British HIV Association (BHIVA) guidelines for the treatment of HIV‐infected adults with antiretroviral therapy

1.0 Synopsis 1.1 When to start treatment 1.2 Initial therapy 1.3.1 When to switch therapy for virological failure 1.3.1 Which drugs to use following failure of initial therapy 1.4 When to switch therapy in the absence of virological failure 1.5 Resistance testing 1.6 Therapeutic drug monitoring (TDM) 1.7 New drugs likely to be used in the near future 1.7.1 Atazanavir 1.7.2 Enfuvirtide (T20) 2.0 Introduction 2.1 Purpose of guidelines 2.2 Basing recommendations on evidence 2.3 Use of evidence published as abstracts 2.4 Implication for research 2.5 Use of surrogate marker data 2.6 Issues concerning design and analysis of clinical trials 2.6.1 Trial designs 2.6.2 Method of analysis 2.6.3 Intention to treat and on treatment analysis 2.6.4 Equivalence 2.6.5 Cross-study comparisons: presentation of data 2.7 Adverse event reporting 3.0 When to start treatment 3.1 Primary HIV infection 3.1.1 Treatment of primary HIV infection to alter the natural history 3.1.2 Use of structured treatment interruption (STI) in acute infection 3.1.3 Treatment during PHI for immediate clinical benefit 3.1.4 Treatment during PHI to reduce onward transmission 3.1.5 Recommendations for starting treatment in PHI (CIII) 3.2 Symptomatic HIV infection 3.3 Asymptomatic HIV infection 3.3.1.1 Individuals with CD4 counts 3 3.3.1.2 Individuals with CD4 counts >350 3.3.1.3 Individuals with CD4 counts 201-350 cells/mm3 3.1.6 Recommendations regarding asymptomatic chronic HIV infection 4.0 What to start with 4.1 Choices of initial therapy 4.2 Which HAART regimen is best? 4.2.1 Two NRTIs plus an NNRTI 4.2.1.1 Efavirenz (EFV) 4.2.1.2 Nevirapine (NVP) 4.2.1.3 Delavirdine 4.2.2 Two NRTIs plus a PI 4.2.2.1 Two NRTIs plus a boosted PI 4.2.3 Three NRTIs 4.3. Choice of NRTI backbone for initial therapy 4.4. Recommendations for initial therapy: conclusions 5.0 Issues concerning antiretroviral use 5.1 Follow up of the HIV patient 5.2 Adherence 5.3 Toxicity 5.3.1 Lipodystrophy 5.3.1.1 Management of lipodystrophy 5.3.1.2 Other therapies 5.3.1.3 Conclusions 5.3.2 Mitochondrial toxicity and lactic acidosis 5.3.2.1 Aetiology of NRTI induced mitochondrial toxicity 5.3.2.2 Lactic acidosis and hyperlactataemia 5.3.2.3 Incidence 5.3.2.4 Clinical and laboratory features 5.3.2.5 Management of hyperlactataemia and lactic acidosis 5.3.2.6 Recommendations for managing lactic acidosis 5.4 Resistance testing 5.4.1 Recommendations 5.5 Therapeutic drug monitoring (TDM) 5.5.1 Drug levels and efficacy 5.5.2 Drug levels and toxicity 5.5.3 Use of TDM 5.5.4 Inhibitory quotients 5.6 Structured treatment interruption (STI) 6.0 Changing or stopping therapy in the absence of virological failure 6.1 Patients started on regimens that are not currently recommended for initial therapy 6.1.1 2NRTI plus unboosted PI regimens 6.1.2 3NRTIs 6.1.3 Regimens containing stavudine 6.1.4 Non-HAART regimens 6.2 Patients on recommended regimens 6.2.1 Switching from PI-based regimens 6.2.2 Switching between NNRTIs 6.2.3 Stopping NNRTI-based regimens in non-emergency situations 6.3 Stopping therapy in individuals with complete viral suppression (STI) 6.3.1 Intermittent on-off therapy cycles of one month or longer 6.3.2 Intermittent on-off therapy cycles of one week 6.3.3 Discontinuation of therapy with re-start based on CD4 count (CII) 7.0 Changing and stopping therapy for virological failure 7.1 Virological failure 7.1.1 Viral load blips 7.1.2 Sustained viral load rebound 7.2 Changing therapy (BII) 7.2.1 Failure of two nucleoside analogues plus a protease inhibitor (BII) 7.2.2 Failure of two nucleoside analogues plus a non-nuceloside reverse transcriptase inhibitor (BIII) 7.2.3 Failure of triple nucleoside analogue therapy (BIV) 7.3 Patients whose therapy fails having used at least three classes of drugs (‘salvage therapy’) 7.3.1 Criteria for success in patients exposed to multiple drug classes 7.3.2 Principle of optimizing success in highly treatment experienced patients 7.3.3 Management of patients with multiple class resistance 7.3.4 Recommendations for subsequent virological failure (third or more regimen) (BIII) 8.0 New therapies 8.1 Enfuvirtide (T20) 8.2 Atazanvir 8.3 Extended release stavudine (d4T) 8.4 Emtricitabine (FTC) 8.5 Tenofovir 8.6 Fosamprenavir 9.0 Reference 10.0 Conflict of interest List of Tables Table 1. Basic net NHS cost of antiretroviral drugs Table 2. Grading of recommendations and levels of evidence Table 3. Recommendation for starting treatment Table 4. Initial HAART regimens Table 5. Currently available non-nucleoside reverse transcriptase inhibitors (NNRTIs) Table 6. Currently advised protease inhibitors (PIs) for initial therapy Table 7. Routine tests and examinations in the HIV patient Table 8. Meta-analysis of trials of HIV resistance testing Table 9. Proposed indications for therapeutic drug monitoring (TDM) Table 10. Potential objectives of structured treatment interruptions (STI) in different clinical settings Table 11. Changing therapy on first virological failure [BIII] Table 12. What to change to after first virological failure: summary of recommendations [BII/IV] This synopsis represents a consensus from the writing committee based upon the evidence given in detail in the full guidelines. It aims to offer a pragmatic approach to some difficult questions concerning HIV therapy. Some aspects of this synopsis are controversial and clinicians are advised to read the current guidelines in full, alongside those regarding adherence to treatment and hepatitis B and C coinfection [1, 2]. In asymptomatic patients this decision should be driven primarily by the CD4 count. A value of 200 cells/μL represents the minimum level at which treatment should be advised. Treatment should be initiated when the CD4 count is between 200 and 350 cells/μL, and the exact timing should depend on individual factors such as symptoms, patient preference, likely adherence and potential toxicity. In this range, the rate of CD4 decline, viral load level and age provide additional information to the CD4 count on the short-term risk of progression. Severely symptomatic disease is unusual at CD4 counts above 350 cells/μL in chronic HIV infection, but provides a rationale for treatment when it occurs. For patients with primary HIV infection, treatment is only recommended for the purpose of resolving severe symptoms. Otherwise the committee feels there is insufficient evidence to recommend treatment outside clinical controlled trials. When choosing therapy, the committee takes the view that in the present state of knowledge more weight should be given to ease of adherence and minimisation of toxicity, including development of lipodystrophy, than to the likely pattern of resistance mutations emerging following treatment failure. There is more evidence and experience available in support of initial regimens which include two nucleoside analogue reverse transcriptase inhibitors (NRTIs) plus either a protease inhibitor (PI) or an non-nucleoside reverse transcriptase inhibitor (NNRTI) than for other combinations, so we generally support the use of such combinations for starting therapy. Combinations including stavudine (d4T) are not recommended for initial therapy due to possible risks of more rapid development of lipodystrophy. Several recent studies have shown excellent surrogate marker responses for combinations containing lamivudine (3TC) and efavirenz. On current data, the nucleoside-only combination abacavir (ABC)/3TC/zidovudine (ZDV) should not be considered as a first line option for initial therapy (but see section 4.2.3 where this is discussed more fully). NNRTI regimens produce equivalent or superior surrogate marker responses compared with unboosted PI regimens. There are no studies of sufficient size that directly compare NNRTIs with boosted PIs, which the committee recommend when a PI containing regimen is used. Therefore, the choice between these needs to be based on other considerations. The advantages of NNRTIs over currently licensed PIs include suitability for use in once daily regimens, lower pill burdens, fewer clinically important lipid abnormalities, and a lower frequency of central fat accumulation. There is no evidence of a difference in clinical effectiveness between PI and NNRTI regimens in late disease. PIs may have advantages in that broad cross class resistance is less likely than for the NNRTIs and have less hepatotoxicity and rash. Hence at present the writing committee considers that NNRTIs are more suitable as the agents of first choice for initial regimens. The choice of NNRTI will need to be made by clinicians interpreting a range of studies, including the 2NN trial. The latter study indicates that there is little difference in potency between nevirapine (NVP) and efavirenz (EFV). NVP treatment was associated with more rash and hepato-toxicity, whereas the disadvantages of EFV are a variable rate of psychological disturbances and the possibility of teratogenesis. A number of options for once a day therapy are now available and these regimens are likely to become increasingly popular. A number of strategic studies are in progress to address this issue. Until they report, the committee takes a pragmatic view. In cases where tolerable treatment options are available that are likely completely to inhibit viral replication, these should be used as soon as virological rebound has been confirmed by two consecutive viral loads >400 copies/mL. The committee emphasises that before therapy is changed, factors other than resistance, such as poor adherence and pharmacokinetic effects reducing plasma levels of drugs to below optimum levels, need to be examined carefully in this group. In cases where it is unlikely that further treatment will produce complete inhibition of viral replication, continuation of current therapy is reasonable if the imminent risk of death is low, judged by the CD4 count. However, there is a rationale for switching – albeit to a less than optimally suppressive regimen – in those individuals in whom the imminent risk of death is high, in order to try to improve viral suppression with the aim of raising the CD4 count. Wherever possible one or two different drugs of the NRTI/PI/NNRTI classes should be included in such a switch regimen. Resistance testing should be used to inform the choice of regimen, particularly for those with prior experience of NRTIs and PIs or NNRTIs. Patients whose therapy fails with adherence problems may benefit from simpler but effective regimens to which their virus is still sensitive. When an individual drug as part of a regimen is causing toxicity the choice of agents to switch to is often self evident and is usually within class, particularly for patients on complex regimes who are finding complete adherence difficult. Regimen simplification, for example to a triple NRTI pill of ZDV, 3TC and ABC appears to be safe in those whose previous antiretroviral treatment has not failed and may improve adherence. Switching because of the development of abnormal lipids or the fat redistribution syndrome is more complex and of less certain benefit. It is dealt with in detail in the guidelines. Any improvements in fat redistribution that occur as a result of such switching are likely to be slow. There is variable evidence for a benefit of resistance testing for treatment-experienced patients in controlled trials. However, the level of resistance to a regimen predicts the response to a new regimen and the cost of such testing is low compared with that of drug treatment. We recommend testing at each virological rebound. Although it will be difficult to design trials to establish the value of resistance testing in drug-naive individuals, such testing prior to therapy may be of crucial value for a proportion of patients who carry mutations, especially in the context of demonstrable transmitted drug resistance. In view of this, resistance testing is recommended for all drug-naive patients prior to commencing treatment using the nearest available sample to seroconversion or presentation. Evidence for the clinical value of TDM remains limited although the committee recognises that it may have a role in reducing toxicity in individual patients and in adjusting doses in those with significant hepatic or renal impairment or at extremes of body weight, and may help to explain treatment failures. It may also be of value in individuals with advanced disease who are limited to using combinations that may have the potential for unknown pharmacokinetic or drug interactions. The main interests in this drug lie in its once-daily dosing, its lack of effect on lipid metabolism and, by implication, low probability of an association with lipodystrophy. Present studies indicate similar potency to an NNRTI or unboosted PI in treatment-naive patients. If these early results are confirmed, many clinicians may find this drug an important option when choosing a first PI. This drug has recently been licensed for patients with treatment failure. It is the first of a new class of compounds interfering with HIV cell entry. Its main disadvantage will be mode of administration (by injection twice a day). Wherever possible it should be reserved for use after second or subsequent treatment failure, in combination with one or preferably two other new drugs that are expected to be active on the basis of resistance data, to give a realistic prospect of suppressing viral replication completely. This may mean keeping some patients on their current regimen while waiting for other new drugs to become available for use alongside T20. In individuals at imminent risk of clinical disease and death, T20 might have a limited role as additional therapy in a failing regimen where there are no other active drugs available. The BHIVA executive is committed to producing updates on the antiretroviral treatment guidelines for adults on a regular basis. Over a 1- to 2-year period much can change in clinical practice as new scientific evidence is published. To reflect these changes the guidelines have been extensively revised and some sections removed and new sections added. The sections on coinfection with HIV and hepatitis B or C now form separate guidelines which can be found at the BHIVA website (http://www.bhiva.org). New guidelines on tuberculosis coinfection, paediatric HIV care, and adherence to treatment are also in preparation, while guidelines on the management of HIV and pregnancy will be updated over the coming year. The principles of consensus that overarched the previous guidelines [3, 4] were upheld in the production of this, the latest version. We believe it is important that the process by which the guidelines have been revised is made clear. As in the previous guidelines each section was designated to two members of BHIVA, and the Executive Committee have taken on the bulk of this work. The members were asked to revise the guidelines in line with new evidence that had been published either in peer reviewed journals or as peer reviewed abstracts at international meetings. The authors were asked to consult widely with colleagues and co-opt other BHIVA members if need be. Guidelines for the treatment and management of HIV infection have been produced in a number of countries in Europe, as well as in Australia and the USA [5-8]. The BHIVA guidelines have a number of important roles which are: To promote a uniformly high standard of care in all HIV treatment centres in the UK. To set out the strengths, weaknesses and relevance of recent research findings. To assist in discussions between purchasers and providers regarding funding for HIV/AIDS diagnostic testing, care and treatments. To act as a basis for clinical audit within clinical governance. To act as a source of reference on antiretroviral treatments for those physicians caring for patients infected with HIV. To act as a source of reference for HIV-positive people. The cost of drugs has not been considered in preparing these guidelines, but basic NHS costs are shown in [Table 1]. Clinicians will be aware that actual costs may vary especially where bulk purchases have been arranged. These guidelines should not be seen as a substitute for research, nor as a manual for managing an individual. While the guidelines attempt to represent the current state of knowledge it is inevitable that, as HIV/AIDS is a rapidly evolving medical field, new data will change therapeutic choices and preferences. Consequently, the guidelines will require modifications as important new data emerge and the website version (found at http://www.bhiva.org) will be amended at regular intervals to reflect these data. Revisions are planned annually. Recommendations made within these guidelines have been graded according to the level of evidence on which they are based [Table 2]. Recommendations range from ‘essential’ to ‘optional’ and the quality of evidence from ‘at least one randomised trial with clinical endpoints’ through to ‘expert opinion’. They are to be found in parentheses in the document, for example (AII). The committee used an evidence-based medicine approach to produce these guidelines. In reality, if only the most reliable form of clinical evidence was taken into account (i.e. results of one or more randomised controlled trials with clinical endpoints), it would be impossible to formulate these guidelines. Many important aspects of clinical practice remain to be formally evaluated and very few trials with clinical endpoints are ongoing or planned. Results from clinical trials with viral load and CD4 count changes as endpoints were included as, in many instances, they are the only source of evidence. However, most such trials have been performed in order to obtain drug approval and are not ideally suited to addressing questions of clinical usage. The most significant drawbacks of such trials are their short duration and the lack of follow-up data on patients who switch therapy. In most cases the only available data on long-term outcomes are from routine clinical cohorts. While such cohorts are representative of routine clinical populations, the lack of randomisation to different regimens means that comparisons between the outcomes of different regimens are highly susceptible to bias [9, 10] opinion forms an important part of all consensus guidelines; however, this is the least valuable and robust form of evidence. The authors of these guidelines recognise that there is often a considerable time lag between initial presentation of important data, whether orally or in abstract/poster format, and full publication. Consequently, there is danger in relying on data that have not been to peer and published in We have research that only in more than 3 (i.e. before guidelines are and and valuable research that might improve of care will be It would be to that certain clinical controlled trials would be if they not to the guidelines, especially when these guidelines are based upon opinion than more reliable evidence. The executive has that clinical guidelines be used to or treatment options CD4 cell counts and plasma viral load are used as of the of antiretroviral therapy in and trials. in viral load to a in CD4 with seen in those with and more viral suppression in these in response to therapy are associated with clinical response CD4 counts in on antiretroviral therapy have been associated with a risk of no than that expected in individuals with similar CD4 counts The CD4 count is a of the immediate risk of than the viral load in those on antiretroviral therapy responses to therapy, a in plasma and in CD4 cell have to of antiretroviral agents it is to for clinical trials to be before drugs are are given full approval on the basis of trials and, in some approval based on data to clinicians would that a drug based on surrogate is reasonable and However, it should be that CD4 count and viral load responses not reflect the expected clinical and are not of the clinical response This is because the drugs have other effects with clinical those in viral load and CD4 count The short of trials to obtain drug approval means that, at the time of little may be the long-term antiretroviral drug trials are performed by as part of their to obtain approval and the designs are often not ideally suited to information on using the drugs in clinical the short duration of their is the lack of data on outcomes in who change from the randomised regimen, with a of those new regimens The results are only as as a high proportion of remain on the regimens. Clinical questions which drugs to start with or switch to require longer trials that changes to the treatment. a clinical it no to to patients after a regimen has been use of a given drug can outcomes after it has been For it may for virus to drugs not or which with those by other However, of such trials is not and account be taken of which drugs were used subsequent to the regimen in each or ongoing trials such an approach include for Clinical on and the trial is in the by the design may the rate of trial trial design may result in levels of from is to be the least effective regimen, while a controlled study may reduce adherence in all because of the pill It is also important to recognise that controlled clinical trials provide an treatment but results from the use of regimens in clinical practice are usually not as The committee generally into trials for patients whose clinical are with a number of in these guidelines. a of trials currently in the at and treatment should to are in to patients to trials at centres within or outside their clinical Several have been used to viral load and CD4 count including the change from at a given time and the change from or the the For virological however, the most approach to whether the viral load is below a certain usually which is the lower of for most viral load in routine The proportion of with viral load at a given time is for the choice of this is that some studies have that if this minimum level is not subsequent viral load rebound is more likely However, when treatment regimens, in viral load between treatment levels are are highly of subsequent in clinical to those with viral load would not other information in the viral load A of response to an initial regimen is time to virological failure. Virological failure is by failure to viral suppression or viral load rebound after but there is no consensus on in a trial in between the treatment at to outcomes for some patients can this and bias in the between the treatment In order to risk of such analysis by to treat outcomes for all randomised patients. For this it is to data on all patients if they have from the regimen. As this is the to treat is by for those patients who have out of the trial. When the is the proportion of with viral load at a given time the approach is to to all patients who have therapy or have the viral load value for This is as the approach an approach failure of a regimen due to potency viral drug resistance with to a regimen due to pill the of these two outcomes are likely to be This approach is often because it a minimum proportion for given treatment over all possible However, the primary purpose of an is to compare treatment and, in this this approach is not in On treatment outcomes only in those still the treatment. In the context of the proportion of with viral load at a given time this little because therapy is in patients who experience viral load rebound during a trial. this management the proportion of on the regimen who have a viral load will reflect the with which clinicians to switch therapy in response to the first viral load copies/mL. It is difficult to see it provides a means to compare the efficacy of different regimens. the context of time to virological failure, the on treatment analysis may be more In situations where there is a high proportion of patients who not have a viral load value at a given time where this due to is difficult and no approach is of patients are usually to between regimens (i.e. to no or a difference in response between Many surrogate marker studies are to that studies have shown no significant difference between the treatment is very different from that the that CD4 cell counts or viral loads in response to therapy may in efficacy between The for the difference in between treatment should be examined carefully in such of adherence to regimens is an in trials because an to treat analysis would to the difference in between treatment and treatment changes during the trial reflect would in clinical the results from to treat analysis would be It is to compare results of individual drug combinations in different trials. comparisons however, difficult to because of in with to viral load and CD4 cell of analysis to treat on of adherence and of viral load Many of antiretroviral therapy have been only after drug It is that as soon as possible so that these are A by the the Committee for of and the in the for reporting to the treatment of HIV currently available antiretroviral of HIV infection is not likely to be possible The main aim of treatment is to and