| |
Protease Inhibitor-Based Regimens for HIV Therapy: Safety and Efficacy
|
| |
| |
JAIDS Journal of Acquired Immune Deficiency Syndromes:Volume 45 Supplement 11 June 2007pp S5-S13
[Supplement Article]
Walmsley, Sharon MD, MsC, FRCPC
From the Division of Infectious Diseases, Toronto General Hospital, University of Toronto, Toronto, Ontario, Canada.
Summary:
Antiretroviral (ARV) treatment strategies for HIV-infected patients continue to evolve. Over the past few years, there was a shift towards the use of nonnucleoside reverse transcriptase inhibitor-based regimens, mostly because of better tolerability, a lower pill burden, and improved adherence relative to using protease inhibitor (PI)-based regimens. Although the 2 strategies do afford similar potency and durability, the PI-based regimens provide a higher genetic barrier to the development of ARV resistance. This has become progressively more important for reasons that include increasing rates of baseline ARV resistance in newly infected patients and the risk of developing ARV resistance in treated populations with suboptimal adherence. With the introduction of novel ARVs and reformulated agents with more convenient dosing requirements, improved tolerability, and unique resistance characteristics, boosted PI-based strategies are increasingly being considered when initiating therapy in ARV-naive patients. In this article, the evidence for the use of boosted PIs as early therapy is reviewed, with emphasis on data available from comparative randomized controlled trials.
The introduction of the protease inhibitor (PI) in 19961 and the beginning of the age of highly active antiretroviral therapy (HAART) had a dramatic impact on the management of HIV infection. Use of these compounds was associated with an improvement in the quality and prolongation of life, with a dramatic fall in HIV-related opportunistic infections and decreased mortality. These breakthroughs were seen not only in clinical trials but in the clinical setting and provided new hope for those with HIV. There was a shift from HIV representing an acute, rapidly fatal, infectious disease to a chronic manageable illness. PI-based strategies were rapidly adopted into first-line antiretroviral (ARV) therapy.2-5
The early PIs were not without problems, however. Many of the compounds had poor oral bioavailability or short half-lives.1,6 Consequently, patients were required to take a large number of tablets at frequent dosing intervals. Life revolved around complicated dosing schedules and careful timing. Fear about missing or delaying doses and the loss of viral control leading to the subsequent emergence of resistance were major concerns. Toxicity then became apparent and problematic; short-term toxicity, including gastrointestinal (GI) side effects (nausea, vomiting, or diarrhea), rash, renal colic, dry skin, and paronychia, significantly affected quality of life. The real hesitation by patients to the initial use of this class of agent occurred, however, as a result of the link between PIs and fat redistribution changes of the lipodystrophy syndrome, which were described in 2% to 83% of patients on PI therapy.7-10 The resulting body fat changes had a negative effect on confidence, self-image, and well-being; additionally, noticeable changes in body shape and weight increased patients' anxiety about their ability to maintain confidentiality regarding their condition.11,12 Therefore, adherence was compromised and resulting durability of response decreased.13
Consequently, there was a major shift toward the use of nonnucleoside reverse transcriptase inhibitor (NNRTI)-based therapies as initial treatment in patients with HIV, and this remains the preferred strategy.14,15 Although NNRTI-based strategies have been shown to have immunologic and virologic responses comparable to those of PI-based therapy, there are some concerns and limitations. Efavirenz (EFV) is now contraindicated in the first trimester of pregnancy because of the potential for neural tube defects and must be used with consideration in women of child-bearing potential.16 Nevirapine use is limited by the black box warning on the label, recognizing the increased risk of hepatotoxicity in women with CD4 counts ≥250 cells/mm3 and men with CD4 counts ≥400 cells/mm3.17 Additionally, although the pill burden is lower, benefiting adherence, virologic failure may result in resistance to 2 drug classes: NNRTIs and nucleoside reverse transcriptase inhibitors (NRTIs).
In light of these issues, and the emergence of newer PIs with improved formulations and better dosing schedules, is it again time to consider PIs as first-line ARV therapy? The most recent International AIDS Society guidelines recommend any of the PI/ritonavir-based regimens (eg, lopinavir/ritonavir [LPV/r], fosamprenavir/ritonavir [FPV/r], saquinavir/ritonavir [SQV/r], atazanavir/ritonavir [ATV/r]) as first-line options in this class.18 The US Department of Health and Human Services recommends LPV/r, FPV/r, and ATV/r as first-line choices.19 What is the evidence supporting these new recommendations?
BOOSTED PROTEASE INHIBITORS
The first major advance for the new PIs came as a consequence of the use of the boosted approach.20 It was recognized that ritonavir in low doses could inhibit the liver enzyme cytochrome p450, which is responsible for the metabolism of many drugs, including the PIs. This resulted in significant increases in the blood plasma levels of other PIs when the drugs were coadministered. Consequently, PI-based strategies could now be administered at decreased dosing intervals and frequently with decreased total PI doses or pill burden, often without the food or fasting requirements attached to older, less forgiving PI regimens. The downside, however, was that the pharmacologic boosting could raise the PI levels too high, increasing the risk of concentration-dependent adverse events; also, ritonavir itself could frequently add to the adverse event profile, particularly GI side effects and hypertriglyceridemia. In addition, ritonavir could have an impact on other agents metabolized through cytochrome p450, consequently making drug interactions a concern.21
The initial PIs prescribed in the boosted form were SQV and indinavir (IDV).1,22 SQV was initially prescribed at 400 mg, together with 400 mg of ritonavir, administered twice daily. This dose was associated with a high incidence of GI intolerance and hypertriglyceridemia, however. Subsequent studies demonstrated the efficacy and improved tolerability of a dosing schedule comprising 1000/100 mg of SQV/r administered twice daily. IDV was also used, boosted with ritonavir at a dose of 800/100 mg of IDV/r administered twice daily. Although this combination proved highly efficacious, its use was inhibited by poor tolerability and an increased risk of renal colic. The 2 boosted PIs were compared in the MaxCmin1 study in a heterogeneous group of patients (ARV-naive, PI-naive, or PI-intolerant), which found both agents similar in terms of efficacy (HIV RNA suppression and CD4 cell count increases), but the overall performance of IDV was lower, largely driven by discontinuations attributable to adverse events.23
Lopinavir/Ritonavir TOP
The first (and only) coformulated boosted PI is LPV/r. In an early dose-finding study, 1 group of patients initiated LPV/r alone for 2 weeks at a dose of 200/100 mg administered twice daily or 400/100 mg administered twice daily, followed by combination therapy with lamivudine (3TC) and stavudine (d4T); the second group initiated all 3 agents simultaneously, with an LPV/r dose of 400/100 mg administered twice daily or 400/200 mg administered twice daily.24 The rate of viral decay from baseline to week 2 was similar regardless of whether LPV/r was used alone or in combination with the 2 NRTIs (-1.73 vs. -1.68 log10 copies/mL). Using a modified HIV RNA assay (Roche Amplicor Ultrasensitive assay; Roche Molecular Systems, Inc., Branchburg, NJ) with a lower limit of detection of 3 copies/mL, 56% of the patients in this study had at least 1 HIV RNA value ≦3 copies/mL through week 72 of treatment,25 further testifying to the relative potency of this agent.
This study was expanded to the first phase 2 study of LPV/r; the M97-720 study was an open-label study of LPV/r in combination with 3TC and d4T in 100 ARV-naive patients. Conducted between 1997 and 2005, the study provided the longest follow-up of any ARV regimen ever reported. The proportions of patients achieving suppression of HIV RNA levels to <400 copies/mL and <50 copies/mL at 48 weeks were 82% and 79%, respectively, in those patients who initiated LPV/r, d4T, and 3TC simultaneously by intention-to-treat (ITT) analysis.24 These results represented improvements over historic outcomes with older PIs, wherein only 41% to 60% and 34% to 49% of patients were able to achieve HIV RNA levels <400 to 500 copies/mL and <50 copies/mL, respectively.26-30 Seven-year follow-up data for the M97-720 study showed that 61% (ITT) of patients maintained suppression of HIV RNA levels to <400 copies/mL (98% by on-treatment [OT] analysis). Using the HIV RNA threshold level of <50 copies/mL, these figures were 59% (ITT) and 95% (OT) (Fig. 1). Further, the CD4 count increase has continued over time, with mean increases of approximately 500 cells/mm3, irrespective of the baseline CD4 cell count.31
Lopinavir/Ritonavir
The first (and only) coformulated boosted PI is LPV/r. In an early dose-finding study, 1 group of patients initiated LPV/r alone for 2 weeks at a dose of 200/100 mg administered twice daily or 400/100 mg administered twice daily, followed by combination therapy with lamivudine (3TC) and stavudine (d4T); the second group initiated all 3 agents simultaneously, with an LPV/r dose of 400/100 mg administered twice daily or 400/200 mg administered twice daily.24 The rate of viral decay from baseline to week 2 was similar regardless of whether LPV/r was used alone or in combination with the 2 NRTIs (-1.73 vs. -1.68 log10 copies/mL). Using a modified HIV RNA assay (Roche Amplicor Ultrasensitive assay; Roche Molecular Systems, Inc., Branchburg, NJ) with a lower limit of detection of 3 copies/mL, 56% of the patients in this study had at least 1 HIV RNA value ≦3 copies/mL through week 72 of treatment,25 further testifying to the relative potency of this agent.
This study was expanded to the first phase 2 study of LPV/r; the M97-720 study was an open-label study of LPV/r in combination with 3TC and d4T in 100 ARV-naive patients. Conducted between 1997 and 2005, the study provided the longest follow-up of any ARV regimen ever reported. The proportions of patients achieving suppression of HIV RNA levels to <400 copies/mL and <50 copies/mL at 48 weeks were 82% and 79%, respectively, in those patients who initiated LPV/r, d4T, and 3TC simultaneously by intention-to-treat (ITT) analysis.24 These results represented improvements over historic outcomes with older PIs, wherein only 41% to 60% and 34% to 49% of patients were able to achieve HIV RNA levels <400 to 500 copies/mL and <50 copies/mL, respectively.26-30 Seven-year follow-up data for the M97-720 study showed that 61% (ITT) of patients maintained suppression of HIV RNA levels to <400 copies/mL (98% by on-treatment [OT] analysis). Using the HIV RNA threshold level of <50 copies/mL, these figures were 59% (ITT) and 95% (OT) (Fig. 1). Further, the CD4 count increase has continued over time, with mean increases of approximately 500 cells/mm3, irrespective of the baseline CD4 cell count.31
The first head-to-head comparison of an unboosted PI versus a boosted PI was the M98-863 study, which remains the largest trial of LPV/r to date.32 This multicenter international trial randomized 653 ARV-naive patients to LPV/r at a dose of 400/100 mg administered twice daily or nelfinavir (NFV) at a dose of 750 mg administered 3 times daily, both in combination with 3TC and d4T. At week 48, 75% (67%) of subjects in the LPV/r arm maintained a viral load (VL) <400 (VL <50) copies/mL versus 63% (52%) in the NFV arm (P < 0.001); the corresponding hazard ratio (95% confidence interval [CI]) was 2.0 (1.5 to 2.7).32 Results at week 60 were similar, at 74% versus 61% (P < 0.001).33 These data on potency and efficacy are the foundation for current recommendations to use LPV/r when initiating PI-based therapy in ARV-naive patients.
Subanalysis of this trial also demonstrated that the efficacy of LPV/r was independent of baseline CD4 cell count or VL. In contrast, patients in the NFV arm fared less well when the baseline CD4 cell count was lower or the VL was higher.34
GI symptoms, including nausea, vomiting, and diarrhea, are the most commonly reported side effects with LPV/r. Diarrhea may prove to be the major drawback of the newly approved 800/200-mg once-daily dosing schedule, which was reported in 17% of these patients and only in 5% of those receiving conventional dosing (P = 0.01).35 Diarrhea occurred with 400/100 mg of LPV/r administered twice daily at rates similar to those for some comparator PIs in the studies described previously: 15.6% versus 17.1% for NFV,32 7% versus 9% for investigator-selected comparator PIs,36 and 11% versus 13% for FPV.37
The most frequently seen laboratory abnormalities with LPV/r use are changes in the lipid profile, with the bulk of changes occurring within the first month of treatment. In the Kaleobs observational cohort of 1315 patients on LPV/r in France, among a subset of patients who completed 18 months of follow-up (n = 127), mean total cholesterol changes from baseline to 1 month were increases of 13 mg/dL, 9 mg/dL, and 21 mg/dL for ARV-naive, PI-naive, and PI-experienced patients, respectively, whereas corresponding changes for triglycerides were increases of 1 mg/dL, 77 mg/dL, and 69 mg/dL.38 In clinical trials with ARV-naive patients, cholesterol elevations to >300 mg/dL were seen in 3% to 22% of patients who received LPV/r-containing regimens, and triglyceride elevations to >750 mg/dL were seen in 4% to 22% of patients who received LPV/r-containing regimens.39 Such changes are generally higher than those seen in clinical trials with comparator PIs, including ATV,40 ATV/r,41 and NFV.32 Elevations in hepatic transaminases are another important metabolic abnormality seen with LPV/r use, particularly in those coinfected with hepatitis B or C virus. Only 3% to 6% of patients discontinued treatment because of drug-related adverse events in a pooled analysis of 4 phase 2 and phase 3 clinical trials, however.42
An important parameter that may limit the effectiveness of therapy is the potential for ARV resistance. Eleven mutations in the viral protease gene contribute to in vitro resistance to LPV. The LPV mutation score, defined as the number of these mutations present in a given isolate, is exponentially related to decreases in the isolate's susceptibility by a factor of 1.74, and it is the cumulative effect of multiple mutations that ultimately leads to phenotypic failure. This property of LPV/r is referred to as its high genetic barrier to resistance.43
To date, in clinical trials of LPV/r plus NRTIs in previously ARV-naive patients, no individual has developed resistance to LPV. Of the patients with virologic failure, the only mutation that has been found consistently is the 184V mutation, conferring resistance to 3TC, or the occasional thymidine analogue mutation. Again, this contrasts with those patients randomized to NFV, in whom 45% with virologic failure had key mutations to PIs, and an increased proportion of patients also had mutations to 3TC and to the thymidine analogues.44
A new formulation of LPV/r is now available in many countries. This decreases the pill burden to 2 pills administered twice daily and does not require refrigeration, a factor of particular importance for the developing world. Pharmacokinetic studies have demonstrated less variability in drug levels with the new formulation and a lesser dependence on food for absorption. In a trial evaluating single-agent LPV/r in ARV-naive patients (IMANI-2), a subgroup (n = 30) of patients who started with the soft-gel capsule (SGC) and switched to the tablet formulation were assessed for tolerability and preference. In addition to a reduction in diarrhea with the LPV/r tablet, 46.7% of patients reported improved overall tolerability with the new formulation.45 In a survey conducted among patients in the United States who switched from LPV/r SGC administered twice daily to the tablet formulation administered twice daily, 82% of patients reported no or improved diarrhea and approximately 20% reported improved satisfaction and overall tolerability.46
Fosamprenavir/Ritonavir
Shortly after the LPV/r trials, once-daily FPV/r was compared with NFV in the SOLO study, an international, open-label, randomized study of 649 ARV-naive patients. At week 48, virologic responses were similar between arms by an ITT rebound/discontinuation = failure analysis. Virologic failure, however, did occur more frequently in the NFV arm compared with the FPV/r arm (17% vs. 7%). Similar to the observations in the LPV/r trials when failure occurred, no mutations to PIs were found in the boosted FPV/r arm, whereas mutations to protease were seen in 50% failing NFV. This, once again, reinforces the high genetic barrier to resistance of the boosted PIs. The main adverse reactions reported for FPV/r were diarrhea (9%), nausea (7%), and vomiting (6%). The incidence of diarrhea was higher in the NFV arm (16%; P = 0.008). Discontinuations attributable to adverse events were uncommon and occurred in 9% patients on the FPV/r arm.47
More recently, the Kaletra versus Lexiva with Epivir and Abacavir in ART-Naive patients (KLEAN) study compared FPV/r with LPV/r in 878 ARV-naive patients. Patients were randomized to receive twice-daily therapy with LPV/r SGCs or twice-daily FPV/r, each combined with fixed-dose combination abacavir/3TC. By ITT time to loss of virologic response (TLOVR) analysis, 73% randomized to FPV/r had a VL <400 copies/mL at week 48 compared with 71% in the LPV/r arm (95% CI: -4.84 to 7.05). In addition, response rates using the <50-copy/mL cutoff were 66% versus 65%, respectively (Fig. 2). It is concluded that FPV/r is not inferior to LPV/r in ARV-naive subjects. In a subanalysis, responses were similar in patients with high baseline VLs and low CD4 cell counts compared with those with less advanced disease. Of the small number of patients who developed virologic failure on their boosted PI regimen, a few developed minor PI-resistant mutations (2 on FPV/r and 2 on LPV/r), whereas 3 and 5 developed NRTI mutations, respectively. No differences between study groups occurred in treatment-related adverse events, including GI symptoms, or in the rate of grade 3 or 4 laboratory abnormalities, including total cholesterol, low-density lipoprotein cholesterol (LDL-C), and triglycerides.37
Results from week 24 of an ongoing 48-week clinical trial assessing an alternative daily dosing of FPV/r (1400/100 mg) combined with fixed-dose abacavir/3TC showed similar treatment responses in ARV-naive patients compared with those who received the standard daily dose of FPV/r (1400/200 mg).48 By ITT: missing = failure analysis, 87% versus 68% and 89% versus 76% achieved HIV RNA levels <50 copies/mL and <400 copies/mL in the 1400/100-mg FPV/r versus 1400/200-mg FPV/r arms, respectively. The median change from baseline in CD4 counts for the 1400/100-mg versus 1400/200-mg arms was 108 cells/mm3 versus 121 cells/mm3, respectively. Adverse events (grades 2-4) were comparable between treatment arms, and 5% of patients developed an abacavir hypersensitivity reaction. A study of similar design comparing the daily regimens of FPV/r (1400/100 mg) versus ATV/r (300/100 mg) combined with fixed-dose tenofovir/emtricitabine (TDF/FTC) in ARV-naive patients (ALERT) also showed similar results at an interim analysis at week 24.49 By ITT: missing = failure analysis, 79% versus 83% and 89% versus 89% achieved HIV RNA levels <50 copies/mL and <400 copies/mL in the FPV/r versus ATV/r arms, respectively. The median change from baseline in CD4 counts for the FPV/r versus ATV/r arms were 110 cells/mm3 versus 134 cells/mm3, respectively. Median fasting total cholesterol and triglycerides increased from baseline by 11% and 20%, respectively, in the FVP/r arm, whereas these same parameters increased by only 6% and 10%, respectively, in the ATV/r arm. Adverse events (grades 3-4) occurred in 49% of patients in the ATV/r arm (mostly hyperbilirubinemia) and 13% of patients in the FPV/r arm, however. These 24-week results of both trials should be interpreted with caution until week 48 data are available.
Saquinavir/Ritonavir
SQV/r was studied in ARV subjects in the induction phase of the Staccato cohort, largely in combination with didanosine and d4T.50 Of the 200 patients in the Thai subanalysis who received SQV/r at a once-daily dose of 1600/100 mg, 89% achieved a VL reduction to <50 copies/mL at week 24 and the median CD4 count rise was 122 cells/mm3. The drug was well tolerated, with only a small number of patients developing adverse events greater than grade 1. None of the 8 (2.9%) of 272 patients with virologic failure developed a major PI mutation.51,52
The MaxCmin2 trial compared LPV/r with SQV/r in a heterogeneous group of patients (ARV-naive, PI-naive, and PI-intolerant). The proportion of individuals achieving VL suppression was similar between the 2 treatment arms, as was the CD4 cell response. The proportion of patients not reaching the study-defined endpoint of failure (combination of viral failure and discontinuation of study drug) was greater in the SQV/r arm, however. The reasons for the difference remain unclear but could be related to differences in potency, tolerability, pill burden, and adherence.53
A second trial, GEMINI, studying the new 500-mg formulation of SQV/r (1000/100 mg administered twice daily) versus LPV/r (400/100 mg administered twice daily) combined with fixed-dose TDF/FTC in ARV-naive patients is currently ongoing.54 An interim analysis of the first 150 patients who completed week 24 showed that 69% and 81% of patients receiving SQV/r and 75% and 84% of patients receiving LPV/r achieved HIV RNA levels <50 copies/mL and <400 copies/mL, respectively. Mean CD4 changes from baseline were 157 cells/mm3 and 140 cells/mm3 in the SQV/r and LPV/r arms, respectively. The percentages of patients with fasting total cholesterol ≥200 mg/dL were 7.9% (SQV/r) versus 25% (LPV/r), and triglycerides ≥400 mg/dL were 0% (SQV/r) versus 9.4% (LPV/r). We must wait for the final 48-week analysis of the trial before concluding the noninferiority of SQV/r relative to LPV/r and to confirm the lipid differences.
Atazanavir/Ritonavir
The newest member of the PI class to be evaluated when boosted with ritonavir in ARV-naive subjects is ATV. The efficacy of unboosted ATV was comparable to that of NFV in 2 trials over 48 weeks.40,55 In a dose-finding trial, the mean changes in HIV RNA (log copies/mL) from baseline to 48 weeks were -2.51, -2.58, and -2.31 for 400 mg/d of ATV, 600 mg/d of ATV, and NFV 1250 mg, administered twice daily, respectively. The proportions with an HIV RNA level <400 copies/mL (ITT, noncompletion = failure) were 64%, 67%, and 53%, whereas the proportions with an HIV RNA level <50 copies/mL were 35%, 36%, and 34%, respectively. The mean increase in CD4 count was comparable at approximately 220 cells/mm3. Adverse events were also similar across treatments, with the exception of an increased rate of diarrhea in the NFV arm (56% vs. 15% to 20% in the ATV arm) and jaundice in the ATV arm (11% to 20% vs. 0% in the NFV arm). The lipid profile also favored the ATV group. At 48 weeks, the mean percent increases in the 400-mg ATV group compared with the NFV group were 5.1% versus 24.6% for total cholesterol, 5.2% versus 23.2% for fasting LDL-C, and 7.2% versus 49.5% for fasting triglycerides, with all comparisons being statistically significant.55
In a subsequent trial, BMS 034, of 810 ARV-naive patients, unboosted ATV was found to be noninferior to EFV when given in combination with fixed-dose 3TC/AZT. At week 48, 32% of the ATV subjects and 37% of the EFV subjects had a VL <50 copies/mL by ITT (TLOVR) analysis, whereas 70% and 64%, respectively, had a VL <400 copies/mL. Both arms of this study underperformed in relation to historic studies, and part of the discrepancy between the <50-copy/mL and <400-copy/mL results were attributed to the collection tubes used for processing the samples.56
With relation to current clinical practice, the relative efficacy of boosted versus unboosted ATV in first-line therapy is a more relevant concern. A study randomized 200 patients to unboosted ATV (400 mg) or boosted ATV (300/100 mg of ATV/r), both in combination with 3TC and the extended-release formulation of d4T, all dosed once daily. At week 48, ATV/r was demonstrated to be noninferior to unboosted ATV, with 86% versus 85% and 75% versus 70% of patients achieving an HIV RNA level <400 copies/mL and <50 copies/mL, respectively (Fig. 3).57 Paired isolates available from 2 of 3 ATV/r-treated patients who experienced virologic failure showed no presence of primary PI mutations, including the ATV-associated mutations I50L and I50I/L. In isolates available from 8 of 10 unboosted ATV-treated patients with virologic failure, however, ATV-associated and other PI-specific mutations were detected in 3 patients.58 Hyperbilirubinemia occurred more frequently with the use of the ATV/r (59%) versus ATV (20%).59 Similarly, the lipid profile was better in those on unboosted ATV versus ATV/r (total cholesterol: 6% vs. 15%; LDL-C: 16% vs. 23%; high-density lipoprotein cholesterol [HDL-C]: 29% vs. 30%; and fasting triglycerides: -3% vs. 26% from baseline, respectively; Fig. 4).57 An evaluation of patients matched at baseline for total cholesterol, HDL-C, and triglycerides who received ATV/r, FPV/r, or LPV/r was conducted at 12 months for changes in plasma lipids. Increases in total cholesterol and triglycerides were similar in those treated with FPV/r and LPV/r, whereas patients who received ATV/r had reductions in both parameters. Slight increases in HDL-C occurred among all 3 treatment groups.60
Studies to evaluate an induction/maintenance strategy are under way, wherein patients are initially treated with boosted ATV and then switched to the unboosted form if suppression is achieved. This should allow a comparison of the longer term durability of boosted compared with unboosted ATV and enable direct comparison of the lipid issues. The results of such studies should be of great interest to the practicing physician.
On a practical note, however, is the interaction between ATV and TDF. When used in combination, the drug concentrations of ATV are found to be lower and subtherapeutic.61 The mechanism for the interaction remains elusive; however, this can be overcome by the use of ATV in the boosted form. Given the increasing frequency of use of FTC/TDF as the backbone NRTI therapy in ARV-naive subjects, ATV is likely to be more commonly used in the boosted form.
Results from week 24 of an open-label single-arm evaluation of ATV/r (300/100 mg) combined with daily abacavir/3TC in ARV-naive patients (SHARE) reported that 85% and 90% of patients achieved HIV RNA levels <50 copies/mL and <400 copies/mL, respectively, by ITT: missing = failure analysis. The median CD4 count change from baseline was 127 cells/mm3 and increases in total cholesterol (17%), LDL-C (7%), HDL-C (34%), and triglycerides (30%) were observed. Suspected abacavir hypersensitivity reaction was reported in 8% of patients, and hyperbilirubinemia (grades 3-4) developed in 41% of patients, all of whom remained in the study.62
Studies are also ongoing comparing ATV/r with LPV/r, SQV/r, and FPV/r in ARV-naive subjects.
PROTEASE INHIBITORS VERSUS NONNUCLEOSIDE REVERSE TRANSCRIPTASE INHIBITORS
The debate remains as to how to initiate ARV therapy and whether to use a PI or NNRTI approach. In comparative trials, EFV was found to be comparable with ATV,56 but EFV was recently found to be superior to LPV/r in the Adult Clinical Trials Group (ACTG) 5142 study.63 In the virologic failures from the NNRTI approach, however, resistance to the NNRTI and NRTIs was frequently observed. In the ACTG 5142 study, 0 of 52 patients with failure with LPV/r plus 2 NRTIs who had resistance testing had evidence of PI mutations. Nevertheless, of the 33 patients with resistance testing after failure on EFV, 48% had resistance to the NNRTI and 10 of 33 had evidence of resistance to 2 classes. It should also be noted that in the 39 patients with resistance testing after failure on EFV with LPV/r, 4 (10%) had NRTI mutations, 27 (69%) had NNRTI mutations, and 2 developed major PI mutations.
Although the development of ARV resistance while on treatment may be a significant concern, the presence of ARV resistance at baseline in treatment-naive patients is a growing concern, particularly with regard to NNRTI use. In ACTG A5095, the presence of baseline NNRTI resistance more than doubled the risk of virologic failure among patients receiving EFV-containing regimens.64 These results illustrate the importance of baseline ARV resistance testing for certain populations of patients.
Other than the resistance problems,65 the other main concern is the issue of lipid abnormalities and the long-term risk of coronary artery disease. Cohorts of HIV-infected individuals are aging in many industrialized countries as a side effect of the successes of the HAART era.66 With this changing demographic comes important challenges relating to cardiovascular health and metabolic risk in HIV-positive individuals. The Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D) study, a multinational cohort including more than 36,199 person-years of follow-up, recently described a 3.5 per 1000 person-year incidence of myocardial infarction (MI) among treated patients, with a relative rate for MI of 1.26 (95% CI: 1.12 to 1.41) per year of ARV therapy.67 The metabolic toxicity of PIs like LPV/r are thus of particular concern, and antilipid pharmacotherapy is only part of the solution. ATV is a competitor agent in the PI class with a more favorable metabolic profile and, importantly, maintains good antiviral efficacy in an unboosted form. Such exclusion of ritonavir from the regimen, even compared with the low boosting doses used with LPV/r, may significantly reduce GI side effects and improve tolerability. Although lipid abnormalities are generally responsive to lipid-lowering therapy, this strategy further adds to the pill burden, and dangerous drug interactions must be carefully avoided.68
In ACTG 5142, the metabolic effects of LPV/r and EFV were assessed when used together (NRTI-sparing) or each in combination with 2 NRTIs. At week 96, LPV/r-treated patients had significantly greater triglyceride increases relative to those who received EFV plus 2 NRTIs. Changes in total, HDL, and non-HDL cholesterol were similar in patients treated with 2 NRTIs plus LPV/r or EFV, but significantly greater in those who received LPV/r plus EFV. Significantly fewer patients treated with LPV/r plus 2 NRTIs had lipoatrophy compared with patients who received EFV plus 2 NRTIs. Lipoatrophy frequency was lowest in the NRTI-sparing group and in NRTI-treated patients who received TDF-containing regimens.69
ALTERNATIVE DOSING AND NOVEL TREATMENT STRATEGIES TOP
Once-daily strategies with LPV/r, APV/r, and SQV/r are also being evaluated in attempts to improve adherence. Of course, it needs to be demonstrated that this does not compromise efficacy. Once-daily FPV/r was used in the SOLO study reported previously.45 This dose has been approved for ARV-naive subjects, although the twice-daily dose is still recommended for ARV-experienced patients. ABT 418 was a study designed to compare the safety and efficacy of once-daily versus twice-daily LPV/r, each combined with TDF and FTC in 190 treatment-naive subjects.70 Antiviral efficacy at 96 weeks was similar between arms, with 53% to 57% of patients achieving a VL <50 copies/mL by ITT analysis. PI mutations were not detected among the 23 genotypes available from 28 patients with virologic failure. There was a higher rate of discontinuations of therapy for adverse events in the once-daily arm (17% vs. 9%), and diarrhea was significantly more common in the once-daily dosing arm (-17% vs. 5%; P = 0.014).35 In another comparison of once-daily versus twice-daily LPV/r, ACTG A5073 showed that, overall, the probability of a sustained virologic response through week 48 was not different between the 2 treatment groups. However, in those patients with HIV RNA ≥100,000 copies/mL, the probability of a sustained virologic response through week 48 was significantly higher in the twice-daily LPV/r treatment group relative to the once-daily arm.71
Given the high genetic barrier to resistance, several groups of investigators have considered novel treatment strategies using ritonavir-boosted PIs, including the use of LPV/r or ATV/r monotherapy as a maintenance regimen in patients who had successfully suppressed HIV replication on a conventional triple-drug induction regimen. The Only Kaletra (OK) study was an open-label pilot study in 42 patients achieving a VL <50 copies/mL on a regimen of LPV/r plus 2 NRTIs, with no history of virologic failure on a PI-containing regimen. Patients were randomized to continue or stop the NRTIs, and similar proportions of patients maintained a VL <50 copies/mL at 48 weeks (95% vs. 81%, respectively; P = 0.16). Those who failed monotherapy had no evidence of PI mutations on genotyping and were successfully reinduced after reintroduction of the NRTIs.72 The OK04 trial, a study of similar design with a larger patient population (N = 198), reported slightly different results. In patients with virologic failure through week 48 who qualified for genotyping, 2% and 1% of those randomized to LPV/r monotherapy and LPV/r plus 2 NRTIs, respectively, developed major PI mutations.73 In a similar approach, patients in the ACTG 5201 study with sustained viral suppression discontinued their NRTIs, although continuing with ATV/r monotherapy. In this 24-week pilot study, 3 of 36 patients failed to maintain viral suppression, none of whom developed PI mutations.74
The induction maintenance strategy was also pursued in the M03-613 trial, in which ARV-naive subjects were randomized to LPV/r plus AZT/3TC induction therapy for at least 24 weeks, followed by maintenance therapy with LPV/r monotherapy after 3 consecutive VLs <50 copies/mL. The comparator groups continued LPV/r plus AZT/3TC or EFV with AZT/3TC. In the primary ITT analysis, 50% of the LPV/r-treated patients and 61% of the EFV-treated patients had a VL <50 copies/mL at 96 weeks of therapy (P = 0.23). There were a number of subjects on LPV/r monotherapy who had VLs between 50 and 500 copies/mL, and the concern about the long-term durability of the response and the risk of resistance remains.75 A slightly different approach is under investigation in the Monotherapy Antiretroviral Kaletra (MONARK) pilot study, in which 136 treatment-naive patients were randomized to initiate LPV/r monotherapy or LPV/r plus AZT/3TC. At 48 weeks, suboptimal virologic responses were seen in 11% and 13% of patients, respectively. At 48 weeks, 2 of 83 failing subjects in the monotherapy arm and 1 of 53 failing subjects in the combination therapy arm had PI mutations on genotyping, suggesting a possible increase in the risk of LPV/r resistance emerging with this strategy.76
SUMMARY
Pill burden has been a challenge for patients and practitioners alike since the start of the AIDS epidemic. Adherence is also of concern. Many PI strategies still require a number of pills and twice-daily schedules. For treatment-naive patients, the highly effective NNRTI-based combination of EFV/TDF/FTC has recently been reduced to just 1 pill per day. The NNRTI pill burden advantage is disappearing with the new agents and formulations, however. The new formulation of LPV/r and once-daily dosing can decrease the pill burden to 4 pills once per day. FPV/r, likewise given once daily, has a pill burden of 5, and the new formulation of SQV/r enables dosing at 3 pills twice daily. The lowest pill burden is with ATV, at 2 pills per day if given boosted or unboosted. Thus, the advantages of the NNRTI approach in terms of pill burden are rapidly fading.
Therefore, although therapy is continually individualized, the pendulum may be shifting back to the PI. The high genetic barrier to resistance and the lack of emergence of PI resistance with early failure, combined with the improved formulations decreasing pill burden, have prompted many physicians to reconsider PI as first-line therapy. Potency, efficacy, and durability of response, combined with safety and convenience, are now all important considerations as the number of options increases and the understanding of limitations evolves. Clinicians are now all in this for the long term, and their initial choices are crucial to ensuring success.
|
|
| |
| |
|
|
|
