Pharmacokinetics in HIV-1-Infected Infants and Children

A phase-I single-dose pharmacokinetic study evaluated two oral abacavir doses (4 and 8 mg/kg of body weight) in 22 HIV-1-infected children ages 3 months to 13 years.³³ Abacavir was rapidly absorbed, with time to peak concentration in plasma occurring within 1.5 h postdosing. Abacavir was rapidly eliminated, with a mean elimination half-life of 0.98–1.13 h. The extent of exposure to abacavir appears to be slightly lower in children than in adults, with the comparable unit doses being based on body weight.³³

In Vitro Effects

Abacavir and carbovir are less myelotoxic than either zidovudine, didanosine, or zalcitabine when tested in vitro using human and murine hematopoietic progenitor cells.^2,5 Carbovir triphosphate has relatively low inhibitory activity against cellular DNA polymerase γ when compared to the active triphosphates of other dideoxynucleoside analogs.⁵ It is the inhibition of cellular DNA polymerase γ that has been associated with drug-induced peripheral neuropathy and mitochondrial toxicity (including lipoatrophy/lipodystrophy). Taken together, the preclinical development of abacavir sulfate was supported by both a relative lack of myelosuppressive effects and less inhibition of cellular enzymes involved in development of peripheral neuropathy and other mitochondrial toxicities.

In Vivo Effects

In clinical trials to date, the majority of adverse events were rated as mild or moderate. No specific hepatic, pancreatic, renal, or bone marrow toxicity patterns have been described. In ACTG 5095, adding abacavir to a backbone regimen of combined zidovudine plus lamivudine plus efavirenz had no consistent metabolic effects in HIV-1-infected adults as measured by fasting triglyceride and total, LDL, and HDL cholesterol, lactate, glucose, and insulin resistance measurements.³⁴ Mitochondrial toxicity (including lipoatrophy) occurring in association with treatment with abacavir may be lower than the risk with many other NRTIs. However, lactic acidosis and severe hepatomegaly with steatosis, including fatal cases, have been reported with the use of nucleoside analogs alone or in combination, including abacavir, zidovudine, lamivudine, and other antiretrovirals. In a Commmunity Program for Clinical Research on AIDS (CPCRA) study, 96 antiretroviral-naive subjects were enrolled and randomized to didanosine plus stavudine (n = 46) or abacavir plus lamivudine (n = 50).³⁵ Those subjects receiving didanosine plus stavudine experienced a decline in regional fat and total body fat when compared to increases in these parameters in subjects in the abacavir plus lamivudine arm (P < 0.05). Abacavir plus lamivudine therapy increased the rate of change of high-density lipoprotein cholesterol, whereas this rate decreased during didanosine plus stavudine therapy. Early and sustained increases in insulin resistance were seen only during didanosine plus stavudine therapy.³⁵ In lipoatrophic HIV-1-infected adults enrolled in the RAVE study, switching from a thymidine analog to abacavir or tenofovir for 48 weeks improved lipoatrophy.³⁶ Replacing the PI component of treatment regimens improved the lipid profile in 90 treatment-experienced subjects with controlled plasma viremia who were receiving nevirapine (n = 29), efavirenz (n = 32), or abacavir (n = 29) in the 24-month Spanish NEFA study, but did not seem to be effective for reversing body fat abnormalities.³⁷ After 1 year, a trend toward a lower incidence of adverse effects leading to discontinuation among those subjects receiving abacavir was seen, but also a trend toward a higher rate of virological failure was noted.

A separate 48-week, randomized, open-label study of HIV-1-infected subjects compared three abacavir-base substitution approaches in the management of dyslipidemia and peripheral lipoatrophy.³⁸ Subjects were receiving initial stavudine with either a PI or NNRTI and had hypercholesterolemia and/or lipoatrophy with well-controlled plasma viremia (plasma HIV-1 RNA ≤50 copies/mL). Thirty subjects were randomized to replace: (1) stavudine with abacavir; (2) PI or NNRTI with abacavir; or (3) stavudine and PI or NNRTI with abacavir plus zidovudine. Replacement of stavudine with abacavir led to modest improvements in fat mass in this study, and replacement of a PI or NNRTI with abacavir led to modest improvements in both cholesterol and triglycerides.

Gastrointestinal and Neurologic Side Effects

The most common side effects seen during abacavir therapy are gastrointestinal (GI) and neurologic. Nausea, vomiting, and diarrhea have been ascribed to abacavir therapy. Nausea is common. All of these GI complaints tend to abate over the first few weeks of abacavir therapy. Neurologic side effects are less common than GI complaints. Headache, malaise, dizziness, insomnia, and rarely paresthesias have been noted during abacavir therapy. In placebo-controlled clinical trials, only nausea, vomiting, and malaise and fatigue were statistically significant and more frequent in abacavir-containing regimens in comparison to the control regimens.

Abacavir Hypersensitivity Syndrome

Hypersensitivity reactions occurred in ∼8% of 2670 patients (n = 206) receiving abacavir for the treatment of HIV-1 infection in nine clinical trials (range: 2–9%) enrolled from Nov 1999 to Feb 2002 (Abacavir FDA prescribing information). The appearance of this syndrome demands immediate discontinuation and an absolute contraindication to reinitiation of abacavir or an abacavir-containing medication (e.g., Trizivir or Epzicom). A total of 1015 cases have been described among 26 769 subjects enrolled in phase II and III clinical trials and expanded access programs.³⁹ The median time to onset was 11 days, with 93% of cases appearing within 6 weeks of initiation of abacavir, although reactions may occur at any time during therapy. In a risk factor meta-analysis for hypersensitivity reactions to abacavir introduction, the incidence observed was 3.7%. The odds ratios (OR) for subjects with previous HAART treatment and those of African descent were significantly lower (OR 0.41, 95% confidence intervals (CIs): 0.23, 0.72; and OR 0.51, 95% CI: 0.28, 0.92, respectively) than HAART-naive patients and other ethnic origins.⁴⁰

The mechanism of abacavir hypersensitivity syndrome is not fully understood, but may involve an immunologic response. GlaxoSmithKline is conducting pharmacogenetic research to determine if cellular genetic DNA polymorphisms (variations) can be identified in HIV-1-infected patients who have developed hypersensitivity reactions following treatment with abacavir, in order to help predict risk of abacavir hypersensitivity in susceptible patients.⁴¹ An association between HLA-B*5701, HLA-DR7, and HLA-DQ3 and the hypersensitivity syndrome has been reported.^42,43 A strong association between this syndrome and HLA-B*5701 carriage in Caucasians has been confirmed.⁴⁴ Conversely, a lower reported rate of suspected abacavir hypersensitivity reactions among African-American patients across five randomized controlled trials has been attributed to a lower prevalence of HLA-B*5701 among African-Americans.⁴⁵ A cutaneous patch test may help identify subjects with true abacavir hypersensitivity syndrome, since the syndrome may be difficult for clinicians to recognize.⁴⁶ HLA-B*5701 was present in all seven patch test-positive subjects with remote abacavir hypersensitivity syndrome versus one of the 11 controls tolerating abacavir (P < 0.001). Five of seven patch tests (71%) versus one of 11 controls (9%) (P = 0.005) showed significant abacavir-specific CD8+ T-cell proliferation, suggesting a direct role for HLA-B*5701-restricted CD8+ T cells in the pathophysiology of abacavir hypersensitivity syndrome.⁴⁷

The usual presentation includes symptoms indicating involvement of multiple organ systems.⁴⁸ The most common symptoms are fever, rash, GI symptoms (nausea, vomiting, diarrhea, or abdominal pain), and malaise or fatigue. Fever is a key feature of the syndrome. Less common symptoms include respiratory symptoms (cough, sore throat, or dyspnea) and musculoskeletal complaints (myalgia, myolysis, arthralgia, edema) and headache and paresthesia. Hypotension may be present in 5% of reactions. In contrast to immunoglobulin E (IgE)-mediated anaphylactic reactions, wheezing is distinctly uncommon, and edema less common. In contrast to described reactions to NNRTIs, the rash is rarely severe, usually appearing later (e.g., 1–3 days) after the onset of the systemic symptoms, and may be absent in up to 30% of cases.^49,50 The rash is usually maculopapular or urticarial. Morbidity is related more to the systemic symptoms rather than the rash. Although the time course is variable, symptoms tend to occur in succession over the course of several days, with fever and GI symptoms appearing early and rash later. The severity of symptoms tends to increase with each dose. A delay in discontinuation of abacavir in the face of an active hypersensitivity reaction may lead to a severe reaction and even death, although the risk of a fatal reaction is far more likely in the setting of a rechallenge (see further ahead).

After discontinuation of abacavir, symptoms improve quickly and resolve over a few days, although the rash, if present, may persist longer. If abacavir is reinitiated (rechallenge), symptoms reappear within hours and may be more severe. Fever, rash, malaise, hypotension, and facial and/or throat swelling or bronchospasm have been described. Symptoms not present with the initial reaction may also appear on rechallenge and temperature elevations to 39–40°C may occur. One case report described severe anaphylactic shock after rechallenge with abacavir without preceding hypersensitivity.⁵¹ Hypotension was present in ∼25% of cases where abacavir was reinitiated after an initial reaction. Support with intravenous fluids with or without vasopressors (dopamine/dobutamine) may be necessary. Fatalities have been reported.⁵² Among the reported cases where a rechallenge has been attempted, 4% of patients have died. The rechallenge reaction has occurred even at reduced doses of 100 mg abacavir. Thus, once hypersensitivity to abacavir has been diagnosed, do not rechallenge with abacavir or an abacavir-containing medication (e.g., Trizivir or Epzicom).

Laboratory abnormalities observed during a hypersensitivity reaction to abacavir include an increase in liver function tests and an acute, transient lymphopenia in some subjects. Occasionally, decreased platelet counts, elevated serum creatinine, and elevated creatine phosphokinase have been described. These laboratory abnormalities return to normal with prompt discontinuation of abacavir. No preexisting condition or screening laboratory abnormality has yet been identified that indicates an increased likelihood of developing this reaction.

Diagnosis of ‘hypersensitivity reaction to abacavir’ occurred in 2–3% of subjects assigned to the nonabacavir-treated subjects in double-blinded studies.⁵³ Most cases had rash alone or rash accompanied by nonspecific, mild symptoms, whereas hypersensitivity reactions to abacavir are part of a multisymptom syndrome with fever as the hallmark. Cutaneous reactions are common with other medications often prescribed to patients with HIV-1 infection, such as NNRTIs or sulfamethoxazole. The appearance or characteristics of a rash without other symptoms are not sufficiently specific to differentiate the cause in most cases. However, the morbidity associated with hypersensitivity reactions to abacavir is not related to rash, and rash is not frequently an early sign of this reaction. Thus, for patients in whom rash is the only clinical symptom, abacavir may be continued with the warning to follow the patient carefully for the development of other symptoms. The patient should be counseled to report the appearance of any additional symptoms immediately. In cases where systemic symptoms and rash are present together, differentiation may not be possible and abacavir should be discontinued. The rash of abacavir hypersensitivity is mild, but may involve mucous membranes. However, the Stevens–Johnson syndrome has only been reported when abacavir has been combined with NNRTIs.

In some cases, the symptoms of a hypersensitivity reaction may resemble acute infections or reactions to other medications. Keiser et al described the clinical features of diagnosed influenza infection with cases of hypersensitivity to abacavir.⁵⁴ Both rash and GI symptoms were strongly associated with the latter. Respiratory symptoms with hypersensitivity to abacavir appeared in conjunction with GI symptoms. Nevertheless, for some cases differentiation of hypersensitivity to abacavir from an acute infection can be difficult. The symptoms will increase in severity if dosing is continued. Moreover, the timing of the onset of symptoms with respect to drug exposure is critical for recognition of the syndrome. Abacavir hypersensitivity starts several days to 6 weeks after starting the drug in most cases. Furthermore, the symptoms improve after stopping abacavir. Severe reaction and fatalities have occurred when symptoms of hypersensitivity to abacavir were ascribed to another cause and the discontinuation of abacavir was delayed or rechallenge was attempted. It is important not to ascribe the symptoms of abacavir hypersensitivity to nonspecific ‘flu’ or ‘gastroenteritis’.^49,50

There is no evidence to indicate that interruption of abacavir administration in the absence of any symptoms (i.e., for a reason other than a side effect) increases the frequency or severity of a subsequent hypersensitivity reaction.^55,56 However, there have been cases of hypersensitivity reactions occurring within 1 day of restarting abacavir where no prior symptoms of an initial hypersensitivity reaction were recorded. This underscores the importance to advise a patient to contact the healthcare provider if abacavir is discontinued. The healthcare provider should then obtain a careful history to assure that symptoms indicating a hypersensitivity reaction were not present at the time of interruption.

Most Common Adverse Events in Adult HIV-1-Infected Subjects Enrolled in CNAA2001

A summary of clinical toxicities associated with abacavir in the CNAA2001 trial is shown in Table 11-1.^57,58 In that study, abacavir was administered as monotherapy for 4 weeks followed by randomization to therapy for 4 weeks followed in turn by randomization to zidovudine or its matching placebo for an additional 8 weeks. The most commonly reported adverse events were nausea, headache, asthenia, diarrhea, insomnia, fever, dizziness, vomiting, abdominal pain, and rash. Nausea was common, occurring in nine of 20 subjects (45%) at 300 mg PO bid. Headache occurred in eight of 20 subjects (40%) at 300 mg PO bid. A possible trend of increase in incidence of nausea and headache with increasing abacavir dose was observed. However, the small numbers of patients and the design of the study make it difficult to draw definitive conclusions about the significance of this preliminary observation. Only eight subjects (10%) withdrew from the study because of adverse events, all of which were considered possibly related to the study drug. The most common reasons for discontinuation were nausea (with or without vomiting) or hypersensitivity reaction. The latter occurred within 4 weeks of dosing. Only four subjects experienced a serious adverse event, none of which were considered related to the study drug.

Table 11-1 Most Common Adverse Events Associated with Abacavir in the CNAA2001 Trial^a

a Adverse events experienced by the 300 mg bid cohort (n= 20 subjects) in the CNAA2001 trial, a phase I/II trial to evaluate the effects of multiple dosing of abacavir alone and in combination with zidovudine capsules. The majority of adverse events were rated as mild or moderate. The number of subjects listed experienced an event (regardless of association with abacavir) (see text).⁵⁷ This table does not specifically list the occurrence rate of the abacavir hypersensitivity reaction, which has been estimated at 3% (range 2–5%) in abacavir-treated subjects.

Abacavir did not demonstrate any evidence of bone marrow suppression, as evidenced by the lack of abnormalities in any hematologic parameters during the study that were attributable to abacavir. Additionally, abacavir did not demonstrate any evidence of hepatic, renal, or pancreatic toxicity by biochemical analysis. Most treatment-emergent hematologic and clinical chemistry toxicities were rated as grade 1 or 2. No distinguishing pattern could be detected between the cohorts or treatment arms when clinical chemistry laboratory parameters were examined. The incidence of abnormal laboratory values was generally low, and no dose-related trends were observed. There was no evidence to indicate that abacavir used in combination with zidovudine was associated with an increase in the severity of clinical or laboratory toxicities compared with administration of the study drug alone.^57,58

Safety in HIV-1-Infected Infants and Children

The side effects observed among children receiving abacavir are similar to those observed among adults. Hypersensitivity reactions to abacavir have been reported in children, and have similar frequency and clinical characteristics to those observed among adults receiving abacavir.

Studies Conducted in Antiretroviral-Naive HIV-1-Infected Adults

Several phase I/II controlled clinical studies were conducted to determine pharmacokinetics, safety, and antiviral activity of abacavir. A parallel, dose-ranging 12-week study (CNAA2001) was completed in 79 antiretroviral-naive HIV-1-infected adults in the United States with limited prior antiretroviral experience (<12 weeks of zidovudine therapy).^57,58 This was the first multiple-dosing study in humans. The doses of abacavir studied were 200 mg tid, 400 mg tid, 600 mg tid, or 300 mg bid. Patients received open-label abacavir monotherapy for 4 weeks, followed by 8 weeks of zidovudine (600 mg/day) or placebo in addition to abacavir. Abacavir given as monotherapy for 4 weeks resulted in median decreases in plasma HIV-1 RNA by 1.11–1.77 log₁₀ copies/mL and median CD4+ T-lymphocyte count increases of 63–111 cells/μL in all groups. At week 12, median plasma HIV-1 RNA levels decreased by 1.02–2.24 log₁₀ copies/mL (for abacavir monotherapy) and by 1.81–2.01 log₁₀ copies/mL (for abacavir plus zidovudine). At week 12, median CD4+ T-lymphocyte counts increased by 79–195 cells/μL (for abacavir monotherapy) and by 93–142 cells/μL (for abacavir plus zidovudine). The percentage of subjects who had plasma HIV-1 RNA levels at week 12 below 400 and 40 copies/mL were 28% and 11%, respectively (abacavir monotherapy) and 69% and 22%, respectively (abacavir plus zidovudine). Eight subjects (10%) discontinued due to adverse events; nausea (n = 4) and hypersensitivity (n = 3) were the most common reasons for withdrawal; there were no fatalities. After 12 weeks of abacavir plus zidovudine or placebo, 72/79 patients were required to interrupt abacavir until essential preclinical studies were completed.

In a preliminary analysis, 43 of 72 subjects elected to restart therapy with abacavir following an interruption of up to 1 year. Six subjects received abacavir uninterrupted. During this extension phase, patients were offered open-label abacavir to use in appropriate combination antiretroviral regimens of abacavir plus either an NNRTI or a PI. The goal of this long-term open-label study was to monitor the long-term antiviral effects of abacavir therapy. Altogether, 16 of 42 subjects had plasma HIV-1 RNA levels of <400 copies/mL as extended therapy began. By the week 12 visit, 27 of 38 subjects had levels <400 copies/mL. Of 22 subjects reaching 36 weeks, 16 had viral loads <400 copies/mL. The proportions undetectable at week 24 for NRTI-only combination therapy and abacavir/PI/NNRTI were 14/18 and 9/10, respectively. Over 48 weeks of therapy, >50% of patients receiving either nucleoside-only therapy including abacavir or PI-containing therapy with abacavir were maintained below a plasma HIV-1 RNA level of 400 copies/mL. Sustained increases in CD4+ cell counts were seen in both abacavir treatment groups while on therapy, with or without PI therapy.

A second randomized, double-blind, dose-ranging clinical trial (CNAB2002) of abacavir 100, 300, and 600 mg bid was conducted in Europe.⁵⁹ A total of 60 antiretroviral-naive patients with CD4+ T-lymphocyte counts of ≥100 cells/μL and plasma HIV-1 RNA levels of ≥30 000 copies/mL received up to 24 weeks of abacavir therapy alone; 55 then went into an open-label randomized study of abacavir 300 mg bid plus other antiretrovirals for 72 weeks. At week 24, all subjects who remained on abacavir alone were switched to abacavir 300 mg bid/lamivudine 150 mg bid/zidovudine 300 mg bid or other licensed antiretrovirals as determined by their treating physician. At week 4, greater reductions in plasma HIV-1 RNA were seen in subjects receiving 300 or 600 mg abacavir twice daily (median changes −1.55 and −1.61 log₁₀ copies/mL, respectively) than subjects receiving 100 mg abacavir twice daily (median change, −0.63 log₁₀ copies/mL). At 24 weeks, the 300 and 600 mg twice-daily groups had a median change of plasma HIV-1 RNA of −0.70 and −1.30 log₁₀ copies/mL, respectively. During the open-label phase in which zidovudine/lamivudine was added to 300 mg abacavir twice daily, a further median reduction in plasma HIV-1 RNA of 1.74 log₁₀ copies/mL was seen. At week 48, a median 2.82 log₁₀ drop in plasma HIV-1 RNA from randomized baseline was seen in pooled data from all abacavir-treated subjects. Sixty-five percent of patients had a plasma viral load of <400 copies/mL after 48 weeks of antiretroviral therapy (ART) containing abacavir, and 43% of patients had a plasma viral load <50 copies/mL at the same timepoint. An additional decrease in HIV-1 RNA of 2.16 log₁₀ was seen during the open-label phase from the reset open-label combination baseline. Most subjects (46/55, or 87%) received abacavir combined with lamivudine 150 mg bid and zidovudine 300 mg bid since their switch to open label. Overall, this study demonstrated that abacavir 100 mg bid was inferior to 300 mg bid or 600 mg bid, with the latter two doses showing similar viral load reductions. The majority (42/47, 90%) of subjects at week 72 remained on the triple combination of abacavir/lamivudine/zidovudine, four subjects were receiving PIs in addition to the triple combination, and one subject had substituted stavudine for zidovudine. By week 72, the median change in plasma HIV-1 RNA was −2.8 log₁₀ copies/mL. At week 72, 72% of subjects had a plasma HIV-1 RNA <400 copies/mL and 50% of subjects had plasma HIV-1 RNA <50 copies/mL.⁶⁰ The most frequently seen adverse experiences reported by 55 subjects during the long-term, open-label phase of the study were nausea/vomiting in 23 of the 55 (41%), malaise/fatigue in 10 (18%), headache in nine (16%), muscle pain in six (11%), GI discomfort/pain in seven (12%), sleep disorders in four (7%), and skin rashes in four (7%). These data included new adverse experiences that began once subjects switched to open-label abacavir, and not those that may have been ongoing from the randomized phase.⁶⁰

The CNAA3003 trial investigated the effect of zidovudine and lamivudine with or without abacavir in antiretroviral-naive adults located in the USA, Spain, Belgium, and the UK.⁶¹ This double-blind, randomized phase III study was designed to assess the safety, tolerance, and antiviral activity of abacavir/lamivudine/zidovudine over 16 weeks, as well as the safety, tolerance, and durability of the abacavir/lamivudine/zidovudine response over 48 weeks. Subjects (n = 173 with CD4+ T-lymphocyte counts >100 cells/μL) were randomized 1:1 to receive abacavir 300 mg bid or placebo/lamivudine 150 mg bid/zidovudine 300 mg bid. Subjects were stratified based on plasma HIV-1 RNA at screening: <10 000 copies/mL, 10 000–100 000 copies/mL, or >100 000 copies/mL. Both treatment groups had marked decreases in plasma HIV-1 RNA by week 4, which was sustained in the abacavir/lamivudine/zidovudine group. At week 16, the proportion of subjects with plasma HIV-1 RNA below the limit of detection (<400 copies/mL) was significantly superior in the abacavir/lamivudine/zidovudine group (75%) when compared to the lamivudine/zidovudine group (35%) (P < 0.0001 analyzed by the Cochran–Mantel–Haenzel test controlling for randomized plasma HIV-1 RNA strata). In addition, the triple combination was effective irrespective of the baseline plasma HIV-1 RNA strata. By contrast, the virologic response to lamivudine/zidovudine began to rebound toward baseline until week 16 when subjects were eligible to add abacavir plus other ART. The virologic response to lamivudine/zidovudine at week 16 was also diminished at the higher plasma HIV-1 RNA baseline strata. Mean increase in CD4+ T-lymphocyte counts was similar between the treatment groups at 16 weeks. The triple combination was generally well tolerated (3% of this group withdrew due to adverse experiences as compared to 2.5% of subjects in the lamivudine/zidovudine group).

The CNAF3007/Ecureuil open-label study evaluated the efficacy and safety of the triple nucleoside combination Combivir/abacavir versus Combivir/nelfinavir as first-line therapy in HIV-1-infected adults.⁶² This randomized, open-label study in 195 HIV-1-infected adults, with plasma HIV-1 RNA 1000–500 000 copies/mL, compared Combivir/abacavir to Combivir/nelfinavir over 48 study weeks. At the intent-to-treat analysis at 48 weeks, 64% and 61% of subjects had plasma viral load <50 copies/mL in the Combivir/abacavir and Combivir/nelfinavir groups, respectively. The baseline viral load was comparable in both groups with a median of 4.2 log₁₀ copies/mL (Combivir/abacavir) and 4.1 log₁₀ copies/mL (Combivir/nelfinavir). The median change from baseline in plasma viral load was −2.3 log₁₀ copies/mL in both groups. Median CD4+ T-lymphocyte count increases were 109 and 120 cells/μL in the Combivir/abacavir and Combivir/nelfinavir arms, respectively. Possible hypersensitivity reactions to abacavir were reported in four subjects (4%). These results suggest that Combivir/abacavir is generally well-tolerated first-line ART in HIV-1-infected adults, with comparable antiviral activity to that of a PI-containing regimen over 48 weeks.

Of 155 subjects completing 48 weeks in the CNAF3007 trial, 92 were enrolled into the CNAF3021 study to look at long-term safety and efficacy: 47 subjects initially randomized to the Combivir/abacavir group and 45 subjects to the Combivir/nelfinavir arm.⁶³ After a follow-up of at least 2 years post-CNAF3007 randomization (median 3 years), there was a higher proportion of patients in the Combivir/nelfinavir group (67%) than in the Combivir/abacavir group (30%) who discontinued the treatment initially attributed by randomization of the CNAF3007. In total, 11/30 patients in the Combivir/nelfinavir switched from nelfinavir to abacavir. In the intent-to-treat analyses (switch included), 76% of subjects in the Combivir/abacavir group and 71% of subjects in the Combivir/nelfinavir group had a plasma HIV-1 RNA <50 copies/mL at the single long-term entry CNAF3021 visit.

The CNA3014 study compared Combivir/abacavir to another PI-containing regimen, indinavir (800 mg tid)/abacavir in a 48-week open-label randomized study in 342 HIV-infected ART naive adults with CD4+ T-lymphocyte counts >100 cells/μL and plasma HIV-1 RNA >5000 copies/mL.⁶⁴ Subjects were stratified based on screening plasma HIV-1 RNA: >5000–100 000 copies/mL or ≥100 000 copies/mL. Results demonstrated that time to treatment failure over 48 weeks by intent-to-treat analysis was significantly longer in the abacavir/Combivir group than in the indinavir/Combivir group (P < 0.001). In the primary analysis (intent-to-treat, missing = failure, plasma HIV-1 RNA <400 copies/mL), the abacavir/Combivir regimen was superior. In the intent-to-treat analysis at week 48 by baseline plasma HIV-1 RNA stratum, abacavir/Combivir was more effective than indinavir/Combivir in patients with >100 000 copies/mL (70% vs 49% with plasma HIV-1 RNA <400 copies/mL). This difference was not observed in patients with plasma HIV-1 RNA >100 000 copies/mL (60% in abacavir/Combivir vs 51% in indinavir/Combivir). Sustained increases in CD4+ T lymphocytes were observed in both treatment groups over 48 weeks. Self-reported adherence to randomized treatment was significantly higher in the abacavir/Combivir group, with 72% of subjects reporting taking all of their doses or missing an average of less than one dose per week during the last month, compared with 45% of subjects in the indinavir/Combivir group (P < 0.001). The percentages of drug-related adverse events in the study were 65% (108/165 in abacavir/Combivir group) and 87% (142/164 in the indinavir/Combivir group). Six percent (10/165) of the subjects in the abacavir/Combivir group had a possible abacavir hypersensitivity reaction, all of which occurred during the first 6 weeks of therapy.

The KLEAN study examined two other PIs, fosamprenavir–ritonavir versus lopinavir–ritonavir, each in combination with abacavir–lamivudine.⁶⁵ This open-label, noninferiority study enrolled 878 antiretroviral-naive HIV-1-infected subjects to receive fosamprenavir–ritonavir 700 mg/100 mg twice daily or lopinavir–ritonavir 400 mg/100 mg twice daily, each with the eco-formulation of abacavir–lamivudine 600 gm/300 mg (Epzicom) once daily. Noninferiority of fosamprenavir–ritonavir to lopinavir–ritonavir was demonstrated for the primary study endpoint of proportion of subjects at week 48 achieving plasma HIV-1 RNA <400 copies/mL (95% CI around the treatment difference −4.84–7.05 was shown). At week 48, 315 of 434 (73%) subjects in the fosamprenavir–ritonavir group and 317 of 444 (71%) in the lopinavir-ritonavir group achieved plasma HIV-1 RNA <400 copies/mL. Treatment discontinuations because of an adverse event, which was another primary study endpoint, were few and occurred with similar frequency in the two treatmentgroups (fosamprenavir–ritonavir 53 (12%); lopinavir–ritonavir 43 (10%)). Suspected abacavir hypersensitivity was reported in 53 cases (6%); 32 (7%) in the fosamprenavir–ritonavir arm and 21 (5%) in the lopinavir–ritonavir arm.

Abacavir versus the NNRTI nevirapine combined with zidovudine/lamivudine was examined as a first-line ART for HIV-1 infection in Uganda, Africa.⁶⁶ The rationale for this safety study was that these agents might avoid drug interactions with tuberculosis therapy as described with HIV-1 PIs, plus the tolerability and toxicity of abacavir in Africa were unknown. This randomized, double-blind, 24-week trial enrolled 599 symptomatic antiretroviral-naive subjects with CD4+ T-lymphocyte counts <200 cells/μL who were either zidovudine/lamivudine (Combivir) with either 300 mg abacavir or nevirapine placebo (n = 300) or nevirapine 200 mg and abacavir placebo (n = 299) twice daily. Of 599 enrolled subjects, 569 (9%) completed 24 weeks; however, 24 (4%) died and six (1%) were lost to follow-up. Twenty subjects had serious adverse reactions (six (2%) on abacavir and 14 (4.7%) on nevirapine) (primary endpoint: hazards ratio = 4.1; 95% CI = 0.16–1.08; log rank P = 0.06). Serious adverse reactions reported in 19 of 20 patients were consistent with potential hypersensitivity reactions: similar proportions on abacavir (four of six) and nevirapine (seven of 13) had respiratory, constitutional (three with abacavir, seven with nevirapine), or GI symptoms (two with abacavir, five with nevirapine), rash (six with abacavir, 10 with nevirapine), fever (six with abacavir, nine with nevirapine), or oral/mucosal involvement (two with abacavir, five with nevirapine); but hepatic involvement was seen only in three patients on nevirapine. The remaining serious adverse reaction was asymptomatic hepatitis (nevirapine arm). In total, 14 (4.7% abacavir) and 30 (10%) nevirapine patients discontinued blinded abacavir/nevirapine (P = 0.01) due to toxicity (six (2%) abacavir vs 15 (5%) nevirapine, respectively, P = 0.05), all rash or possible hypersensitivity reactions (six with abacavir, 13 with nevirapine), or hepatotoxicity (four with nevirapine). This study concluded that abacavir had a lower discontinuation rate and a trend toward a lower rate of serious adverse reactions when compared with nevirapine in African patients initiating ART with low CD4+ T-lymphocyte counts.

The CNA30024 trial studied abacavir versus zidovudine combined with lamivudine and the NNRTI, efavirenz, in an international randomized, double-blind noninferiority clinical trial in 654 antiretroviral-naive HIV-1-infected patients in the USA, Costa Rica, Chile, Portugal, and Denmark.⁶⁷ Of 654 enrolled subjects, 649 received more than one dose of randomized study medication. A total of 503 (77%) completed ≥48 weeks of study (78% in the abacavir group and 77% in the zidovudine group), and 145 (24%) of the subjects prematurely discontinued the study (similar for both treatment groups). The incidence of reported serious adverse events was 20% in the abacavir group and 14% in the zidovudine group. Of the 30 cases of suspected hypersensitivity reactions reported in the abacavir group, only five met the definition of a serious event. The primary study objective was the comparison of the proportion of subjects achieving plasma HIV-1 RNA ≤50 copies/mL through 48 weeks of study: 70% of subjects in the abacavir arm versus 69% of subjects in the zidovudine group maintained confirmed plasma HIV-1 RNA levels ≤50 copies/mL. Therefore, abacavir was found to be noninferior to zidoduvine when combined with lamivudine and efavirenz (stratified 95% CI of −6.3% to 7.9%). Furthermore, the proportion of subjects with plasma HIV-1 RNA ≤50 copies/mL at week 48 was similar between treatment arms when stratified by screening plasma HIV-1 RNA levels of either <100 000 copies/mL (72% abacavir group vs 70% zidovudine group) or >100 000 copies/mL (67% in both treatment arms).

The ACTG 5095 study compared the triple-nucleoside regimen of abacavir plus Combivir versus NNRTI-containing (efavirenz) regimens for the initial treatment of HIV-1-infection.^68,69 This randomized, double-blind study evaluated three antiretroviral regimens for the initial treatment of subjects infected with HIV-1: zidovudine/lamivudine/abacavir, zidovudine/lamivudine plus efavirenz, and zidovudine/lamivudine/abacavir plus efavirenz. Combivir/abacavir (Trizivir) versus Combivir plus efavirenz versus Combivir/abacavir plus efavirenz. A total of 1147 subjects with a mean baseline HIV-1 RNA level of 4.85 log₁₀ copies/mL and a mean CD4+ T-lymphocyte count of 238/μL were enrolled. Prespecified stopping rules led to the recommendation to stop the triple-nucleoside regimen (abacavir plus Combivir) due to virologic inferiority and allowed presentation of results in the triple-nucleoside group in comparison to pooled data from the efavirenz groups.⁶⁸ The time to virologic failure was found to be significantly shorter in the triple-nucleoside group (P < 0.001). After a median follow-up of 32 weeks, 82 of 382 subjects (21%) in the triple-nucleoside group and 85 of 765 (11%) in the pooled efavirenz groups had virologic failure. These differences persisted despite stratification for baseline plasma HIV-1 RNA levels in subjects either entering the study with <100 000 copies/mL (P = 0.001) or >100 000 copies/mL (P < 0.001). These differences also persisted despite stratification for baseline pretreatment CD4+ T-lymphocyte subjects of ≥100 copies/mL (P = 0.001) or lower (P < 0.001), as well as baseline pretreatment CD4+ T-lymphocyte subjects of ≥200 copies/mL (P = 0.004) or lower (P < 0.001). In a post hoc analysis, the 195 subjects in the triple-nucleoside arm who reported 100% adherence at week 12 had a higher rate of virologic failure than the 382 subjects in the pooled efavirenz arms with the same level of adherence (P < 0.001). Suspected abacavir hypersensitivity reactions were reported in 27 (7%) subjects in the triple-nucleoside group and 59 (8%) subjects in the pooled efavirenz arms. The results of the completed three versus four drug A5095 trial demonstrated no difference between the three-drug regimen of zidovudine/lamuvudine plus efavirenz versus the four-drug regimen of zidvoudine/lamivudine/abacavir plus efavirenz as initial treatment in HIV-1-infected subjects.⁶⁹ After a median 3 year follow-up, 99 (26%) of 382 and 94 (25%) of 383 patients receiving the three-drug and four-drug regimens, respectively, reached protocol-defined virologic failure; time to virologic failure was not significantly different (hazard ratio, 0.95; 97.5% CI, 0.69–1.33, P = 0.73). Overall, at least 80% of subjects had plasma HIV-1 RNA levels <50 copies/mL through 3 years of treatment. This study demonstrated that adding abacavir as a fourth initial drug to this dual-nucleoside plus NNRTI-containing regimen provided no additional benefit in a direct three versus four initial drugs comparison.

Not all abacavir-containing triple-nucleoside regimens are recommended for HIV-1-infected antiretroviral-naive subjects. Early virologic nonresponse to abacavir, tenofovir, and lamivudine was reported in the ESS30009 Study.⁷⁰ This randomized open-label, multicenter study compared tenofovir disoproxil fumarate versus the NNRTI efavirenz when both administered once daily with abacavir/lamivudine fixed combination as first-line antiHIV triple-drug therapy. A total of 340 subjects were randomized with a median baseline HIV-1 RNA level of 4.7 log₁₀ copies/mL and median baseline CD4+ T-lymphocyte count of 251 cells/μL. After reports of early virologic nonresponse, an unplanned interim analysis was performed on 194 of the enrolled 340 subjects who had ≥8 weeks of plasma HIV-1 RNA results available. Several definitions of virologic nonresponse were employed: (1) a <2.0 log₁₀ copies/mL decrease in HIV-1 RNA level by study week 8; (2) a rebound of plasma HIV-1 RNA ≥1.0 log₁₀ copies/mL above the nadir; or (3) for subjects with two consecutive plasma HIV-1 RNA measurements <50 copies/mL, a subsequent increase to >400 copies/mL on two consecutive occasions. Interim results revealed that virologic nonresponse occurred in 50 (49%) of 102 subjects in the tenofovir disoproxil fumarate arm when compared to five (5%) of 92 subjects in the efavirenz arm (P < 0.001). By study week 12, genotyping of plasma HIV-1 isolates from nonresponders in the tenofovir disoproxil fumarate/abacavir/lamivudine arm detected the HIV-1 RT M184V or M184I/M/V mixtures in 40 (98%) of 41 subjects and HIV-1 RT K65R and M184V or mixtures in 22 (54%) of 41 subjects (see section on HIV-1 Drug Resistance). Subjects initially randomized to the tenofovir disoproxil fumarate arm were then permitted to switch to an investigator-determined second-line regimen and 131 (77%) of 171 subjects remained in follow-up. Of the 61 subjects with genotype data available at or before switching, 31 had a detectable HIV-1 RT K65R mutation. The efavirenz arm continued unaltered, and by contrast to the abacavir arm, 120 (71%) of 169 subjects achieved a plasma HIV-1 RNA level of <50 copies/mL after 48 study weeks. As mentioned previously (see section on In Vitro Activity and Mechanisms of Action), no drug–drug interactions to explain the virologic nonresponse of abacavir, tenofovir disoproxil fumarate, and lamivudine were found. Rather, the low genetic barrier to development of HIV-1 drug resistance is most likely to blame since this three-drug regimen exerts selective pressure with rapid development of M184V, followed by K65R in some patients. Therefore, the combined regimen of abacavir, tenofovir disoproxil fumarate, and lamivudine for HIV-1-infection should not be used.

The COL40263 trial evaluated the once-daily regimen of trizivir and tenofovir disoproxil fumarate in antiretroviral-naive HIV-1-infected subjects with plasma HIV-1 RNA >30 000 copies/mL.⁷¹ The primary endpoints of this open-label, single-arm, multicenter trial were the proportion of subjects with plasma HIV-1 RNA <50 copies/mL at week 48 and the proportion of subjects with grade 3 or 4 adverse events and laboratory toxicities. Of 123 enrolled subjects, 52 (42%) discontinued study through week 48, including 14 due to adverse events, 13 lost to follow-up, and 12 with virologic nonresponse, as well as other reasons. Suspected abacavir hypersensitivity occurred in eight subjects (6.5%) within the first 3 weeks of study drug administration. Of subjects who remained on therapy over 48 weeks, 49 of 64 (77%) of subjects had plasma HIV-1 RNA levels <50 copies/mL at week 48.

The ESS40013 study tested fourdrug induction with abacavir/lamivudine/zidovudine plus the NNRTI efavirenz followed by maintenance treatment with the simplified regimen of abacavir/lamivudine/zidovudine alone in antiretroviral-naive HIV-1-infected subjects.⁷² All 488 subjects received abacavir/lamivudine/zidovudine plus efavirenz during the 48-week induction phase followed by a 1:1 randomization to either the continued regimen of abacavir/lamivudine/zidovudine plus efavirenz or to simplified regimen of abacavir/lamivudine/zidovudine for the subsequent 48-week maintenance phase. No significant differences were noted in proportion of subjects with plasma HIV-1 RNA level <50 copies/mL at week 96: 79% abacavir/lamivudine/zidovudine plus efavirenz versus 77% abacavir/lamivudine/zidovudine; P = 0.75). Virologic failure occurred in 22 patients during induction and in 24 patients (16 in abacavir/lamivudine/zidovudine arm and eight in abacavir/lamivudine/zidovudine plus efavirenz group; P = 0.134) during maintenance. More subjects reported perfect adherence at week 96 in the triple therapy arm than in the four-drug arm (88.8% vs 79.6%; P = 0.057).