Antimicrobial Chemotherapy, 5th Edition

Part 1 - General Properties of Antimicrobial Agents

Chapter 7

Antiretroviral Agents

The first human immunodeficiency virus (HIV) was identified and characterized as a retrovirus in 1983. Long-term infection with HIV leads to inexorable destruction of the host immune system, and the development of the acquired immune deficiency syndrome (AIDS). Since that momentous discovery, considerable progress has been made in devising novel antiretroviral agents, which, by inhibiting HIV replication, can prolong the time to the development of AIDS. Unfortunately, the concept of ‘curing’ a patient of HIV infection remains an elusive and as yet unattainable goal.

The fact that HIV belongs to the retroviral family identified the first and most obvious target in the search for anti-HIV drugs—the enzyme reverse transcriptase. This enzyme, which copies an RNA template into DNA, is not present within host cells, which have no need to perform this ‘backwards’ synthetic process. The enzyme is encoded within the HIV genome, and the protein itself is an essential component of infectious viral particles.

Initially, anti-HIV drugs were all nucleoside analogues. Subsequently, it became clear that compounds with other structures could also act as reverse transcriptase inhibitors. Further advances came with the development of drugs that target a completely different virally encoded enzyme—HIV protease—or prevent the virus from entering the host cell by blocking the initial fusion step. Thus, there are currently four categories of antiretroviral drug licensed for clinical use: nucleoside analogues, non-nucleoside reverse transcriptase inhibitors, protease inhibitors, and fusion inhibitors (Table 7.1). Various other potential targets have been identified, and it is hoped that the number of anti-HIV drugs will continue to expand.

Antiretroviral drugs are commonly formulated as combination products in the hope of suppressing the development of resistance, to simplify the often complex dosage regimens that are required for effective therapy and, in some cases, to exploit possible synergic interactions. However, the frequent occurrence of toxic interactions between various antiretroviral compounds demands careful selection of such drug permutations.


Table 7.1 Antiretroviral agents in clinical use in the UK and/or US (2006)

Nucleoside reverse transcriptase inhibitors

Non-nucleoside reverse transcriptase inhibitors

Protease inhibitors

Fusion inhibitors





Didanosine (ddI)








Lamivudine (3TC)




Stavudine (d4T)








Zalcitabine (ddC)




Zidovudine (AZT)






A pro-drug formulation;

Strictly, a nucleotide rather than a nucleoside analogue.

Abbreviations in brackets are often used to describe these compounds. They are derived from the chemical structures: didanosine is 2′,3′-dideoxyinosine; lamivudine is 2′-deoxy-3′-thiacytidine; stavudine is 2′,3′-didehydro-3′-deoxythymidine; zalcitabine is 2′,3′-dideoxycytidine; zidovudine is 3′-azido-2′,3′-dideoxythymidine.

Nucleoside analogues


Zidovudine (azidothymidine; often simply called AZT; Fig. 7.1) was originally investigated for use as an anticancer agent. However, when the HIV epidemic arose in the 1980s, it was tested along with many other drugs for activity against the virus. It was found to inhibit the reverse transcriptase activity of retroviruses at concentrations considerably lower than those needed to interfere with synthesis of host cell DNA.

Molecules such as zidovudine are referred to as 2′-3′-dideoxy nucleoside analogues, since they lack hydroxyl groups at both the 2′ and 3′ positions of the deoxyribose ring. Like other nucleoside analogues, these compounds are activated by phosphorylation to the triphosphate form, these steps being carried out by cellular enzymes. As with aciclovir (p. 99), incorporation of the triphosphate into a growing DNA chain will result in chain termination, as there is no 3′ hydroxyl group available for formation of the next 5′-3′ phosphodiester linkage.



Fig. 7.1 Structure of azidothymidine (AZT).

Non-specific side effects such as headache, anorexia, and nausea are common with zidovudine therapy, but these effects often abate after 2-3 weeks. More serious toxicity to bone marrow cells is dose-related, and hence much effort has been directed at defining the minimum dose that exhibits effective antiviral action. Combination therapy with ganciclovir, although desirable, since many HIV-infected patients also suffer serious CMV infection, is complicated by an additive effect on the bone marrow.

Other nucleoside analogues

The success of zidovudine, albeit with limitations, at least showed that it was possible for anti-HIV drugs to be therapeutically useful. The pharmaceutical industry was therefore encouraged to design and test other potential nucleoside analogues. To date, seven further drugs of this type have been licensed for use in the USA and some other countries (Table 7.1). One of these compounds, tenofovir, is phosphorylated and is thus, like cidofovir (p. 101), a nucleotide, rather than a nucleoside analogue, but its mode of action is similar to that of true nucleoside analogues: it is converted to a triphosphate form that acts as a chain terminator and competitive inhibitor of HIV-derived reverse transcriptase.

The most frequent adverse reaction with all of these drugs is gastrointestinal disturbance, including nausea, vomiting, abdominal pain, and diarrhoea. More serious and potentially life-threatening effects include lactic acidosis and hepatomegaly. Among the many side effects more commonly associated with individual compounds are: peripheral neuropathy (didanosine, zalcitabine, stavudine); pancreatitis (didanosine, zalcitabine); elevation of liver transaminases (stavudine); hypersensitivity reactions (abacavir); increased serum phosphate levels and renal failure (tenofovir); and pruritus (emtricitabine).


Non-nucleoside analogue reverse transcriptase inhibitors

The recognition that compounds with a range of dissimilar structures could also act as inhibitors of reverse transcriptase led to the categorization of a new class of anti-HIV agents: the non-nucleoside analogue reverse transcriptase inhibitors. In contrast to the nucleoside analogues, these agents bind specifically to a non-substrate binding site of reverse transcriptase located close to the substrate-binding site. These drugs inhibit the reverse transcriptase of HIV-1 but not that of HIV-2. Many chemical classes of non-nucleoside compounds that act as reverse transcriptase inhibitors have been described, but only three drugs have so far been used extensively in the treatment of HIV infection: nevirapine (Fig. 7.2), delavirdine (which is not licensed in Europe), and efavirenz.

Rashes, which can be severe, hepatotoxicity, and numerous other side effects of varying frequency and severity are associated with the use of this class of compounds.

Protease inhibitors

The development of drugs acting on a viral target other than reverse transcriptase represented a major breakthrough in therapy for HIV-infected patients. When HIV replicates, it produces polycistronic mRNA, which is translated into a series of polyproteins that must be cleaved to yield the component proteins for infectious viral particles. This function is performed by a virally encoded aspartyl protease. Although mammalian cells also contain aspartyl proteases, these do not cleave HIV polyproteins efficiently.

Nine drugs that specifically inhibit HIV protease are currently licensed for use (Table 7.1). They are complex molecules and most are structurally related. All have important side effects. Besides gastrointestinal problems (particularly prominent with ritonavir), the most worrying adverse effect of protease inhibitors arises from their interference with fat and  carbohydrate metabolism, leading to hyperlipidaemia, glucose intolerance (even frank diabetes mellitus), and peripheral lipodystrophy (abnormal body fat distribution) with increased abdominal fat, ‘buffalo humps’, and breast hypertrophy. The frequency and pathophysiology of these reactions are unclear, as is the long-term risk of other complications such as ischaemic heart disease. Indinavir may cause nephrolithiasis, by precipitation in the renal tubules.


Fig. 7.2 Structure of nevirapine.

Ritonavir has the useful property of boosting the activity of other protease inhibitors by prolonging plasma concentrations, possibly by competing for liver enzymes that metabolize these drugs. The effect is obtained with low doses of ritonavir that lack intrinsic antiviral activity.

Fusion inhibitors

The process of fusion whereby viral particles gain entry into the CD4 T cell following the initial attachment process is an attractive target for selectively active antiretroviral drugs. Various synthetic peptides have been designed to mimic part of the viral envelope glycoprotein 41, which is involved in the fusion of the virus to the membrane of the target cell, and several of them efficiently prevent viral infection of cells in vitro. Only one of these compounds, enfuvirtide (formerly pentafuside) has so far been approved for therapeutic use. As a peptide enfuvirtide is digested when taken by mouth and it is administered by subcutaneous injection. It is metabolized in the body and injections need to be given twice daily. Enfuvirtide is licensed for use in patients who are intolerant of other antiretroviral drugs or are not responding to preferred regimens. Side effects of the drug are common, but are not normally serious.

Resistance to anti-HIV drugs

Unfortunately, optimism encouraged by the emergence of effective anti-HIV drugs has had to be tempered by the realization that the virus can rapidly acquire resistance to these agents. The enzyme reverse transcriptase is considerably more error-prone than other DNA polymerases. Thus, generation of large numbers of viral mutants is part of the natural replication cycle of HIV: it has been estimated that, within a single patient, every possible base substitution could occur at every possible nucleotide position in the viral genome every day! Clearly, the vast majority of these mutations will be deleterious and result in non-viable virus. However, the potential is there for mutations to occur that do confer benefit to the virus, especially if those mutations result in decreased efficacy of an antiviral drug. Experience with all anti-HIV drugs used thus far indicates that resistance is an inevitable consequence of drug usage, at least if the suppression of virus replication is not absolute.

Isolates of HIV derived from patients who have been taking zidovudine for at least 6 months are invariably less sensitive to the drug in vitro than isolates taken from the same patient at the initiation of therapy. This arises from mutations in the gene coding for reverse transcriptase, leading to reduced binding of zidovudine triphosphate. Resistance arises as a series of five sequential point mutations within the reverse transcriptase gene, each additional mutation resulting in an increase in the dose of drug necessary to inhibit the virus. This relative resistance is of considerable clinical importance. Trials of zidovudine monotherapy conducted in the late 1980s showed early promise: increasing CD4 cell counts and prolonging survival in treated patients. However, as these patients were followed for longer periods, it became clear that the therapeutic efficacy of the drug was limited to a period of about 6 months, after which it was of no benefit. This loss of efficacy coincided with emergence of highly resistant viruses in treated patients.

Resistance to other nucleoside analogues similarly arises through mutations in the reverse transcriptase gene. In general, cross-resistance between zidovudine and the other nucleoside analogues is not a problem, as the positions of the mutations conferring resistance to zidovudine differ from those giving rise to resistance to, for instance, didanosine or zalcitabine. In contrast, mutations causing resistance to didanosine and zalcitabine are similar, and also overlap with those causing lamivudine resistance.

The news concerning resistance mutations may not be all bad, however. Mutation at position 184 in the reverse transcriptase gene results in increasing resistance to lamivudine, but paradoxically causes an increase in sensitivity to zidovudine in virus previously resistant. Thus, in theory at least, a combination regimen of zidovudine plus lamivudine should be of benefit, a hope reflected by the availability of a tablet that contains both agents.

Yet another set of mutations in the reverse transcriptase gene confers resistance to the non-nucleoside reverse transcriptase inhibitors, and several such changes have been reported. Some, but not all, lead to cross-resistance among the different drugs of this type, but they are distinct from those leading to resistance to the nucleoside analogues.

As with the reverse transcriptase inhibitors, resistance to the protease inhibitors may arise through point mutations in the gene coding for the target protein. Various mutations in the protease gene have been described.

Many of these map at similar points in the genome of virus from patients receiving different protease inhibitors, indicating that cross-resistance between these drugs is common. Continued use of a given protease inhibitor leads to the accumulation of mutations, with a concomitant increase in resistance. Tipranavir was introduced because it retains activity against strains resistant to other protease inhibitors.

Although the use of combination therapy substantially reduces the chance of treatment failure due to the emergence of drug resistance there are disturbing reports of simultaneous resistance developing to two or more classes of antiretroviral drug. The spread of multiresistant strains of the virus clearly constitutes a major threat and it is important to minimize this possibility. Tests for drug resistance on isolates from patients receiving treatment are valuable in directing appropriate changes in therapy. Since resistant strains may revert to susceptibility when therapy is discontinued, a prompt switch to alternative agents is often helpful.

Anti-HIV drugs in development

Despite the clinical successes achieved by use of the above agents, there is a desperate need for new drugs with different sites of action, and different toxicity profiles in the fight against HIV infection. Much investigation continues into safer and more effective alternatives for existing classes of antiretroviral compounds.

A novel target for anti-HIV drugs that has attracted particular interest is the virus-encoded enzyme, integrase, which is responsible for integration of the DNA provirus into host cell chromosomes. Inhibition of the integrase would prevent virus from becoming latent in resting T cells. Optimal therapy for patients with HIV infection may involve treatment with other drugs that do not necessarily exhibit antiviral activity. Immunomodulatory agents such as interleukin-2 have their advocates. Initially encouraging reports of the use of hydroxyurea, which blocks cellular activation necessary for viral replication in resting CD4-positive T cells have unfortunately not been confirmed.