HIV Articles  
Back 
 
 
Stopping antiretroviral therapy [OPINION]
 
 
  AIDS:Volume 21(13)20 August 2007p 1673-1682
 
Taylor, Stephena,b; Boffito, Martac; Khoo, Sayed; Smit, Erasmusb,e; Back, Davidd From the aDirectorate of Sexual Medicine and HIV, Birmingham Heartlands Hospital, Birmingham, UK
bDivison of Immunity and Infection, University of Birmingham, Birmingham, UK
cSt Stephen's Centre, Chelsea and Westminster Hospital, London, UK
dPharmacology Department, University of Liverpool, Liverpool, UK
eHealth Protection Agency, Birmingham, Birmingham Heartlands Hospital, Birmingham UK.
 
Combination antiretroviral regimens continue to fail in some patients [1,2]. Here we consider a factor that may be important; i.e. how patients stop therapy. We previously postulated that if one drug in a regimen has a significantly longer half-life than others, and all agents are stopped simultaneously [3] then the patient will be taking essentially 'functional monotherapy' after the shorter half-life drugs are eliminated (Fig. 1a). The probability of resistance emerging will depend on: (i) the drug's genetic barrier; (ii) the magnitude of viral replication when the drug remains at a concentration capable of inducing resistance; and (iii) the length of time the drug remains in the zone of resistance selection.
 
Figure 1. (pictures of graphs/charts) Strategies for stopping ARTs at end of report.
 
To prevent resistance emerging, stopping the drug with the longer half-life before the other regimen components has been suggested [3-5], but for how long? In some cases resistance to the longer half-life agent still occurs despite this approach [6]. With the development of drugs with extended half-lives, e.g. tenofovir and emtricitabine, it will become more difficult to provide guidance on stopping drug combinations, especially those with very disparate half-lives.
 
Stopping antiretroviral therapy (ART) occurs for many reasons, e.g. drug toxicity, intercurrent illness, after pregnancy, virological failure, patient choice, and in some situations 'stocks out' or cost [7]. Guidance is needed to minimize the probability of resistance developing and to preserve future treatment options.
 
The evidence so far
Anti-HIV activity relies on the presence of all components in the regimen; resistance develops more readily when replicating virus is exposed to only one or two drugs. The half-life (Table 1, see table 1 at end of article) of each individual component is therefore critical. For example, a 'balanced' regimen comprising three drugs with comparable half-lives would give plasma drug concentrations (or intracellular drug triphosphates of nucleoside reverse transcriptase inhibitors; NRTI) decaying at approximately the same rate, potentially rendering the regimen less vulnerable to resistance emerging when it is stopped. In contrast, an 'unbalanced' regimen may give functional monotherapy if all drugs are discontinued simultaneously.
 
Indirect evidence that the simultaneous discontinuation of unbalanced regimens may lead to resistance derives from case reports [5,48,49] and trials assessing structured treatment interruption (STI) (Table 2). In one report [5] a Caucasian man simultaneously discontinued a regimen (stavudine/lamivudine/nevirapine) because of hepatotoxicity. A subsequent regimen in which efavirenz replaced nevirapine was also stopped simultaneously. Genotyping subsequently confirmed the acquisition of the K103N mutation conferring broad non-nucleoside reverse transcriptase inhibitor (NNRTI) resistance. In the DART trial, plasma nevirapine concentrations were determined in 18 African adults discontinuing a nevirapine-based regimen. In that study there was a staggered stop (7 days). A conservative median plasma half-life of 39 h was estimated, and nevirapine concentrations were below the therapeutic range by one week [49]. It is important to note that after a single dose of nevirapine, half-lives of 60 h or more have been observed [27,68]. There is therefore particular vulnerability to developing resistance in women given single dose nevirapine for the prevention of mother-to-child transmission [69-71]. The most likely reason for the difference in half-life (i.e. single dose versus multiple dose) is the autoinduction of nevirapine during chronic therapy, leading to a shortened half-life [72]. When a 40-year-old African woman simultaneously stopped efavirenz, tenofovir, lamivudine and didanosine [52] her viral load (previously < 50 copies/ml) became detectable at 12 weeks (with wild type and the M230L/M mutation) and therapeutic plasma efavirenz concentrations were still present 8 weeks after stopping therapy.
 
In the Staccato study [65] 19 out of 36 patients randomly assigned to an STI strategy developed virological failure, with the acquisition of resistance mutations in patients receiving NNRTI. In another STI study [73] new drug-resistance mutations were observed in three out of eight patients discontinuing an efavirenz-based regimen. Both studies were terminated prematurely.
 
Furthermore, most other STI studies have reported resistance emerging in the stopping arm of the study, with resistance levels as high as 45% (Table 2) [55,56,63].
 
STI studies also provide important information on the time to virological rebound after stopping, as this will influence resistance development. Although somewhat variable, viral rebound is usually rapid and exponential, with an HIV doubling time of approximately one and a half days. The average patient will have detectable virus present by 2 weeks, whereas almost all patients will have a high viral load by 4 weeks [50]. Importantly, STI studies usually only enrol patients already fully suppressed
 
Stopping strategies
We have classified stopping strategies as: (i) 'simultaneous stop'; (ii) 'staggered stop'; (iii) 'switch' or 'exchange stop'; and (iv) 'protected stop'. We also consider stopping therapy in some special situations.
 
The simultaneous stop
Early treatment guidelines [74] recommended patients to stop all agents at once to ensure they were not taking mono or dual therapy by only discontinuing one agent in the regimen. Generally, for 'balanced' regimens comprising two short half-life NRTI (Fig. 1b) plus a protease inhibitor (boosted or unboosted) or else two long half-life NRTI/nucleotide reverse transcriptase inhibitors (NtRTI) plus nevirapine/efavirenz (Fig. 1d) this strategy is likely to be correct.
 
The 'simultaneous stop' may, however, be problematical when using longer half-life NNRTIs with short half-life NRTIs. Here the imbalance in half-lives is considerable. As nevirapine and efavirenz can persist at resistance-inducing concentrations for 2 weeks or more Fig. 1a [3,5,27,52], here the 'simultaneous stop' strategy is not advisable.
 
More problematical is stopping regimens containing longer half-life NRTI/NtRTI (i.e. emtricitabine and tenofovir) in combination with boosted protease inhibitors. This may lead to functional dual (Fig. 1c) or single (Fig. 1a) agent therapy if these agents are stopped together. In this situation, a different method of stopping may be more appropriate. When longer half-life NRTI/NtRTI are combined with nevirapine or efavirenz the matching of half-lives may support a simultaneous stop (Fig. 1d). We recognize that individual variation in drug metabolism [75] means that matching half-lives of all agents in the regimen is difficult. Clinical data are required to determine whether the simultaneous stopping of these agents will lead to resistance. This is important as fixed dose combination tablets become available [76].
 
The staggered stop
In attempting to overcome the potential for resistance development with nevirapine or efavirenz and short half-life drugs, some guidelines suggest a 'staggered stop' [77]. Here the long half-life drug is discontinued before stopping the other agents in the regimen. Three days were initially suggested, but then at least 5 days were recommended [5]. Lyons and colleagues [6] reported that for some individuals 5 days appeared too short. The correct duration still remains a matter of opinion. We recommend that nevirapine (after multiple doses) should be stopped at least 7 days (perhaps even 2 weeks) before other shorter half-life drugs, or a different stopping strategy should be employed. Fidler et al. [78] showed that when stopping efavirenz 5-7 days before a predominantly zidovudine/lamivudine-containing regimen, no NNRTI resistance was documented at week 4 after stopping treatment. Our own data [2], and that of others [52,79] have demonstrated that in some individuals the half-life of efavirenz can be greater than 100 h, and efavirenz concentrations can persist at resistance-selecting concentrations for 2 weeks or more. There are clearly ethnic differences in drug handling so it may be more appropriate to stop efavirenz at least 2 weeks before other shorter half-life drugs, or adopt a different stopping strategy (Fig. 1e).
 
The 'staggered stop' approach is problematical if another agent in the regimen has a long half-life, e.g. efavirenz, tenofovir, or lamivudine. In that situation, if efavirenz was stopped 2 weeks before the other two agents, the shorter half-life of lamivudine relative to tenofovir would potentially give functional tenofovir monotherapy.
 
The replacement or exchange stop
In this strategy the drug with the longest half-life is exchanged for a drug with a higher genetic barrier to resistance and a shorter half-life (e.g. lopinavir/ritonavir) for a period of time before the other agents and the new agent are stopped simultaneously (Fig. 1e). The advantage is that continued viral suppression is likely as the discontinued drug passes through the zone of resistance selection. Four weeks is probably advisable with this strategy to maintain viral suppression. The extended period will allow clearance of the NNRTI at the time of stopping other drugs. This method is likely to work when NNRTI are combined with the shorter-acting NRTI. Nevertheless, it may still be problematical with long half-life NRTI/NtRTI, i.e. functional monotherapy with either tenofovir or emicitrabine.
 
The protected stop
In an attempt to provide a universal stopping strategy, we are currently investigating the 'protected stop' (Fig. 1f), in which single agent lopinavir/ritonavir is given on the day patients stop all of the other ART agents. The regimen then continues for 4 weeks before stopping. The rationale comes from single agent lopinavir/ritonavir studies in which maintaining viral suppression was achieved with the minimal development of resistance [80,81]. The potential for resistance developing in the short time of single agent therapy is small, thus preserving future treatment options. Clinical data are, however, required before this can be recommended as a standard strategy.
 
Unplanned stopping of therapy and stopping in special situations
Sometimes stopping ART will be unplanned and will be the result of unpredictable circumstances such as emergency surgery, intercurrent illness, or severe drug toxicity. Clearly, the overall safety of the patient is foremost over the need to preserve future treatment options. Generally, stopping all the drugs simultaneously would be recommended; however, there may be differences according to the situation:
 
Severe or life-threatening situations caused by drug toxicity
We recommend stopping all agents simultaneously irrespective of the drug half-life.
 
Short-term interruption of therapy: surgical or medical conditions With balanced half-life drugs that do not require food for absorption, a simultaneous stop is advised with recommencement of the same regimen as soon as possible.
 
If the patient is allowed water then these drugs may be continued if it is anticipated that drug absorption can occur. If the drug requires food for optimal absorption and all components have similar half-lives, then temporary discontinuation of all drugs simultaneously is warranted, with recommencement of the same regimen as soon as the patient can resume food.
 
Patient taking an unbalanced regimen with unplanned stop
Either a simultaneous stop or a limited staggered stop could be considered, depending upon the severity of the illness. The 'exchange stop' or 'protected stop' are probably best avoided as it would mean starting a new drug of which the patient has had no previous experience in an emergency situation.
 
Stopping therapy after pregnancy
Some women receive ART in pregnancy to prevent mother-to-child transmission, but may not require therapy for their own health. If the regimen taken by the mother does not lend itself to simultaneous stopping, deferring the stop until after the mother and baby have established a stable routine may be preferable. This is only advisable if it is anticipated that the mother will take therapy until a planned stop date is agreed.
 
Discontinuation of emtricitabine, lamivudine or tenofovir in hepatitis B co-infected patients
If patients are hepatitis B surface antigen or e antigen positive, then discontinuation of these agents may lead to an exacerbation of hepatitis or a 'hepatitic flare' [82]. These patients require close monitoring. Some suggest replacing the agents with activity against both HIV and hepatitis with drugs with a predominantly hepatitis B activity, i.e. adefovir, entecavir or telbivudine. In such cases expert opinion should be sought. It would not be recommended to continue lamivudine/emicitrabine or tenofovir alone for their antihepatitis B activity because of the risk of selecting HIV-resistant variants [1].
 
Conclusion
Whatever the reasons for stopping ART, pharmacological and virological issues must be considered in order for a clinician to give guidance. Understanding the pharmacokinetics of each drug is critical, because there is the potential for functional monotherapy if drugs with different half-lives are stopped simultaneously. Although there are few prospective clinical trial data to guide clinicians on how best to stop ART, there are clearly different strategies.
 
Fig. 1. Strategies for stopping antiretroviral therapy. (a) Representation of three drugs A, B, and C with very disparate half lives. When stopping the three agents simultaneously drugs A and B are quickly cleared from the system. In contrast, drug C remains at therapeutic/subtherapeutic concentrations for a prolonged period creating a state of 'functional monotherapy'. (b) Simultaneous stopping of three drugs A, B and C with 'balanced' short half-lives. The degree of 'forgiveness' for missed doses will be low, but on stopping simultaneously the drugs may have cleared from the body before viral rebound occurs. The potential for resistance developing may be less than the scenario depicted in (a). (c) Simultaneous stopping of drugs A, B and C when drug A has a significantly shorter half-life than the other two agents. Whether resistance develops depends on the degree of viral suppression maintained by drugs B and C together and the time spent in the 'zone of resistance selection' when viral rebound occurs. (d) Simultaneous stopping of three drugs with relatively 'balanced' long half-lives. All three drugs will continue at therapeutic concentrations for a prolonged period of time after stopping these agents simultaneously. This will provide a large degree of 'forgiveness' for missed doses. Whether resistance will be induced as the three agents fall through the zone of resistance selection is currently unclear. It is theoretically less likely than with the scenario in (a). (e) Two stopping strategies: first, the 'staggered stop', in which drug C, a drug with a long half-life and a low genetic barrier to resistance, is discontinued for a specified period of time before agents A and B (drugs with relatively balanced short half-lives); second, the 'replacement or exchange stop', in which the drug with the longest half-life is stopped and exchanged for drug D, which has a shorter half-life (and higher genetic barrier to resistance). After a specified period of time the three drugs B, C, and D are stopped simultaneously. (f) The 'protected stop' strategy. In this scenario all drugs A, B, and C are stopped simultaneously (irrespective of their individual half-lives) at the same time as drug D is started. This strategy requires drug D to be potent enough to maintain viral suppression while drugs A, B, and C are cleared from the body. Furthermore, it should not induce resistance to itself when it is stopped a specified number of weeks later.
 

Mono-1.gif

laDos-2.gif

replacing-3.gif

CCR5-4.gif

Vicriv-5.gif

 
 
 
 
  icon paper stack View Older Articles   Back to Top   www.natap.org