geometric mean ratios for maraviroc Cmax and AUCinf were 2.4-fold and 3.2-fold higher, respectively, for subjects with severe renal impairment, and 1.7-fold and 2.0-fold higher, respectively, for subjects with ESRD as compared with subjects with normal renal function in this trial. Hemodialysis had a minimal effect on maraviroc clearance and exposure in subjects with ESRD. Exposures observed in subjects with severe renal impairment and ESRD were within the range observed in previous 300-mg single-dose trials of CELSENTRI in healthy volunteers with normal renal function. However, maraviroc exposures in the subjects with normal renal function in this trial were 50% lower than those observed in previous trials. Based on the results of this trial, no dose adjustment is recommended for patients with renal impairment receiving CELSENTRI without a potent CYP3A inhibitor or inducer. However, if patients with severe renal impairment or ESRD experience any symptoms of postural hypotension while taking CELSENTRI 300 mg twice daily, their dose should be reduced to 150 mg twice daily [see Recommended Dosage Recommendations in Patients with Renal Impairment under Dosage & Administration; Cardiovascular Events under Precautions].
In addition, the trial compared the pharmacokinetics of multiple-dose CELSENTRI in combination with saquinavir/ritonavir 1,000/100 mg twice daily (a potent CYP3A inhibitor combination) for 7 days in subjects with mild renal impairment (CrCl greater than 50 and less than or equal to 80 mL per minute, n = 6) and moderate renal impairment (CrCl greater than or equal to 30 and less than or equal to 50 mL per min, n = 6) with healthy volunteers with normal renal function (n = 6). Subjects received 150 mg of CELSENTRI at different dose frequencies (healthy volunteers - every 12 hours; mild renal impairment - every 24 hours; moderate renal impairment - every 48 hours). Compared with healthy volunteers (dosed every 12 hours), geometric mean ratios for maraviroc AUCtau, Cmax, and Cmin were 50% higher, 20% higher, and 43% lower, respectively, for subjects with mild renal impairment (dosed every 24 hours). Geometric mean ratios for maraviroc AUCtau, Cmax, and Cmin were 16% higher, 29% lower, and 85% lower, respectively, for subjects with moderate renal impairment (dosed every 48 hours) compared with healthy volunteers (dosed every 12 hours). Based on the data from this trial, no adjustment in dose is recommended for patients with mild or moderate renal impairment [see Recommended Dosage Recommendations in Patients with Renal Impairment under Dosage & Administration].
Geriatric Patients: Pharmacokinetics of maraviroc have not been fully evaluated in the elderly (aged 65 years and older). Based on population pharmacokinetic analyses, age did not have a clinically relevant effect on maraviroc exposure in subjects up to age 65 years [see Use in the Elderly under Precautions].
Race and Gender: Based on population pharmacokinetics and 2 clinical CYP3A5 genotype analyses for race, no dosage adjustment is recommended based on race or gender.
Drug Interaction Studies: Effect of Concomitant Drugs on the Pharmacokinetics of Maraviroc: Maraviroc is a substrate of CYP3A and P-gp and hence its pharmacokinetics are likely to be modulated by inhibitors and inducers of these enzymes/transporters. The CYP3A/P-gp inhibitors ketoconazole, boceprevir, lopinavir/ritonavir, ritonavir, darunavir/ritonavir, saquinavir/ritonavir, and atazanavir ± ritonavir all increased the Cmax and AUC of maraviroc (see Table 8). The CYP3A and/or P-gp inducers rifampin, etravirine, and efavirenz decreased the Cmax and AUC of maraviroc (see Table 8). While not studied, potent CYP3A and/or P-gp inducers carbamazepine, phenobarbital, and phenytoin are expected to decrease maraviroc concentrations. Based on in vitro study results, maraviroc is also a substrate of OATP1B1 and MRP2; its pharmacokinetics may be modulated by inhibitors of these transporters.
Tipranavir/ritonavir (net CYP3A inhibitor/P-gp inducer) did not affect the steady-state pharmacokinetics of maraviroc (see Table 8). Cotrimoxazole and tenofovir did not affect the pharmacokinetics of maraviroc. (See Table 8.)
Click on icon to see table/diagram/imageEffect of Maraviroc on the Pharmacokinetics of Concomitant Drugs: Maraviroc is unlikely to inhibit the metabolism of coadministered drugs metabolized by the following cytochrome P enzymes (CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C19, and CYP3A) or to inhibit the uptake of OATP1B1 or the export of MRP2 because maraviroc did not inhibit activity of those enzymes or transporters at clinically relevant concentrations in vitro. Maraviroc does not induce CYP1A2 in vitro. Additionally, in vitro studies have shown that maraviroc is not a substrate for, and does not inhibit, any of the major renal uptake inhibitors (organic anion transporter [OAT]1, OAT3, organic cation transporter [OCT]2, novel organic cation transporter [OCTN]1, and OCTN2) at clinically relevant concentrations.
In vitro results suggest that maraviroc could inhibit P-gp in the gut. However, maraviroc did not significantly affect the pharmacokinetics of digoxin in vivo, indicating maraviroc may not significantly inhibit or induce P-gp clinically.
Drug interaction trials were performed with maraviroc and other drugs likely to be coadministered or commonly used as probes for pharmacokinetic interactions (see Table 8).
Coadministration of fosamprenavir 700 mg/ritonavir 100 mg twice daily and maraviroc 300 mg twice daily decreased the Cmin and AUC of amprenavir by 36% and 35%, respectively. Coadministration of fosamprenavir 1,400 mg/ritonavir 100 mg once daily and maraviroc 300 mg once daily decreased the Cmin and AUC by 15% and 30%, respectively. No dosage adjustment is necessary when CELSENTRI is dosed 150 mg twice daily in combination with fosamprenavir/ritonavir dosed once or twice daily. Fosamprenavir should be given with ritonavir when coadministered with CELSENTRI.
Maraviroc had no significant effect on the pharmacokinetics of elvitegravir, boceprevir, zidovudine, or lamivudine. Maraviroc decreased the Cmin and AUC of raltegravir by 27% and 37%, respectively, which is not clinically significant. Maraviroc had no clinically relevant effect on the pharmacokinetics of midazolam, the oral contraceptives ethinylestradiol and levonorgestrel, no effect on the urinary 6β-hydroxycortisol/cortisol ratio, suggesting no induction of CYP3A in vivo. Maraviroc had no effect on the debrisoquine metabolic ratio (MR) at 300 mg twice daily or less in vivo and did not cause inhibition of CYP2D6 in vitro until concentrations greater than 100 microM. However, there was 234% increase in debrisoquine MR on treatment compared with baseline at 600 mg once daily, suggesting potential inhibition of CYP2D6 at higher doses.
Toxicology: Nonclinical Toxicology: Carcinogenesis: Long-term oral carcinogenicity studies of maraviroc were carried out in rasH2 transgenic mice (6 months) and in rats for up to 96 weeks (females) and 104 weeks (males). No drug-related increases in tumor incidence were found in mice at 1,500 mg per kg per day and in male and female rats at 900 mg per kg per day. The highest exposures in rats were approximately 11 times those observed in humans at the therapeutic dose of 300 mg twice daily for the treatment of HIV-1 infection.
Mutagenesis: Maraviroc was not genotoxic in the reverse mutation bacterial test (Ames test in Salmonella and E. coli), a chromosome aberration test in human lymphocytes, and mouse bone marrow micronucleus test.
Impairment of Fertility: Maraviroc did not impair mating or fertility of male or female rats and did not affect sperm of treated male rats at approximately 20-fold higher exposures (AUC) than in humans given the recommended 300-mg twice-daily dose.
Microbiology: Mechanism of Action: Maraviroc is a member of a therapeutic class called CCR5 co-receptor antagonists. Maraviroc selectively binds to the human chemokine receptor CCR5 present on the cell membrane, preventing the interaction of HIV-1 gp120 and CCR5 necessary for CCR5-tropic HIV-1 to enter cells. CXCR4-tropic and dual-tropic HIV-1 entry is not inhibited by maraviroc.
Antiviral Activity in Cell Culture: Maraviroc inhibits the replication of CCR5-tropic laboratory strains and primary isolates of HIV-1 in models of acute peripheral blood leukocyte infection. The mean EC50 value (50% effective concentration) for maraviroc against HIV-1 group M isolates (subtypes A to J and circulating recombinant form AE) and group O isolates ranged from 0.1 to 4.5 nM (0.05 to 2.3 ng per mL) in cell culture.
When used with other antiretroviral agents in cell culture, the combination of maraviroc was not antagonistic with non-nucleoside reverse transcriptase inhibitors (NNRTIs: delavirdine, efavirenz, and nevirapine), NRTIs (abacavir, didanosine, emtricitabine, lamivudine, stavudine, tenofovir, zalcitabine, and zidovudine), or protease inhibitors (PIs: amprenavir, atazanavir, darunavir, indinavir, lopinavir, nelfinavir, ritonavir, saquinavir, and tipranavir). Maraviroc was not antagonistic with the HIV-1 gp41 fusion inhibitor enfuvirtide. Maraviroc was not active against CXCR4-tropic and dual-tropic viruses (EC50 value greater than 10 microM). The antiviral activity of maraviroc against HIV-2 has not been evaluated.
Resistance in Cell Culture: HIV-1 variants with reduced susceptibility to maraviroc have been selected in cell culture, following serial passage of 2 CCR5-tropic viruses (CC1/85 and RU570). The maraviroc-resistant viruses remained CCR5-tropic with no evidence of a change from a CCR5-tropic virus to a CXCR4-using virus. Two amino acid residue substitutions in the V3-loop region of the HIV-1 envelope glycoprotein (gp160), A316T, and I323V (HXB2 numbering), were shown to be necessary for the maraviroc-resistant phenotype in the HIV-1 isolate CC1/85. In the RU570 isolate a 3-amino acid residue deletion in the V3 loop, △QAI (HXB2 positions 315 to 317), was associated with maraviroc resistance. The relevance of the specific gp120 substitutions observed in maraviroc-resistant isolates selected in cell culture to clinical maraviroc resistance is not known. Maraviroc-resistant viruses were characterized phenotypically by concentration-response curves that did not reach 100% inhibition in phenotypic drug assays, rather than increases in EC50 values.
Cross-Resistance in Cell Culture: Maraviroc had antiviral activity against HIV-1 clinical isolates resistant to NNRTIs, NRTIs, PIs, and the gp41 fusion inhibitor enfuvirtide in cell culture (EC50 values ranged from 0.7 to 8.9 nM [0.36 to 4.57 ng per mL]). Maraviroc-resistant viruses that emerged in cell culture remained susceptible to enfuvirtide and the protease inhibitor saquinavir.
Clinical Resistance: Virologic failure on maraviroc can result from genotypic and phenotypic resistance to maraviroc, through outgrowth of undetected CXCR4-using virus present before maraviroc treatment (see Tropism as follows), through resistance to background therapy drugs (see Table 9), or due to low exposure to maraviroc [see Pharmacology: Pharmacodynamics as previously mentioned].
Antiretroviral Treatment-Experienced Adult Subjects (Trials A4001027 and A4001028): Week 48 data from treatment-experienced subjects failing maraviroc-containing regimens with CCR5-tropic virus (n = 58) have identified 22 viruses that had decreased susceptibility to maraviroc characterized in phenotypic drug assays by concentration-response curves that did not reach 100% inhibition. Additionally, CCR5-tropic virus from 2 of these treatment-failure subjects had greater than or equal to 3-fold shifts in EC50 values for maraviroc at the time of failure.
Fifteen of these viruses were sequenced in the gp120 encoding region and multiple amino acid substitutions with unique patterns in the heterogeneous V3 loop region were detected. Changes at either amino acid position 308 or 323 (HXB2 numbering) were seen in the V3 loop in 7 of the subjects with decreased maraviroc susceptibility. Substitutions outside the V3 loop of gp120 may also contribute to reduced susceptibility to maraviroc.
Antiretroviral Treatment-Naive Adult Subjects (Trial A4001026): Treatment-naive subjects receiving CELSENTRI had more virologic failures and more treatment-emergent resistance to the background regimen drugs compared with those receiving efavirenz (see Table 9).
Click on icon to see table/diagram/imageIn an as-treated analysis of treatment-naive subjects at 96 weeks, 32 subjects failed a maraviroc-containing regimen with CCR5-tropic virus and had a tropism result at failure; 7 of these subjects had evidence of maraviroc phenotypic resistance defined as concentration-response curves that did not reach 95% inhibition. One additional subject had a greater than or equal to 3-fold shift in the EC50 value for maraviroc at the time of failure. A clonal analysis of the V3 loop amino acid envelope sequences was performed from 6 of the 7 subjects. Changes in V3 loop amino acid sequence differed between each of these different subjects, even for those infected with the same virus clade, suggesting that there are multiple diverse pathways to maraviroc resistance. The subjects who failed with CCR5-tropic virus and without a detectable maraviroc shift in susceptibility were not evaluated for genotypic resistance.
Of the 32 maraviroc virologic failures failing with CCR5-tropic virus, 20 (63%) also had genotypic and/or phenotypic resistance to background drugs in the regimen (lamivudine, zidovudine).
Tropism: In both treatment-experienced and treatment-naive subjects, detection of CXCR4-using virus prior to initiation of therapy has been associated with a reduced virologic response to maraviroc.
Antiretroviral Treatment-Experienced Subjects (Trials A4001027 and A4001028): In the majority of cases, treatment failure on maraviroc was associated with detection of CXCR4-using virus (i.e., CXCR4- or dual/mixed-tropic) which was not detected by the tropism assay prior to treatment. CXCR4-using virus was detected at failure in approximately 55% of subjects who failed treatment on maraviroc by Week 48, as compared with 9% of subjects who experienced treatment failure in the placebo arm. To investigate the likely origin of the on-treatment CXCR4-using virus, a detailed clonal analysis was conducted on virus from 20 representative subjects (16 subjects from the maraviroc arms and 4 subjects from the placebo arm) in whom CXCR4-using virus was detected at treatment failure. From analysis of amino acid sequence differences and phylogenetic data, it was determined that CXCR4-using virus in these subjects emerged from a low level of pre-existing CXCR4-using virus not detected by the tropism assay (which is population-based) prior to treatment rather than from a coreceptor switch from CCR5-tropic virus to CXCR4-using virus resulting from mutation in the virus.
Detection of CXCR4-using virus prior to initiation of therapy has been associated with a reduced virological response to maraviroc. Furthermore, subjects failing twice-daily maraviroc at Week 48 with CXCR4-using virus had a lower median increase in CD4+ cell counts from baseline (+41 cells per mm3) than those subjects failing with CCR5-tropic virus (+162 cells per mm3). The median increase in CD4+ cell count in subjects failing in the placebo arm was +7 cells per mm3.
Antiretroviral Treatment-Naive Subjects (Trial A4001026): In a 96-week trial of antiretroviral treatment-naive subjects, 14% (12 of 85) who had only CCR5-tropic virus at screening with an enhanced sensitivity tropism assay (TROFILE) and failed therapy on maraviroc had CXCR4-using virus at the time of treatment failure. A detailed clonal analysis was conducted in 2 previously antiretroviral treatment-naive subjects enrolled in a Phase 2a monotherapy trial who had CXCR4-using virus detected after 10 days' treatment with maraviroc. Consistent with the detailed clonal analysis conducted in treatment-experienced subjects, the CXCR4-using variants appear to emerge from outgrowth of a pre-existing undetected CXCR4-using virus. Screening with an enhanced sensitivity tropism assay reduced the number of maraviroc virologic failures with CXCR4- or dual/mixed-tropic virus at failure to 12 compared with 24 when screening with the original tropism assay. All but one (11 of 12; 92%) of the maraviroc failures failing with CXCR4- or dual/mixed-tropic virus also had genotypic and phenotypic resistance to the background drug lamivudine at failure and 33% (4 of 12) developed zidovudine-associated resistance substitutions.
Subjects who had only CCR5-tropic virus at baseline and failed maraviroc therapy with CXCR4-using virus had a median increase in CD4+ cell counts from baseline of +113 cells per mm3 while those subjects failing with CCR5-tropic virus had an increase of +135 cells per mm3. The median increase in CD4+ cell count in subjects failing in the efavirenz arm was +95 cells per mm3.
