| |
Adherence to Antiretroviral Therapy:
How Much Is Enough? Editorial, Trip Gulick, MD
|
| |
| |
Clinical Infectious Diseases Oct 1, 2006;43:942-944
Roy M. Gulick
Cornell HIV Clinical Trials Unit, Division of International Medicine and Infectious Disease, Weill Medical College of Cornell University, New York, New York
The development of effective antiretroviral therapy in the mid-1990s led to significant virologic suppression and immune reconstitution and resulted in dramatic reductions in HIV-related morbidity and mortality [1, 2]. However, initial potent antiretroviral regimens were inconvenient: for example, a standard regimen of zidovudine, lamivudine, and indinavir that was associated with dramatic decreases in HIV RNA levels [3] and that significantly delayed progression of HIV disease [4] required that a patient take 20 pills per day, divided every 8 h, in a fasting state. In the late 1990s, Paterson et al. [5] assessed adherence in 99 largely treatment-experienced patients who were receiving regimens containing nucleosides and protease inhibitors (PIs) (mostly indinavir and nelfinavir), and they found that more than one-half of their patients experienced virologic failure if their rate of adherence, as assessed by an electronic monitoring system that was attached to the pill bottle, was <95%. This widely cited article led to the principle that >95% adherence to antiretroviral medications was required for optimal treatment effect.
Over the past 10 years, the convenience of potent antiretroviral regimens has markedly improved. The nonnucleoside reverse-transcriptase inhibitors (NNRTIs; e.g., efavirenz and nevirapine) that were originally sidelined because of the rapid emergence of drug resistance have demonstrated their potency as easy-to-take, first-line regimens when combined with dual nucleoside analogues; in head-to-head comparisons, NNRTIs outperformed indinavir-based [6], nelfinavir-based [7], and triple-nucleoside-based [8] regimens. Because low doses of ritonavir inhibited the cytochrome P450 CYP 3A4 enzyme system and thereby increased the levels of other PIs [9], the use of twice-daily or even once-daily "ritonavir-boosted" PIs also became standard antiretroviral therapy. The development and approval of coformulated antiretroviral drugs (twice-daily zidovudine/lamivudine in 1997, lopinavir/ritonavir and zidovudine/lamivudine/abacavir in 1999, and once-daily abacavir/lamivudine and tenofovir/emtricitabine in 2004) led to decreased regimen pill counts. Improved formulations of antiretroviral drugs, including enteric-coated, once-daily didanosine (approved in 2000), the amprenavir prodrug fosamprenavir (approved in 2003), and the tablet formulation of lopinavir/ritonavir (approved in 2005; that both reduced the pill count and eliminated the need for refrigeration), also improved convenience. Because of these improvements, every-8-h regimens yielded to twice-daily and even once-daily regimens. In fact, earlier this year, a 1-pill, once-daily potent coformulated combination regimen of tenofovir-emtricitabine-efavirenz was approved by the US Food and Drug Administration.
Because of the widespread use of these newer, more convenient regimens, several groups have returned to the question of what level of adherence is required for optimal virologic effect. Maggiolo et al. [10] assessed adherence with the AIDS Clinical Trial Group adherence questionnaire, administered to 543 patients who were taking first-, second-, or third-line antiretroviral regimens that consisted of PI- (primarily indinavir and nelfinavir), NNRTI-, or triple-nucleoside-based treatment and in whom viral loads were suppressed to <50 copies/mL for >6 months [11]. They observed that patients who were receiving an NNRTI-based regimen had significantly higher rates of adherence than did those who were receiving PI-based regimens (94% vs. 90%; P = .02). Also, adherence to an NNRTI-based regimen of >75% was associated with maintained virologic suppression, with a <10% virologic failure rate. In contrast, with the PI-based regimen, at least 85% adherence was required to maintain virologic suppression at the same level. They concluded that NNRTI-based regimens were "more forgiving" than PI-based regimens.
In this issue, Bangsberg [12] reports the results of an assessment of adherence in a marginally housed population of 110 patients (50 with prior nucleoside therapy experience) who were starting their first potent antiretroviral regimen, one-half of whom received a PI (primarily indinavir and nelfinavir) and one-half of whom received an NNRTI (nevirapine or efavirenz). Using unannounced pill counts, the best rates of virologic suppression were observed when adherence was at least 95%. However, there was a significant difference between the 2 regimens because a majority of patients taking NNRTI-based regimens reached virologic suppression when adherence was at least 54%, whereas a majority of patients who were receiving PI regimens experienced virologic suppression only when adherence rates exceeded 73%. They concluded that virologic suppression was possible with moderate levels of adherence to NNRTI-based regimens.
From prior clinical trials and from these assessments, it appears that NNRTI-based regimens offer sustained virologic suppression even with some degree of imperfect adherence. This is likely because of several factors, including the inherent antiretroviral potency of NNRTI drugs (in combination with nucleoside analogues), as well as their convenience, tolerability, and long plasma drug half-lives [6, 7, 13]. Together, these factors have been deemed "forgiveability"\that is, how likely an antiretroviral regimen will be able to durably suppress viremia in the setting of imperfect adherence. An illustration of this concept can be seen in a pilot study reported by Dybul et al. [14] that observed 8 patients who were receiving a once-daily efavirenz-based regimen who underwent short-cycle intermittent therapy, with 7 days on therapy followed by 7 days off therapy. By employing this strategy, they found that patients maintained virologic suppression for up to 84 weeks and that efavirenz levels were detectable at day 7 of the off-therapy period. This pilot study clearly illustrates the increased forgiveability of NNRTI-based regimens.
Several caveats are in order. The optimal way to assess adherence to antiretroviral therapy is not known. The Paterson et al. [5], Maggiolo et al. [10], and Bangsberg [12] studies each used 3 different methods to assess adherence: electronic monitoring medication caps, a self-reported questionnaire, and unannounced pill counts (as well as electronic caps in a subset of patients), respectively. The ACTG 359 team previously reported that they assessed adherence in 254 treatment-experienced patients using 4 different methods (pill counts, a self-reported questionnaire, electronic monitoring medication caps, and PI drug level testing) and concluded that only patient self-report and the PI level were significant predictors of virologic response [15]; however, the best method to assess adherence remains unclear. In addition, the length of assessment of adherence was relatively short in all 4 studies, at only 6-9 months. It is probable that adherence will change over time, particularly as patients are asked to take antiretroviral drugs for periods of years (or decades) and long-term treatment fatigue becomes an issue.
In addition, the Paterson et al. [5], Maggiolo et al. [10], and Bangsberg [12] studies focused primarily on PI-based regimens that used indinavir or nelfinavir (without ritonavir boosting); these regimens have demonstrated inferior potency compared with both NNRTI- [6, 7] and ritonavir-boosted regimens [16] and, as a result, are much less commonly recommended or used today [17, 18]. Current PI regimens, including once- or twice-daily lopinavir/ritonavir or fosamprenavir boosted with ritonavir and once-daily atazanavir, with or without ritonavir, offer potency, convenience, and tolerability. Because of these factors, boosted PI regimens are likely to provide forgiveability as well.
The consequence of suboptimal adherence with antiretroviral regimens is intermittent or rebound viremia and, ultimately, the development of drug resistance. Despite their potency, lamivudine (or emtricitabine) and NNRTI-containing regimens rapidly select for resistance to lamivudine (or emtricitabine), to the NNRTI, or to both in the presence of ongoing viral replication [19]. In contrast, lamivudine (or emtricitabine) and ritonavir-boosted PI regimens select for lamivudine (or emtricitabine) resistance but not for PI resistance in the presence of ongoing viral replication [20, 21]. Head-to-head comparative studies with NNRTI- versus boosted PI-based regimens are underway and will help to further inform the field.
The choice of the optimal antiretroviral regimen depends on the consideration of a number of factors, including potency, durability of virologic and immunologic responses, tolerability, convenience, drug-resistance barrier, and pharmacokinetic issues; ultimately, the choice must be individualized. For patients who are taking antiretroviral therapy, the treatment goal is maximal virologic suppression, maintenance or enhancement of immune function, and prevention of HIV-related disease and death; this is best accomplished with (near) perfect adherence to antiretroviral medications. Newer data with NNRTI-based regimens (which also likely apply to ritonavir-boosted, PI-based regimens) reassure us that our current antiretroviral regimens can tolerate some degree of imperfect adherence while maintaining the beneficial effects of HIV therapy in most patients.
|
|
| |
| |
|
|
|
