In the treatment of human immunodeficiency virus type 1 (HIV-1) infection, clinical trials have demonstrated that combinations of two nucleoside reverse transcriptase inhibitors (NRTIs) and a protease inhibitor (PI) improve clinical outcome, suppress viral load, and increase CD4+ cell counts to a greater degree and for a longer duration than either NRTI monotherapy or double therapy [1–4]. In the longest study reported to date, a triple combination of lamivudine, zidovudine, and indinavir suppressed HIV-1 RNA below a 500 copy/ml lower limit of detection in 78% of patients at 100 weeks . The effectiveness of triple combination therapy in maximally suppressing viral loads in most patients has prompted the International AIDS Society–USA  and the United States Department of Health and Human Services  to recommend a combination of two NRTIs and a potent PI as one of the preferred treatments of established HIV-1 infection.
As many combination antiretroviral therapies involve complicated dosing schedules and high pill counts, good adherence to such regimens is often difficult to achieve . Only 50 to 70% of patients with HIV-1 infection report taking 80% or more of their antiretroviral doses [8–10]. AIDS Clinical Trials Group (ACTG) Study 175 showed that over a 90- to 360-day period patients took just 55–76% of all antiretroviral doses at the prescribed daily frequency and 27% of all doses at the prescribed dosage intervals . Adherence to antiretroviral therapy is critically important because it is directly related to clinical efficacy [12–14]. Analysis of data from the INCAS, AVANTI 2, and AVANTI 3 clinical trials revealed that non-adherent patients (those taking < 75% of their antiretroviral doses prior to one or more study visits) do not achieve as low a plasma HIV-1 nadir over a 48-week period as is achieved in adherent patients . Furthermore, the missing of antiretroviral doses for as few as 3 days has been linked to more rapid development of resistance due to mutant HIV-1 strains [15,16].
A new combination tablet that contains 150 mg lamivudine and 300 mg zidovudine has been developed to simplify antiretroviral therapy by allowing patients on lamivudine and zidovudine to take just one tablet twice daily. Moore et al. have previously established the bioequivalence of lamivudine and zidovudine from the combination tablet and individual 150 mg lamivudine tablets and 300 mg zidovudine tablets administered concurrently. However, to date, there have been no clinical trials to evaluate the efficacy, safety, or adherence with a combination tablet/PI regimen compared with a regimen where lamivudine, zidovudine, and a PI are given as separate components. The purpose of this clinical trial was three-fold: (1) to show that a combination lamivudine 150 mg/zidovudine 300 mg tablet, given twice daily in combination with recommended doses of a PI, is clinically equivalent (i.e. not inferior) with respect to virologic and immunologic efficacy to a conventional regimen of 150 mg lamivudine twice daily, 200 mg zidovudine three times daily, plus a PI in antiretroviral-experienced patients with HIV-1 infection; (2) to assess the safety and tolerability of each of these regimens; and (3) to assess adherence to each regimen by using patient diary card data and information derived from the Patient Medication Adherence Questionnaire (PMAQ) .
Male and female outpatients were eligible for study enrollment if they were 18 years or older, had documented HIV-1 infection with plasma HIV-1 RNA levels < 10 000 copies/ml and CD4+ cell counts (≥ 300 × 106/l, and had been receiving 150 mg lamivudine twice daily, 600 mg zidovudine daily, and a Food and Drug Administration (FDA)-approved PI for at least 10 weeks immediately prior to screening. Patients with the following characteristics were excluded from the study: a clinical diagnosis of AIDS; inability to comply with the study schedule; concurrent enrollment in another investigational drug study; female who was pregnant or breast feeding; presence of malabsorption syndrome affecting drug absorption; clinically significant laboratory abnormalities; a requirement for radiation therapy or cytotoxic chemotherapy within 4 weeks prior to entry, or anticipated need for these treatments within 20 weeks; treatment with immunomodulating agents; or previous treatment with an HIV immunotherapeutic vaccine.
Study design and treatment
In this open-label, randomized, parallel-group, comparative trial, patients were stratified by PI prior to randomization. Eligible patients were randomly assigned to receive either their current conventional triple combination regimen of 150 mg lamivudine twice daily, 200 mg zidovudine three times daily, plus an FDA-approved PI, or a regimen consisting of one combination tablet, administered twice daily, plus their currently prescribed PI, for 16 weeks. Lamivudine was given as 150 mg tablets of Epivir, zidovudine as two 100 mg capsules of Retrovir, and the combination lamivudine 150 mg/zidovudine 300 mg tablet as Combivir (all products manufactured by Glaxo Wellcome Inc., Research Triangle Park, North Carolina, USA). The PI regimens consisted of 800 mg indinavir [two 400 mg capsules; Crixivan (Merck & Co., Inc., West Point, Pennsylvania, USA)] every 8 h; 600 mg saquinavir mesylate [three 200 mg capsules; Invirase (Roche Laboratories, Nutley, New Jersey, USA)] three times daily; 600 mg ritonavir [six 100 mg capsules; Norvir (Abbott Laboratories, North Chicago, Illinois, USA)] twice daily; or 750 mg nelfinavir mesylate [three 250 mg tablets; Viracept (Agouron Pharmaceuticals, Inc., La Jolla, California, USA)] three times daily. The study was conducted between May 1997 and June 1998 in the United States and Puerto Rico. All patients provided written informed consent to participate, and the protocol for the study was approved by the institutional review boards at each of 19 outpatient treatment sites.
The primary efficacy measure was quantitative plasma HIV-1 RNA levels, as determined by Roche Amplicor polymerase chain reaction [lower limit of quantitation (LLOQ), 400 copies/ml]. The primary study endpoint was treatment failure, defined as an increase in HIV-1 RNA ≥ 0.5 log10 above baseline in patients with viral load > LLOQ at randomization and as HIV-1 RNA increasing to ≥ 1250 copies/ml in patients with viral load < LLOQ at randomization. Virologic failures had to be confirmed by a repeat HIV-1 RNA measurement within 4 weeks of the initial measurement. Secondary efficacy measures were an assessment of changes in the number and percentage of CD4+ lymphocyte counts.
All study candidates underwent a medical history and physical examination at the pre-entry visit (week −2) for determination of study eligibility. HIV-1 RNA levels and CD4+ cell counts were determined at this time, and at weeks 0 (baseline), 4, 8, and 16. The mean of week −2 and week 0 constituted the baseline HIV-1 RNA, and CD4+ values used in data analysis.
Assessment of adherence
At study entry, patients were given preprinted diary cards and asked to record daily each dose of study medication taken, including PIs. Diary cards were specific for each treatment arm and formatted to indicate number of doses taken over each assessment period. Diary cards were checked by investigators at weeks 4, 8, and 16.
The PMAQ version 1.0  was administered prospectively to patients to complete at baseline and at weeks 8 and 16 of the study. This 31-item patient self-administered questionnaire is designed to assess a patient's treatment adherence behavior and its determinants by providing a quantitative measure of adherence and information on why a patient is or is not adherent. The PMAQ is divided into two sections: Section 1 contains six items pertaining to medication-taking behaviors and asks patients to rank on an ordinal scale how often they missed individual doses or days of study medication. Section 2 contains 25 items pertaining to barriers and motivators to taking medications. Items, such as `I don't want people to see me taking my medicines for HIV' are rated on a Likert scale from 1 (`definitely true') to 5 (`definitely false'). These items are then scored to produce a patient profile across five hypothesized dimensions: Memory and Recall, Scheduling and Timing, Physical Effects, Social Support, and Knowledge and Attitudes. The PMAQ 1.0 was scored as follows. After reverse coding the positively worded items, the actual raw score was calculated as the sum of the reversed-coded responses. The transformed dimension score was then calculated as: 100 × (raw score –lowest possible raw score)/(highest possible raw score – lowest possible raw score). A total score was derived in the same fashion using the individual items. The dimension and adherence scores thus ranged from 0 to 100, with higher scores indicating fewer barriers and more motivators to adherence.
Mean corpuscular volume (MCV) which has been used as a laboratory marker of zidovudine adherence was also measured.
The patients were monitored for adverse events, laboratory abnormalities, and any HIV-1-related illnesses at weeks 4, 8, and 16. Patients who experienced ACTG grade 4 toxicities  were withdrawn from the study, as were patients who failed to achieve resolution of ACTG grade 3 toxicities  with interruption of therapy.
This study sought to establish that the combination tablet regimen was not inferior [20,21] to the conventional regimen with respect to proportion of patients who remained successfully treated as defined under the heading Efficacy Assessment above. The null hypothesis was that treatment with the combination tablet regimen was worse than with the conventional regimen, and the alternative hypothesis was that the former was not inferior to the latter. A final sample size of 223 patients was deemed necessary to provide approximately 80% power to conclude non-inferiority of the combination tablet regimen compared with the conventional regimen assuming a treatment success proportion of 0.90 (90%) on each arm, using a non-inferiority margin of 0.10 (10%) and significance level of 0.048. The latter significance level for concluding lack of inferiority reflected a penalty for two administrative interim analyses.
The combination tablet regimen compared to the conventional regimen was said not to be inferior if the lower limit of the one-sided 95.2% confidence interval (CI) for the difference between the combination tablet regimen and conventional regimen in proportion of treatment successes was −0.10 or higher. The CI was constructed using the normal approximation with a continuity correction. In addition, a two-sided 90.4% CI was constructed to consider the hypothesis of the two regimens being clinically equivalent, applying a level of significance of 0.096. All analyses were performed on the intent-to-treat (ITT) population using observed data. However, to examine the sensitivity of the conclusions to missing data, another analysis was performed on the ITT population assuming that treatment-related dropouts (those not `lost to follow-up') were virologic failures. Due to the relatively small number of virologic failures, no comparison of the treatments was carried out adjusting for PI regimens. In another analysis, patients' HIV-1 RNA was classified as `detectable' or `undetectable' according to viral load being ≥ LLOQ or < LLOQ. For each time of measurement, changes from baseline detectability status (`worsened', `no change', `improved') were compared between groups using a Wilcoxon rank sum test.
Average area under the curve minus baseline for CD4+ cell counts were compared between groups using an analysis of covariance controlling for baseline PI stratum, with the corresponding baseline mean CD4+ value as covariate. Estimates of treatment difference were based on the least squares adjusted means. A similar procedure was used for the analysis of MCV data. No formal statistical comparisons were performed for adverse event data. The responses to the medication-taking behaviors section of the PMAQ 1.0 were compared across the two treatment groups using the Cochran–Mantel–Haenszel row mean scores statistic controlling for baseline PI stratum. The van-Elteren test controlling for PI at baseline was used to evaluate treatment differences in mean scores on the PMAQ for Memory and Recall, Scheduling and Timing, Physical Effects, Knowledge and Attitudes, and Social Support. Spearman rank correlation and logistic regression were used to assess the correlation between PMAQ results and virologic outcome (HIV-1 RNA levels). For all secondary endpoints, analyses were performed using a 5% level of significance and corresponding 95% CI. Because of the multiple comparisons being performed, P-values for PMAQ scores were to be interpreted as descriptive statistics.
In the rest of the manuscript, where we discuss virologic efficacy and CIs and refer to 95 and 90%, these should be understood to mean 95.2 and 90.4%, respectively. Correspondingly, where we discuss significance levels of 5 and 10%, these should be understood to mean 4.8 and 9.6%.
Patient characteristics and disposition
All 223 patients were randomized to treatment and were evaluable, including 113 in the conventional lamivudine/zidovudine/PI regimen group and 110 in the combination tablet regimen group. The two treatment groups did not differ with respect to baseline characteristics (Table 1). Of the PIs taken as part of the conventional and combination tablet regimens, indinavir was the most frequently used (60 and 63% of patients, respectively). During the trial, all patients remained on the PI they were receiving at study entry. A similar number of patients in the conventional regimen and combination tablet regimen groups [19 (17%) and 20 (18%), respectively] withdrew prematurely from the study for reasons delineated in Table 1.
Virologic measurements of efficacy
The combination tablet regimen was associated with a 3.5% greater success rate than the conventional regimen (96.4 versus 92.9%), with four and eight patients failing treatment due to increases in HIV-1 RNA levels, respectively (Table 2). The lower limit of the associated CI for the difference was −2.4%, which was well within the −10% margin predefined as clinically unimportant. Thus, the null hypothesis that the combination tablet regimen is worse than the conventional regimen must be rejected (level of significance, 5%) in favor of the alternative (non-inferiority). This establishes the clinical equivalence (non-inferiority) of the combination tablet and conventional regimens regarding virologic response. If strict clinical equivalence of the two regimens were of interest, the associated two-sided 90% CI would be −0.024 to 0.093, indicating that treatment with the combination tablet regimen could be as much as 9% better, but no more than 2% less effective, than the conventional tablet regimen. Therefore, there was 95% confidence that combination tablet therapy is no more than 2% less effective and is possibly more effective than conventional therapy when comparing the proportion of patients who were successfully treated. For the analysis assuming that dropouts not due to `lost to follow-up' were virologic failures, the lower bound of the one-sided 95% CI was −0.065 (or 6.5%), confirming non-inferiority of the combination tablet regimen to the conventional tablet regimen. Due to the relatively small number of virologic failures, no comparison of the treatments was done adjusting for PI regimens.
Throughout the study, HIV-1 RNA levels and the percentage of patients with undetectable HIV-1 RNA remained high in each treatment group; more than 90% of patients maintained HIV-1 RNA levels within 0.5 log10 of baseline (Fig. 1a, b). At 16 weeks, the combination tablet and conventional regimen groups did not differ with respect to mean (± SD) HIV-1 RNA levels (2.74 ± 0.35 versus 2.78 ± 0.39 log10 copies/ml), changes in HIV-1 RNA from baseline (−0.02 ± 0.24 versus 0.04 ± 0.24 log10 copies/ml), percentage of patients with undetectable HIV-1 RNA levels (74 versus 80%), or percentage of patients maintaining HIV-1 RNA levels < LLOQ at the end of study or improving from baseline to undetectability (94 versus 91%). Seven patients [six on the combination tablet regimen (6%) and one on the conventional regimen (1%)] who were detectable at baseline, became undetectable at the end of the study, whereas 16 patients [six on the combination tablet regimen (6%) and 10 on the conventional regimen (9%)] who were undetectable at baseline became detectable by the end of the study. The shift from baseline detectability status was not statistically significant between groups (P = 0.063). The difference was most pronounced at the week 8 visit: none of combination tablet regimen patients compared to 12% of conventional regimen patients changed from undetectable to detectable, 95 versus 86% had no change, and 5 versus 2% improved from detectable to undetectable.
Logistic regression analysis indicated that patients with detectable viral load at baseline were about three times as likely to become treatment failures as those with undetectable viral load levels [ITT population, dropouts = failure;P = 0.002; odds ratio 3.26, 95% CI (1.53–6.96)].
Immunologic measurements of efficacy
The combination tablet and conventional regimens were similar with respect to CD4+ cell counts [treatment difference, 5.9 cells (95% CI from −15.8 to 27.6 × 106 cells/l)], analysis of average area under the curve minus baseline. When analyzed according to which PI was administered, patients on saquinavir had a decrease in CD4+ cell counts from baseline (42 × 106 cells/l), whereas patients in the indinavir, nelfinavir, and ritonavir groups had marginal increases in CD4+ cell counts. Median CD4+ cell counts and changes in these counts from baseline were also similar between the treatment arms at all times during the study (Fig. 2). At Week 16, the median CD4+ cell counts in the conventional regimen and combination tablet regimen groups were 504 and 530 × 106 cells/l, respectively.
Diary card and MCV findings regarding adherence
Diary card data showed that the adherence rate with zidovudine and lamivudine was ≥ 98% throughout the clinical trial. Based on these self reports, it appears that adherence was not compromised by administering zidovudine and lamivudine as part of a combination tablet regimen compared with administering them as separate components of the conventional regimen. Adherence with PIs was also equally high for morning and evening doses (97–99%), although there was a trend for lower adherence to noon doses (94–99%).
Changes in MCV did not differ significantly between the combination tablet and conventional regimens (Table 2). When analyzed by treatment group, a small but statistically significant reduction (P = 0.023) in MCV levels from baseline was observed in patients receiving the conventional regimen (P = 0.023), whereas the change in MCV levels from baseline in patients receiving the combination tablet regimen was not statistically significant (P = 0.176).
Patient medication adherence questionnaire findings
Patients in the combination tablet regimen group were less likely to miss doses of nucleoside analogues than patients in the conventional therapy arm at weeks 8 (P = 0.007) and 16 (P = 0.046) (Table 3). Mean scores for Scheduling and Timing, Social Support, and the Total Score were significantly higher for the combination tablet regimen compared with the conventional regimen (Table 3) at weeks 8 (P < 0.001, P = 0.029, and P = 0.002, respectively) and 16 (P = 0.022, P = 0.027, and P = 0.020, respectively). Mean scores for Physical Effects were significantly higher in the combination tablet arm than the conventional therapy arm at week 16 (P = 0.039), but not week 8. No other comparisons of PMAQ results across treatment arms reached statistical significance. Memory and Recall barriers and motivators were most highly associated with missed doses and drug holidays (P < 0.001). Patients who reported missing at least one dose of study drug were 2.5 times more likely to have detectable HIV-1 RNA than those who reported no missed doses (P = 0.002), and there was an association between higher total adherence scores and better virologic outcomes (P = 0.001).
The total incidence of adverse events was 61% in each treatment group. Adverse events that investigators considered drug-related occurred with a similar frequency in the combination tablet arm [23 patients (21%)] and the conventional therapy arm [22 patients (19%)]. The most frequently reported drug-related adverse events were diarrhea (seven and two patients, respectively), malaise and fatigue (seven patients in the conventional group only), decreased white cells (three patients in each group), and abnormal bilirubin levels (three patients in each group). Adverse events prompted premature withdrawal of treatment in four patients treated with the conventional regimen and five patients treated with the combination tablet regimen. Serious adverse events, none of which were considered drug-related, were observed in one patient in the conventional regimen group (acetaminophen overdose) and four patients in the combination tablet group (abnormal liver function tests, acute hepatitis and renal calculi; abnormal bilirubin levels; and thrombophlebitis).
Viral load, virologic endpoint, and CD4+ cell count data over the 16 weeks of this study indicate that a regimen employing the twice-daily lamivudine 150 mg/zidovudine 300 mg combination tablet plus a PI has similar activity to a conventional regimen of 150 mg lamivudine twice daily, 200 mg zidovudine three times daily, and a PI in the treatment of antiretroviral-experienced, patients with HIV-1 infection. The success rate was 3.5% higher with the combination tablet regimen. The lower limit of the associated CI for this difference was −2.4%, well within the −10% difference predefined as clinically unimportant. This establishes the clinical equivalence (non-inferiority) of the regimens, and supports the conclusion, with 95% certainty, that the combination tablet regimen is similar to and possibly better than the conventional regimen in patient populations similar to the one studied. If strict clinical equivalence of the two regimens were of interest, the associated two-sided 90% CI would be −0.024 to 0.093, indicating that treatment with the combination tablet regimen could be as much as 9% better, but no more than 2% less effective, than the conventional tablet regimen. The combination tablet regimen maintained virologic suppression in 3.5% more subjects than the conventional regimen in this study. The specific PI and PI dosage administered to each patient were not altered during this study. Therefore, if differences between the two triple combination regimens had been evident, they would have been due to the lamivudine/zidovudine component administered. Preliminary reports on the bioequivalence and antiviral and immunologic activity of the combination lamivudine/zidovudine tablet with lamivudine and zidovudine administered at the same daily dose as separate regimens support the results of this study [17,22].
During this study, HIV-1 RNA levels were maintained consistently within 0.5 log10 (threefold) of baseline in both treatment groups. The overall percentage of patients in each treatment group who had undetectable HIV-1 RNA levels at baseline and throughout the treatment period stayed between 70 and 82%. This overall percentage takes into account the fact that during the study a few patients (seven) who had detectable HIV-1 RNA levels at baseline achieved undetectable levels by the end of treatment, and that in a few other patients (16) initially undetectable HIV-1 RNA levels rose above the 400 copies/ml limit of detection. In view of the rapid replication characteristics of HIV-1 , virologic differences between the regimens, had they existed, would probably have been reflected in HIV-1 RNA changes during the 16 weeks of the study. One limitation of the present study was that the assay used to measure plasma HIV-1 RNA had a LLOQ of 400 copies/ml. It is unknown whether different virologic results would be seen if a more sensitive assay (e.g. one with a 50-copies/ml LLOQ ) were used. However, patients with detectable HIV-1 RNA at baseline were at greater risk of becoming virologic failures. Median CD4+ cell counts in the combination tablet regimen and conventional regimen treatment groups also changed only insignificantly over the study period, indicating stabilization of this marker.
The combination tablet and separate regimens of lamivudine/zidovudine/PI were equally well tolerated, and the majority of adverse events reported were related to the underlying HIV-1 disease and associated infections. The most frequently reported drug-related adverse events were malaise and fatigue and events related to the gastrointestinal tract. There were no apparent clinically significant differences between regimens with respect to hematology or blood chemistry values. These tolerability/safety findings are in agreement with those reported previously [1,2]. The fact that the combination tablet had similar safety and tolerability compared with administration of its separate components also confirms previous work .
The lamivudine/zidovudine combination tablet was developed with the goal of helping to simplify the complex combination antiretroviral regimens and enhancing adherence. Mehta et al.  have suggested that the more complicated the treatment regimen, the more likely it is for one or more doses to be missed. Forgetfulness is a major cause of missed doses, a problem potentially exacerbated in patients whose memory is affected by AIDS-related cognitive disorder . In a study involving 504 ambulatory HIV-infected patients on antiretroviral therapy, patients requiring more than 13 medication doses per day had significantly lower adherence rates than patients requiring fewer daily doses . Using the twice-daily lamivudine/zidovudine combination tablet regimen, patients in our study were able to eliminate taking three extra two-capsule doses of zidovudine per day (i.e. 2190 capsules per year) that otherwise would have been needed with a 200 mg three-times-daily zidovudine regimen. Patients using 300 mg zidovudine twice daily with lamivudine would have two doses per day or 730 fewer tablets per year. In terms of simplifying antiretroviral therapy, use of the twice-daily lamivudine/zidovudine combination tablet reduced the total number of daily antiretroviral doses to 8, 11, 11, and 14 in patients receiving concurrent indinavir, nelfinavir, saquinavir, and ritonavir regimens, respectively.
Diary card data showed that reported adherence was very high (generally, 98–99%) regardless of the regimen administered in this study. Adherence was slightly lower with regimens that involved a mid-day dose of a PI (94–99%). A trend for lower adherence with the PI component of a combination antiretroviral regimen also has been reported by Tebas et al. . Outside the strict controls of a clinical trial, missing a mid-day dose might be expected to occur more frequently. This underscores the importance of developing twice-daily regimens for PIs that can synchronize with twice-daily lamivudine/zidovudine combination tablet or other twice-daily antiretroviral administration. Although no significant differences were observed in MCV changes between the conventional and combination tablet regimens, only the combination tablet arm was associated with an increase in MCV levels from baseline, suggesting better adherence to zidovudine with the combination tablet .
The PMAQ was previously tested in a focus group of 67 patients with HIV-1 infection who were receiving various antiretroviral regimens . These patients indicated that regimens involving multiple daily doses and dietary restrictions were the most problematic with respect to achieving total adherence, that multiple regimens with different dosing frequencies were hardest to remember, and that most patients prefer twice-daily dosing over three-times-daily dosing of antiretrovirals. The present study is the first comparative clinical efficacy trial to use the PMAQ to evaluate behaviors and determinants affecting adherence. The PMAQ was instrumental in finding adherence problems that could not be detected using diary card data. This lack of agreement between the two monitoring tools may have been due to patients feeling more comfortable and less threatened by the administration of the PMAQ compared with the diary card method. Unlike the diary card, the PMAQ provided information about patient recall, scheduling problems, physical effects, knowledge/attitudes, and social support issues that could allow the clinician to pinpoint the specific problem(s) most associated with non-adherence in individual patients. Previous efforts to monitor adherence have relied on pill counts, Medical Event Monitoring System (MEMS) technology, and prescription refill records, all of which can tell, with varying accuracy, how often non-adherence is occurring, but not why . In view of the fact that the twice-daily lamivudine/zidovudine combination tablet regimen is easier to remember and to administer, it is not surprising that the PMAQ Scheduling and Timing scores were significantly better for patients who received this regimen than for patients who received the conventional regimen. It is unclear why the Social Support scores were better for patients receiving the combination tablet regimen, although this may have been related to this simpler regimen incurring less interference, interruption, and inconvenience in the patients' daily social activities, or toe in the patients' daily social activities, or to the comparatively higher Social Support scores in the combination tablet arm that were present at baseline. The PMAQ also showed the relationship of fewer barriers and more motivators to adherence to better virologic outcomes independent of treatment assignment. This finding is consistent with those of earlier studies that reported lower HIV-1 RNA levels in more adherent patients [12,13]. Overall higher adherence to either therapy was significantly associated with virologic outcomes (P = 0.001)
In conclusion, this 16-week, randomized, open-label, multicenter clinical trial demonstrated that in antiretroviral-experienced patients infected with HIV-1, a fixed-dose tablet combining 150 mg lamivudine and 300 mg zidovudine, administered twice daily with a PI regimen, is clinically equivalent (non-inferior) to a conventional regimen of lamivudine, zidovudine, and the PI, administered as separate components. The possibility that the combination tablet regimen is superior to the conventional regimen cannot be ruled out with certainty. By reducing the number of dosage forms required to achieve a therapeutic response, regimens including the combination lamivudine/ zidovudine tablet may simplify antiretroviral treatment, making it less cumbersome, more acceptable, and easier to remember, thus possibly facilitating improved adherence.
The authors wish to thank Greg T. Platek for statistical programming support and Gary E. Pakes for editorial assistance in the writing of this manuscript.
We also wish to thank the study participants and to acknowledge the following study investigators and study team personnel: John A. Bartlett, MD (Duke Medical Center, Durham, North Carolina); Douglas Dieterich, MD (Liberty Medical LLP, New York, New York); Neil Flynn, MD, MPH (UC Davis Medical Center, Sacramento, California); Wolfe Gerecht, MD (Ferrell-Duncan Clinic, Springfield, Missouri); Frank Graziano, MD (Univ. of Wisconsin, Madison, Wisconsin); James Jones, Jr, MD (Manhattan Medical Care, New York, New York); Russell Petrak, MD (Innovative Ventures LLC, Hindsale, Illinois); Donald Poretz, MD (Infectious Diseases Physicians, Inc., Annandale, Virginia); Stephen Raffanti, MD (Comprehensive Care Center, Nashville, Tennessee); Jorge Santana, MD (San Juan AIDS Program, Guaynabo, Puerto Rico); Gladys Sepulveda, MD (Ponce, Puerto Rico); David Weinrib, MD (Carolinas Medical Center, Charlotte, North Carolina; Joseph Racine, MD (Specialty Medical Care Centers of South Florida, Lauderhill, Florida); Jeffrey Nadler, MD (USF College of Medicine, Saint Petersburg, Florida); Melanie Thompson, MD (AIDS Research Consortium of Atlanta, Atlanta, Georgia); Jerry W. Snidow, PharmD, Sissi V. Pham, PharmD, Ralph A. DeMasi, PhD, and Neil M.H. Graham, MD (Glaxo Wellcome Inc., Research Triangle Park, North Carolina).
1. Hammer S, Squires K, Hughes M. et al
. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. N Engl J Med 1997, 337: 725–733.
2. Gulick RM, Mellors JW, Havlir D. et al
. Simultaneous vs sequential initiation of therapy with indinavir, zidovudine, and lamivudine for HIV-1 infection: 100-week follow-up. JAMA 1998, 280: 35–41.
3. Hirsch M for the Protocol 039 (Indinavir) Study Group, Meibohm A, Rawlins S, Leavitt R. Indinavir (IDV) in combination with ZDV and 3TC in ZDV-experienced patients with CD4 cell counts (50 cells/mm3 – 60-week follow-up.Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, Illinois, February 1998 [abstract No. 383].
4. Hoen B, Harzic M, Dumon C, et al
. Efficacy of zidovudine, lamivudine, and ritonavir combination in patients with symptomatic primary HIV-1 infection: the ANRS 053/53B trial. Can eradication be obtained?Fifth Conference on Retroviruses and Opportunistic Infections
. Chicago, Illinois, February 1998 [abstract 524].
5. Carpenter CCJ, Fischl MA, Hammer SM. et al
. Antiretroviral therapy for HIV infection in 1998.Updated recommendations of the International AIDS Society-USA Panel.
JAMA 1998, 280: 78–86.
6. National Institutes of Health. Report of the NIH panel to define principles of therapy of HIV infection and guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents.http://www.hivatis.org/guidelines/AAMay599.
7. Tseng AL-I. Compliance issues in the treatment of HIV infection. Am J Health-Sys Pharm 1998, 55: 1817–1824.
8. Eldred LJ, Wu AW, Chaisson RE, Moore RD. Adherence to antiretroviral and pneumocystis prophylaxis in HIV disease. J Acquir Immune Defic Syndr Hum Retrovirol 1998, 18: 117–125.
9. Muma RD, Ross MW, Parcel GS, Pollard RB. Zidovudine adherence among individuals with HIV infection. AIDS Care 1995, 7: 439–447.
10. Singh N, Squier C, Sivek C, Wagener M, Hong Nguyen M, Yu VL. Determinants of compliance with antiretroviral therapy in patients with human immunodeficiency virus: prospective assessment with implications for enhancing compliance. AIDS Care 1996, 8: 261–269.
11. Hammer SM, Katzenstein DA, Hughes MD. et al
. A trial comparing nucleoside monotherapy with combination therapy in HIV-infected adults with CD4 cell counts from 200 to 500 per cubic millimeter. N Engl J Med 1996, 335: 1081–1090.
12. Raboud JM, Montaner JSG, Rae S, et al
. Predictors of duration of plasma viral load suppression.37th Interscience Conference on Antimicrobial Agents and Chemotherapy
. Toronto, September –October 1997 [abstract A14].
13. Montaner JSG, Raboud JM, Rae S for the INCAS and AVANTI Study Groups. Adherence to treatment increases duration of virologic suppression regardless of pVL nadir38th Interscience Conference on Antimicrobial Agents and Chemotherapy
. San Diego, California, September 1998 [abstract LB-10].
14. Knobel H, Carmona A, Grau S, Pedro-Botet J, Díez A. Adherence and effectiveness of highly active antiretroviral therapy (letter). Arch Intern Med 1998, 158: 1953.1953.
15. Blaschke TF. Noncompliance and resistance to protease inhibitors.Fourth National Conference on Retroviruses and Opportunistic Infections.
Washington, DC; January 1997 [abstract S43].
16. Vanhove GR, Schapiro JM, Winters MA. et al
. Patient compliance and drug failure in protease inhibitor monotherapy (letter). JAMA 1996, 276: 1955–1956.
17. Moore KHP, Shaw S, Laurent AL. et al
. Lamivudine/zidovudine as a combined formulation tablet: bioequivalence compared with lamivudine and zidovudine administered concurrently and the effect of food on absorption. J Clin Pharmacol 1999, 39: 593–605.
18. Childress JK, Jhingran P, Scott-Lennox J, Tolson J. Development and initial evaluation of the Patient Medication Adherence Questionnaire (PMAQ).
Presented at the Sixth European Conference on Clinical Aspects and Treatment of HIV Infection
. Hamburg, Germany, October 11–15, 1997.
19. Anon. Table for Grading of Adult Adverse Experiences:
Washington, DC: National Institute of Allergy and Infectious Diseases, Division of AIDS; August 1992.
20. Food and Drug Administration. E9 statistical principles for clinical trials. Federal Register 1998, 63: 49583–49598.
21. Committee for Proprietary Medicinal Products (CPMP) European Working Party. Points to Consider on Biostatistical/methodological Issues arising from Recent CPMP Discussions on Licensing Applications: Superiority, Non-inferiority and Equivalence.
London, UK: Committee for Proprietary Medicinal Products (CPMP) European Working Party; September 23, 1999.
22. Rozenbaum W, Chauveau E. Phase 3 study of the antiviral activity of AZT-3TC given as separate regimens versus a new fixed dose combination (Combivir1
). Fifth Conference on Retroviruses and Opportunistic Infections. Chicago, Illinois, February 1998 [aAbstract No. 669].
23. Ho DD, Neuman AU, Perelson, Perelson, AS, Chen W, Leonard JM, Markowitz M. Rapid turnover of plasma virions and CD4 lymphocytes in HIV-1 infection. Nature 1995, 373: 123–126.
24. Mehta S, Moore RD, Graham NMH. Potential factors affecting adherence with HIV therapy. AIDS 1997, 11: 1665–1670.
25. Corless IB, Holzemer WL, Turner JG, et al
. Cognition and HIV adherence.XIIth International Conference on AIDS.
Geneva, Switzerland, June–July 1998 [abstract 32387].
26. Von Bargen J, Moorman A, Holmberg S. How many pills do patients with HIV infection take? [letter]. JAMA 1998, 280: 29.29.
27. Tebas P, Royal M, Fichtenbaum C, et al
. Relationship between adherence to HAART and disease state.Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, Illinois, February 1998 [abstract 149].
28. Kim HM, Lagakos SW. Assessing drug compliance using longitudinal marker data, with application to AIDS. Stat Med 1994, 13: 2141–2153.
29. Frick PA, Gal P, Lane TW, Sewell PC. Antiretroviral medication compliance in patients with AIDS. AIDS Patient Care and STDs 1998, 12: 463–470.