| |
Efficacy and Safety of a Once-Daily Fixed-Dose Combination of Abacavir/Lamivudine Compared With Abacavir Twice Daily and Lamivudine Once Daily as Separate Entities in Antiretroviral-Experienced HIV-1-Infected Patients (CAL30001 Study)
|
| |
| |
JAIDS Journal of Acquired Immune Deficiency Syndromes: Volume 41(5) 15 April 2006 pp 598-606
[Clinical Science]
LaMarca, Anthony MD*; Clumeck, Nathan MD, PhD ; Plettenberg, Andreas MD, PhD ; Domingo, Pere MD, PhD ; Fu, Kaisong MD, PhD ; Craig, Charles PhD ; Zhao, Henry PhD∥; Watson, Maria PhD ; Gordon, David MB, ChB#; Scott, Trevor PhD ; on behalf of the CAL30001 Study Team
From the *Therafirst Medical, Fort Lauderdale, FL; CHU Saint-Pierre, Bruxelles, Belgium; Institut fuer Interdisziplinaere Infektiologie und Immunologie GmbH, Hamburg, Germany; Hospital Sant Greu y Sant Paublo, Barcelona, Spain; GlaxoSmithKline, Research Triangle Park, NC; GlaxoSmithKline, Stevenage, UK; and #GlaxoSmithKline, Greenford, UK.
CAL30001 compared two 4-drug regimens containing ABC, once daily (QD) or twice daily (BID), in combination with 3TC, TDF, and either a protease inhibitor (PI) or a nonnucleoside reverse transcription inhibitor (NNRTI) based upon past treatment history. The ABC QD arm received ABC + 3TC as a fixed-dose combination (FDC) tablet. The target population was likely to have a dominant plasma virus carrying some mutations that may have altered virologic susceptibility to the NRTIs proposed but was not expected to abolish clinically significant antiviral activity of the antiretroviral agents tested in this trial. The noninferiority of the QD dosing of ABC was tested relative to BID dosing, in the setting of a 4-drug regimen for VF.
Abstract
Background: A one-tablet, once-daily abacavir/lamivudine fixed-dose combination (FDC) has been recently approved to treat HIV-1 infection.
Methods: A randomized, open-label, parallel-group, multicenter study to compare the efficacy and safety of the FDC group to the separate entities (SE) group, in combination with tenofovir and a new protease inhibitor or nonnucleoside reverse transcription inhibitor in antiretroviral-experienced adults experiencing virologic failure (VF). Eligible subjects had viral loads >1000 copies/mL with ≦3 nucleoside reverse transcription inhibitor-associated mutations. The primary efficacy end point was time-average changed from baseline (average area under the curve minus baseline) in plasma HIV-1 RNA over 48 weeks.
Results: A total of 186 subjects were enrolled. The average area under the curve minus baseline was -1.65 and -1.83 log10 copies/mL in the FDC and SE groups, respectively (intention to treat; 95% confidence interval: -0.13, 0.38). Patients in the FDC (50%) and SE groups (47%) achieved viral loads <50 copies/mL based on the time to loss of virologic response algorithm. VF was low and similar in both groups (FDC, 16%; SE, 18%). Tolerability was similar between the 2 groups.
Conclusions: The FDC group had noninferior efficacy over 48 weeks to the SE group in treatment-experienced subjects with VF.
Combination antiretroviral therapy (ART) has dramatically reduced the morbidity and mortality associated with HIV infection.1 However, current combination therapies have a high rate of failure due to issues involving potency, toxicity, patient adherence, individual pharmacokinetics, and/or viral resistance.2 Regardless of the cause for failure, emergence of viral resistance to one or more of the antiretroviral agents being taken is a frequent outcome. The anticipated pattern of nucleoside reverse transcription inhibitor (NRTI) resistance in the average subject failing first- or second-line ART therapy is the M184V mutation plus 1-2 thymidine analogue-associated mutations (TAMs).3
Cross-resistance among NRTIs is an important consideration in the selection of subsequent regimens. Among the most important mutations to consider are the 6 major TAMs (D67N, K70R, T215Y/F, M41L, L210W, and K219Q/E), as well as the K65R, L74V, and M184V mutations.4 The multinucleoside-resistant complexes marked by the Q151M or T69S mutations may abolish most of the antiviral activity of the NRTIclass.5,6 Mutations selected by zidovudine (ZDV) or stavudine (d4T) can have a negative effect on the susceptibility of virus to all the NRTIs in vitro, but both abacavir (ABC) and tenofovir (TDF) have been shown to confer clinical benefit in subjects with virus harboring some of these mutations.7-9 The K65R mutation decreases susceptibility to ABC, didanosine (ddI), and TDF.10,11 The L74V mutation decreases susceptibility to ddI and ABC while increasing susceptibility to ZDV and possibly TDF.12-14 Lamivudine (3TC) rapidly selects for the M184V mutation upon virologic rebound, leading to a large increase in phenotypic resistance to 3TC.15 However, HIV-1 with the M184V mutation has a reduced replication rate compared with the wild type, and reverse transcription with this mutation exhibits enzymatic defects such as reduced pyrophosphorolysis and decreased processivity.16 The mutation reverses the effect of the TAMs in vitro and can also delay the emergence of these mutations.15 Additionally, although the M184V mutation is associated with a small (2- to 3-fold) reduction in susceptibility to ABC,17,18 it enhances susceptibility to TDF, d4T, and ZDV.19
There is currently limited clinical data to address the use of ABC and 3TC as part of combination therapy in ART-experienced subjects. Therapy becomes more individualized in later lines of therapy based upon a subject's treatment history and/or viral genotype, which may require regimens with a higher degree of potency sufficient to achieve viral control. The sustained benefit of 3TC in the presence of the M184V mutation, such as the antagonistic effect of M184V on TAMs, the increased susceptibility of M184 virus to ZDV, d4T, and TDF, and the antiviral efficacy of ABC in subjects with M184V virus, provides a rationale for including 3TC, ABC, and TDF in subsequent treatment regimens after virologic failure (VF) with the M184V mutation.
Study Participants
Male or nonpregnant female out-patients were eligible for study enrollment if they were ≥18 years; had HIV infection, as documented by HIV-1 antibody ELISA and confirmed by Western blot test; were ART experienced and receiving a stable drug regimen containing 3 NRTIs, or 2 NRTIs plus 1 PI or NNRTI for at least 3 months, but naive to TDF; had plasma HIV-1 RNA levels ≥1000 copies/mL and CD4 cell counts >50 cells/mL at least once within 21 days of study entry; and harbored no more than 3 NRTI-related mutations, with no K65R, L74V, 69S insertion, or Q151M mutations by genotyping of subjects plasma from screening.
Study Design and Treatment Regimens TOP
This 48-week trial was a Phase 3, randomized, open-label, parallel-group study conducted at 57 centers in Europe and North America. The primary efficacy outcome was to test the noninferiority of ABC in an FDC tablet administered QD versus ABC administered BID with 3TC QD as separate entities (SE), as measured by plasma HIV-1 RNA average area under the curve minus baseline (AAUCMB) over 48 weeks. Eligible subjects were stratified according to study entry plasma HIV-1 RNA level (<5000 vs. ≥5000 copies/mL) and viral genotype (presence vs. absence of the M184V/I mutation).
After a 21-day screening phase, subjects were randomized (1:1) to 1 of 2 treatment groups:
FDC arm: ABC/3TC FDC QD + TDF QD + new NNRTI or PI;
SE arm: ABC BID + 3TC QD + TDF QD + new NNRTI or PI.
The ABC/3TC combination was administered as a single FDC tablet containing 600 mg of ABC (GlaxoSmithKline, Research Triangle Park, NC) and 300 mg of 3TC (GlaxoSmithKline), or as one 300-mg tablet of ABC BID and two 150-mg tablets of 3TC QD. TDF was administered as a single 300-mg tablet of TDF (Gilead Sciences, Foster City, CA). All background PIs and NNRTIs were initiated along with the above study drugs and were supplied to patients as prescribed medications by their doctor.
RESULTS
A total of 186 subjects were randomized to either the FDC group (n = 95) or the SE group (n = 91). Of the subjects randomized, 4 did not initiate treatment (1 in the FDC group; 3 in the SE group). In total, 25% (45/182) of the subjects in the ITT-exposed population discontinued the study. The proportion of subjects discontinuing study drug was similar in the FDC and the SE groups (Table 1).
Demographic and Baseline Characteristics
Demographic characteristics and results of assessment of baseline characteristics for the ITT-exposed population are summarized in Table 2. Overall, the treatment groups were well matched with regard to the demographic characteristics. Most subjects were male (76%) and white (66%). The 2 treatment groups were generally comparable with regard to the baseline characteristics. There was, however, a statistically significant difference between the 2 treatment groups in the median plasma HIV-1 RNA levels at baseline (FDC group = 3.92 log10 copies/mL; SE group = 4.22 log10 copies/mL; P = 0.017), which was not observed at screening.
Previous ART and Fourth ART at Baseline
All subjects were taking highly active ART before entry into the study (Table 3). The most commonly prescribed NRTIs before the study start were 3TC (98%), composed of 3TC, Combivir (3TC/ZDV), and Trizivir (ABC/3TC/ZDV); ZDV (63%), composed of ZDV, Combivir (3TC/ZDV), and Trizivir (ABC/3TC/ZDV); and d4T (55%). ABC-containing regimens (including Trizivir) were used before study entry by 13% and 19% of subjects in the FDC and SE groups, respectively.
Fifty-six percent of subjects took PI as their fourth drug during the course of the study. Overall, the most commonly used fourth drug was lopinavir/ritonavir (FDC, 35%; SE, 42%) or efavirenz (FDC, 35%; SE, 26%). Of subjects taking PIs, the majority continued taking PI as their fourth drug throughout the study period (FDC group, 51%; SE group, 52%). Only 1 subject in the SE group switched from PI to NNRTI, and 7 subjects switched to a different PI (2 in the FDC group; 5 in the SE group). More subjects initiated NNRTI therapy in the FDC group (45%) than in the SE group (35%).
Baseline Genotype
Overall, baseline NRTI mutations were balanced between the 2 treatment groups. The most common baseline NRTI resistance-associated mutation was at reverse transcription residue 184 (ITT-exposed population: 124/182 subjects, 68% with M184V and/or I mutation; VF population: 22/31 subjects, 71% with M184V and/or I mutation). TAMs were present in significant numbers in the subjects of the ITT-exposed population (132 TAMs in 72 of 182 subjects; 40%) but were fewer in the subjects of the VF population (12 TAMs in 6 of 31 subjects; 19%). The findings suggest that subjects with baseline TAMs may not be particularly susceptible to VF. In addition, there were differences in the distributions of individual TAMs between treatment groups, with more M41L and T215Y/F mutations in the FDC group and more of the K70R mutation in the SE group. Finally, subjects treated with PI tended to have more TAMs at baseline than those treated with NNRTI (total baseline TAMs: NNRTI subjects, n = 35; PI subjects, n = 92). For the most part, those subjects treated with NNRTI were assigned to this therapy without the presence of a genotype predictive of NNRTI resistance (NNRTI treated and susceptible: FDC = 37/42, 88%; SE = 30/31, 97%). In the PI-treated subjects, there was a higher proportion of subjects with primary PI mutations in the SE arm (12 of 51 subjects taking PI, 50%) than in the FDC arm (12 of 51 subjects taking PI).
Antiviral Activity
Table 4 presents median plasma HIV-1 RNA AAUCMB data at week 48 for the ITT-exposed population (observed data), stratified by baseline viral load and the presence or the absence of the M184V mutation at baseline. At week 48, the median plasma HIV-1 RNA AAUCMB was -1.65 and -1.83 log10 copies/mL for the FDC and SE groups, respectively. The stratified median difference between the 2 treatment groups at week 48 was 0.13 log10 copies/mL (95% CI: -0.13, 0.38). Because the 95% confidence limit was below the predefined noninferiority margin (0.4 log10 copies/mL), the noninferiority of the FDC group compared with the SE group over a 48-week treatment period was confirmed.
Overall, the median number of days contributing to AUC during the 48-week treatment period was similar in the 2 treatment groups (FDC group: 360 days, range 10-378; SE group, 362 days, range 22-378).
Plasma HIV-1 RNA AAUCMB was summarized for the ITT-exposed population (observed data) by gender, age, race, country, and fourth drug (data not shown). No statistically significant (P ≥ 0.4194) differences were found between the treatment groups in relation to any of the subgroups analyzed.
Median plasma HIV-1 RNA values are presented for the ITT-exposed population (observed data) in Figure 1. Five subjects were noted as having prematurely discontinued from the study at their week 48 visit; however, these subjects did have week 48 efficacy data available (plasma HIV-1 RNA) that was included in Figure 1.
The slope of viral decay was similar between the 2 treatment groups, with reductions in median plasma HIV-1 RNA at week 48 of greater than 1.8 log10 copies/mL from baseline in both treatment groups.
During the 48-week treatment period, a similar proportion of subjects maintained viral suppression in the FDC group relative to the SE group, based on the TLOVR algorithm. At week 48, the proportion with viral load <50 copies/mL was 50% in the FDC group compared with 47% in the SE group (Fig. 2).
In the study population stratified by baseline viral load and presence or absence of M184V/I, the proportions of subjects in the FDC and SE groups with viral loads <50 and <400 copies/mL at week 48 were similar. The proportion of subjects defined as VFs was similar between the groups (FDC group, 16%; SE group, 18%).
Immunologic Changes
At baseline the median CD4 cell counts were 304 cells/mL in the FDC group and 309 cells/mL in the SE group. During the 48-week treatment period, both treatment groups were associated with nominal increases in median CD4 cell counts from baseline. At week 48, the median change from baseline in CD4 cell counts was +47.5 cells/mL (25th and 75th percentiles, -18 and 117) in the FDC group and +95.0 cells/mL (25th and 75th percentiles, 7 and 169) in the SE group.
The incidence of HIV-1-associated conditions during the 48-week treatment period was low and comparable between the treatment groups (2% in each group). Furthermore, no subjects had progression of their HIV disease state during the study period.
Adverse Events
Incidences of drug-related AEs by body system were generally comparable between treatment groups (FDC group, 48%; SE group, 46%; Table 5). In both treatment groups, the body system with the highest incidence of drug-related AEs was gastrointestinal disorders (FDC group, 26%; SE group, 26%). Serious adverse events (SAEs) were reported in 16% (FDC) and 9% (SE) of subjects during the study; none of the SAEs reported were fatal. This difference was mainly driven by reports of drug hypersensitivity, which was reported in 9% of subjects in the FDC group and 4% in the SE group. The incidences of specific treatment-emergent grades 3 and 4 clinical chemistry and hematology abnormalities were low and generally comparable between treatment groups.
The incidence of all AEs and laboratory abnormalities were significantly lower during the second 24-week phase of the study in both treatment groups, especially severe AEs, drug-related AEs, and AEs leading to treatment terminations (data not shown). The safety profiles of the second 24-week phase were comparable between the 2 treatment groups.
Virology
Most VFs defined with confirmed viral load >400 copies/mL were successfully analyzed at failure by viral genotype and phenotype (n = 29/31). The K65R mutation was the most commonly selected mutation (FDC, n = 4; SE, n = 3). It did not occur in the presence of TAMs but was positively associated with the M184V/I mutation (7/7) and with nevirapine use (5/7). The L74V mutation was selected in 1 subject per arm and the Y115F mutation was selected for in 3 subjects (FDC, n = 2; SE, n = 1). The selection of the M184V/I mutation was low, probably due to the high incidence of this mutation at baseline. Phenotype was generally predictable from genotype, although in some instances partial deselection of the M184V/I mutation resulted in resensitization to 3TC.
Adherence
The coprimary adherence analyses were comparison of treatment groups at LTORT and comparison of groups across all time points in the randomized treatment phase. In the LTORT analysis, 67% (45/67) of patients in the FDC group were ≥95% adherent to ABC + 3TC compared with 53% (39/73) of patients in the SE group. Over the entire randomized treatment phase, 59% (49/83) of patients in the FDC group were ≥95% adherent to ABC + 3TC compared with 53% (42/79) of patients in the SE group. The differences between treatment groups for adherence to the ABC and 3TC portion of the regimens were not statistically significant (at LTORT P = 0.104; over randomized treatment phase, P = 0.417). Over the entire randomized treatment phase, the proportions of subjects who achieved ≥95% overall adherence to ABC + 3TC + TDF were 65% (44/68) versus 48% (35/73) at LTORT and 54% (46/85) versus 41% (33/80) in the FDC and SE groups, respectively.
DISCUSSION
In ART-naive subjects, ABC + 3TC has been successfully used as an NRTI backbone in a variety of regimens in both QD and BID dosing schedules.24-29 In 2 pilot studies, the elimination half-life of intracellular carbovir triphosphate (CBV-TP; the active moiety of ABC) was >12 hours, and inhibitory concentrations were maintained over 24 hours with either 300 mg BID or 600 mg QD dosing.30,31 Additionally, data from a clinical pharmacology study demonstrated a prolonged (21.64 hours) intracellular half-life of CBV-TP, which supports ABC 600 mg QD dosing.32 The bioequivalence analysis also indicated that ABC/3TC FDC QD was similar to ABC BID + 3TC BID based on AUC and Cmax.33 Therefore, ABC has received approval for QD dosing in the United States, the European Union, and other regions recently. The study is one of the first trials of ABC + 3TC in HIV therapy-experienced patients. The efficacy results of this study demonstrated that over 48 weeks of treatment, the FDC group was noninferior to the SE group with respect to plasma HIV-1 RNA AAUCMB, the primary end point of the study. AAUCMB provides a useful measure of antiviral activity in therapy-experienced subjects with VF,23 as it allows measurements at multiple time points to contribute to the overall measurement of efficacy. The end point is less sensitive to missing data, and this has a lower variability compared with absolute change in plasma HIV-1 RNA. However, its value as an end point is dependent on treatment groups being comparable with respect to baseline plasma HIV-1 RNA levels. Although AAUCMB was the primary end point in this trial, secondary end points included a comparison of the proportions of subjects with plasma HIV-1 RNA <50 and <400 copies/mL, as these are important surrogate markers of long-term efficacy. In the analyses of these secondary parameters of efficacy, there were comparable proportions of subjects with plasma HIV-1 RNA values <50 copies/mL in the FDC group compared with the SE group (50% vs. 47%) based on the TLOVR algorithm. Similar results were also observed in the proportions of subjects with plasma HIV-1 RNA values <400 copies/mL. Overall, these results supplement and are consistent with the observations based on AAUCMB. Subsequent to the enrollment of this study, frequent early failure leading to treatment termination was documented in several clinical trials including studies in ART-naive HIV-1-infected subjects treated with ABC + 3TC + TDF once daily34-36 and ddI + 3TC + TDF.37 These cases were characterized by the presence of M184V or M184I in nearly all instances, with the selection of K65R in 50%-82% of genotyped viruses. Although ABC, 3TC, and TDF have been used as part of the treatment regimens in this study, it is reassuring that similar high failure rates were not observed. This suggests that the fourth drug (new PI or NNRTI) had a modulating effect on efficacy with this triple NRTI regimen. More detailed analyses of the virologic data of this study are ongoing. Analyses of baseline genotypes observed no gross differences of numbers in subjects with NRTI resistance-associated mutations between the 2 treatment groups. However, several discrete differences within and between the groups may underlie the variable virologic response observed within this study. There were markedly more TAMs present in subjects treated with PIs across both treatment groups, and it is not clear if the presence of these TAMs was associated with the fewer VFs observed among the subjects treated with PIs. In addition, small differences in specific baseline viral mutations suggest a slight bias against the FDC group, where there were more subjects with resistance to 3TC, more subjects with M41L and T215Y/F mutations, and more subjects treated with NNRTI. This might have been offset to some extent by the slightly higher mean plasma HIV-1 RNA concentration in the SE group, and although the intent of the stratified randomization by study entry plasma HIV-1 RNA was to ensure an equal distribution between the groups, this may not have occurred. More detailed analyses of the virologic data from the treatment phase of the study are ongoing Both treatment groups were associated with small increases in CD4 cell counts from baseline, which was expected in this subject population. The levels of CD4 cell count increases over 48 weeks were not significantly different between the 2 treatment groups. The clinical relevance of the small difference between the 2 treatment groups is therefore unknown. In the primary adherence analysis, the proportion of subjects who were at least 95% adherent to the ABC and 3TC portion of the regimen was numerically higher in the FDC group than in the SE group (59% vs. 53% based on average adherence over the first 48 weeks; 67% vs. 53% based on LTORT). Although these differences were not statistically significant (P = 0.417 and 0.104, respectively), the study was not powered to detect adherence differences. It is notable that both methods of analyzing adherence indicate a trend favoring the FDC group. Moreover, the adherence benefit of an FDC product is likely to be more pronounced in the real world, where it not only reduces pill burden but also reduces the number of prescriptions to be filled. The real-world adherence benefits of FDC products are suggested by 2 observational database studies comparing another FDC product containing 3TC and ZDV (Combivir, GlaxoSmithKline) to its components. In these studies, patients receiving the FDC were 2-3 times more likely to achieve optimal adherence than patients given the component medications.38,39 SAEs were reported in 16% of subjects in the FDC group and 9% in the SE group. This difference was mainly driven by drug hypersensitivity. The rate of ABC HSR was 9% in the FDC group and 4% in the SE group. However, the significance of the observed differences was difficult to account for and most likely related to the relatively small size of the overall treatment population. Recent label changes for ABC-containing products are reflective of the rates observed in this study. In addition, hypersensitivity was reported as an SAE regardless of whether the event met the regulatory definition. The presentation of reported hypersensitivity did not change when ABC was administered QD or BID, and most cases were reported within 6 weeks of initiating ABC. No subject required hospitalization because of hypersensitivity. It is possible that overlapping toxicities associated with the use of NNRTI as the fourth drug may have contributed to a slightly higher proportion of subjects in the FDC group reporting hypersensitivity. The low absolute number of subjects reporting hypersensitivity makes it difficult to draw conclusions about the difference in reporting rate in each study group. Indeed, there were only 4 additional cases in the FDC group, and therefore a chance difference is likely.40 It has already been established that the safety and tolerability of the FDC regimen is comparable with the SE regimen during the first 24-week treatment period of this study.41 The significantly lower incidences of total AEs and laboratory abnormalities during the second 24-week period suggest that FDC-containing regimens are likely to have favorable safety profiles in the long-term treatment of HIV-1-infected patients, as has previously been demonstrated for SE-containing regimens. In conclusion, the results of this study demonstrated that the FDC arm had noninferior efficacy over 48 weeks to the SE arm in ART-experienced subjects with prior VF. The 2 regimens were generally well tolerated with similar safety profiles. The once-daily ABC/3TC FDC is a simple and effective NRTI backbone for both ART-naive and ART-experienced patients.
|
|
| |
| |
|
|
|
