Stavudine has shown potent antiviral activity when used as monotherapy in patients with asymptomatic or advanced HIV-1 infection [5-8], as well as clinical efficacy in zidovudine-experienced patients [9]. Stavudine plus didanosine has shown potent antiviral and immunologic efficacy [10]. Lamivudine has been studied primarily in combination therapy with zidovudine in both treatment-naive and treatment-experienced patients and has been shown to have clinical benefit due to its potent antiviral activity [3, 11-14]. Although treatment with lamivudine results in the rapid selection of the mutation on the reverse transcriptase gene at codon 184 and the development of phenotypic resistance [15, 16], only a partial loss of viral activity occurs [11-1417, 18]. Furthermore, this mutation has been shown to cause zidovudine-resistant virus to reacquire susceptibility to zidovudine [19]. Stavudine and lamivudine have additive or synergistic activity against both zidovudine-sensitive and zidovudine-resistant HIV-1 strains in vitro [20]. In vitro studies have also shown that prolonged treatment with stavudine is generally not associated with the development of high levels of viral resistance to the drug [21].

Our open-label pilot study, which is known as ALTIS, was designed to evaluate therapy with stavudine plus lamivudine in patients who had never before received an antiretroviral agent and those who had previously received antiretroviral agents other than stavudine or lamivudine.

Methods

Top Methods Results Discussion Author & Article Info References

Study Design and Patients

We conducted a 24-week open-label pilot study at three clinical centers in Paris to evaluate the virologic and immunologic efficacy and safety of stavudine plus lamivudine. Major inclusion criteria were age 18 years or older, seropositivity for HIV-1 confirmed by Western blot, CD4⁺ cell count between 50 and 400 cells/mm³, HIV-1 RNA plasma viral load of more than 15 000 copies/mL, Karnofsky performance score of at least 70, and no previous use of stavudine or lamivudine. Patients were also required to have a hemoglobin level of 80 g/L or more, a total neutrophil count of more than 750 cells/mm³, a platelet count of more than 50 000 cells/mm³, a liver enzyme level no more than 5 times the upper limit of normal, a serum creatinine concentration no more than 1.5 times the upper limit of normal, and a lipase level less than 2 times the upper limit of normal. Patients at any stage of HIV disease were recruited and were divided into two subgroups: those who had not previously received any antiretroviral drugs and those who had received zidovudine, didanosine, zalcitabine, or any combination of two of these drugs.

The study protocol was approved by the institutional review board of the Groupe Hospitalier Pitie-Salpetriere. Each patient provided written informed consent.

Treatment Regimen

Patients were given stavudine, 30 mg twice daily if their body weight was 60 kg or less or 40 mg twice daily if their body weight was more than 60 kg, and lamivudine, 150 mg twice daily. This combination was administered for 24 weeks. Concomitant therapies, except for other antiretroviral or immune-based therapies, were allowed if they were required by the patient's condition. Prophylaxis against Pneumocystis carinii and Toxoplasma gondii was recommended for all patients with a CD4⁺ count less than 200 cells/mm³; the decision to institute such prophylaxis and the type of prophylaxis were left to the physician and patient.

Clinical and Laboratory Evaluations

Patients were evaluated within the 2 weeks before the start of the study; at day 0, week 2, and week 4; and then every 4 weeks through week 24. Screening and baseline evaluations included a full medical history. At each visit, patients were questioned about any new clinical or physical events that may have indicated clinical progression or adverse events; adverse events were judged according to the AIDS Clinical Trials Group system. At each visit, vital signs, clinical signs or symptoms, and Karnofsky score were recorded, and complete blood cell counts, biochemical tests, CD4⁺ and CD8⁺ lymphocyte counts, and quantitative plasma viral load were measured. The HIV-1 RNA viral load was measured by using a quantitative polymerase chain reaction assay with a lower limit of detection of 200 copies/mL (Roche Amplicor monitor system, Roche, Branchburg, New Jersey). The HIV-1 RNA viral load was measured on acid citrate dextrose plasma samples that were immediately frozen and stored at –80°C.Specimens for each patient were batch tested with a negative control and two positive controls (low, 10 000 to 20 000 copies/mL; high, 300 000 to 400 000 copies/mL). All analyses were performed with the same lot of reagents.

Criteria for Efficacy

Measures of virologic efficacy were evaluated by using the difference between the baseline HIV-1 RNA value (geometric mean of the screening and the entry value) and the value at week 24. Initially, the protocol stated that the treatment would be considered effective if the median decrease in plasma viral load at week 24 was at least 0.6 log₁₀. However, in accordance with standard current practice, we chose to consider a decrease in viral load of at least 1 log₁₀ as clinically significant. Other measures included the percentage of patients with a viral load of 3000 HIV-1 RNA copies/mL or less or 400 copies/mL or less and the percentage of patients with an increase in CD4⁺ cell count of 50 cells/mm³ or more or 50% or more compared with baseline values.

Statistical Analysis

Data were analyzed from all patients who received at least one dose of the study treatment. For the six treatment-experienced patients who stopped taking the study treatment before week 24, the last-observation-carried-forward technique was used to analyze viral load and CD4⁺ count [22]. With this technique, the missing value at a given week is replaced by the last value observed during the trial treatment. This procedure was followed because most of these patients were switched to a more potent antiretroviral regimen, and intention-to-treat analysis tended to overestimate the treatment effect of stavudine plus lamivudine. Analyses of virologic and immunologic variables and of tolerance were mainly descriptive in nature. A graph representing the absolute difference between viral load at week 24 and at baseline was drawn to ascertain the percentage of patients in whom the maximal efficacy was observed (the difference between the baseline viral load and the threshold of the technique was used to measure the viral load).

Role of Study Sponsor

Stavudine (Zerit) was provided by Bristol-Myers Squibb Paris (France), and lamivudine (Epivir) was provided by Glaxo Wellcome Paris (France). Data were gathered, analyzed, and interpreted independently of the pharmaceutical companies that provided the drugs, according to the monitoring and operating procedures of the Agence Nationale de Recherche sur le SIDA.

Results

Top Methods Results Discussion Author & Article Info References

Demographic and Clinical Characteristics

A total of 83 patients were recruited into the study (42 treatment-naive patients and 41 treatment-experienced patients). Baseline characteristics are listed in Table 1. Most of the patients were male (85.5%), and 83% had acquired HIV-1 infection through sexual contact. Treatment-naive patients were predominantly asymptomatic (71%) and had a median baseline CD4⁺ cell count of 258 cells/mm³ and a median viral load of 76 502 HIV-1 RNA copies/mL (4.88 log₁₀). Treatment-experienced patients were more frequently symptomatic; 59% had stage B or C symptoms according to Centers for Disease Control and Prevention criteria. The median CD4⁺ cell count was 172 cells/mm³, and the median viral load was 91 255 HIV-1 RNA copies/mL (4.96 log₁₀). Among the treatment-experienced patients, the median duration of previous treatment was 35 months (range, 4 to 94 months); 49% of patients had received monotherapy, and 51% had received nucleoside combination therapy.

Of the 83 patients enrolled in the trial, 77 completed 24 weeks of therapy (42 of 42 in the treatment-naive group and 35 of 41 in the treatment-experienced group). Discontinuation or modification to the allocation of study drugs was due to loss to follow-up at week 4 (1 patient), lymphoma at week 9 (1 patient), adverse events at week 15 (1 patient) and week 20 (1 patient), and biological progression of HIV disease at week 16 (2 patients).

Changes in Viral Load and CD4⁺ Cell Count

Viral load decreased and CD4⁺ cell counts increased after treatment with stavudine plus lamivudine in both treatment-naive and treatment-experienced patients. The virologic and immunologic response profile differed between the groups. In the treatment-naive patients, a median reduction in plasma HIV-1 RNA of 1.96 log (₁₀) (range, –2.76 to –0.70)was observed after 4 weeks of stavudine plus lamivudine therapy, which was sustained at a median reduction of 1.66 log₁₀ (range, –3.04 to –0.79)at week 24 (Figure 1). In the treatment-experienced patients, the median maximum reduction in plasma HIV-1 RNA was 1.32 log₁₀ (range, –3.13 to 0.49) after 4 weeks, followed by a rebound in virus production, resulting in a final median plasma HIV-1 RNA reduction of 0.55 log₁₀ (range, –2.86 to 0.52) below baseline at week 24. After 24 weeks of therapy, 86% of patients (95% CI, 75% to 97%) in the treatment-naive group and 34% of patients (CI, 19% to 49%) in the treatment-experienced group had a decrease in plasma HIV-1 RNA of 1 log₁₀ or more. A viral load of less than 3000 HIV-1 RNA copies/mL was recorded in 57% of patients (CI, 41% to 72%) in the treatment-naive group and 22% of patients (CI, 10% to 38%) in the treatment-experienced group. A viral load of less than 400 HIV-1 RNA copies/mL was found in 26% of patients (CI, 12% to 40%) in the treatment-naive group and 5% of patients (CI, 1% to 17%) in the treatment-experienced group. In patients whose viral load reached 400 copies/mL at week 24, the mean baseline viral load was 32 700 copies/mL (CI, 16 160 to 66 150 copies/mL) compared with 99 980 copies/mL (CI, 62 360 to 160 320) in patients whose viral load remained above 400 copies/mL (P = 0.05). The treatment-naive patients who attained optimal viral treatment efficacy were those who had a lower viral load before initiation of combination therapy (Figure 2).

View larger version (11K): [in this window] [in a new window]   Figure 1. Median decrease in plasma viral load for treatment-naive (circles) and treatment-experienced (squares) patients over 24 weeks of therapy with stavudine plus lamivudine. The numbers in parentheses are the numbers of treatment-naive patients/treatment-experienced patients.

View larger version (13K): [in this window] [in a new window]   Figure 2. Individual decrease in HIV-1 RNA viral load after 24 weeks measured against baseline HIV-1 RNA in treatment-naive patients. The solid diagonal line represents the maximum benefit in viral load reduction that can be measured by using a technique with a lower limit of detection of 400 copies/mL. Thus, no data point () can be located beyond this line. The closer the data point is to the line, the greater the efficacy of treatment.

The immunologic response to stavudine plus lamivudine mirrored the virologic response (Figure 3). Among the treatment-naive patients, a rapid CD4⁺ cell count increase that reached a median of 90 cells/mm³ (range, –85 to 302 cells/mm³) was seen at week 4, followed by a progressive, slower increase that reached a median of 108 cells/mm³ (range, –58 to 406 cells/mm³) at week 24. In treatment-experienced patients, a median increase in CD4⁺ cell count of 45 cells/mm³ (range, –64 to 166 cells/mm³) was found at week 4, which was sustained at 46 cells/mm³ (range, –188 to 311 cells/mm³) above baseline at week 24.

View larger version (9K): [in this window] [in a new window]   Figure 3. Median changes in CD4+ cell counts for treatment-naive (circles) and treatment-experienced (squares) patients over 24 weeks of therapy with stavudine plus lamivudine. The numbers in parentheses are the numbers of treatment-naive patients/treatment-experienced patients.

After 24 weeks of therapy, 71% (CI, 54% to 85%) of treatment-naive patients and 34% (CI, 19% to 51%) of treatment-experienced patients had a minimum increase of 50 CD4 cells/mm³. A minimum increase in CD4 cells of 50% was observed at week 24 in 50% (CI, 34% to 66%) of treatment-naive and 37% (CI, 22% to 53%) of treatment-experienced patients.

Safety

Adverse events are reported in Table 2. Overall, 17% (CI, 9% to 27%) of the 82 patients with more than 4 weeks of follow-up experienced grade 3 or 4 toxicity. The most common grade 3 or 4 laboratory adverse events were elevated aspartate aminotransferase and alanine aminotransferase levels (7% [CI, 3% to 15%]), increased creatine phosphokinase levels (5% [CI, 1% to 12%]), increased lactate dehydrogenase levels (1% [CI, 0% to 6%]), and increased levels of amylase and lipase (1% [CI, 0% to 6%]). Eight patients experienced grade 1 or 2 paresthesia during the study. Two patients discontinued therapy because of intolerance.

Discussion

Top Methods Results Discussion Author & Article Info References

Nucleoside analogues are the core of all antiretroviral therapeutic regimens. Because antiretroviral combinations that do not include zidovudine are increasingly used in clinical practice, it is important to identify and characterize potent two-nucleoside combinations with favorable antiviral efficacy and safety profiles as alternatives to regimens containing zidovudine.

Our 24-week, open-label study showed that stavudine plus lamivudine is a potent antiviral two-nucleoside combination, resulting in a median viral reduction of 1.66 log₁₀ in treatment-naive patients and a reduction of 0.55 log₁₀ in patients previously exposed to a nucleoside analogue other than stavudine or lamivudine. We did not compare stavudine plus lamivudine with other therapies or placebo.

In treatment-naive patients, the greatest reduction in plasma HIV-1 RNA occurred at week 4 (1.96 log₁₀). A reduction in plasma HIV-1 RNA of 1.66 log (₁₀) was sustained over the 24-week study period. Of interest, although direct comparison among trials is only indicative, the kinetics of this virologic response over 24 weeks in treatment-naive patients differs slightly from that observed in two trials evaluating zidovudine plus lamivudine in treatment-naive patients [11, 12]. In those studies, a similar maximum peak reduction in viral load of approximately 1.5 log₁₀ was followed by a rebound in viral production leading to a median reduction of 1 log₁₀ over 24 weeks.

One of the main interpretations of the rebound in viral load after 4 to 8 weeks of therapy with zidovudine plus lamivudine has been the rapid selection of viral populations containing the methionine-to-valine mutation at amino acid 184 of reverse-transcriptase enzyme. However, the preliminary data in our study on reverse-transcriptase sequencing show that five of six treatment-naive patients had the M184V mutation at week 12; the sixth patient had a mixed wild and mutant viral population (data not shown). Therefore, the absence of viral rebound is unlikely to be explained by an absence of M184V mutation with lamivudine therapy. In treatment-naive patients, there seem to be two slopes of the increasing CD4⁺ cell count curve-an initial rapid increase of 90 cells/mm³ above baseline during the first 4 weeks, followed by a slightly progressive increase up to 108 cells/mm³ above baseline at week 24. This pattern may be due to the increase of two different lymphocyte populations: the acutely infected CD4⁺ cells that are no longer being killed by the viral production, leading to a rapid increase in CD4⁺ cells, followed by a less pronounced increase of CD4⁺ cells that may correspond to a subset of the chronically infected cell population, which has a less rapid turnover [23].

In patients who had previously received other nucleoside analogues, the virologic and immunologic responses to therapy followed a different pattern. The greatest reduction in plasma HIV-1 RNA was achieved at week 4 (1.32 log₁₀ copies/mL) and was followed by a rebound, bringing the final reduction in viral load at week 24 to 0.55 log₁₀. This pattern of virologic rebound after 4 weeks of therapy is similar to data reported in patients pretreated with zidovudine plus lamivudine [13, 14], zidovudine plus didanosine, or zidovudine plus zalcitabine [23]. The increase in CD4⁺ cell count at week 4 (45 cells/mm³) was sustained at week 24.

The difference in the efficacy profile of stavudine plus lamivudine in treatment-naive patients compared with treatment-experienced patients (a difference of approximately 1 log₁₀ in the viral load reduction and 60 CD4⁺ cells/mm³ in the CD4⁺ cell count increase in favor of treatment-naive patients) is intriguing. Because neither population had previously received stavudine or lamivudine, one can speculate that preexposure to nucleoside analogues that have been unsuccessful in profoundly reducing the viral load reduces the potency of a new combination of nucleoside analogues. Preliminary data on complete sequencing of the virus suggest that primary resistance of HIV-1 to stavudine in the pretreated patients can be ruled out because the sequencing of the virus at baseline did not show genotypic resistance to any of the viral mutations associated with stavudine resistance (those on codon 75 or codon 50). Furthermore, these data suggest that the rapid appearance of lamivudine resistance due to the high frequency (>90%) of acquisition of the M184V mutation within a few weeks after initiating any lamivudine-containing regimen occurs similarly in both treatment groups (data not shown). Another explanation may be that previous exposure to other nucleoside analogues modifies the intracellular pharmacodynamics of lamivudine's metabolism, leading to changes in intracellular phosphorylation that may decrease the blood concentration of the active metabolite of nucleoside analogue [24]. Stavudine has an interesting resistance profile: To date, it has been difficult to clearly identify any mutations associated with phenotypic resistance [21]. This suggests that stavudine may play an important role as a part of first-line multiple antiretroviral therapy.

Until now, few direct comparisons of nucleoside analogue combinations have been made, despite their extensive use in clinical practice. The most extensively studied combinations have been zidovudine plus didanosine, zidovudine plus zalcitabine (which has shown an approximate reduction in HIV-1 RNA of 0.8 to 1 log₁₀), and zidovudine plus lamivudine [11, 12, 23-26]. In the absence of a true comparative trial, it is impossible to stratify therapies according to their antiviral power. However, the performance of stavudine plus lamivudine in this trial, with a viral load reduction of about 1.5 log₁₀ in treatment-naive patients and 0.5 log₁₀ in pretreated patients, compares favorably with that of the standard, widely studied zidovudine plus lamivudine combination or didanosine plus stavudine regimens.

The combination of stavudine and lamivudine was well tolerated by both treatment-naive and treatment-experienced patients. Only two patients discontinued therapy as a result of intolerance, and no patients experienced greater than grade 2 neurologic symptoms. Seventeen percent of patients experienced at least one grade 3 or grade 4 event, most often elevations of aminotransferase levels. The high tolerability of this combination may result from largely nonoverlapping toxicity profiles and the good safety profile of each drug.

In summary, the strong antiretroviral activity of stavudine plus lamivudine, the simplicity of the dosing schedule, and the good safety profile make this two-nucleoside combination potentially attractive as an option in multidrug therapy for HIV.

Drs. Coutellier and Bonmarchand: Department of Medicine, Hopital Pitie-Salpetriere, 47, Boulevard de l'Hopital, 75013 Paris, France.

Drs. Matheron and Longuet: Department of Infectious Diseases, Hopital Bichat-Claude Bernard, 46, rue H. Huchard, 75018 Paris, France.

Drs. Calvez and Agut: Department of Virology, Hopital Pitie-Salpetriere, 47, Boulevard de l'Hopital, 75013 Paris, France.

Drs. Descamps and Brun-Vezinet: Department of Virology, Hopital Bichat-Claude Bernard, 46, rue H. Huchard, 75018 Paris, France.

Mr. Lancar and Dr. Costagliola: INSERM SC4, 27, rue de Chaligny, 75012 Paris, France.