Intended for healthcare professionals

Clinical Review

Antiretroviral treatment of HIV infected adults

BMJ 2006; 332 doi: (Published 22 June 2006) Cite this as: BMJ 2006;332:1489
  1. Steven G Deeks, associate professor of medicine (sdeeks{at}
  1. 1 University of California, San Francisco and San Francisco General Hospital, San Francisco, CA 94110, USA
  1. Correspondence to: S G Deeks
  • Accepted 18 May 2006


It has been about 10 years since the first report that three drug combination antiretroviral therapy can durably suppress HIV replication.1 Subsequent studies have confirmed that when used appropriately highly active antiretroviral therapy (see box 1) can suppress viral replication to such low levels that the virus is unable to generate drug resistance mutations. Theoretically, once this level of viral suppression is achieved, treatment should work indefinitely, and the long term risk of morbidity and morality related to HIV associated immunodeficiency becomes negligible. Experience to date suggests that lifelong suppression of HIV is feasible.

This review is aimed at informing clinicians about the current management of HIV infection. Authoritative and continuously updated reviews are available on the web (for example, the US Department of Health and Human Services treatment guidelines at; this review does not attempt to exhaustively summarise the literature or to provide guidance to clinicians with expertise in HIV. Rather, I summarise those issues that are likely to confront clinicians, including clinicians who do not routinely treat people infected with HIV.

Sources and selection criteria

I searched PubMed databases for studies pertaining to antiretroviral therapy and its complications, lipodystrophy and lipoatrophy, and immune reconstitution. I also consulted recently published national and international treatment guidelines and considered unpublished data presented at international meetings.

What is the goal of therapy?

HIV seems to be designed to mutate and evolve as rapidly as possible. This evolutionary capacity is the result of at least three properties: an extremely high rate of virus turnover (at least 1010 new virions are produced per day), a high mutation rate (about one mutation per new virion), and an impressive capacity of many HIV proteins to function in the face of multiple amino acid changes (for example, at least 33 of the 99 amino acids in HIV protease can mutate and cause drug resistance without lethal loss of function).

Theoretical considerations predict that the only manner in which to ensure long term effectiveness of therapy is to durably suppress replication to below the threshold necessary for systemic HIV evolution. Clinical experience suggests that this threshold is likely at or near the level of HIV RNA quantification with most viral load assays (about 50-200 copies of RNA/ml). Although the goal of any treatment is to restore health and to prolong life, from a practical perspective the goal of antiretroviral treatment is to achieve and maintain undetectable plasma levels of HIV RNA. This goal determines almost all of the critical decisions.

Summary points

Resistance to antiretroviral drugs can emerge quickly (in as short as one week for some drugs)

Once resistance to antiretroviral drugs is established it persists indefinitely

Drug resistance testing is recommended before starting therapy as about 10% of newly infected patients harbour drug resistant variants

Antiretroviral drugs can initially result in paradoxical worsening of some pre-existing conditions (“immune reconstitution syndrome”)

The addition of some commonly used drugs to a pre-existing antiretroviral regimen can result in reduced antiretroviral drug levels, leading to suboptimal drug exposures and virological failure

Interruption of antiretroviral therapy is commonly associated with rapid declines in CD4 T cell counts, an increased risk of short term complications, and increased risk of developing drug resistant variants

All nucleoside and nucleotide analogues can cause lactic acidosis or severe hepatic steatosis

Most nucleoside reverse transcriptase inhibitors may need to be dose adjusted in patients with impaired renal function

All non-nucleoside reverse transcriptase inhibitors and protease inhibitors need to be used with care in patients with significant hepatic disease

When to start therapy

Perhaps the single most important question facing a treatment naive patient is when to start antiretroviral therapy. Ideally such a decision would be informed by a randomised clinical study comparing immediate treatment with deferred treatment. No such study has ever been done and it is unlikely that such a study is even feasible (the duration of follow-up necessary to detect a clinical benefit of one strategy over the other would probably take many years and many thousands of patients).

Data from several well carried out prospective observational studies have consistently shown that the pretreatment CD4 T cell count rather than the plasma HIV RNA level is the single best predictor of morbidity and mortality in patients starting therapy. These same studies have shown that deferring therapy until CD4 T cell counts are less than 200×106/l is associated with increased risk of progressing to AIDS or death, compared with starting therapy with a CD4 T cell count above this threshold.24

The clinical benefit of starting therapy at higher CD4 T cell counts has not been established. However, from a pathogenesis perspective, untreated HIV infection is associated over time with many untoward effects, including an increased risk for serious complications such as lymphoma and perhaps tuberculosis5; a progressive and probably irreversible loss of immunological function6; an increasing diversity of the HIV quasi-species, which may be associated with the de novo emergence of drug resistance mutations; increased risk for the development of potentially more virulent viruses; and progressive loss of neurological function. HIV is a virulent virus that causes clear if not dramatic harm throughout its course.

Despite these well accepted risks of untreated HIV infection, current guidelines generally do not recommend early intervention, in part because the toxicity associated with antiretroviral drugs may be greater than the risks of HIV replication. As discussed in detail below, antiretroviral therapy can be associated with a variety of short term and long term toxicities, including peripheral neuropathy and potentially disfiguring redistributions of body fat (lipodystrophy). Also, the ongoing requirement for strict adherence to a daily treatment regimen can negatively affect quality of life, at least for some.

Given the lack of clarity about the risks and benefits of starting therapy in patients with CD4 T cell counts greater than 200×106/l, most guidelines are conservative and generally suggest that patients only consider therapy.79 However, as the safety and tolerability of first line regimens improve and as the regimens become increasingly convenient, it is expected that future guidelines will recommend treatment earlier and earlier in the clinical course of HIV disease.


A critical component in any decision to start therapy is the patient's perceived ability to adhere to drugs indefinitely. Starting therapy in patients who are not fully committed may lead to poor adherence and virological failure. Since resistance is likely to emerge in such patients, it is often better to defer therapy until access and adherence to drugs can be guaranteed.7 These same considerations argue for stopping therapy in patients who become non-adherent; however, any decision to stop therapy to prevent emergence of drug resistance must also consider the rapid immunological progression that can occur when interrupting a partially effective regimen.10

What to start

In contrast to the “when to start” question, the “what to start” question is based on a growing number of rigorous, well carried out randomised clinical studies. Collectively, these studies support the recommendations that a first line regimen should include a “backbone” of two nucleoside reverse transcriptase inhibitors and a third “anchor” drug that can be either a non-nucleoside reverse transcriptase inhibitor or a ritonavir boosted protease inhibitor (boxes 1-4).79 Three options are generally recommended for the nucleoside analogue backbone, all available as fixed dose combination pills: once daily tenofovir plus emtricitabine, once daily abacavir plus lamivudine, or twice daily zidovudine plus lamivudine (although the latter may no longer be preferred given the association of zidovudine with lipoatrophy and anaemia).11 12

Box 1 Definitions of terms related to HIV medicine

Highly active antiretroviral therapy (HAART)

Any regimen that is expected to result in durable viral suppression

This is generally a three drug regimen including two nucleoside reverse transcriptase inhibitors and either a non-nucleoside reverse transcriptase inhibitor or a protease inhibitor

Salvage therapy

Any regimen designed to suppress drug resistant HIV in patients who have previously exhibited virological failure on at least two regimens

Virological failure

Incomplete viral suppression with highly active antiretroviral therapy, generally defined as persistent plasma HIV RNA levels greater than 200 copies/ml after four to six months of therapy

Immune reconstitution syndrome

Paradoxical worsening of pre-existing condition soon after the initiation of an effective regimen

Also referred to as immune reconstitution inflammatory syndrome (IRIS)

Nucleoside reverse transcriptase inhibitors (NRTI)

Also referred to as nucleoside analogues—this classification generally includes the closely related nucleotide analogues such as tenofovir

These drugs prevent HIV DNA synthesis and often form the backbone of all treatment regimens

Non-nucleoside reverse transcriptase inhibitors (NNRTI)

These drugs are potent inhibitors of DNA synthesis and often are the anchor drug for initial regimens

They have fewer long term toxicities than other drug classes

When used inappropriately, these drugs can select for high level resistance in days to weeks

Protease inhibitors

These drugs prevent cleavage of HIV proteins, thereby preventing viral maturation

These drugs may be used as an anchor drug in first line regimens but are more often used in second line and salvage regimens

Entry inhibitors

These drugs prevent entry of HIV into CD4 cells

Enfuvirtide was the first entry inhibitor approved for use—this drug is often used in salvage therapy given its costs and need for parental administration

Other entry inhibitors include those that block HIV binding to CCR5 (“R5 inhibitors”)

A growing and impressive database supports the use of efavirenz as the preferred first line anchor drug. Efavirenz is a highly effective and generally well tolerated non-nucleoside reverse transcriptase inhibitor that is taken once daily.1315Because of a potential for neural tube defects, efavirenz should be used with caution in women of childbearing age. Also, efavirenz causes short term side effects of the central nervous system and should be used with caution in patients with severe psychiatric illnesses or active substance misuse. Nevirapine is a reasonable alternative for efavirenz but should not be used in women with a CD4 T cell count greater than 250×106/l or in men with a CD4 T cell count greater than 400×106/l owing to greatly increased risk of severe hepatotoxicity at higher CD4 T cell counts.

Box 2 Nucleoside and nucleotide reverse transcriptase inhibitors


Associated with potentially life threatening hypersensitivity reactions in about 5% of people

Didanosine (ddI)

Must be given in fasting state

Associated with peripheral neuropathy and pancreatitis

Avoid or use with caution in regimens containing tenofovir

Emtricitabine (FTC)

Similar characteristics as lamivudine

May be associated with reversible skin pigmentation

Generally used as coformulation with tenofovir

Lamivudine (3TC)

Cornerstone of most first line regimens owing to its well established safety and potency

Often used in “salvage” because resistance to this drug enhances activity of zidovudine, stavudine, and tenofovir, and because resistance mutations reduce fitness


Popular first line drug owing to its well established tolerability and potency

May cause renal dysfunction when used in combination with other potentially nephrotoxic drugs or in patients with other risk factors for renal disease

Stavudine (d4T)

Associated with development of lipoatrophy

Other toxicities include peripheral neuropathy, pancreatitis, and high risk of lactic acidosis (compared with other nucleotide reverse transcriptase inhibitors)

Zidovudine (ZDV, AZT)

Associated with anaemia and neutropenia, particularly in patients with advanced disease

May cause lipoatrophy

All nucleoside and nucleotide analogues have been associated with lactic acidosis and hepatic steatosis

Protease inhibitors may also be used as the anchor drug. Ten protease inhibitors are currently widely available (box 4). Most are rapidly metabolised by the cytochrome P-450 metabolic system. As ritonavir is a potent inhibitor of cytochrome P-450 CYP3A it is often co-administered with other protease inhibitors. The ritonavir mediated increase in serum half lives of the co-administered protease inhibitor results in both less frequent dosing schedules and improved long term efficacy. On the basis of randomised clinical trials, ritonavir plus lopinavir is currently the preferred option for a first line protease inhibitor.16 However, ritonavir plus atazanavir is often used in clinical practice as it can be given once daily and seems to be well tolerated and effective.

Triple nucleoside reverse transcriptase inhibitor regimens have been used as first line regimens in the past. Although simple to administer, these combinations seem to be less effective than the preferred approaches outlined above.14

Population specific recommendations

Uncertainty remains about the optimal management of some patient populations, including women (both pregnant and non-pregnant), patients co-infected with hepatitis C or tuberculosis, patients with advanced immunodeficiency, and patients with acute HIV infection. Because drug associated toxicity clearly depends in part on sex, it is reasonable to assume that the optimal management of women may differ from that of men. This is particularly relevant to efavirenz, which should not be used in women who may become pregnant owing to the risk of neural tube defects in the developing fetus. Details of the management of HIV during pregnancy can be found in the latest version of the Public Health Service Task Force recommendations for the use of antiretroviral drugs in pregnant women infected with HIV-1 for maternal health and interventions to reduce perinatal HIV-1 transmission in the United States ( The management of other patient populations can be found at the DHHS guidelines for the use of antiretroviral agents in HIV-1 infected adults and adolescents ( as well as in other published guidelines.79

Transmitted drug resistance and its effect on what to start

One of the public health consequences of widespread access to drugs is the increased incidence of transmitted drug resistance mutations. In most regions where treatment is widely available, about 10% of newly infected patients harbour virus that is at least partially resistant to one or more antiretroviral drugs.1720 Since resistance to one drug risks rapid failure of the entire regimen, genotypic resistance testing is recommended before starting therapy. Details on the interpretation and use of resistance testing can be found in published guidelines,21 22 reviews,23 24 and on the web (,

When to switch therapy

There are at least three major reasons to modify or switch therapy once it has been started: drug toxicity; availability of new agents with improved tolerability, safety, or efficacy; and virological failure. Modifying therapy for the first two reasons is typically straightforward, although clinicians need to realise that even in patients with undetectable HIV RNA levels any pre-existing drug resistant virus will likely persist in cellular or tissue reservoirs indefinitely, and that any new treatment regimen needs to remain effective against such archived viruses.25

Box 3 Non-nucleoside reverse transcriptase inhibitors


Associated with rash

Rarely used owing to requirement for three times daily dosing


Potent and highly effective when used in treatment naive patients

Associated with central nervous system toxicity that typically resolves by week 4

Teratogenic in humans


Potent and effective when used in treatment naive patients

Associated with rash (and potentially Stevens-Johnson syndrome)

Also associated with severe hepatotoxicity in women with CD4 T cells counts greater than 250×106/l and in men with CD4 T cell counts greater than 400×106/l

Management of virological failure

Although combination antiretroviral therapy is highly effective, a major proportion of patients in clinical practice is unable to achieve and maintain an undetectable HIV RNA level.26 At least two broadly defined types of patients have experienced virological failure. The first are those who were exposed to sequential suboptimal regimens since the early era of highly active antiretroviral therapy. Such patients often received nucleoside reverse transcriptase inhibitors in the early 1990s and later started a protease inhibitor and then eventually a non-nucleoside reverse transcriptase based regimen (with “recycled” or partially effective nucleoside reverse transcriptase inhibitors used in each new regimen). This exposure to sequential suboptimal regimens has led to a large cohort of people with highly resistant virus. Although such patients are no longer being created de novo, the number of patients who are in the state of chronic virological failure may be stable, in part because the rate of disease progression in these patients has always been low.27 The mechanism for this delayed disease progression associated with drug resistant virus is complicated and beyond the scope of this review.28

Box 4 Protease and fusion inhibitors


May be administered once daily with or without ritonavir

Popular first line option owing to once daily dosing and to limited effect on lipid levels

May cause elevations in indirect (unconjugated) bilirubin


Must be co-administered with ritonavir

Highly effective for drug resistant HIV

Generally well tolerated


May be administered with or without ritonavir

Associated with gastrointestinal symptoms and rash


May be administered with or without ritonavir

Must be given after fasting every eight hours unless co-administered with ritonavir

May cause kidney stones


Only available as a capsule or tablet co-formulated with ritonavir

Preferred first line option owing to robust nature of clinical trial data

Causes gastrointestinal symptoms and lipid abnormalities


Can not be administered with ritonavir

Less effective than ritonavir boosted protease inhibitor options

Causes diarrhoea


Rarely used at anti-HIV therapeutic doses owing to unfavourable adverse event profile

Potent inhibitor of certain P-450 isoenzymes and drug transporters

Often used as a pharmacological enhancer to boost exposure of others drugs metabolised by P-450 CYP3A

All ritonavir based regimens cause hyperlipidaemia, gastrointestinal distress, and perhaps insulin resistance


Soft gel formulation recently taken off market

Hard gel formulation must be co-administered with ritonavir


Must be co-administered with ritonavir

Highly effective for drug resistant HIV

Often only used in patients with limited therapeutic options owing to risk for severe hepatotoxicity


Fusion inhibitor

Must be given parenterally

Associated with severe local site reactions

Generally used in patients with highly resistant HIV

The second group of patients with virological failure is those who have failed optimal treatment strategies—that is, treatment naive patients who were prescribed a potentially fully suppressive regimen. Most of these individuals presumably failed therapy owing to non-adherence. The prevalence and clinical course of such patients remains unknown.

Management of drug resistant HIV

The management of patients exhibiting incomplete viral suppression on their first highly active antiretroviral therapy regimen is not controversial. The reason for failure needs to be defined and corrected. Given the potency of current first line regimens, the major reason for failure is non-adherence. A resistance test should be considered21 and a new regimen started as soon as possible (see for a detailed summary on how to interpret data on genotypic resistance).29

The management of patients with highly active antiretroviral therapy failure is far more complicated. Such patients have three therapeutic options: switch to a new regimen on the basis of resistance testing (“salvage therapy”), interrupt therapy and then resume therapy once options become available, or remain on a stable regimen.28 In the past, premature switching to a new regimen often failed to achieve complete viral suppression, in part because only one new drug emerged at any given time.30 Fortunately the specialty is now witnessing the emergence of several promising drugs that seem to be both safe and highly effective against drug resistant HIV. These novel drugs include those that prevent integration of HIV DNA into the host genome (“integrase inhibitors”) as well as those that prevent HIV binding to CCR5 (“R5 inhibitors”).

Interrupting therapy is usually associated with rapid loss of CD4 T cells and is not currently recommended.31 32 Thus many patients and their clinicians choose to continue a stable regimen despite the risk of continued viral evolution.

When to stop

Many patients (perhaps the majority) eventually interrupt therapy. Within clinical practice this is often done for one of many reasons, including intolerance to drugs, a desire for a break from taking drugs on a daily basis, a sense of futility (particularly among those with virological failure), a sense that the toxicity of drugs outweighs the benefits, and an inability to access drugs owing to loss of insurance or any means to pay for them.

In addition to these patient specific reasons, intense interest has been shown in the potential therapeutic benefits of controlled or “structured” interruptions to treatment. This interest has focused on two distinct groups: those with well controlled virus receiving therapy and those with poorly controlled highly drug resistant HIV. With regard to those with suppressed viral replication, it has been proposed that brief interruptions of therapy in patients with fully suppressed virus will expose the immune system to HIV antigen in a controlled manner and that this will lead to the expansion of high affinity HIV specific CD4 and CD8 T cells. Done repeatedly this approach could lead to prolonged periods of immunological control when patients are not receiving therapy. This so called “autovaccination” approach has generally not worked, or has only worked for transient periods, perhaps because HIV is able to efficiently and rapidly escape most immunological pressures just as it escapes drug pressure.33 34 With regard to those with non-suppressed virus replication, it has been observed that interruption of therapy in patients with drug resistant HIV leads within a few months to the rapid emergence of a more “fit” wild type HIV.26 Although some studies have suggested a benefit associated with this approach, most studies have either shown no benefit or evidence of significant harm.10 35

Intense interest remains in strategies that seek to maintain clinical and immunological health while minimising exposure to expensive and potentially toxic antiretroviral drugs. One common approach is to interrupt therapy once CD4 T cell counts increase to above a predetermined level (for example, 350×106/l) and then resume therapy once cell counts decline to below a lower threshold (for example, 250×106/l). One large international randomised study of continuous compared with CD4 T cell count driven intermittent therapy was stopped early because of an excessive number of clinical events in those interrupting therapy.36 These studies, plus our increasing understanding of HIV pathogenesis, indicate that the risk of interrupting therapy often outweighs the risk of continuing therapy. Any decision to interrupt therapy needs to be made in the context of these studies and should only be considered in those with a strong reason for not remaining on therapy.

Risk of resistance during interruption of treatment and how to stop drugs

Any clinician may confront a situation in which treatment needs to be discontinued for medical reasons. Such reasons include important drug associated toxicity or the emergence of a comorbid condition preventing strict adherence to treatment, such as substance misuse or surgery. The primary short term risk associated with an interruption is the de novo generation of drug resistance. Theoretically, during an interruption the levels of some drugs may persist at concentrations that permit viral replication while still exerting an antiviral effect. This is particularly true for efavirenz and nevirapine, which persist for days to weeks after interruption of treatment. Also, high level resistance to these drugs requires only a single mutation and can therefore emerge rapidly. If a regimen using a non-nucleoside reverse transcriptase inhibitor must be interrupted then clinicians might consider either interrupting the inhibitor a few days earlier than the other drugs or switching it to a protease inhibitor before the interruption. When feasible, any decision to interrupt therapy should be made in consultation with an expert in HIV.

How to manage drug toxicities and other adverse events

While the short term toxicities of each antiretroviral drug are well known (boxes 2-4) the long term toxicities remain poorly described, in part because randomised clinical studies are either short term (24-48 weeks) or become uncontrolled as patients failing placebo switch to open label drug. Many of the most important toxicities of treatment only become evident in long term observational studies, and the quality of such data has left many questions unanswered.37 38

How to manage drug toxicities and other adverse events

In contrast to specifics on when and how to start and switch drugs, it is imperative that clinicians are aware of the presentation and management of antiretroviral toxicities. Those drug associated toxicities that are likely to present to a non-HIV expert are discussed here.

Abacavir hypersensitivity

Abacavir is a potent nucleoside analogue that is commonly used in one of three formulations either as abacavir only or in fixed dose combinations with lamivudine or lamivudine and zidovudine. A small proportion (about 5%) of abacavir treated patients develops a hypersensitivity reaction. Abacavir hypersensitivity is more common in patients carrying the HLAB5701 allele.39

Most symptoms are non-specific and include fever, nausea, abdominal pain, diarrhoea, malaise, and rash. This reaction typically presents during the first six weeks of treatment but may occur after months of exposure to abacavir. Continued administration of abacavir leads to progressive symptoms that only resolve once the drug is discontinued. Subsequent re-exposure to abacavir can lead to an immediate life threatening reaction characterised by hypotension and respiratory failure. As a decision to stop abacavir for toxicity will prevent a patient from ever using this drug again, the management of patients presenting with symptoms consistent with hypersensitivity reaction should be handled by an expert whenever possible. Rechallenge with abacavir in those with previous evidence of any abacavir related toxicity should not be done. Resuming an abacavir based regimen in those who stopped drugs for vague reasons should only be done in consultation with an expert.

Nucleoside analogue associated lactic acidosis

As analogues of endogenous nucleosides, all nucleoside reverse transcriptase inhibitors have the potential of becoming incorporated into human DNA. Although this does not seem to occur to any appreciable degree in chromosomal DNA, it does occur in mitochondrial DNA. As a consequence, long term exposure to nucleoside reverse transcriptase inhibitors is increasingly being associated with mitochondrial dysfunction.40

The Food and Drug Administration issued the following warning about all nucleoside reverse transcriptase inhibitors: “lactic acidosis and severe hepatomegaly with steatosis, including fatal cases, have been reported with the use of nucleoside analogs alone or in combination with other antiretrovirals.” This complication is more common in women and obese people and may be more common with stavudine than with other nucleoside or nucleotide analogues. These drugs should be discontinued in any patient presenting with unexplained lactic acidosis.

Tenofovir associated renal dysfunction

Tenofovir is a potent generally well tolerated and highly effective nucleotide reverse transcriptase inhibitor. Because of its chemical similarities to adefovir and cidofovir—both of which have been associated with major renal toxicity—there have been persistent concerns about the drug's long term renal and metabolic safety. In large randomised clinical trials of tenofovir compared with either stavudine or zidovudine, tenofovir has proved to be as safe if not safer than these other drugs (when given to treatment naive patients who at study entry had normal renal function and who were not receiving any potentially nephrotoxic drugs).12 13 Still, most cohort studies indicate that tenofovir is associated with a consistent but mild decrease in the estimated glomerular filtration rate.12 This risk is greater in those with diabetes, hypertension, lower body weight, lower CD4 T cell counts, and pre-existing renal dysfunction. Renal failure and other metabolic complications such as Fanconi's syndrome have, however, been uncommonly reported with tenofovir. Current guidelines recommend that tenofovir should be dose adjusted or not used in patients with renal impairment. Of note, tenofovir has also been associated with short term reductions in bone mineral density, but no clear increase in bone fractures.13

Nevirapine associated hepatotoxicity

Nevirapine may cause a rash during the fist few weeks of dosing. This rash can be severe and life threatening. Nevirapine is also associated with increased risk of drug associated hepatitis (about 1% to 2% of patients in one study had grade 3 or 4 increase in transaminases).41 For reasons that are unclear the risk of severe hepatoxicity is higher in patients with higher CD4 T cell counts. Nevirapine is therefore not recommended for women with CD4 T cell counts greater than 250×106/l or in men with CD4 T cell counts greater than 400×106/l.

Lipodystrophy and abnormal fat redistribution syndromes

HIV associated lipodystrophy generally refers to a vaguely defined syndrome that may include fat redistribution (lipoatrophy or central fat accumulation, or both); hyperlipidaemia; and insulin resistance or diabetes mellitus. These latter metabolic abnormalities are more common in patients receiving protease inhibitors. Indeed, administration of protease inhibitors to HIV negative people can lead to rapid onset of insulin resistance, decreased glucose tolerance, and hyperlipidaemia.42 43 Some non-nucleoside reverse transcriptase inhibitors and nucleoside reverse transcriptase inhibitors (particularly stavudine) can abnormally affect lipid levels as well.

Although rarely life threatening, treatment associated redistribution of body fat is probably the single most dominant concern among patients. In the absence of treatment, HIV infection is associated with progressive loss of subcutaneous fat (both peripherally and centrally).44 In the presence of drugs, progressive facial and limb lipoatrophy may occur, resulting in a disfiguring appearance that is unique but difficult to quantify. The mechanism underlying drug associated lipoatrophy is unknown. The use of stavudine and perhaps zidovudine is associated with increased risk of lipoatrophy. The concurrent use of protease inhibitors may accelerate the underlying process.45 Besides surgical correction, the only proved treatment for lipoatrophy is switching from stavudine or zidovudine to another nucleoside reverse transcriptase inhibitor.11 Lipoatrophy (and perhaps other treatment associated adverse events) may be more common in patients with advanced immunodeficiency who start therapy. This important observation argues for starting therapy earlier.46 47

Antiretroviral therapy can also cause an abnormal accumulation of body fat. Various unique presentations have been described, including increased abdominal girth (caused by accumulation of visceral rather than subcutaneous fat), breast enlargement, and the appearance of a dorsocervical fat pad (“buffalo hump”). The mechanism for fat accumulation is not known. Growth hormone and perhaps metformin seem to partially reverse abnormal accumulation of central fat.37

Immune reconstitution syndrome

A major proportion of patients starting an effective combination antiretroviral regimen may have a paradoxical worsening of a pre-existing condition or may present in the first few weeks of therapy with a new opportunistic infection. This syndrome is believed to be the result of the rapid expansion of antigen specific immune responses in patients with clinical or subclinical disease. Immune reconstitution syndrome, or immune reconstitution inflammatory syndrome, is more common in patients who start therapy with a low CD4 T cell count (< 50×106/l) and in patients who have a potent virological response to therapy.48 Several complications are associated with immune reconstitution inflammatory syndrome (IRIS), including Mycobacterium tuberculosis, Mycobacterium avium complex, cytomegalovirus, herpes simplex virus, hepatitis C, Pneumocystis carinii, and Cryptococcus neoformans. Kaposi's sarcoma may also worsen as a consequence of the syndrome.

The management of immune reconstitution inflammatory syndrome has not been carefully defined. In addition to aggressive treatment of the presenting condition, options include interrupting therapy or starting anti-inflammatory drugs, or both. Both of these options have potential negative consequences and should be considered only in severe cases.

How to manage drug-drug interactions

An extensive list of potential drug-drug interactions is associated with antiretroviral therapy. Several websites provide continued updates for these drug interactions (for example, the database of antiretroviral drug interactions maintained at However, HIV experts and non-experts should be aware of the more common and serious interactions.

Ritonavir is a potent P-450 CYP3A inhibitor and is often used to “boost” other protease inhibitors. Ritonavir also decreases the metabolism of several other drugs, including the statins (simvastatin in particular), benzodiazepines (midazolam in particular), and most drugs used for erectile dysfunction (for example, sildenafil). Important and potentially life threatening drug interactions can occur between protease inhibitors and immunosuppressants (for example, ciclosporin, tacrolimus), antiarrhythmics, and ergot derivatives. Other protease inhibitors may have similar effects on P-450 and therefore should be used with caution even if given without ritonavir.

The absorption and metabolism of antiretroviral drugs can be altered by other drugs in important ways. For example, the widely used proton pump inhibitors reduce atazanavir levels through unknown mechanisms. This occurs even if ritonavir is given with atazanavir. Also, rifampin (and to a lesser degree rifabutin) can induce P-450 CYP3A, thereby resulting in increased metabolism of the non-nucleoside reverse transcriptase inhibitors and some protease inhibitors. In either case, suboptimal exposure to the antiretroviral drug may occur, leading to virological failure and the emergence of drug resistance.

Global perspectives

Combination antiretroviral therapy is now becoming more widely available throughout the world, including resource poor regions such as Africa and South East Asia. The goal of therapy is the same in these regions as they are in more industrialised countries. Specifically, a potent regimen containing two nucleoside reverse transcriptase inhibitors and either a non-nucleoside reverse transcriptase inhibitor or a protease inhibitor should be given with the goal of reducing HIV RNA levels to below the level of detection. However, when to start therapy and what to start may vary from region to region, depending on both drug availability and costs.

Unanswered research questions

When should antiretroviral therapy be started?

When should therapy be modified for patients with incomplete viral suppression (“when to switch”)?

Can therapy be administered safely in an intermittent manner (“when to stop”)?

Is viral load monitoring and resistance testing needed in resource constrained regions?

What are the best treatment strategies for patients with active tuberculosis, malaria, hepatitis C, and other significant co-infections?

What is the best initial regimen for patients presenting with advanced disease or with renal or hepatic dysfunction?

What is the best initial regimen for women of childbearing potential?

What is the best therapeutic strategy for patients at risk for non-adherence?

What is the mechanism for body fat redistribution (“lipodystrophy”) and can it be reversed once it has occurred?

Does treatment accelerate or delay atherosclerosis?

Can immunomodulatory drugs be used to delay the need for treatment or to prevent the immune reconstitution syndrome in those starting therapy with advanced disease?

Can therapy be used to prevent infection either before or soon after exposure (pre-exposure and post-exposure prophylaxis)?

Given the limited resources available in many areas of the world, it is unlikely that frequent viral load monitoring and resistance testing will be widely available in the near future. It is also unlikely that the drugs necessary to construct complicated, fully suppressive salvage regimens will be limited in many regions. Thus, although the goal of therapy for treatment naive patients will remain similar, the management of treatment failure will likely vary dramatically between regions. For a more detailed discussion of these issues see “Antiretroviral treatment of HIV infection in infants and children in resource-limited settings, towards universal access: recommendations for a public health approach,” updated at


  • Competing interests The author has received research support or honorariums from Boehringer Ingelheim, Bristol-Myers Squibb, GlaxoSmith Kline, Pfizer, Roche, Tibotec, and Trimeris.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
View Abstract

Log in

Log in through your institution


* For online subscription