BHIVA Writing Committee on behalf of the BHIVA Executive Committee*

1.0 Introduction77

1.1 Purpose of the guidelines

1.2 Achieving consensus

1.3 The process of consensus

1.4 Basing recommendations on evidence

1.5 Use of evidence published as abstracts

1.6 Implications for research

1.7 Use of surrogate marker data

1.8 Surrogate markers in early vs. late disease

1.9 Issues concerning design and analysis of clinical


 1.9.1 Study design

 1.9.2 ‘Intention to treat’ vs. ‘on treatment’


 1.9.3 Equivalence

 1.9.4 Standards for presenting data from clinical

     trials using surrogate endpoints

 1.9.5 Cross-study comparisons –- presentation of


 1.9.6 Long-term follow-up

1.10 Adverse event reporting

2.0 When to start treatment80

2.1 Primary HIV infection

 2.1.1 Eradication

 2.1.2 Recommendations for starting treatment in

     primary HIV infection

2.2 Asymptomatic HIV infection

 2.2.1 Treating ‘early’ or ‘late’

 2.2.2 Surrogate markers as a guide to starting


    Viral load and viral replication

    CD4 cell count

 2.2.3 Role of adherence and potential toxicity

 2.2.4 Recommendations for starting treatment in

     asymptomatic individuals

2.3 Symptomatic HIV infection

 2.3.1 Recommendations for starting treatment in

     patients with symptomatic HIV infection or


3.0 What to start with84

3.1 Choices of initial therapy

3.2 Which HAART regimen is best?

 3.2.1 Two nucleoside analogues plus a protease


 3.2.2 Combinations of protease inhibitors

 3.2.3 Two nucleoside analogues plus a non-nucleo-

     side reverse transcriptase inhibitor




    Advantages and disadvantages of NNRTIs

 3.2.4 Three nucleoside analogues

    Advantages and disadvantages of NAs

3.3 Recommendations for initial therapy

 3.3.1 Two nucleoside analogues plus one or two

     protease inhibitors

 3.3.2 Two nucleoside analogues plus a non-nucleo-

     side reverse transcriptase inhibitor

 3.3.3 Choice of nucleoside analogue backbone for

     initial therapy

    Nucleoside analogue toxicity

    Hydroxyurea and mycophenolate

4.0 Issues concerning protease inhibitor use89

4.1 Adherence

4.2 Pharmacokinetics

4.3 Toxicity

 4.3.1 Lipid abnormalities

 4.3.2 Lipodystrophy

 4.3.3 Recommendations for monitoring lipids

4.4 Haemophilia

4.5 Viral hepatitis

5.0 Changing therapy on first virological failure90

5.1 Initial failure

5.2 Changing therapy

 5.2.1 Recommendations for changing therapy

5.3 Failure of two nucleoside analogues plus a protease


5.4 Failure of two nucleoside analogues plus a non-

  nucleoside reverse transcriptase inhibitor

5.5 Failure of triple nucleoside analogue therapy

6.0 Therapy after more than one previous failure (‘salvage’


6.1 Salvage therapy in PI-experienced patients

6.2 Salvage therapy in NNRTI-experienced patients

6.3 Salvage therapy in patients with multiple class


6.4 Structured treatment interruption

6.5 Stopping therapy

6.6 Viral fitness

6.7 Recommendations for subsequent virological failure

  (third or more regimen)

7.0 Resistance testing94

7.1 Genotypic testing

 7.1.1 Detection of specific point mutations 

 7.1.2 Nucleic acid sequencing of the reverse

     transcriptase and/or protease genes of plasma

     viral RNA

7.2 Phenotypic testing

7.3 Advantages and disadvantages of resistance testing

7.4 Practical issues of resistance testing and


7.5 Primary resistance

8.0 Issues not addressed in these guidelines96

8.1 Pregnancy

8.2 Post-exposure prophylaxis for healthcare workers

9.0 BHIVA/IAS-USA comparisons96

9.1 When to start treatment

9.2 Choices of initial regimen

9.3 What to change to after first virological failure


11.0 Conflict of interest101

1.0 Introduction

  1. Top of page
  2. 1.0 Introduction
  3. 2.0 When to start treatment
  4. 3.0 What to start with
  5. 4.0 Issues concerning protease inhibitor use
  6. 5.0 Changing therapy on first virological failure
  7. 6.0 Therapy after more than one previous failure (‘salvage’ therapy)
  8. 7.0 Resistance testing
  9. 8.0 Issues not addressed in these guidelines
  10. 9.0 BHIVA/IAS–USA comparisons
  11. References
  12. 11.0 Conflict of interest

1.1 Purpose of the guidelines

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[1–4]. The British HIV Association (BHIVA) guidelines have been extensively revised[5] and now include a detailed discussion of their recommendations.

The BHIVA guidelines have a number of important roles which are:

(i)  To promote a uniformly high standard of care in all

  HIV treatment centres in the UK.

(ii)  To set out the strengths, weaknesses and relevance

  of recent research findings.

(iii) To assist in discussions between purchasers and

   providers regarding funding for HIV/AIDS

   diagnostic testing, care and treatments.

(iv) To act as a basis for clinical audit within clinical


(v)  To act as a source of reference on AIDS treatments

    for those physicians caring for patients infected

    with HIV.

These guidelines should not be seen as a substitute for research, nor as a manual for managing an individual, and should be interpreted and applied sensibly and appropriately. 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 will be amended at regular intervals to reflect these data. Complete updates are planned at least annually.

Recommendations made within these guidelines have been graded according to the level of evidence on which they are based (see Table 1). Recommendations range from ‘essential’ to ‘not recommended’ and the quality of evidence from ‘at least one randomized trial with clinical endpoints’ through to ‘expert opinion’. They are to be found in parentheses in the document, for example (AII).

Table 1.  Grading of recommendations and levels of evidence
Strength of recommendationQuality of evidence for recommendations
(A) (B) (C) (D) Essential should always be offered Desired should usually be offered Optional Should in general not be offered (I) (II) At least one randomized trial with clinical endpoints Clinical trials with laboratory endpoints
(E)Should never be offered(III)Expert opinion

1.2 Achieving consensus

The BHIVA guidelines have attempted to provide consensus across a range of healthcare workers including physicians, virologists, people living with HIV disease and special interest voluntary organizations and have been developed following an agreed process outlined below.

1.3 The process of consensus

In order to produce the guidelines the following process was agreed:

(a) A writing committee was selected from the BHIVA executive which comprises specialists in HIV care from a variety of training backgrounds including general medicine and laboratory-based medicine. Physicians practising genitourinary medicine were included, from both high and low prevalence areas for HIV.

(b) A formal consultation process was developed with individuals living with HIV and advocacy groups. Representatives of these groups became members of the writing committee.

(c) Draft guidelines were discussed at the annual BHIVA conference (28 March 1999) and then put on to the BHIVA Internet website ( http:// with a request for comments about facts, interpretations and opinions, so as to inform the final version of the guidelines. They were discussed again at the autumn BHIVA meeting on 9 October 1999.

(d) The final version of the guidelines will be subject to peer review. Reviewers will be allowed to publish their views anonymously, or as an acknowledged comment, both in the draft and the final document.

(e) An active web discussion will continue for 6 months after the final guidelines are posted.

1.4 Basing recommendations on evidence

In producing this document the committee used an evidence-based medicine approach. In reality, it is recognized that if only the best available clinical evidence is taken into account (i.e. the result of at least one or more randomized controlled trial with clinical endpoints) it would be impossible to formulate these guidelines, as many important aspects of clinical practice remain to be formally evaluated. Data based on both the biological plausibility of surrogate markers and expert opinion form an important part of all consensus guidelines; however, such expert opinion is the least valuable and robust form of evidence.

1.5 Use of evidence published as abstracts

In writing these updated guidelines, it has been recognized that there is often a considerable time lag between initial presentation of important data (whether oral or in abstract/poster format) and full publication. Consequently, there is a danger in relying on data that have not been subjected to formal peer review and published in full. We have therefore avoided citing any research findings which appeared only in abstract format more than 3 years ago (i.e. before mid-1996).

1.6 Implications for research

Unless guidelines are interpreted and applied sensibly and with caution valuable research initiatives that might improve standards of care would be stifled. It would be wrong to suggest that certain clinical controlled trials would be unethical if they did not conform to the guidelines, especially when these guidelines were based mainly upon opinions rather than facts. The National Health Service executive has stated that clinical guidelines cannot be used to mandate, authorize or outlaw treatment options[6].

1.7 Use of surrogate marker data

CD4 cell counts and plasma HIV-1 RNA are used as markers of the biological activity of antiretroviral therapy in Phase I and II trials. Favourable responses to therapy, i.e. a decline in plasma HIV-1 RNA and increases in CD4 cell counts, have led to accelerated licensing of antiretroviral agents on the grounds that it is impracticable to wait years for large clinical endpoint trials to be completed before drugs are approved[7].

The antiretroviral activity of a new drug in individuals with advanced HIV disease who are treatment naive can be established in clinical trials by using surrogate marker data at 24 weeks, and durability of response can be reasonably assessed at 48 weeks. Most clinicians would agree that a drug-licensing policy based on surrogate markers is reasonable and humane.

It is important to realize, however, that the assessment and evaluation of drugs based on short-term biological activity and safety profiles means that clinical efficacy cannot be claimed. Clinical efficacy is based on clinical endpoint data and can only be extrapolated from surrogate marker data if the surrogate fulfils two criteria. First, the surrogate must be a correlate of the true clinical outcome and secondly, it must capture the net effect of treatment on clinical outcome, i.e. the effect aggregated over all mechanisms of action of the treatment on the clinical endpoint[8]. As most drugs operate through several mechanisms, surrogate markers cannot capture all their effects[9]. These criteria have largely but not completely been satisfied for both CD4 cell count and plasma HIV-1 RNA. In the Medical Research Council's (MRC) Delta trial, although suppression of plasma HIV-1 RNA was highly predictive of clinical progression and overall survival, it consistently overestimated the clinical benefit from combination therapy[10]. An explanation for this is that important biological processes were not captured by use of these markers.

1.8 Surrogate markers in early vs. late disease

The relative importance and precise meaning of changes in surrogate markers in response to therapy is likely to have different implications in early vs. late disease.

In late disease, the CD4 cell count is of greater prognostic significance than viral load, whereas in early disease the reverse is true[11]. Most of the large clinical controlled trials in which it has been possible to assess the value of surrogate markers have been performed in patients whose CD4 cell counts were around 300 cells/μL. In this situation, relatively transient falls in viral load can produce a sustained increase or stabilization in CD4 cell count and can still influence the rate of development of clinical events 2 or 3 years after the start of the study.

1.9 Issues concerning design and analysis of clinical trials

1.9.1 Study design

Study design may have an important influence on the rate of discontinuation in drug trials. An open trial design may result in higher levels of discontinuation from what is perceived to be the least effective regimen, while a double-blind, placebo-controlled study may reduce adherence in all groups because of the large pill burden. It is important to recognize, however, that controlled clinical trials provide an optimal treatment setting and results in the ‘real world’ are usually not as good.

1.9.2 ‘Intention to treat’ vs. ‘on treatment’ analyses

The apparent potency of a particular drug regimen (in terms of viral load change in response to treatment) is likely to be influenced by the way in which studies are analysed. The only analysis which is unbiased is the ‘intention to treat’. In this analysis, all randomized patients are included if they have been followed-up and have results. Such an analysis might favour a treatment arm which has a higher rate of adherence than the other treatment arm(s), unless they are placebo-controlled. It approximates to real clinical practice and gives an overall view of how effective a therapy might be.

In some analyses, when patients do not attend for follow-up they are either treated as failures or the ‘last observed data point is carried forward’.

‘On treatment’ analysis, whereby only data for those subjects continuing to take the drugs are analysed, allows an understanding of the best that might be achieved by a therapy. It suffers from a number of potentially serious biases by excluding all patients who have dropped out for any reason and represents a situation far removed from real clinical practice. In order to inform clinicians fully, it is preferred that both types of analyses are presented.

1.9.3 Equivalence

Large numbers of patients are usually required to show equivalence between regimens, and many surrogate marker studies are under-powered to demonstrate this. Stating that studies have shown no significance difference between the treatment arms is very different from saying that the arms show equivalence. Graphical representations of CD4 cell count rises or viral load declines in response to therapy that appear to be overlapping may hide differences in efficacy between drugs. The differences in confidence limits should be examined carefully in such studies.

1.9.4 Standards for presenting data from clinical trials using surrogate endpoints

Another important method of presenting data is to measure the proportion of individuals in whom viral replication is suppressed such that the viral load falls below the detectable limit of a particular assay. Until recently, this detection limit was 400 copies/mL but, more recently, assays measuring to a threshold of 50 copies/mL have been used. Therefore, when reporting the proportion of subjects in whom viral load has fallen to below the detectable limit, it is important to state the sensitivity of the particular assay used.

It has been proposed that the proportion of patients who achieve a plasma HIV-1 RNA load of less than 50 copies/mL at 24 weeks using an intention-to-treat analysis (whereby people who do not complete the trial are termed study endpoint non-completers or failures) should be the standard for presenting data from clinical trials[12]. Several studies have suggested that patients who achieve a more modest initial viral load decline in response to drug treatment (below 400 copies/mL but above 50 copies/mL) are more likely to fail virologically, i.e. for the virus to reappear in the circulation, than those whose viral load falls below 50 copies/mL[13].

Time to treatment failure (TTF) is becoming a standard analysis for many trials. TTF is an intention-to-treat analysis that provides information on the comparative effectiveness of treatments over time and has been used extensively in oncology. The definition of treatment failure may vary between studies and, in the context of HIV therapy, usually encompasses discontinuation for any reason, clinical progression (e.g. AIDS events) or virological rebound.

1.9.5 Cross study comparisons –- presentation of data

It is tempting to compare results of individual drug combinations assessed in different trials. Such comparisons are fraught with difficulties because of differences in entry criteria (particularly with respect to viral load and CD4 cell counts), methods of analysis (e.g. intention to treat vs. on treatment), degrees of adherence and sensitivities of viral load assays.

1.9.6 Long-term follow-up

For long-term survival, the suppression of viral load over many years is likely to be important both for morbidity and mortality, i.e. predicting clinical benefits. Most studies are designed to examine short-term surrogate marker changes obtained in response to therapy at 16 or 24 weeks and some to assess how well results are sustained to 48 weeks. Few studies include long-term monitoring of patients, as most are terminated at 48 weeks or, more exceptionally, at 72 weeks.

It is becoming clear that sequential therapy may be needed for long-term control of the disease. A particular combination may produce the best initial decline in viral load which is also sustained for the longest period. However, that same combination may preclude the use of a subsequent therapy, either because of the emergence of a virus strain which is cross-resistant to other drugs of the same class, or because of overlapping toxicity. Studies evaluating long-term strategies of non-overlapping sequential therapy are just beginning (e.g. in Europe, the INITIO trial and in the USA, AIDS Clinical Trials Group [ACTG] protocol 384). Long-term follow-up (preferably lifelong) should be incorporated into such studies.

1.10 Adverse event reporting

Many previously unsuspected side-effects of antiretroviral therapy have been reported only after drug licensing. It is vital that any possible adverse events be reported rapidly by prescribers in order that recognition of adverse events occurs swiftly. A blue-card scheme, organized by the Medicines Control Agency, the Committee for Safety of Medicines (CSM) and the MRC operates in the UK for reporting adverse events relating to the treatment of HIV.

2.0 When to start treatment

  1. Top of page
  2. 1.0 Introduction
  3. 2.0 When to start treatment
  4. 3.0 What to start with
  5. 4.0 Issues concerning protease inhibitor use
  6. 5.0 Changing therapy on first virological failure
  7. 6.0 Therapy after more than one previous failure (‘salvage’ therapy)
  8. 7.0 Resistance testing
  9. 8.0 Issues not addressed in these guidelines
  10. 9.0 BHIVA/IAS–USA comparisons
  11. References
  12. 11.0 Conflict of interest

The three groups of treatment-naive patients for whom treatment guidelines are required are: (i) patients with primary HIV infection; (ii) patients with asymptomatic HIV infection and (iii) patients with symptomatic HIV disease/AIDS. The recommendations are summarized in Table 2.

Table 2.  When to start treatment: summary of recommendations
PresentationSurrogate markersRecommendation
  1. 1It is essential to discuss the merits of treatment if patients present early (i.e. within 6 months). Store blood sample before starting therapy.

Primary HIV infection  If treatment considered, start as soon as possible,1
preferably within 6 months
of contracting HIV
AsymptomaticCD4 count > 500 cells/μL,Defer treatment
HIV infectionany viral load 
CD4 count 350–500 cells/μL,Defer treatment
viral load < 30 000 copies/mL 
CD4 count 350–500 cells/μL,Consider treatment or
viral load > 30 000 copies/mLdefer and monitor at least
CD4 count 0–350 cells/μL,Treat
any viral load 
Symptomatic HIV infection Treat

2.1 Primary HIV infection

The period prior to complete seroconversion, which is often symptomatic, is called primary HIV infection (PHI) or ‘acute seroconversion syndrome’. The advantages and disadvantages of early treatment may be different in an HIV-infected individual at the time of PHI compared with established HIV infection. It is likely that, at the time of PHI, there is a narrowing of the genetic diversity of the infecting virus compared with the virus in the index case[14–15]. It is thought that the initial round of viral replication following sexual infection occurs within the macrophage/monocyte cell lineage in the genital tract or rectum, as most virus detectable shortly after PHI is of the non-syncytium-inducing phenotype, which uses the CCR5 chemokine co-receptor for cell entry[14].

If the initial round of replication in macrophages does reduce genetic diversity, PHI may represent a unique window of opportunity for treatment when the virus is at its most vulnerable because (i) its ability to infect different cell types may be limited; (ii) the immune response is at its height; and (iii) the number of replicating viral particles is still small. Therefore, drug treatment may be particularly effective during this time.

A variety of triple-drug therapy regimens appear able to suppress viral replication in the plasma, lymph nodes and gut in the majority of patients treated within a few months of PHI[16–17]. Recent studies demonstrated that shortly after PHI there is a specific and strong CD4 helper HIV response[18–19]. This is in contrast to chronic infection where, with the exception of long-term non-progressors[20], the HIV-specific CD4 helper response is generally lost[21]. These CD4 helper responses may be important in maintaining an adequate CD8 response. Such immune responses appear to be maintained in people treated with potent antiretroviral therapy shortly after PHI and perhaps represent the best biological evidence that treatment at this time is valuable. The measurement of CD4 helper responses is, however, difficult and there is no evidence to date that these responses persist indefinitely with continuing treatment or after treatment is withdrawn.

Control of viral replication with no return of viraemia has apparently occurred in a few patients treated very soon after PHI. Some were receiving hydroxyurea as part of their regimen, which is known to inhibit CD4 activation, whereas others were poorly adherent with treatment[22]. Although it has been postulated that such intermittent exposure to viraemia might lead to enhanced anti-HIV immunity, these preliminary findings have yet to be confirmed. It is unclear whether hydroxyurea, treatment interruption or both/neither may have contributed or whether the patients would have been long-term non-progressors even without treatment.

2.1.1 Eradication

On the basis of small preliminary studies, it was hoped that treatment at seroconversion might be curative (or at least change the viral load set-point to a lower level). Unfortunately, the likelihood of eradication of HIV now appears small and there is evidence for the existence of long-lived cells, latently infected with HIV, which maintain the viral pool[23–25]. If this is the true scenario, and HIV release from this cell compartment cannot be controlled by the immune system, it has been calculated that treatment will need to be continued for many years if eradication is ever to be achieved[26].

2.1.2 Recommendations for starting treatment in primary HIV infection (CIII)

It is very important at the time of PHI that patients and physicians make the most appropriate decision based on the limited data available.

  • cirf
    & ; As there are few data to guide us as to the best therapeutic option, entering patients into a clinical trial is the committee's preferred choice.
  • cirf
    & ; If patients are treated outside a clinical trial setting, a regimen appropriate for treating chronic HIV infection should be used (see later).
  • cirf
    & ; The biological plausibility that early treatment may be beneficial for the immune system should be balanced against considerations of adherence to long-term therapy, potential toxicity and development of resistance.
  • cirf
    & ; If treatment is not started within a few months of recognized PHI, a potential advantage may be permanently lost. The theoretical advantage of starting treatment should be weighed against the uncertainty of efficacy, the problems of adherence and the emergence of drug resistance.
  • cirf
    & ; If a decision to start treatment is made, this can be reviewed (i.e. therapy can be stopped or continued) in the light of evolving data.
  • cirf
    & ; It is appreciated that many patients will choose not to be treated at this time and, because of the inherent uncertainties, this choice should be respected and supported.

2.2 Asymptomatic HIV infection

There are few data to support an absolute time point for starting therapy in asymptomatic patients based on surrogate markers. The arguments for starting early or late are outlined below.

2.2.1 Treating ‘early’ or ‘late’

For the purpose of these guidelines, we have defined treating early as when a patient's CD4 count has fallen to around 500 cells/μL and their plasma HIV-1 RNA level has risen above 55 000 copies/mL (measured by reverse transcriptase polymerase chain reaction [RT-PCR] assay) or 30 000 copies/mL (by branched DNA [bDNA] assay). We would define treating ‘late’ as when the CD4 count has fallen to 350 cells/μL or below, irrespective of the plasma HIV-1 RNA copy number.

Once HIV infection has become chronic/established (1 year to 18 months following PHI), treatment with potent antiretroviral therapy appears to restore HIV-specific immune responses poorly or not at all[27]. However, such therapy can at least restore and enhance antigen-specific immunity against pathogens such as mycobacteria and cytomegalovirus. Substantial rises in CD4 cell counts occur and memory CD4 cell numbers increase soon after the initiation of highly active antiretroviral therapy (HAART), due to redistribution of the cells from the lymphoid system into the circulation. It has also been shown that naive CD4 T cells, which are important for responding to new antigens, can be restored gradually with prolonged viral suppression. Theoretical arguments for starting therapy early are that the amount of viral-induced immune destruction is limited, immune responses to various pathogens are preserved, and the ability of the immune system to recover is maximized. There are few convincing data that the immune consequences or prognosis are worse for those who start therapy at a CD4 count of around 350 cells/μL, compared with those who start at 500 cells/μL. If therapy can be safely delayed until the CD4 count is lower, then the duration of drug exposure and potential long-term complications of therapy might be reduced and quality of life improved.

For any patient, the correct balance needs to found between the benefit of starting treatment early (e.g. when their CD4 count is 500 cells/μL) and the potential long-term adverse effects of therapy such as toxicity, metabolic abnormalities and resistance, as well as the psychological impact of therapy and its effect on quality of life. It is important that patients are informed of the risks and benefits of such treatment.

Unfortunately, the question of optimal timing of initial treatment has only been addressed in controlled clinical trials of antiretroviral monotherapy. Although ACTG 016[28] was stopped prematurely because zidovudine (ZDV) monotherapy appeared to improve the clinical outcome of those patients with a CD4 count below 500 cells/μL, the benefit was not sustained. The Concorde trial suggested some short-term benefit of ZDV, but there was no long-term benefit of monotherapy for the group starting therapy early compared with the deferred-therapy group[29]. Other clinical trials have shown major benefits of dual vs. monotherapy at different CD4 counts, but offer little help with deciding whether treatment with three or more drugs (the present standard of care) would be more beneficial if started early rather than late, and the argument therefore centres around biological plausibility ( Table 3).

Table 3.  Treatment for asymptomatic HIV-infected individuals: arguments for and against early treatment
Arguments for
(1) (2) (3) (4) HIV is an infectious disease and should be treated as such Control of viral replication leads to reduction of viral burden and decreases the risk of evolution of drug-resistant virus Delays progression to AIDS and death Possible greater potential for immune reconstitution
(5)Possible lower incidence of short-term drug toxicity
(6)Possible decreased risk of viral transmission
Arguments against
(1) (2) (3) (4) (5) Adverse drug effects –- problems with adherence, effects on lifestyle and psychology Earlier development of drug resistance Possible transmission of drug-resistant virus Limits future choices of antiretroviral agents if resistance occurs Long-term toxicity –- some of this is currently unknown/ill defined
(6)Unknown duration of efficacy of current antiretroviral therapy

In deciding when to start therapy, surrogate markers, potential toxicity and adherence issues all need to be addressed.

2.2.2 Surrogate markers as a guide to starting therapy

Surrogate marker measures have prognostic value and are important when deciding whether or not to start treatment, the goal of therapy being to suppress viral load and cause a rise in CD4 cell count.

Viral load and viral replication. The plasma viral load represents the net result of steady-state viral replication and viral clearance. The former is estimated to be around 1010 particles per day and the number of virions produced increases as the size of the viral pool increases[30].

There are no definitive data to support the commencement of treatment at any particular viral load. Patients with a high baseline viral load tend to progress more rapidly to symptomatic disease and AIDS than those with lower baseline viral load values.

Sufficiently potent drugs would be expected to inhibit viral replication, however high the viral load. For most of the available drugs, the ability to suppress viral replication to below the detection limit of current sensitive assays appears to depend, in part, on the initial viral load prior to therapy[31–35]. Patients with plasma HIV-1 RNA levels above 100 000 copies/mL appear to have a lower rate of successful suppression of viral replication with some triple-drug regimens. This may not hold true for some combinations of the newer drugs[36]. The effectiveness of more aggressive therapy, starting with more than three drugs, is unknown. Consequently, it has been argued that treatment given early, at a point when the viral load is still at a relatively low level and the CD4 count is still reasonably high, would best benefit the patient. Data from the Multicentre AIDS Cohort Study suggested that patients with HIV RNA > 30 000 copies/mL have an increased probability of progressing to AIDS-defining complications of HIV disease within 3 years, even while the CD4 count is relatively well preserved (> 500 copies/μL). For these patients, the pros and cons of starting antiretroviral therapy should be discussed[11].

Recent data suggest that, in patients with equivalent CD4 counts and time to progression to AIDS, the viral load is 24–40% lower in women than in men[37–38]. This difference has not been consistently observed[39–40] and, for the present, recommendations for initiation of treatment should be the same for men and women.

CD4 cell count. Earlier treatment may limit the progressive destruction of the immune system and thereby reduce the subsequent risk of infection by opportunistic pathogens (e.g. Mycobacterium tuberculosis) and development of tumours (e.g. Kaposi's sarcoma and lymphoma). However, there is considerable redundancy within the immune system, so many opportunistic infections (the so-called ‘hallmark’ of immunodeficiency) are not usually observed until the CD4 count falls below 200–250 cells/μL[11–41]. Moreover, recent data have confirmed that immune reconstitution is possible when treatment is initiated in late disease (i.e. CD4 count below 250 cells/μL[42]). Thus, it could be argued that treatment should start when there is moderate immunodeficiency, as measured by a CD4 count of around 350 cells/μL, but before patients run the risk of succumbing to opportunistic infections and tumours, which may lead to disability or death.

2.2.3 Role of adherence and potential toxicity

The pivotal role of adherence in the success of antiretroviral therapy has been amply demonstrated[43–46]. Numerous studies which have attempted to identify characteristics that are predictive of poor adherence have largely failed to produce consistent findings, aside from an association with psychological and psychiatric problems[47–48]. This argues against a policy of denying treatment to individuals on the basis of socio-demographic factors. The potential to experience problems with adherence should be viewed objectively for all patients, and this is supported by recent research suggesting that doctors are relatively poor at judging their patients' adherence levels[44–49].

Data from a large UK cohort[50], supported by several studies[51–55], suggest that the most common reasons for missing doses are simple forgetfulness and drug toxicity. Other than the negative impact of dietary restrictions, regimen complexity and high pill burden have not been found to predict poor adherence consistently, although they undoubtedly have great potential to impair quality of life[50]. Understanding the purpose of therapy and the relationship between poor adherence and viral resistance[50] has also been associated with improved adherence.

Adherence is potentially more difficult in asymptomatic patients who might lack the reinforcement of an improved sense of wellbeing to stimulate regular pill-taking. It will be significant for these patients that their first HIV-related symptoms may be side-effects from treatments which they have been told will help them.

All patients starting treatment need a commitment to therapy. Nevertheless, effective treatment needs to be easy to adhere to and relatively free of adverse effects. None of the current regimens fulfil these criteria.

When initiating therapy, at least the following should be considered:

The motivation of an individual to begin and continue therapy.

The impact of therapy on an individual's lifestyle and psychology, including the need to establish and maintain a pill-taking routine, perhaps using memory aids.

The potential risks and benefits of therapy in the short-and long-term.

The provision of written information to provide support outside the clinic setting.

Many clinicians feel that a multidisciplinary ‘starting therapy’ clinic is important for patients contemplating treatment for the first time. In such a clinic, time can be devoted to issues of adherence, adverse effects, peer support, dietary advice, psychological support and so on, presenting patients with additional opportunities to raise concerns about their use of treatments.

2.2.4 Recommendations for starting treatment in asymptomatic individuals
  • cirf
    & ; Given the lack of compelling evidence to decide when treatment should begin, the pragmatic view must be to ensure that the benefit of presently available treatment outweighs the risk of deferring therapy.
  • cirf
    & ; Asymptomatic patients should be offered treatment before clinical progression becomes likely as a result of (potentially irreversible) immune damage.
  • cirf
    & ; Currently, our recommendation is that a CD4 count of around 350 cells/μL is a reasonable level of immunodeficiency at which therapy may be started (AIII).
  • cirf
    & ; Although this recommendation is based on the absolute CD4 count alone, we would suggest close monitoring of patients who are likely to progress rapidly. Three such groups of patients can be identified: (i) those with progressive and rapidly declining CD4 count; (ii) those with rapidly rising viral load; and (iii) those with a high viral load (> 30 000 copies/mL [bDNA] or 55 000 copies/mL [RT-PCR]).
  • cirf
    & ; If patients to whom these recommendations apply choose not to go on treatment, it is suggested that their CD4 counts and viral load be monitored intensively (e.g. every 2–3 months) and the decision to start treatment reviewed at regular intervals.
  • cirf
    & ; Patients who have a stable CD4 count below 350 cells/μL and stable viral load, and not wishing to commence therapy, should also be closely monitored.

2.3 Symptomatic HIV infection

2.3.1 Recommendations for starting treatment in patients with symptomatic HIV infection or AIDS (AI)
  • cirf
    & ; All patients with clinical symptoms of underlying immune suppression or with a diagnosis of AIDS (according to the Centers for Disease Control 1993-modified case definition[56]) should be offered therapy regardless of CD4 count or viral load.

3.0 What to start with

  1. Top of page
  2. 1.0 Introduction
  3. 2.0 When to start treatment
  4. 3.0 What to start with
  5. 4.0 Issues concerning protease inhibitor use
  6. 5.0 Changing therapy on first virological failure
  7. 6.0 Therapy after more than one previous failure (‘salvage’ therapy)
  8. 7.0 Resistance testing
  9. 8.0 Issues not addressed in these guidelines
  10. 9.0 BHIVA/IAS–USA comparisons
  11. References
  12. 11.0 Conflict of interest

3.1 Choices of initial therapy

There is overwhelming evidence from cohort studies that the very dramatic fall in AIDS-related mortality and frequency of AIDS events seen in the developed world over the last 3 years coincides with the introduction of highly active antiretroviral therapy (HAART)[57]. The choice of any HAART regimen should be individualized in order to maximize adherence, and particular attention should be paid to potency, tolerability, potential toxicity and likely drug–drug interactions. Since the first-line option may limit future choices in the event of virological failure, it should be selected with extreme care. Table 4 summarizes the pros and cons of available options.

Table 4.  Choices of initial therapy: summary of recommendations
  • 1

    Hard-gel saquinavir should not be used as the sole PI. There are fewer data concerning use of saquinavir soft-gel in this context than for other PIs.

  • 2

    Primary reason for combining PIs is to improve pharmacokinetics. Suggested regimens: low-dose ritonavir (i.e. 100–400 mg) with saquinavir, indinavir or amprenavir.

  • 3

    Recommended NNRTIs are efavirenz or nevirapine. In one controlled trial, efavirenz was as effective in patients with viral loads > 100 000 copies/mL as in those with < 100 000 copies/mL. There are fewer data from controlled trials to address this issue for nevirapine.

  • 4

    May be suitable for patients with viral load ≤100 000 copies/mL. Two regimens have been studied: abacavir + lamivudine + zidovudine and stavudine + didanosine + lamivudine.

Primary HIV infection
 Clinical trialRecommended  
 No therapyConsider  
Chronic HIV infection
 2NAs + PI 1Recommended(1) RCT evidence with(1) Toxicity common
clinical endpoints(2) High pill burden
(2) Evidence of efficacy(3) Drug interactions
in late disease 
(3) Long-term follow-up 
 2NAs + 2PIs 2Recommended(1) Easier adherence(1) No clinical endpoint
(2) Better pharmacokineticsdata (2) Less comparative surrogate marker data (3) Possible increased toxicity and drug interactions
 2NAs + NNRTI 3Recommended(1) Equivalent or superior(1) No clinical endpoint
efficacy in surrogate markerdata
trials at 72 weeks(2) Lack of surrogate
(2) Easier adherencemarker data in late
(3) Less known toxicity thandisease
PI-containing regimen(3) Shorter follow-up (4) Little evidence of immune reconstitution (5) Single mutations may lead to cross-class resistance
 3NAs 4Under evaluation(1) Spares PIs and NNRTIs(1) No clinical endpoint
(2) Fewer drug interactionsdata (2) Short-term surrogate marker data only (3) Less effective at high viral load

3.2 Which HAART regimen is best?

There have been no definitive clinical controlled trials to demonstrate superiority of any one potent (HAART) regimen, used as initial treatment, over another. Patients should be informed about and encouraged to participate in any such trials. Several regimens will be discussed below and the advantages and disadvantages of each will be assessed. For a detailed discussion of the issues surrounding protease inhibitor (PI) use, see section 4.

3.2.1 Two nucleoside analogues plus a protease inhibitor

A dramatic decline in clinical progression and HIV-related deaths followed the introduction of the protease inhibitor (PI) class of antiretrovirals ( Table 5). These agents have shown clinical and surrogate marker efficacy in clinical practice. Sustained suppression of plasma HIV-1 RNA levels for more than 3 years has been seen in some patients within the Merck 035 study taking indinavir and two nucleoside analogues[58]. ACTG 320 was a clinical endpoint study which demonstrated long-term virological suppression and improved clinical outcome in patients taking ZDV/3TC/indinavir[32]. Improvement in clinical outcome has also been seen in studies with combinations of ritonavir added to background antiretroviral therapy (containing NAs) in late disease[59]. The hard-gel formulation of saquinavir has also shown clinical benefit when used in combination[60], but is not recommended as the sole PI in a HAART regimen because of its poor bioavailability. There has been no conclusive demonstration, either in clinical endpoint or surrogate marker studies, of superiority of one PI-containing combination over another. Surrogate marker data on all PIs in combination with two nucleoside analogues show similar reductions in viral load and proportions of patients achieving suppression to below 50 copies/mL at 48 weeks.

Table 5.  Currently available protease inhibitors
Protease inhibitorDoseFrequencyDaily pill burdenDietary restrictionsMajor side-effects
  • 1

    Dose currently unlicensed.

  • 2

    2 Recent data suggest that the bd regimen is less effective in suppressing viral load.

  • 3

    3 Progressive deterioration of renal function may be associated with long-term use.

  • 4

    Often used at lower doses (e.g. 100–400 mg bd) as part of a dual PI-containing HAART regimen.

  • 5

    5 The hard-gel formulation is still available for use in combination with ritonavir.

  • 6

    6 Not yet licensed in Europe. bd, twice a day; tds, three times a day.

  • *

    *Can take after food to prevent nausea.

Nelfinavir750 mgtds9 tabletswith foodMild to moderate
or 1250 mgbd 110 tabletswith fooddiarrhoea
Indinavir 800 mg tds 2 6 capsules empty stomach Renal stones, crystalluria & sludge, hyperbilirubinaemia 3
Ritonavir Saquinavir (soft-gel) 5 Amprenavir 6 600 mg4 1200 mg 1 or 1800 mg1 1200 mg bd tds bd bd 12 capsules 18 capsules 18 capsules 16 capsules none * with food with food none * Taste perversion, nausea, diarrhoea, perioral tingling Nausea, diarrhoea, abdominal pain, headache Nausea, diarrhoea, rash, headache, perioral tingling
3.2.2 Combinations of protease inhibitors

Combinations of PIs in clinical use are: (i) ritonavir/saquinavir; (ii) nelfinavir/saquinavir; (iii) ritonavir/indinavir; and (iv) ritonavir/amprenavir. The data on these combinations are limited, but they appear to have advantages over either agent used alone. First, they have improved 24-h pharmacokinetics, with higher trough levels and reduced fluctuations in plasma concentration over time. This allows lower doses to be taken at longer intervals, i.e. twice instead of three times a day, and without dietary restrictions. Secondly, the improved pharmacokinetics should assist adherence. Thirdly, the additive or synergistic antiviral effects of PIs in combination may improve potency relative to a regimen including a single PI. One study suggested that the ritonavir/saquinavir combination is more likely to reduce the viral load below 50 copies/mL than a triple combination containing either ritonavir or indinavir as the only PI[61]. This may be related to tolerance and adherence rather than intrinsic antiviral potency. Another study showed no statistical difference in viral load suppression between combinations containing either nelfinavir or soft-gel saquinavir as the sole PI and those containing a combination of both[62].

Possible disadvantages of combining PIs are an increased risk of lipodystrophy and severe lipid abnormalities. Unfavourable pharmacokinetic interactions between the PIs or with other drugs may also occur.

3.2.3 Two nucleoside analogues plus a non-nucleoside reverse transcriptase inhibitor

These drug combinations have not been as well evaluated as PIs combination in controlled trials using clinical endpoints[63]. A number of studies using these combinations have shown impressive surrogate marker results. Unfortunately, there are no direct data comparing the three currently available non-nucleoside reverse transcriptase inhibitors (NNRTIs), efavirenz, nevirapine and delavirdine, in terms of their effects on surrogate markers. The choice between them is therefore likely to be based on clinical trial data, toxicity and adherence (see Table 6).

Table 6.  Currently available non-nucleoside reverse transcriptase inhibitors (NNRTIs)
NNRTIDoseFrequency/dayDaily pill burdenDietary restrictionsMajor side-effects
  • 1

    At night.

  • 2

    The initial dose is 200 mg/day for 2 weeks, increasing to 400 mg/day.

  • 3

    3 Delavirdine is not yet licensed in Europe.

  • 4

    4 Larger dose tablets and a twice-daily regimen are expected to be introduced shortly.

  • 5

    Dose may be dissolved in cola. C/I, contraindicated.

Efavirenz600 mgonce 13 capsulesnoneDysphoria C/I: pregnancy
Nevirapine 2200 mgtwice2 tabletsnoneRash, hepatitis
Delavirdine 3400 mgthree 412 tablets 4none 5Rash (usually mild), headache

Efavirenz. Impressive surrogate marker results have been obtained using efavirenz with two nucleoside analogues (ZDV and lamivudine[3TC]). In a randomized open-label study, this combination was compared with ZDV, 3TC and indinavir and showed either no difference or superior surrogate marker endpoints at the 16–48-week time-point when analysed in a variety of ways (including intention to treat and on treatment analysis)[36]. The major drawback of this study was the high discontinuation rate in both arms. Thus, by 48 weeks, 35% of the indinavir group and 25% of the efavirenz group had discontinued, which may have biased the intention-to-treat analysis in favour of efavirenz. These discontinuation rates probably mimic what is seen in clinical practice. One of the chief advantages of an efavirenz-containing regimen is the once daily dosing of efavirenz, which may improve adherence. This is made possible by its long plasma half-life, which ensures that drug levels far exceed the IC90 (the concentration, determined in vitro, required to inhibit 90% of viral growth) throughout the day. This advantage may become a problem on stopping combinations with efavirenz; the long half-life of efavirenz, in comparison to some nucleoside analogues, means the drug may remain in the plasma for several days effectively exposing the patient to monotherapy and the associated risk of drug resistance.

The major side-effect is dysphoria; manifestations include vivid dreams, depression, drowsiness and, in some, insomnia. This is usually self-limiting and it is unusual for patients to have to discontinue the drug because of it. Although rashes can occur, severe rashes with efavirenz are unusual. Lipid abnormalities, mainly rises in cholesterol above baseline values, have been observed in patients on efavirenz-containing combinations[64]. Efavirenz appears to be potent when used in patients with high viral titres (> 100 000 copies/mL). As efavirenz may be teratogenic, women of child-bearing age should be warned of the potential risks if they become pregnant while taking this drug.

Nevirapine. Nevirapine with ZDV/didanosine (ddI) has been compared with ZDV/ddI alone in antiretroviral-naive patients[65]. This study, which led to the licensing of nevirapine, showed superior HIV-1 RNA suppression at 48 weeks in the triple-therapy arm. Virological and immunological data are also available from a head-to-head comparison of nevirapine with a PI (indinavir) in combination with ddI and stavudine (d4T), with most of the data extending to 48 weeks (the Atlantic study[66]).

Nevirapine is currently used twice a day, but the pharmacokinetics indicate that once a day would be reasonable. Major side-effects of nevirapine are rash, which occasionally manifests as Stevens–Johnson syndrome, and hepatitis, both of which can be fatal. Whether the drug rash with nevirapine can be reduced by the co-administration of steroids[67] or antihistamines remains to be proved in prospective clinical trials.

Delavirdine. A number of studies with delavirdine have been performed comparing dual NA combinations with a delavirdine-containing triple regimen in both treatment-naive and treatment-experienced patients[33–69]. The degree of viral load suppression at 1 year was higher in the delavirdine-containing arms. There are no data comparing delavirdine-containing triple regimens with other NNRTI-or PI-containing regimens.

The major side-effect of delavirdine use is rash. The present form of delavirdine requires four tablets to be taken three times a day, which may have an impact on adherence. Larger dose tablets in a twice-daily regimen will be introduced shortly. This agent is not currently licensed in Europe.

Advantages and disadvantages of NNRTIs. The side-effect profile of all the NNRTIs appears to be more favourable than that of PIs. This fact, coupled with their good pharmacokinetic properties and potential ease of adherence, is the major reason why many clinicians may choose to use nucleoside analogues and an NNRTI as first-line therapy. A major disadvantage of the use of NNRTIs as part of such a regimen is that a single mutation in the reverse transcriptase gene can produce a virus with markedly reduced sensitivity, which probably extends to all members of the class[70].

Studies showing continuing success of HAART regimens in the long term are particularly important, but as yet there are very few data for patients on NNRTIs. For example, patients on PI-containing regimens who experience virological failure may maintain a stable or rising CD4 count with low-level plasma viraemia for long periods of time and a slow rate of clinical progression[71]. Whether the same will hold true for NNRTI-based regimens remains to be seen.

3.2.4 Three nucleoside analogues

The strategy used in the Caesar study is no longer regarded as appropriate because, in the majority of patients, when 3TC was added to an established dual NA regimen, durable suppression of viral replication was not achieved. However, this study showed improvement in clinical outcome compared with continuing the dual therapy[72]. It has been suggested that 3TC-resistant virus is less replication-competent than wild-type virus, because those patients in whom 3TC treatment is continued have a reduced viral load compared with pretreatment levels and improved outcome despite the presence of resistance[73].

Preliminary 48-week data from the Atlantic study showed no significant difference in viral load suppression (to below the 50 copies/mL threshold) for the triple NA arm (d4T/ddI/3TC) compared with the other two arms (d4T/ddI/nevirapine or d4T/ddI/indinavir) by intention-to-treat or on-treatment analysis[66]. A post-hoc analysis showed that the triple NA arm did not suppress viral load to undetectable (< 50 copies/mL) as well as the other arms in subjects with baseline viral loads above 51 000 copies/mL, but the difference was not statistically significant (P = 0.08).

Studies indicate that the combination of abacavir (a guanosine analogue and inhibitor of reverse transcriptase) with two NAs is capable of producing long-term suppression of virus for up to 72 weeks[74]. Experience with triple NA combinations is limited, and more long-term data are required before these combinations can be recommended. The results of two trials comparing triple NA combinations with a combination of two NAs plus one PI show no significant difference in the degree of viral load suppression at 48 weeks[66–75]. There is, however, a suggestion from these two studies that triple NA combinations are less able to suppress viral replication completely in those with an initially high viral load. Data from the double-blind placebo-controlled trial of abacavir/ZDV/3TC vs. indinavir/ZDV/3TC at 48 weeks showed that, for patients with starting viral loads exceeding 100 000 copies/mL, the triple NA arm was significantly less likely to suppress virus to below 50 copies/mL, although the two regimens were equivalent for viral load reduction to below 400 copies/mL[75].

Advantages and disadvantages of NAs. Major advantages of triple NA regimens are good tolerability, a relative lack of drug–drug interactions and a low incidence of side-effects. The exception to this is abacavir hypersensitivity, a multiple system disorder often presenting with fever and rash, which occurs in up to 3% of patients. Re-challenging affected individuals has resulted in severe toxicity and death on occasion[75]. Therefore, the use of abacavir requires careful physician and patient education to avoid continuing or re-starting therapy if such reactions occur.

The other potential advantages of the triple NA therapy regimen are that virological failure will not be associated with the development of resistance to the other two currently available classes of drugs (NNRTIs and PIs). The negative side of this is the potential difficulty of finding further regimens containing new NAs for individuals who experience failure of their first-line therapy, because of overlapping toxicity or resistance profiles.

3.3 Recommendations for initial therapy

Insufficient data are available to make a clear recommendation as to initial treatment. Initial therapy should be individualized for each patient and the risks and benefits of the treatment considered including toxicity, adherence, resistance, immunological benefit, long-term safety, clinical trial data and stage of disease.

3.3.1 Two nucleoside analogues plus one or two protease inhibitors (BI)

There are immunological and clinical endpoint data supporting the use of PIs both in early and advanced disease.

If PIs are considered as the initial (or even subsequent regimen) for patients, using a combination of PIs may have an advantage over single PI regimens. Although there is no definitive evidence, these combinations should be regarded as first choice when a PI regimen is being chosen. Physicians prescribing such combinations should be aware that, for some patients, there may be more pronounced effects on lipid metabolism and lipodystrophy syndrome when ritonavir/saquinavir or ritonavir/indinavir are used together.

3.3.2 Two nucleoside analogues plus a non-nucleoside reverse transcriptase inhibitor (BII)

This combination is likely to have advantages as an initial regimen in terms of adherence, fewer drug interactions and improved pharmacokinetic profile. Short-term toxicity, although affecting a high proportion of patients, has been relatively easy to manage in the setting of clinical trials. Few studies have addressed the issue of long-term toxicity, but available data are reassuring and more information concerning durability of viral load suppression and possible improvement in clinical outcome associated with NNRTIs should soon be forthcoming.

The theoretical disadvantages of NNRTIs are that no studies have shown clinical benefit, and there are fewer data than for PI-containing combinations to indicate that immune reconstitution with these agents is possible.

In the committee's view there are insufficient data to choose between efavirenz or nevirapine as the preferred initial NNRTI. Similarly, analyses of the virological response in individuals with a viral load above 100 000 copies/mL at initiation of treatment indicate an equivalent response with efavirenz as in those with a lower viral load. These data are less complete for nevirapine. Delavirdine requires up to 12 tablets to be taken in a thrice-daily dosing regimen and, as yet, no trials have compared it with the current standard of care.

3.3.3 Choice of nucleoside analogue backbone for initial therapy

The choice of which NAs to use together is governed by overlapping toxicity, for example d4T/zalcitabine. ZDV/d4T are not used together as they share a common phosphorylation pathway for intracellular activation and one small study has shown a deleterious effect on CD4 counts when the latter combination was used[76]. Primary drug resistance, especially to ZDV, may affect the choice of initial NAs. This may be of particular relevance when considering a combination of ZDV with 3TC or an NNRTI, because of the relative ease with which resistant mutants are acquired. Although abacavir has been used most often in triple NA combinations, it may also be used in a ‘nucleoside analogue backbone’.

Nucleoside analogue toxicity. NA usage has been associated with numerous toxicities including lactic acidosis and hepatic steatosis, both of which have been linked to inhibition of mitochondrial DNA synthesis[77]. Lactic acidosis is often fatal. It can start insidiously, with nausea, weight loss, bloating and abdominal pain. In its full-blown form, it may closely resemble severe Gram-negative sepsis. Treatment with riboflavin has been suggested[78], without any clear evidence of benefit. Routine monitoring of serum bicarbonate anion gap and pyruvate/lactate ratios has been proposed in order to detect this syndrome early, but no data exist regarding a suitable early marker. The prevalence of this condition is unknown, but since it was not apparent in most of the large clinical controlled trials of NA therapy, it is likely to be rare or to occur only after prolonged therapy.

Hydroxyurea and mycophenolate. Hydroxyurea (HU) has been used as an antineoplastic agent and as therapy for sickle cell anaemia. It depletes the endogenous stores of natural nucleotides used by HIV to synthesize viral DNA prior to its integration into the host, thus increasing the likelihood of incorporation of administered NAs into the growing viral DNA chain and inhibition of viral DNA production[79]. The inhibitory effect appears to be more pronounced with ddI. However, there are data suggesting that the action of other drugs, especially abacavir and the nucleotide analogue PMPA, can also be enhanced by HU. HU also prevents CD4 cell activation, a necessary prerequisite for infection by HIV[80]. There are some data showing that the reduction in viral HIV RNA with ddI is enhanced by 0.4–0.6 log10 copies/mL by the addition of HU[81]. Other data suggest that the virological response to a NA combination of ddI and d4T is also enhanced by HU and long-term data are now emerging[82]. HU is being used at all stages of HIV infection in clinical trials and its role remains to be well defined. Toxicity of HU includes myelosuppression and hair loss. ddI-induced pancreatitis may result from HU enhancement of ddI intracellular drug concentrations, and HU might also be expected to increase mitochondrial toxicity.

The CD4 response to a regimen which includes HU is difficult to interpret as the drug may cause lymphopenia and a fall in absolute CD4 cell count. The CD4 percentage is probably a better surrogate marker to use in this situation.

Mycophenolate is an immunosuppressant used in renal transplantation and is a specific inhibitor of lymphocyte proliferation. It has been shown to enhance intracellular concentrations of abacavir by inhibition of inosine monophosphate dehydrogenase, an essential enzyme of guanosine metabolism[83]. Such treatment is potentially toxic and unproven. Patients should not be started on this therapy without being part of a closely monitored controlled trial.

4.0 Issues concerning protease inhibitor use

  1. Top of page
  2. 1.0 Introduction
  3. 2.0 When to start treatment
  4. 3.0 What to start with
  5. 4.0 Issues concerning protease inhibitor use
  6. 5.0 Changing therapy on first virological failure
  7. 6.0 Therapy after more than one previous failure (‘salvage’ therapy)
  8. 7.0 Resistance testing
  9. 8.0 Issues not addressed in these guidelines
  10. 9.0 BHIVA/IAS–USA comparisons
  11. References
  12. 11.0 Conflict of interest

The main drawbacks to using PIs in initial HAART regimens, in spite of their potency, are difficulties associated with (i) adherence; (ii) pharmacokinetics; (iii) toxicity (including lipid abnormalities and lipodystrophy syndrome); and (iv) drug–drug interactions. Major characteristics of the PIs are summarized in Table 5.

4.1 Adherence

In clinical practice, all the currently available PIs, except indinavir, are taken twice daily (bd). (Nelfinavir has not yet been licensed for bd dosing but, in practice, most patients are on bd dosing. Recent data support this schedule[84].) Many patients are now combining indinavir with ritonavir so that the former need only be taken twice daily. Gastrointestinal disturbance, dietary restrictions, pill burden and size and the need for strict timing of doses may all have an impact on adherence.

4.2 Pharmacokinetics

Considerable intersubject variation in the absorption and metabolism of PIs means that, in some individuals, despite full adherence to therapy, plasma concentrations fall below the IC90 (a measure of the amount of drug required to inhibit 90% of virus production in vitro)[85]. Consequently, viral replication may occur during certain parts of the day and is more likely to occur if doses are missed, badly spaced or taken incorrectly. Concomitant therapies may induce the liver enzyme cytochrome P450, which metabolizes PIs, and could lead to low PI plasma levels. It is also possible that intracellular levels of PIs may be reduced by other drugs that induce the activity of P-glycoprotein, a cellular efflux transporter[86]. The use of therapeutic drug monitoring may resolve some of these difficulties, but available data have shown little evidence of clinical benefit[87]. The growing use of dual PI therapy is likely to reduce the need for such monitoring.

Some occurrences of virological failure in the Trilege and ACTG 343 studies may have been due to suboptimal adherence and/or low plasma concentrations of indinavir despite good adherence[43–88]. Compartmentalization of effect may also be an issue, because although PIs penetrate lymph nodes[89], they penetrate seminal fluid only poorly[90].

4.3 Toxicity

A number of toxicities have become apparent since the licensing of PIs. It is not always clear which of these are due specifically to the PI and which are due to other agents in the regimen.

4.3.1 Lipid abnormalities

During the initial PI development programmes, increases in cholesterol and triglycerides were not widely appreciated, despite being reported[92]. However, it is now well known that some degree of plasma lipid elevation is to be expected shortly after the start of treatment with virtually all PIs and, in a small proportion of patients, very large increases have occurred[92]. It remains unclear whether these abnormalities are more severe with one PI than another. There is some evidence to suggest that ritonavir induces greater increases and saquinavir lesser increases in triglycerides[93]. A proposed mechanism by which these abnormalities arise invokes homology with lipid receptors and carrier proteins, as well as inhibition of cytochrome P450[94].

Lipid abnormalities may be associated with abnormal glucose tolerance and diabetes in some patients[95]. Abnormal carbohydrate metabolism associated with a reduced sensitivity to insulin is thought by some to be the primary reason for the development of the lipid abnormality.

Elevated plasma lipid levels may prove life-threatening in a few individuals; they may have a synergistic effect on the cardiovascular risk associated with HIV itself or with other risk factors such as smoking, hypertension, obesity, age, family history and so on. Very high triglyceride levels (10–20 times above normal) increase the risk of pancreatitis and some cases of coronary artery disease have been reported[96].

4.3.2 Lipodystrophy

The lipodystrophy, or fat redistribution syndrome, includes elements of both lipo-atrophy and lipohypertrophy. While apparently most common in patients on PIs, it has also been seen in patients on non-PI-containing HAART regimens[97]. It is possible that some elements of lipodystrophy or of its causation may result from PI use, and others from use of other drug classes (particularly NAs) or even from HIV infection itself.

Lipodystrophy occurs in a proportion of patients on PIs and other HAART regimens, with or without other lipid abnormalities. The syndrome may include fatty accumulation around the belly or upper torso, particularly the breasts in women, and is often combined with fat loss, especially from the limbs and face[95–98]. Rarer symptoms of abnormal fat distribution, such as ‘buffalo hump’ and symmetrical lipomatosis, are also observed[99].

The reported occurrence of lipodystrophy varies widely in different studies. This variation is due largely to inconsistency in definition of the syndrome between series, and different periods of follow-up. The prevalence of the condition in a population of patients on PIs increases with time on therapy, and changes usually become clinically apparent only after several months[100]. Ritonavir, or the combination of ritonavir/saquinavir, may produce more severe or rapid changes than other PIs[95], but lipodystrophy appears to be associated with use of all currently licensed PIs.

No standard method of measuring lipodystrophy currently exists and trials are ongoing to evaluate computer-assisted tomography (CAT), dual-energy X-ray absorptiometry and other methods. A single CAT scan taken through the umbilicus may show an increase of visceral compared with subcutaneous fat in individuals with this condition[101].

The causes of elements of the lipodystrophy syndrome remain controversial. According to one theory, yet to be confirmed, NAs may play a role by inducing mitochondrial toxicity through binding to DNA polymerase gamma, an essential enzyme for the replication of mitochondrial DNA[102]. This might lead to apoptosis of peripheral fat cells and subsequent lipoatrophy. Mitonchondria in visceral fat cells may become dysfunctional, resulting in accumulation of fat released into the circulation from peripheral stores (as a result of apoptosis). The situation may be exacerbated by abnormalities in lipid metabolism induced by PIs.

4.3.3 Recommendations for monitoring lipids (BIII)
  • cirf
    & ;   Baseline lipids should be measured in all patients with HIV prior to starting HAART. They should be monitored regularly during treatment.
  • cirf
    & ;   Patients should be forewarned of the possibility of developing the lipodystrophy syndrome. There are some data to suggest that metformin and other agents (e.g. growth hormone) may decrease buffalo hump and abdominal fat[103–104]. Exercise may improve the lipo-atrophy. It may be useful to switch from a PI- to a nevirapine-containing regimen[105]. This may reverse lipid abnormalities but, because of the difficulties with defining lipodystrophy cases, it remains to be proven whether abnormal fat distribution can be reversed.

4.4 Haemophilia

Haemophiliacs who have taken PIs have experienced episodes of spontaneous bleeding[106]. While this side-effect seems to be relatively common, its aetiology is uncertain. It appears that the bleeding tendency may settle with time[107]. Physicians prescribing PIs to haemophiliac patients should be aware of this potential problem.

4.5 Viral hepatitis

Patients co-infected with hepatitis C should have their liver enzymes carefully monitored as there have been reports of disturbed liver function in such patients taking PI-containing HAART regimens.

5.0 Changing therapy on first virological failure

  1. Top of page
  2. 1.0 Introduction
  3. 2.0 When to start treatment
  4. 3.0 What to start with
  5. 4.0 Issues concerning protease inhibitor use
  6. 5.0 Changing therapy on first virological failure
  7. 6.0 Therapy after more than one previous failure (‘salvage’ therapy)
  8. 7.0 Resistance testing
  9. 8.0 Issues not addressed in these guidelines
  10. 9.0 BHIVA/IAS–USA comparisons
  11. References
  12. 11.0 Conflict of interest

The committee believes it is important to differentiate between failure following first-line therapy, when other available options are likely to produce a fully suppressive regimen, from failure in patients who have received multiple previous therapies. The chances of further combinations succeeding in preventing viral replication are less likely in these patients.

5.1 Initial failure

Viral rebound is usually associated with clinical progression in patients on treatment whose viral load continues to rise above 50 copies/mL. In spite of apparent virological failure, those who maintain viral loads within the range 50–500 copies/mL do not seem to progress in the short term and the CD4 count may not fall. There is some short-term evidence of differential clinical progression with viral load copy numbers in this range. One cohort study, which controlled for CD4 cell count and number and type of antiretroviral agents used, has demonstrated a virological relapse rate nearly four times greater in those achieving a viral load < 400 copies/mL but > 50 copies/mL[108]. It is likely that the higher the copy number, the more probable the development of resistance. For some drugs (e.g. 3TC or NNRTIs), a mutation at one position in the reverse transcriptase gene can cause resistance. Thus, if viral replication is shown to be persisting, and other options are available which can completely suppress it, then therapy should be changed. Once a patient shows a rise in viral load just above detectable, the viral load should be re-checked within 2–4 weeks; patients who are developing virological failure will show further increases in viral load, whereas those whose viral load is detectable because of assay-related problems will show no further rise or revert to undetectable. This latter pattern may also be seen in patients who experience transient viral/bacterial infections or receive a vaccination.

5.2 Changing therapy (BII)

The following groups of patients should be considered for changing therapy:

(i) Patients who are receiving incompletely suppressive antiretroviral therapy (for historical reasons, usually single or double NA therapy). Some physicians and patients may not want to change such therapy if the viral load is less than 10 000 copies/mL and the CD4 count is above 300 cells/μL and stable.

(ii) Patients who have been on a HAART regimen and whose viraemia was initially suppressed to undetectability but has now become consistently detectable.

(iii) Patients started on HAART whose viraemia was never suppressed to below detectable limits.

5.2.1 Recommendations for changing therapy (BII)

In general, good virological responses are mostly likely to be obtained when adherence is optimized and as many of the drugs as possible in the regimen are changed. The more components of the regimen that can be changed, the more likely it is that clinical events will be delayed because of complete viral suppression by the second regimen ( Tables 7 and 8).

Table 7.  Changing therapy on first virological failure: summary of recommendations
·Only consider virological failure after assessment of adherence and pharmacokinetics.
·Diagnosis of virological failure requires two viral load tests at least 2 weeks apart.
·If changes are being considered because of toxicity, changing a single agent is acceptable in the absence of virological failure.
·If a patient is poorly adherent, it is important to simplify the regimen and provide support and counselling.
·Testing for resistance is recommended (on stored pretreatment sample as well if appropriate).
Table 8.  What to change to after first virological failure: summary of recommendations
Initial regimen
  • 1

    Both NAs should be changed, but potential exists for cross-resistance to other NAs and, if present, could lead to rapid development of resistance to NNRTIs.

  • 2

    2 Should be strongly considered if primary reason for failure is poor adherence or pharmacokinetics. (Resistance to PIs will often not be found on testing.) May be more successful if nelfinavir is the initial PI.

  • 3

    Initial studies with ABT-378 + ritonavir + nevirapine have shown good results. No obvious salvage regimen following use of all three classes.

  • 4

    4 This regimen is still being evaluated.

2NAs + PI2NAs 1 + NNRTI
3NAs 42NAs 1 + NNRTI

5.3 Failure of two nucleoside analogues plus a protease inhibitor

As stated above, the causes of failure are complex, but it has rarely been attributed to PI resistance in the studies of such triple regimens reported to date[109–110]. There is no clear evidence to guide the choice of drugs and it is likely that the best regimen, and choice of subsequent therapy, will depend on whether failure is due to lack of potency, lack of adherence or poor pharmacokinetics of an individual drug. Resistance testing might be particularly important at this stage, possibly identifying which NA(s) will be of most benefit in the new regimen.

Many physicians would change both NAs and switch the PI to an NNRTI. Others would argue that there is a high risk of virological failure with this strategy and would either intensify with two PIs in the first instance or use a regimen combining two PIs and an NNRTI, switching the NAs. If resistance testing is available and no resistance is found, alternative approaches might be to intensify the regimen by adding abacavir, or to change from a single to a dual PI combination to improve pharmacokinetics.

5.4 Failure of two nucleoside analogues plus a non-nucleoside reverse transcriptase inhibitor

No clinical endpoint study has addressed this issue. A PI-based regimen improves the clinical outlook after NA therapy[58] and is likely to do so after two NAs and an NNRTI. Most physicians would treat virological failure of this type of regimen by discontinuing the NNRTI, changing the two NAs if possible and adding a PI component (single or dual).

5.5 Failure of triple nucleoside analogue therapy

There are too few data as to the cause(s) of failure on a triple NA regimen to comment on what would be the optimal regimen to use in this situation. Most physicians would probably switch to a regimen comprising two new NAs plus a PI or an NNRTI. Alternatively, both a PI and an NNRTI might be used with new NAs[111].

6.0 Therapy after more than one previous failure (‘salvage’ therapy)

  1. Top of page
  2. 1.0 Introduction
  3. 2.0 When to start treatment
  4. 3.0 What to start with
  5. 4.0 Issues concerning protease inhibitor use
  6. 5.0 Changing therapy on first virological failure
  7. 6.0 Therapy after more than one previous failure (‘salvage’ therapy)
  8. 7.0 Resistance testing
  9. 8.0 Issues not addressed in these guidelines
  10. 9.0 BHIVA/IAS–USA comparisons
  11. References
  12. 11.0 Conflict of interest

The definition of ‘salvage’ therapy varies. Here, it is taken to mean treatment following exposure to multiple drugs and, usually, all three classes of antiretroviral agents. However, many so-called ‘salvage studies’ have been carried out in patients who are naive with respect to one class of drugs.

The reasons for drug failure are complex. Most studies of salvage therapy to date have not distinguished between virological failure due to poor adherence and failure due to poor pharmacokinetic parameters. In individuals who have been poorly adherent to previous therapy but have not developed resistant virus, subsequent treatment may be effective if adherence is improved. Low blood levels of PI, either because of poor absorption or unforeseen pharmacokinetic interactions, may also lead to failure without the development of virological resistance to PIs. If the regimen is adjusted, again the outcome can be expected to be better than when resistance to PIs is present.

The criteria for success of salvage studies also vary. Suppressing viral load to below detectability (i.e. to below 50 copies/mL) has become the accepted measure of success. However, data from a number of large clinical endpoint studies show that much more modest declines in viral load correlate with improvement in clinical outcome. Viral load reductions of greater than 0.5 log10 copies/mL may be responsible for clinical improvement and may imply that such a regimen is worth pursuing.

Many of the salvage studies have been of short duration with few follow-up data, making it hard to judge whether or not viral suppression will be maintained in the long term. In late disease, the immediate risk of death is much more closely associated with the CD4 count than with the viral load and thus, perhaps, an additional criterion in salvage studies is the degree to which the CD4 count rises.

Despite these difficulties, both cohort and clinical controlled studies identify a number of general principles to consider when deciding upon a salvage regimen.

Firstly, salvage will have the greatest likelihood of success if individuals are naive to one class of drugs. For historical reasons, this is most likely to be the NNRTI class. It seems to be particularly important to give these agents as part of a fully suppressive regimen to avoid the rapid emergence of resistance.

Secondly, improved outcome is more likely with concurrent use of drugs to which the patient has either not been exposed or to which resistance is unlikely or proven to be absent.

Thirdly, salvage therapy is more successful in those who commence at a lower viral load (e.g. < 5000–10 000 copies/mL). This is assumed to be because of the accumulation of additional mutations within the viral population in those who continue on failing therapy, thus increasing the likelihood of cross-resistance to new agents tried subsequently.

Finally, resistance testing is strongly recommended in all cases where there are difficult choices to make concerning the most beneficial treatment. Genotypic and phenotypic resistance both predict responses to using salvage regimens, although the absence of resistance is less predictive of success[112]. It is possible that, even when resistance to each of the drugs has been shown individually (and for all drug classes), a combination might still retain some ability to reduce viral load.

6.1 Salvage therapy in PI-experienced patients

Both cohort data and clinical controlled trials suggest that 60–70% of subjects who are experiencing failure of PI-based therapy and who are NNRTI naive will achieve viral load reductions to below 400 copies/mL at 16 weeks by switching to appropriate (i.e. new) NAs plus two PIs and an NNRTI[113–114]. The two PIs most commonly used in such studies are ritonavir and saquinavir. The primary reason for their efficacy is probably improved pharmacokinetic behaviour although it is possible, in some cases, that saquinavir may have additional direct antiviral effects (i.e. if cross-resistance to saquinavir is absent from the viral population).

Individuals who experience failure of ritonavir or indinavir as their initial therapy are less likely to respond virologically to the introduction of ritonavir/saquinavir than those who took nelfinavir as their first PI. A proportion of individuals exposed to nelfinavir only will harbour virus with a codon 30 mutation, which will continue to be sensitive to other PIs. Thus, most clinicians would recommend a rapid change from nelfinavir, once failure has occurred, to a dual PI-containing regimen. However, one study found that the median time to virological failure prior to change was of the order of 48 weeks[115]. In spite of such long-term virological failure, over 60% of patients achieved a viral load reduction to below 400 copies/mL.

6.2 Salvage therapy in NNRTI-experienced patients

There are few data to guide the appropriate choice of salvage therapy in such cases. Most results have shown only very poor virological responses. Thus, in one single-arm study, where amprenavir, efavirenz and NAs including abacavir were used in a group of patients who were multiple PI-experienced, nearly half were also NNRTIs-experienced[116]. Success was largely dependent on being NNRTI naive and having a low viral load. Only a very small number of NNRTI-experienced patients responded virologically when the viral load at the time of switch exceeded 40 000 copies/mL.

6.3 Salvage therapy in patients with multiple class resistance

A number of approaches have been tried in this situation with varying degrees of success. Some have tried combining large numbers of drugs (five or more)[113–117], so-called mega-HAART, despite resistance to many of the individual components. A number of small cohort studies reported successful maintenance of viral load below 400 copies/mL for up to 2 years. Such studies are difficult to analyse because a proportion of patients were NNRTI-naive, up to one-third suffered from serious toxicity (often leading to withdrawal) and treatment was not assigned randomly. The role of hydroxyurea and its potential toxicity in this situation remains unclear. Although the regimens contained multiple drugs, adherence was often relatively good because of the possibility of twice-daily dosing.

6.4 Structured treatment interruption

The strategy of structured treatment interruption (STI) (or ‘drug holiday’) has been tested at various stages of HIV infection. In early disease or PHI, when T-helper responses are still present, STI might theoretically be a logical step to allow the immune system to control viral replication without HAART. In late disease, however, such responses are lost and there are usually either minimal or no HIV-specific cytotoxic T lymphocyte responses. Thus, it is unlikely that STI would result in immunological control of HIV disease. However, in a number of cohorts stopping all drug therapy resulted in rapid reversion of existing viral variants to wild-type. Importantly, STI is accompanied by a major loss of CD4 cells and a substantial increase in viral load at 1 month[118]. Although resuming drug therapy in such individuals meets with a surprisingly high rate of virological response, which may persist long-term, it takes about a year for the CD4 cell count to be restored to its pre-STI level. While such studies are very interesting and somewhat of an enigma because resistant virus does not immediately reappear when drug pressure is re-instituted, the rise in viral load and decline in CD4 cell count means that such an approach should be used cautiously. This strategy should be the subject of a rigorous controlled trial before it can be recommended as a way of proceeding.

6.5 Stopping therapy

Most drug therapy discontinuations have been associated with a rapid rise in viral load and fall in CD4 cell count. Thus, for the majority of individuals, continuation of therapy is possibly associated with a better prognosis than discontinuation. However, it is most important to weigh up the risks of continuing any medication (potential toxicity) against the benefits (potential for improved outcome).

6.6 Viral fitness

There are some data to suggest that viruses harbouring PI-associated resistance mutations, and possibly those conferring 3TC resistance, are less replicatively fit and thus multiply more slowly. It is theoretically possible, therefore, that a simplified regimen comprising a PI and 3TC may continue to provide selective pressure so that these less fit mutants predominate in the circulation and lead to an improved prognosis. There are cohort data to suggest that individuals with low CD4 count (< 50 cells/μL) and high viral load (> 106 copies/mL) do better on a PI-containing regimen than other patients who were only treated with dual NAs[119].

6.7 Recommendations for subsequent virological failure (third or more regimen)

See also Table 7.

  • cirf
    & ;  Test for genotypic resistance; phenotypic assay may be necessary if genotype difficult to interpret.
  • cirf
    & ;  Change as many drugs as possible.
  • cirf
    & ;  Introduce a new class if possible, but not if chance of success by combining new class with other drugs is small. May be better to defer change until new treatment options available.
  • cirf
    & ;  Structured treatment interruption (drug holiday) is not recommended as standard of care; needs to be evaluated in clinical trials.
  • cirf
    & ;  Mega-HAART or recycling of nucleoside analogues ± hydroxyurea may be of value to some patients.
  • cirf
    & ;  Consider stopping all drug treatment if toxicity outweighs any likely benefit.

7.0 Resistance testing

  1. Top of page
  2. 1.0 Introduction
  3. 2.0 When to start treatment
  4. 3.0 What to start with
  5. 4.0 Issues concerning protease inhibitor use
  6. 5.0 Changing therapy on first virological failure
  7. 6.0 Therapy after more than one previous failure (‘salvage’ therapy)
  8. 7.0 Resistance testing
  9. 8.0 Issues not addressed in these guidelines
  10. 9.0 BHIVA/IAS–USA comparisons
  11. References
  12. 11.0 Conflict of interest

HIV drug resistance is defined as reduced susceptibility to one or more drugs. Phenotypic resistance is measured in the laboratory by propagating the virus in the presence of a range of drug concentrations, and is expressed as the IC50 or IC90. This phenotype is determined by specific mutations within the HIV genome, which can be detected by genotypic analysis (nucleotide sequencing).

Two lines of evidence support the use of such assays in management of HIV-infected patients. First, retrospective studies have demonstrated that the phenotype, or presence/absence of specific mutations at the time of antiretroviral failure, can predict short-term response to new therapies[120–123]. More importantly, two prospective, randomized studies of genotyping in patients for whom therapy is failing have shown a significant virological benefit of testing vs. not testing[124–125]. Nevertheless, all these studies have been short-term and it may be that any benefit of resistance testing is lost over a longer period of time. Further studies are required to identify those patients most likely to benefit from such laboratory interventions.

All currently available assays (genotypic and the recombinant phenotype assay) are based on detection of the predominant viral RNA species in plasma. They will not detect minor populations or variants represented within proviral DNA in latently infected cells. These assays may not, therefore, detect the presence of ‘archived’ resistant species, which have been generated previously. Thus, failure to detect a resistant species within an assay may not imply drug sensitivity. Interpretation of these data must take account of drug history.

7.1 Genotypic testing

Specific data on the relationship between genotypic mutations and reduced drug susceptibility have been generated through in vitro experiments as well as clinical drug trials. Much of the latter data are based on monotherapy studies. There range of techniques available for determining the viral genotype is growing, but they fall into two general categories.

7.1.1 Detection of specific point mutations

This approach may involve detection of a mutation or wild-type base at a specific position of interest within the genome, or the use of a number of probes to detect mutations at more than one key position.

7.1.2 Nucleic acid sequencing of the reverse transcriptase and/or protease genes of plasma viral RNA

This approach generates a large volume of information for each patient and therefore requires suitable software programs for analysis and identification of possible resistance-associated mutations. The use of combination therapy has increased the complexity of genotypes identified with resistance. This more comprehensive approach is now the technique of choice.

Assays for detecting specific mutations (e.g. the line probe assay, LiPA[126]) and for genome sequencing have been commercialized, and incorporate specific methods for amplifying virus. These have the advantage of standardization, and thus consistency between lab-oratories. However, since many such assays are validated against clade B virus, they may be suboptimal for genotypic analysis of other subtypes. This mirrors previous problems with viral load assays[127]. Further assay development is currently being undertaken to address this issue. In the meantime, ‘in house’ sequencing methodologies remain important as a reference function.

7.2 Phenotypic testing

The phenotype defines drug susceptibility of a virus. This is expressed an IC50 or IC90 against a particular drug. The difficulty in undertaking phenotypic assays rapidly and routinely by traditional culture methods has led to the development of the recombinant virus assay, whereby the reverse transcriptase and/or protease genes of plasma virus are recombined with a laboratory strain of virus deficient in these relevant genes, which can then be assayed rapidly for drug susceptibility. The IC50 of the recombinant virus is compared with the IC50 of the original laboratory strain, and thus results are expressed as the fold increase in value. Currently, these assays are only available on a routine basis from commercial companies.

7.3 Advantages and disadvantages of resistance testing

Genotypic assays: advantages:

  • cirf
    & ;  Rapid assay turnaround (within 2–3 weeks).
  • cirf
    & ;  Commercialized assays available for use in virology laboratories.
  • cirf
    & ;  Allows detection of genetic mutations which may occur before phenotypic changes can be noted.
  • cirf
    & ;  Sequencing of reverse transcriptase and/or protease genes allows detection of a wide range of resistance-associated mutations.

Genotypic assays: disadvantages:

  • cirf
    & ;  Insensitive to the presence of minor variants (i.e. comprising < 20% of plasma population).
  • cirf
    & ;  Interpretation requires knowledge of the relationship between single and multiple mutations and phenotype.
  • cirf
    & ;  Assays generate a large amount of nucleotide sequence data which require sophisticated analysis for use as a high throughput test.

Phenotypic assays: advantages:

  • cirf
    & ;  Determines the overall impact of the array of drug-resistance-associated mutations on virus phenotype.
  • cirf
    & ;  Provides information on cross-resistance.

Phenotypic assays: disadvantages:

  • cirf
    & ;  May not reflect early emergence of resistance or subpopulations of virus.
  • cirf
    & ;  More time-consuming and expensive than genotypic assays.
  • cirf
    & ;  Unlikely that diagnostic virology laboratories will be able to provide a routine service in the near future.

7.4 Practical issues of resistance testing

  • cirf
    & ;  Genotypic assays and the recombinant virus assay require a viral load of at least 1000–5000 copies/mL to ensure a result.
  • cirf
    & ;  All assays require careful interpretation before being utilized to guide therapy. This interpretation should be undertaken in the light of a full drug history. Expert opinion should be sought.
  • cirf
    & ;  At present, no recognized internal or external quality control or quality assurance scheme is operating for these assays within diagnostic laboratories, although some are being piloted. Clinicians utilizing these assays should be aware of the quality control criteria used by their laboratory.

7.5 Primary resistance

It is clear that resistant virus can be transmitted but how commonly this occurs is unclear, as most of the studies examining this issue have involved small numbers of patients[128]. The transmission of virus genotypically resistant to ZDV is relatively common, although in one study such virus was usually phenotypically sensitive to the drug[129].

Up to 10% of patients in a number of American studies have shown possible genotypic resistance to PIs, but in most patients only polymorphisms (not associated with drug resistance) have been observed. Phenotypic resistance rates have been much lower. In one study fairly high rates of NNRTI genotypic resistant polymorphisms were found, but a proportion of such individuals also had phenotypic resistance to NNRTI. These individuals clustered in a geographical area where NNRTI monotherapy had been used[129].

It is our view, based on the available data, that routine resistance testing should be recommended prior to therapy in chronically infected HIV patients, and the clinician should be aware that treatment of naive patients may be complicated by pre-existing drug resistance. Alternatively, we would advise checking virological responses at 4–8 weeks of therapy and storing pretreatment samples for future resistance testing if initial responses are unsatisfactory. In cases of primary HIV infection, a resistance test should always be performed as it might guide therapy.

8.0 Issues not addressed in these guidelines

  1. Top of page
  2. 1.0 Introduction
  3. 2.0 When to start treatment
  4. 3.0 What to start with
  5. 4.0 Issues concerning protease inhibitor use
  6. 5.0 Changing therapy on first virological failure
  7. 6.0 Therapy after more than one previous failure (‘salvage’ therapy)
  8. 7.0 Resistance testing
  9. 8.0 Issues not addressed in these guidelines
  10. 9.0 BHIVA/IAS–USA comparisons
  11. References
  12. 11.0 Conflict of interest

8.1 Pregnancy

Pregnancy guidelines have recently been published elsewhere[130].

8.2 Post-exposure prophylaxis for healthcare workers

In the UK, such guidelines have been published by the Expert Advisory Group on AIDS[131]. There are no available data to show when post-exposure prophylaxis should be given in other situations (e.g. following sexual exposure), but issues of timeliness of taking therapy, cost–benefit, toxicity and adherence should be addressed.


  1. Top of page
  2. 1.0 Introduction
  3. 2.0 When to start treatment
  4. 3.0 What to start with
  5. 4.0 Issues concerning protease inhibitor use
  6. 5.0 Changing therapy on first virological failure
  7. 6.0 Therapy after more than one previous failure (‘salvage’ therapy)
  8. 7.0 Resistance testing
  9. 8.0 Issues not addressed in these guidelines
  10. 9.0 BHIVA/IAS–USA comparisons
  11. References
  12. 11.0 Conflict of interest
  • 1
    Delfraissy, JF et al. Prise en charge thérapeutique des personnes infectées par le VIH. recommandations du groupe d'experts. France, Flammarion, 1999.
  • 2
    Smith, D, Whittaker, B, Crowe, S et al. Antiretroviral therapy for HIV infection: principles of use (standard of care guidelines). HIV/AIDS Clinical trials and treatments advisory committee booklet (October), 1997.
  • 3
    Carpenter, CCJ, Cooper, DA, Fischl, MA et al. Antiretroviral therapy in adults. Updated recommendations of the International AIDS Society–USA Panel. JAMA 2000; 283, 381 390.
  • 4
    Centers for Disease Control and Prevention. Report of the NIH Panel to define principles of therapy of HIV infection and guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. MMWR 1998; 47, 1 91.
  • 5
    Gazzard, B & Moyle, G. on behalf of the BHIVA Guidelines Writing Committee. 1998 revision to the British HIV Association guidelines for antiretroviral treatment of HIV seropositive individuals. Lancet 1998; 352, 314 316.
  • 6
    Hurwitz, B. Legal and political considerations of clinical practice guidelines. BMJ 1999; 318, 661 664.
  • 7
    Pozniak, A. Surrogacy in HIV-1 clinical trials. Lancet 1998; 351, 536 537.
  • 8
    Fleming, TR & Demets, DL. Surrogate end points in clinical trials: are we being misled? Ann Intern Med 1996; 125, 605 613.
  • 9
    De Gruttola, V, Fleming, T, Lin, DY, Coombs, R. Perspective: validating surrogate markers – are we being naive? J Infect Dis 1997; 175, 237 246.
  • 1
    Babiker, A. for the Delta Co-ordinating Committee and Virology Group. Can HIV-1 RNA viral load be used as a surrogate for clinical endpoints in HIV disease? 6th European Conference on Clinical Aspects and Treatment of HIV Infection, Hamburg, Germany. (abstract 103)., 1997.
  • 11
    Mellors, JW, Munoz, A, Giorgi, JV et al. Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med 1997; 126, 946 954.
  • 12
    Gazzard, B, Hill, A, Gartland, M. for the AVANTI Study Group. Different analyses give highly variable estimates of HIV-1 RNA undetectability and log10 reduction in clinical trials. AIDS 1998; 12, S36((Abstract) P77).
  • 13
    Raboud, JM, Montaner, JSG, Conway, B et al. Suppression of plasma viral load below 20 copies/ml is required to achieve a long-term response to therapy. AIDS 1998; 12, 1619 1624.
  • 14
    Zhu, T, Mo, H, Wang, N et al. Genotypic and phenotypic characterization of HIV-1 patients with primary infection. Science 1993; 261, 1179 1181.
  • 15
    Zhang, LQ, Mackenzie, P, Cleland, A et al. Selection for specific sequences in the external envelope protein of human immunodeficiency virus type 1 upon primary infection. J Virol 1993; 67, 3345 3356.
  • 16
    Markowitz, M, Vesanen, M, Tenner-Racz, K et al. The effect of commencing combination antiretroviral therapy soon after human immunodeficiency virus type-1 infection on viral replication and antiviral immune responses. J Infect Dis 1999; 179, 527 537.
  • 17
    Zhang, L, Ramratnam, B, Tenner-Racz, K et al. Quantifying residual HIV-1 replication in patients receiving combination antiretroviral therapy. N Engl J Med 1999; 340, 1605 1613.
  • 18
    Koup, RA, Safrit, JT, Cao, Y et al. Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome. J Virol 1994; 68, 4650 4655.
  • 19
    Borrow, P, Lewicki, H, Hahn, BH, Shaw, GM, Oldstone, MB. Virus-specific CD8+ cytotoxic T-lymphocyte activity associated with control of viremia in primary human immunodeficiency virus type 1 infection. J Virol 1994; 68, 6103 6110.
  • 20
    Rosenberg, ES, Billingsley, JM, Caliendo, AM et al. Vigorous HIV-1-specific CD4+ T cell responses associated with control of viremia [see comments]. Science 1997; 278, 1447 1450.
  • 21
    Walker, BD, Rosenberg, ES, Hay, CM, Basgoz, N, Yang, OO. Immune control of HIV-1 replication. Adv Exp Med Biol 1998; 452, 159 167.
  • 22
    Lisziewicz, J, Rosenberg, E, Lieberman, J et al. Control of HIV despite the discontinuation of antiretroviral therapy. N Engl J Med 1999; 340, 1683 1684.
  • 23
    Wong, JK, Hezareh, M, Gunthard, HF et al. Recovery of replication-competent HIV despite prolonged suppression of plasma viremia. Science 1997; 278, 1291 1295.
  • 24
    Finzi, D, Hermankova, M, Pierson, T et al. Identification of a reservoir for HIV-1 in patients on highly active antiretroviral therapy. Science 1997; 278, 1295 1300.
  • 25
    Chun, T-W, Carruth, L, Finzi, D et al. Quantification of latent tissue reservoirs and total body viral load in HIV-1 infection. Nature 1997; 387, 183 187.
  • 26
    Ho, DD. Toward HIV eradication or remission: the tasks ahead. Science 1998; 280, 1866 1867.
  • 27
    Pontesilli, O, Kerkhof-Garde, S, Notermans, DW et al. Functional T cell reconstitution and human immunodeficiency virus-1-specific cell-mediated immunity during highly active antiretroviral therapy. J Infect Dis 1999; 180, 76 86.
  • 28
    Fischl, MA, Richman, DD, Hansen, N et al. The safety and efficacy of zidovudine (AZT) in the treatment of subjects with mildly symptomatic human immunodeficiency virus type 1 (HIV) infection. A double-blind, placebo-controlled trial. The AIDS Clinical Trials Group. Ann Intern Med 1990; 112, 727 737.
  • 29
    Concorde Coordinating Committee. Concorde. MRC/ANRS randomised double-blind controlled trial of immediate and deferred zidovudine in symptom-free HIV infection. Lancet 1994; 343, 871 881.
  • 30
    Perelson, AS, Neumann, AU, Markowitz, M et al. HIV-1 dynamics in vivo: virion clearance rate, infected cell life-span, and viral generation time . Science 1996; 271, 1582 1586.
  • 31
    Yu, G, Clendennin, NJ, Quart, B et al. Viral load levels during treatment with VIRACEPT combination treatment as predictor of response and guidance for treatment choice strategy. 12th World AIDS Conference, Geneva. (abstract 12300)., 1998.
  • 32
    Gulick, RM, Mellors, JW, Havlir, D et al. Treatment with indinavir, zidovudine, and lamivudine in adults with human immunodeficiency virus infection and prior antiretroviral therapy. N Engl J Med 1997; 337, 734 739.
  • 33
    Green, S, Para, MF, Daly, PW et al. Interim analysis of plasma viral burden reductions and CD4 increases in HIV-1 infected patients with Rescriptor (DLV) + Retrovir (ZDV) + Epivir (3TC). 12th World AIDS Conference, Geneva. (abstract 129/12219)., 1998.
  • 34
    Fischl, M, Greenberg, S, Clumeck, N et al. Safety and activity of abacavir (1592, ABC) with 3TC/ZDV in antiretroviral naive subjects. 12th World AIDS Conference, Geneva. (abstract 127/12230)., 1998.
  • 35
    Gisolf, EH, De Wolf, F, Van Wanzeele, F et al. Treatment of ritonavir (RTV) /saquinavir (SQV) versus RTV/SQV/stavudine (d4T) (the Prometheus study): preliminary results. 12th World AIDS Conference, Geneva. (abstract 12274)., 1998.
  • 36
    Tashima, K, Staszewski, S, Stryker, R et al. A phase III, multicenter, randomized, open-label study to comapre the antiretroviral activity and tolerability of efavirenz + indinavir, versus EFV + zidovudine + lamivudine, versus IDV + ZDV + 3TC at 48 weeks. 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract LB-16)., 1999.
  • 37
    Farzadegan, H, Hoover, DR, Astemborski, J et al. Sex differences in HIV-1 viral load and progression to AIDS. Lancet 1998; 352, 1510 1514.
  • 38
    Katzenstein, DA, Hammer, SM, Hughes, MD et al. The relation of virologic and immunologic markers to clinical outcomes after nucleoside therapy in HIV-infected adults with 200–500, CD4 cells per cubic millimeter. N Engl J Med 1996; 335, 1091 1098.
  • 39
    Bush, CE, Donovan, RM, Markowitz, N et al. Gender is not a factor in serum human immunodeficiency virus type 1 RNA levels in patients with viremia. J Clin Microbiol 1996; 34, 970 972.
  • 40
    Moore, RD, Cheever, L, Keruly, JC, Chaisson, RE. Lack of sex difference in CD4 to HIV-1 RNA viral load ratio. Lancet 1999; 353, 463 464.
  • 41
    Stein, DS, Korvick, JA, Vermund, SH. CD4+ lymphocyte cell enumeration for prediction of clinical course of human immunodeficiency virus disease: a review [see comments]. J Infect Dis 1992; 165, 352 363.
  • 42
    Autran, B, Carcelain, G, Li, T et al. Positive effects of combined antiretroviral therapy on CD4+ T cell homeostasis and function in advanced HIV disease. Science 1997; 277, 112 116.
  • 43
    Brun-Vezinet, F et al. Virologic failure, resistance and plasma drug measurements in an induction maintenance therapy trial (ANRS 072 Trilege). International Conference on the Discovery and Clinical Development of Antiretroviral Therapies, St Thomas, US Virgin Islands. (abstract 15)., 1998.
  • 44
    Paterson, D, Swindells, S, Mohr, J et al. How much adherence is enough? A prospective study of adherence to protease inhibitor therapy using MEMSCaps. 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract 92)., 1999.
  • 45
    Demasi, R, Tolson, J, Pham, S et al. Self-reported adherence to HAART and correlation with HIV RNA. initial results with the Patient Medication Adherence Questionnaire. 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract 94)., 1999.
  • 46
    Woodward, J, Wareham, PS, Grohskopf, L, Madigan, D, Hooton, TM. Protease inhibitor adherence and HIV-1 RNA response. 12th World AIDS Conference, Geneva. (abstract 60363)., 1998.
  • 47
    Ferrando, SJ, Wall, TL, Batki, SL, Sorensen, JL. Psychiatric morbidity, illicit drug use and adherence to zidovudine (AZT) among injection drug users with HIV disease. Am J Drug Alcohol Abuse 1996; 22, 475 487.
  • 48
    Broers, B, Morabia, A, Hirschel, B. A cohort study of drug users' compliance with zidovudine treatment. Arch Intern Med 1994; 154, 1121 1127.
  • 49
    Miller, L, Liu, H, Beck, K et al. Providers' estimates of adherence overestimate reports from Medication Event Monitoring System (MEMS) for patients on protease inhibitors (PIs). 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract 97)., 1999.
  • 50
    Anderson, W et al. Taking Heart? The impact of combination therapy on the lives of people with HIV (Phase 2). London, Sigma Research, 1999.
  • 51
    Arabe, J, Rubini, NPM, Rodrigues, ACA et al. Factors which influence adherence to the use of protease inhibitors. 12th World AIDS Conference, Geneva. (abstract 32404)., 1998.
  • 52
    Bachiller, P, Arrando, FR, Liceaga, G et al. Adherence to antiretroviral therapy in HIV-infected patients. 12th World AIDS Conference, Geneva. (abstract 32392)., 1998.
  • 53
    Gifford, AL, Shively, MJ, Bormann, JE et al. Self-reported adherence to combination antiretroviral medication (ARV) regimens in a community-based sample of HIV-infected adults. 12th World AIDS Conference, Geneva. (abstract 32338)., 1998.
  • 54
    Lorenzen, T, Stoehr, A, Weitner, L et al. Compliance with antiretroviral multidrug therapy in HIV-infected patients and reasons for non-compliance. 12th World AIDS Conference, Geneva. (abstract 32371)., 1998.
  • 55
    Nakashima, AK, Jones, JL, Burgess, DA, Ward, JW. Adherence to currently prescribed antiretroviral therapies: results from a multisite interview project. 12th World AIDS Conference, Geneva. (abstract 32326)., 1998.
  • 56
    Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. JAMA 1993; 269, 729 730.
  • 57
    Palella, FJ, Delaney, KM, Moorman, AC et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 1998; 338, 853 860.
  • 58
    Hammer, SM, Squires, KM, Hughes, MD et al. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. N Engl J Med 1997; 337, 725 733.
  • 59
    Cameron, DW, Heath-Chiozzi, M, Danner, S et al. Randomised placebo-controlled trial of ritonavir in advanced HIV-1 disease. Lancet 1998; 351, 543 549.
  • 60
    Haubrich, R, Lalezari, J, Follansbee, SE et al. Improved survival and reduced clinical progression in HIV-infected patients with advanced disease treated with saquinavir plus zalcitabine. Antiviral Ther 1998; 3, 33 42.
  • 61
    Kirk, O, Katzenstein, TL, Gerstoft, J et al. Combination therapy containing ritonavir plus saquinavir has superior short-term antiretroviral efficacy: a randomized trial. AIDS 1999; 13, F9 F16.
  • 62
    Moyle, G & on behalf of the Spice Study team. Study of protease inhibitor combination in Europe (Spice) ; saquinavir soft gelatin capsule (SQV-SGC) plus nelfinavir (NFV) in HIV-infected individuals. 12th World AIDS Conference. Geneva. (abstract 12222)., 1998.
  • 63
    D'aquila, RT, Hughes, MD, Johnson, VA et al. Nevirapine, zidovudine, and didanosine compared with zidovudine and didanosine in patients with HIV-1 infection. A randomized, double-blind, placebo-controlled trial. National Institute of Allergy and Infectious Diseases AIDS Clinical Trials Group Protocol, 241 Investigators. Ann Intern Med 1996; 124, 1019 1030.
  • 64
    Moyle, GJ, Baldwin, C, Dent, N et al. Management of protease inhibitor (PI) -associated lipodystrophy by substitution with efavirenz in virologically controlled HIV-infected persons. 39th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco. (abstract 2064)., 1999.
  • 65
    Montaner, JSG, Reiss, P, Cooper, D et al. A randomized, double-blind trial comparing combinations of nevirapine, didanosine, and zidovudine for HIV-infected patients. The INCAS Trial. JAMA 1998; 279, 930 937.
  • 66
    Murphy, RL, Katlama, C, Johnson, V et al. The Atlantic Study: a randomized, open-label trial comparing two protease inhibitor (PI) -sparing antiretroviral strategies versus a standard PI-containing regimen, 48 week data. 39th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco. (abstract LB-22)., 1999.
  • 67
    Kaspar, R. Prednisone during the induction phase of nevirapine therapy appears to reduce the incidence of nevirapine-associated rash. 38th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego. (abstract I-64)., 1998.
  • 68
    Hawkins, D, Wood, R, Horban, A et al. Final 54-week analysis of a study comparing the safety and antiretroviral activity of dual nucleoside analogues plus placebo versus dual nucleoside analogues plus delavirdine (DLV) in an advanced, treatment-naive HIV-infected population. 7th European Conference on Clinical Aspects and Treatment of HIV Infection, Lisbon. (abstract 550)., 1999.
  • 69
    Hawkins, D, Wood, R, Horban, A et al. Final 54-week analysis of a study comparing the safety and antiretroviral activity of dual nucleoside analogues plus placebo versus dual nucleoside analogues plus delavirdine (DLV) in an advanced, nucleoside analogue-experienced HIV-infected population. 7th European Conference on Clinical Aspects and Treatment of HIV Infection, Lisbon. (abstract 552)., 1999.
  • 70
    Larder, BA. 3′-Azido-3′-deoxythymidine resistance suppressed by a mutation conferring human immunodeficiency virus type 1 resistance to nonnucleoside reverse transcriptase inhibitors. Antimicrob Agents Chemother 1992; 36, 2664 2669.
  • 71
    Kaufmann, D, Pantaleo, G, Sudre, P, Telenti, A. for the Swiss HIV Cohort Study. CD4-cell counts in HIV-1-infected individuals remaining viraemic with highly active antiretroviral therapy (HAART). Lancet 1998; 351, 723 724.
  • 72
    Caesar Coordinating Committee. Randomised trial of addition of lamivudine or lamivudine plus loviride to zidovudine-containing regimens for patients with HIV-1 infection: the CAESAR trial. Lancet 1997; 349, 1413 1421.
  • 73
    Larder, BA, Kemp, SD, Harrigan, PR. Potential mechanism for sustained antiretroviral efficacy of AZT−3TC combination therapy. Science 1995; 269, 696 699.
  • 74
    Staszewski, S, Katlama, C, Herrer, T et al. Long-term, 72 week experience with a triple reverse transcriptase nucleoside therapy regimen containing abacavir (1592, ABC): a follow-up of protocol CNA2002. AIDS 1998; 12, S34((Abstract) 69).
  • 75
    Staszewski, S, Keiser, P, Gathe, J et al. Comparison of antiviral response with abacavir/combivir to indinavir/combivir in therapy-naïve adults at 48 weeks (CNA3005). 39th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco. (abstract 505)., 1999.
  • 76
    Havlir, DV, Friedland, G, Pollard, R et al. Combination zidovudine and stavudine therapy versus other nucleosides: report of two randomized trials (ACTG 290 and 298). 5th Conference on retroviruses and Opportunistic Infections, Chicago. ((Abstract) 2)., 1998.
  • 77
    Brinkman, K, Ter Hofstede, HJ, Burger, DM et al. Adverse effects of reverse transcriptase inhibitors: mitochondrial toxicity as common pathway [editorial]. AIDS 1998; 12, 1735 1744.
  • 78
    Fouty, B, Frerman, F, Reves, R. Riboflavin to treat nucleoside analogue-induced lactic acidosis. Lancet 1998; 352, 291 292.
  • 79
    Lori, F, Malykh, A, Cara, A et al. Hydroxyurea as an inhibitor of human immunodeficiency virus Type-1 replication. Science 1994; 266, 801 805.
  • 80
    De Boer, RJ, Boucher, CAB, Perelson, AS. Target cell availability and the successful suppression of HIV by hydroxyurea and didanosine. AIDS 1998; 12, 1567 1570.
  • 81
    De Antoni, A, Foli, A, Lisziewicz, J, Lori, F. Mutations in the pol gene of human immunodeficiency virus type 1 in infected patients receiving didanosine and hydroxyurea combination therapy. J Infect Dis 1997; 176, 899 903.
  • 82
    Rossero, R, Mckinsey, D, Green, S et al. Open label combination therapy with stavudine, didanosine and hydroxyurea in nucleoside experienced HIV-1 infected patients. 5th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract 653)., 1998.
  • 83
    Margolis, D, Heredia, A, Gaywee, J et al. Abacavir and mycophenolic acid, an inhibitor of inosine monophosphate dehydrogenase, have profound and synergistic anti-HIV activity. J Acquir Immune Defic Syndr 1999; 21, 362 370.
  • 84
    Petersen, A, Antunes, F, Arasteh, KN et al. A comparison of the long-term antiviral efficacy of bid and tid dosing of nelfinavir in combination with stavudine (d4T) and lamivudine (3TC) beyond 48 weeks. Seventh European Conference on Clinical Aspects and Treatment of HIV-Infection, Lisbon, October. (abstract 205)., 1999.
  • 85
    Barry, MG, Merry, C, Lloyd, J et al. Variability in trough plasma saquinavir concentrations in HIV patients – a case for therapeutic drug monitoring [letter]. Br J Clin Pharmacol 1998; 45, 501 502.
  • 86
    Back, D, Eagling, V, Profit, L. Saquinavir absorption: role of P-glycoprotein. AIDS.; 12 (abstract P53)., 1998.
  • 87
    Fletcher, CV, Kakuda, TN, Anderson, PL et al. Viral dynamics of concentration-targeted (C) vs. standard-dose (S) therapy with zidovudine (ZDV), lamivudine (3TC), and indinavir (IDV). 39th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, September. (abstract 322)., 1999.
  • 88
    Havlir, DV, Marschener, IC, Hirsch, MS et al. Maintenance antiretroviral therapies in HIV-infected subjects with undetectable plasma HIV RNA after triple-drug therapy. N Engl J Med 1998; 339, 1261 1268.
  • 89
    Ruiz, L, Van Lunzen, J, Arno, A et al. Protease inhibitor-containing regimens compared with nucleoside analogues alone in the suppression of persistent HIV-1 replication in lymphoid tissue. AIDS 1999; 13, F1 F8.
  • 90
    Taylor, S, Back, DJ, Workman, J et al. Poor penetration of the male genital tract by HIV-1 protease inhibitors. AIDS 1999; 13, 859 860.
  • 91
    Danner, SA, Carr, A, Leonard, JM et al. A short-term study of the safety, pharmacokinetics, and efficacy of ritonavir, an inhibitor of HIV-1 protease. European–Australian Collaborative Ritonavir Study Group. N Engl J Med 1995; 333, 1528 1533.
  • 92
    Di Perri, G, Del Bravo, P, Concia, E. HIV-protease inhibitors. N Engl J Med 1998; 339, 773 774.
  • 93
    Nelson, MR, Moyle, G, Newell, A. The incidence of hyperlipidaemia with protease inhibitors (PI). 38th Interscience Conference in Antimicrobial Agents and Chemotherapy, San Diego,1998.
  • 94
    Carr, A, Samaras, K, Chisholm, DJ, Cooper, DA. Pathogenesis of HIV-1-protease inhibitor-associated peripheral lipodystrophy, hyperlipidaemia, and insulin resistance. Lancet 1998; 352, 1881 1883.
  • 95
    Carr, A, Samaras, K, Burton, S et al. A syndrome of peripheral lipodystrophy, hyperlipidaemia and insulin resistance in patients receiving HIV protease inhibitors. AIDS 1998; 12, F51 F58.
  • 96
    Henry, K, Melroe, H, Huebsch, J et al. Severe premature coronary artery disease with protease inhibitors. Lancet 1998; 351, 1328.
  • 97
    Madge, S, Kinloch -d, E-Loes, S, Mercey, D, Johnson, MA, Weller, IV. Lipodystrophy in patients naive to HIV protease inhibitors. AIDS 1999; 13, 735 737.
  • 98
    Dong, K, Flynn, MM, Dickinson, BP et al. Changes in body habitus in HIV+ women after initiation of protease inhibitor therapy. 12th World AIDS Conference, Geneva. (abstract 12373)., 1998.
  • 99
    Lo, J, Mulligan, K, Tai, V, Algren, H, Schambelen, M. ‘Buffalo hump’ in men with HIV-1 infection. Lancet 1998; 351, 867 870.
  • 100
    Veny, A, Bonjoch, A, Romeu, J et al. Cumulative risk for developing protease inhibitor-associated lipodystrophy (PI-AL) in HIV-infected patients. 38th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego. (abstract I-92)., 1998.
  • 101
    Miller, K, Jones, E, Yanovski, J et al. Visceral abdominal-fat accumulation associated with use of indinavir. Lancet 1998; 351, 871 875.
  • 102
    Brinkman, K, Smeitink, JA, Romijn, JA, Reiss, P. Mitochondrial toxicity induced by nucleoside-analogue reverse-transcriptase inhibitors is a key factor in the pathogenesis of antiretroviral-therapy-related lipodystrophy. Lancet 1999; 354, 1112 1115.
  • 103
    Saint-Marc, T & Touraine, JL. Effects of metformin on insulin resistance and central adiposity in patients receiving effective protease inhibitor (PI) therapy. 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract 672)., 1999.
  • 104
    Torres, R & Unger, K. The effect of recombinant human growth hormone on protease-inhibitor-associated fat maldistribution. 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract 675)., 1999.
  • 105
    Ruiz, L, Bonjoch, A, Paredes, R et al. A multi-center, randomized, open-label comparative trial of the clinical benefit of switching the protease inhibitor by nevirapine in HAART-experienced patients suffering lipodystrophy. 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract LB-14)., 1999.
  • 106
    Yee, TT, Amrolia, PJ, Lee, CA et al. Protease inhibitors and unusual bleeding in haemophiliacs. Haemophilia 1997; 3, 219 221.
  • 107
    Wilde, JT, Lee, CA, Collins, P et al. Increased bleeding associated with PI therapy in HIV-positive patients with bleeding disorders. Br J Haematol 1999; 107, 556 559.
  • 108
    Pilcher, CD, Miller, WC, Beatty, ZA, Eron, JJ. Detectable HIV-1 RNA at levels below quantifiable limits by amplicor HIV-1 monitor is associated with virologic relapse on antiretroviral therapy. AIDS 1999; 13, 1337 1342.
  • 109
    Gulick, RM, Mellors, JW, Havlir, D et al. Simultaneous vs. sequential initiation of therapy with indinavir, zidovudine and lamivudine for HIV-1 infection. JAMA 1998; 280, 35 41.
  • 110
    Goebel, F & the Avanti Study Group. Avanti, 2. A randomised, double-blind, comparative trial to evaluate combination antiretroviral regimens for the treatment of HIV-1 infection: AZT/3TC vs. AZT/3TC/Indinavir in antiretroviral naive patients. 6th European Conference on Clinical Aspects and Treatment of HIV Infection, Hamburg. (abstract 211)., 1997.
  • 111
    Riddler, S, Kahn, J, Hicks, C et al. Durable anti-HIV-1 activity (72 weeks) and tolerability for efavirenz (DMP 266) in combination with indinavir (IDV) [DMP 266–003, Cohort IV]. 12th World AIDS Conference, Geneva. (abstract 12359)., 1998.
  • 112
    Lorenzi, P, Opravil, M, Hirschel, B et al. Impact of drug resistance mutations on virologic response to salvage therapy. Swiss HIV Cohort Study. AIDS 1999; 13, F17 F21.
  • 113
    Youle, M, Mocroft, A, Johnson, M et al. Surrogate marker responses to multidrug combination hydroxyurea, efavirenz, double protease inhibitors and analogues in protease inhibitor failures. 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract 400)., 1999.
  • 114
    Deeks, SG, Hecht, FM, Swanson, M et al. HIV RNA and CD4 cell count response to protease inhibitor therapy in an urban AIDS clinic: response to both initial and salvage therapy. AIDS 1999; 13, F35 F43.
  • 115
    Tebas, P, Patick, AK, Kane, EM et al. Virologic responses to a ritonavir – saquinavir-containing regimen in patients who had previously failed nelfinavir. Aids 1999; 13, F23 F28.
  • 116
    Eron, J, Falloon, J, Masur, H et al. Activity of Combination Abacavir/Amprenavir/Efavirenz Therapy in HIV-1 Infected Subjects Failing Their Current Protease Inhibitor Containing Regimen 1998; 12, S13((Abstract) OP52).
  • 117
    Workman, C, Mussen, R, Sullivan, J. Salvage therapy using six drugs in heavily pretreated patients. 5th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract 426)., 1998.
  • 118
    Garcia, F, Plana, M, Vidal, C et al. Dynamics of viral load rebound and immunological changes after stopping effective antiretroviral therapy. Aids 1999; 13, F79 F86.
  • 119
    Ledergerber, B, Egger, M, Opravil, M et al. Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Lancet 1999; 353, 863 868.
  • 120
    Deeks, SG, Parkin, N, Petropoulos, CJ et al. Correlation of baseline phenotypic drug susceptibility with 16 week virologic response in a pilot combination therapy study in HIV-infected patients who failed indinavir therapy. Antiviral Ther 1998; 3, 36((Abstract) 53).
  • 121
    Patick, A, Zhang, M, Hertogs, K et al. Correlation of virological response with genotype and phenotype of plasma HIV-1 variants in patients treated with nelfinavir in the US expanded access program. Antiviral Ther 1998; 3, 39((Abstract) 57).
  • 122
    Harrigan, PR, Montaner, JS, Hogg, RS et al. Baseline resistance profile predicts response to ritonavir/saquinavir therapy in a community setting. Antiviral Ther 1998; 3, 38((Abstract) 55).
  • 123
    Zolopa, AR, Shafer, RW, Warford, A et al. Predictors of antiviral response to saquinavir/ritonavir therapy in a clinical cohort who have failed prior protease inhibitors: a comparison of clinical characteristics, antiretroviral drug history and HIV genotype. Antiviral Ther 1998; 3, 37((Abstract) 54).
  • 124
    Baxter, JD, Mayers, DL, Wentworth, DN, Neaton, JD, Merigan, TC & the CPCRA, 046 Study Team. . A pilot study of the short-term effects of antiretroviral management based on plasma genotypic antiretroviral resistance testing (GART) in patients failing antiretroviral therapy. 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract LB8)., 1999.
  • 125
    Durant, J, Clevenbergh, P, Halfon, P et al. Drug-resistance genotyping in HIV-1 therapy: the VIRADAPT randomised controlled trial. Lancet 1999; 353, 2195 2199.
  • 126
    Stuyver, L, Wyseur, A, Rombout, A et al. Line probe assay for rapid detection of drug-selected mutations in the Human Immunodeficiency Virus Type 1 reverse transcriptase gene. Antimicrob Agents Chemother 1997; 41, 284 291.
  • 127
    Arnold, C, Barlow, KL, Kaye, S et al. HIV type 1 sequence subtype G transmission from mother to infant: failure of variant sequence species to amplify in the Roche Amplicor Test. AIDS Res Hum Retroviruses 1995; 11, 999 1001.
  • 128
    Perrin, L. Transmission of drug-resistant HIV. 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract S35)., 1999.
  • 129
    Little, S, Daar, E, Keiser, P et al. The spectrum and frequency of reduced antiretroviral drug susceptibility with primary HIV infection in the United States. 6th Conference on Retroviruses and Opportunistic Infections, Chicago. (abstract LB-10)., 1999.
  • 130
    Taylor, GP, Lyall, EGH, Mercey, D et al. British HIV Association Guidelines for Prescribing Antiretroviral Therapy in Pregnancy 1998. Sex Transm Inf 1999; 75, 90 97.
  • 131
    Expert Advisory Group on, AIDS. Guidelines on Post-Exposure Prophylaxis for Health Care Workers Occupationally Exposed to HIV. London, Department of Health, 1997.

11.0 Conflict of interest

  1. Top of page
  2. 1.0 Introduction
  3. 2.0 When to start treatment
  4. 3.0 What to start with
  5. 4.0 Issues concerning protease inhibitor use
  6. 5.0 Changing therapy on first virological failure
  7. 6.0 Therapy after more than one previous failure (‘salvage’ therapy)
  8. 7.0 Resistance testing
  9. 8.0 Issues not addressed in these guidelines
  10. 9.0 BHIVA/IAS–USA comparisons
  11. References
  12. 11.0 Conflict of interest

Dr Anton Pozniak has received unrestricted educational and travel grants from Glaxo Wellcome, Bristol-Myers Squibb, Roche, Boehringer Ingelheim, DuPont, Merck Sharp & Dohme, Gilead and Abbott. He is a member of the UK Medical Advisory Boards of Boehringer Ingelheim, DuPont and Abbott.

Professor Brian Gazzard has received research grants, unrestricted educational and travel grants and been a member of the international advisory boards of Glaxo Wellcome, Bristol-Myers Squibb, Roche, Boehringer Ingelheim, DuPont, Merck Sharp & Dohme, Gilead and Abbott.

Dr Duncan Churchill has received research grants and unrestricted travel grants from Roche, Glaxo Wellcome, Bristol-Myers Squibb and Abbott. He is on the UK Medical Advisory Boards of Glaxo Wellcome, Bristol-Myers Squibb, Boehringer Ingelheim, DuPont and Merck Sharp & Dohme.

Dr Margaret Johnson has received research and travel grants from Glaxo Wellcome, Bristol-Myers Squibb, Roche, Boehringer Ingelheim, DuPont, Merck Sharp & Dohme, Gilead and Abbott, and is a member of the UK Medical Advisory Boards of Abbott, Bristol-Myers Squibb, Merck Sharp & Dohme and DuPont.

Dr Ian Williams has received funding for clinical grants from Roche, Agouron and Gilead. He is on the Medical Advisory Boards of Boehringer Ingelheim and Glaxo Wellcome.

Dr James Deutsch has no conflict of interest.

Dr Alison Gray has received grants for conference attendance and/or honoraria for speaking at meetings from Glaxo Wellcome, Bristol-Myers Squibb and Merck Sharp & Dohme.

Dr Deenan Pillay's laboratory has received research and travel grants and consultancy fees from Glaxo Wellcome, DuPont, Pharmacia & Upjohn and Roche. He is a member of Medical Advisory Boards of Glaxo Wellcome, DuPont and Merck Sharp & Dohme.

Dr Martin Wiselka has received research grants from Roche and Pharmacia & Upjohn and unrestricted travel and educational grants from Roche, Glaxo Wellcome, Bristol-Myers Squibb, Abbott, Boehringer Ingelheim and DuPont. He has served on the UK Medical Advisory Board of Bristol-Myers Squibb.

Dr Graeme Moyle has received funding for clinical grants from Agouron, Bristol-Myers Squibb, Glaxo Wellcome, Roche, Boehringer Ingelheim, DuPont and Pharmacia & Upjohn. He holds consulting agreements with Abbott, DuPont, Bristol-Myers Squibb, Roche and Pharmacia & Upjohn.

Professor Christine Lee has received educational grants for research from Glaxo Wellcome.