Stopping antiretroviral therapy: easier said than done

HAART is now well established as a powerful means of reducing morbidity and mortality in HIV-infected individuals [1,2]. The recent rapid expansion of therapeutic options has allowed for a renewed consensus on the definition of the goal of therapy. As a result, it is now widely accepted that the goal of HAART is to achieve sustained full suppression of plasma HIV-1 RNA (i.e. a plasma HIV-1-RNA level < 50 copies/ml) in treatment-naive and treatment-experienced patients [3,4]. HIV-1 viral characteristics, namely the high replication rate, high mutation rate, and ability of viral proteins to function despite the presence of mutations, allow the virus to evolve rapidly when confronted with incompletely suppressive regimens [5]. Suboptimal drug exposure, often generated through incomplete adherence or altered drug metabolism in individual patients, has been identified as a key determinant of treatment failure, and thus the selection of drug-resistant virus [6]. Near-perfect levels of adherence, typically over 95%, are generally required to achieve optimal responses to HAART [7]. A substantial body of evidence supports current guidelines, focused primarily on how to choose an optimal initial regimen for the treatment of therapy-naive and, to a lesser extent, therapy-experienced patients [8–10]. Limited attention has, however, been focused on the issue of how to stop antiretroviral therapy appropriately. Recent data have shown that anything less than the continued use of HAART can be associated with increased morbidity and mortality, and as such intermittent therapy is generally discouraged [11]. As a result, it is unlikely that a body of new evidence will become available in the foreseeable future regarding how best to interrupt therapy. Treatment interruptions in the clinical setting remain a frequent event, however, typically in patients suffering from toxicity, concomitant illnesses or treatment fatigue. Until prospective evidence is available, therefore, expert opinion will remain the only possible way to close this gap. The report by Taylor and colleagues [12] in this issue of AIDS represents a significant step forward in this regard. The authors provide a comprehensive review of the available evidence, predominantly from pharmacokinetic data and trials involving structured treatment interruptions, and propose rational strategies for discontinuing therapy designed to avoid inadvertent suboptimal drug exposure, which has been shown to promote the development of viral drug resistance. In brief, the authors divide HAART regimens based on their pharmacological profiles into balanced regimens, which include agents with similar half-lives, and unbalanced regimens, which include agents with differing half-lives. Then, the discontinuation of HAART event is classified as planned or unplanned. Using this two-by-two matrix they propose a rational approach to treatment discontinuations in the clinical setting. From a conceptual standpoint, a planned simultaneous discontinuation of a balanced regimen should be preferred, whenever possible. Similarly, for a balanced regimen, an unplanned discontinuation would be less of a concern. The latter event points to the need to encourage the preferential use of balanced regimens whenever possible. Similarly, this also illustrates the need to encourage the development of fixed-dose combinations with balanced pharmacological profiles. If the regimen is unbalanced and the discontinuation is planned, an intermediate step could be construed to balance the regimen first and interrupt it later as a planned simultaneous discontinuation of a now-balanced regimen. How to balance the regimen best remains to be tested, and although the authors provide some valuable suggestions in this regard, there are potential pitfalls to this approach. For example, the ‘replacement stop’ approach is essentially designed to deal with the non-nucleoside reverse transcriptase inhibitor (NNRTI) component of a HAART regimen, as these agents generally have relatively long half-lives. There may, however, be drug interactions between the residual NNRTI and the protease inhibitor used to replace it, leading to lower protease inhibitor exposure. This effect can be counteracted by starting with a higher protease inhibitor dose, but the resulting increased risk of side effects or other drug interactions may make this strategy more complicated and unacceptable to the patient. Emerging side effects to the new drug may also represent a challenge to be considered. Finally, it is not clear whether previous decreased susceptibility to the ‘replacement’ agent may have an effect on the outcome of this strategy, particularly among extensively pretreated patients. The recent availability of new and well-tolerated drugs, some of them also active in new targets, may circumvent some of these limitations. An alternative replacement strategy discussed by Taylor et al.[12], the ‘protected stop’, involves discontinuing the full HAART regimen and beginning monotherapy with lopinavir/ritonavir for 4 weeks. This approach is supported by the results of the OK trial [13], and is currently under evaluation in clinical trials in the United Kingdom. Some of the same limitations outlined above may also, however, apply to this approach, particularly among extensively pretreated patients. A ‘staggered stop’ is another proposed alternative, suitable for the discontinuation of an unbalanced NNRTI-containing regimen. The optimal duration of the staggered phase is, however, poorly defined, with most recommendations supporting a period of 5–7 days [8–10]. Supportive evidence is available from a clinical trial in the setting of the prevention of mother-to-child-transmission, in which the addition of zidovudine and lamivudine for 4 or 7 days after nevirapine dosing reduced the risk of the development of an NNRTI mutation from 60 to 10–12% [14]. Individual and ethnic differences in NNRTI clearance caused by genetic polymorphisms create significant difficulty in accurately predicting the time period required for drug clearance [15,16]. Finally, the remaining and more significant challenge is how to deal with unplanned discontinuations of unbalanced regimens. Unfortunately, as outlined before, this challenge tends to arise quite frequently in clinical practice. In this instance, the ‘simultaneous stop’ approach will continue to occur whenever patients experience significant toxicity, regardless of the composition of the regimen (balanced or unbalanced), in which the risk of death or lasting harm, real or perceived, may outweigh the potential risk of the loss of drug (or drug class) activity secondary to resistance. Again, this emphasizes the need to plan for eventual treatment discontinuations proactively, and preferentially to use balanced regimens and fixed-dose combinations with balanced pharmacological profiles whenever possible.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.