HIV, in common with other retroviruses, possesses the enzyme reverse transcriptase and consists of a lipid bilayer membrane surrounding the capsid (Fig. 41.1). Its surface glycoprotein molecule (gp120) has a strong affinity for the CD4 receptor protein found predominantly on the T-helper/inducer lymphocytes. Monocytes and macrophages may also possess CD4 receptors in low densities and can therefore also be infected. The process of HIV entry is more complex than originally thought, and in addition to CD4 attachment, subsequent binding to co-receptors such as CCR-5 or CXCR-4 and membrane fusion also occur (Fig. 41.2).

Fig. 41.1 Structure of the human immunodeficiency virus (HIV).

Fig. 41.2 Lifecycle of HIV and the sites of action of currently available antiretroviral agents (in bold).

After penetrating the host cell, the virus sheds its outer coat and releases its genetic material. Using the reverse transcriptase enzyme, the viral RNA is converted to DNA using nucleosides. The viral DNA is then integrated into the host genome in the cell nucleus, where it undergoes transcription and translation, enabling the production of new viral proteins. New virus particles are then assembled and bud out of the host cell, finally maturing into infectious virions under the influence of the protease enzyme.

Immediately after primary HIV Infection (PHI, also known as ‘seroconversion’), there is a very high rate of viral turnover. Equilibrium is then reached, at which stage the infection may appear to be clinically latent, but in fact, as many as 10,000 million new virions are produced each day.

Over time, as chronic infection ensues, cells possessing CD4 receptors, particularly the T-helper lymphocytes, are depleted from the body. The T-helper cell is often considered to be the conductor of the ‘immune orchestra’ and thus, as this cell is depleted, the individual becomes susceptible to a myriad of infections and tumours. The rate at which this immunosuppression progresses is variable and the precise interaction of factors affecting it is still not fully understood. It is well recognised that some individuals rapidly develop severe immunosuppression, whilst others may have been infected with HIV for many years whilst maintaining a relatively intact immune system. It is likely that a combination of viral, host (genetic) and environmental factors contributes to this variation.

Clinical manifestations

The sequelae of untreated HIV infection can be broadly considered in five categories:

• Opportunistic infections, that is, infections that would not normally cause disease in an immunocompetent host, for example, P. jiroveci pneumonia and cytomegalovirus (CMV)

• Infections that can occur in immunocompetent patients but tend to occur more frequently, more severely and often atypically in the context of underlying HIV infection, for example, Salmonella, herpes simplex and Mycobacterium tuberculosis

• Malignancies, particularly those that occur rarely in the immunocompetent population, for example, Kaposi’s sarcoma and non-Hodgkin’s lymphoma

• Direct manifestations of HIV infection per se, for example, HIV encephalopathy, HIV myelopathy and HIV enteropathy

• Consequences of chronic immune activation including premature cardiovascular disease, neurocognitive dysfunction, bone mineral density loss.

In addition, approximately 70% of individuals develop a flu-like illness at seroconversion. This primary HIV infection (PHI) is characterised by fever, arthralgia, pharyngitis, rash and lymphadenopathy. Rarely, the degree of associated CD4 count depletion may be sufficient to result in development of an opportunistic illness such as oropharyngeal/oesophageal candidiasis or P. jiroveci pneumonia.

Opportunistic infections generally fall into two categories:

• DNA viruses, for example, CMV and JC virus

• Intracellular pathogens, for example, P. jiroveci, Toxoplasma gondii and Mycobacterium avium.

Although the clinical course of HIV disease varies with each individual, there is a fairly consistent and predictable pattern that enables appropriate interventions and preventive measures to be adopted. Patients can be classified into one of three groups according to their clinical status: asymptomatic, symptomatic or AIDS. Symptomatic disease is characterised by non-specific symptomatology such as fevers, night sweats, lethargy and weight loss, or by complications including oral candidiasis, oral hairy leucoplakia, and recurrent herpes simplex or herpes zoster infections. AIDS is defined by the diagnosis of one or more specific conditions including P. jiroveci pneumonia, M. tuberculosis infection and CMV disease.

Investigations and monitoring

Current and previous infections

The initial diagnosis of HIV infection is made by the detection of antibodies against HIV. With improved technology, it is usually possible to detect antibodies within 3–4 weeks of infection, although individuals are advised that a ‘window period’ of up to 3 months after exposure is required before infection can be excluded. After confirmation of HIV infection, the patient is usually tested for prior exposure to a number of potential pathogens, including syphilis, hepatitis A, B and C, CMV, varicella zoster (VZV), and T. gondii. This can enable subsequent treatment (in the case of undiagnosed syphilis), vaccination (if no prior exposure to hepatitis A, B, or VZV), prevention (if no prior exposure to Toxoplasma and CMV), prophylaxis (if previous exposure to Toxoplasma) and can aid subsequent diagnosis (according to CMV or Toxoplasma status).

CD4 count

The level of immunosuppression is most easily estimated by monitoring a patient’s CD4 count. This measures the number of CD4-positive T-lymphocytes in a sample of peripheral blood. The normal range can vary between 500 and 1500 cells/mm³. As HIV disease progresses, the number of cells falls. Particular complications of HIV infection usually begin to occur at similar CD4 counts (Fig. 41.3) which can assist in differential diagnoses and enable the use of prophylactic therapies. For example, patients with a CD4 count of less than 200 cells/mm³ should always be offered prophylaxis against P. jiroveci pneumonia. Similarly, both patient and clinician are likely to use the CD4 count as the major indicator of when to consider starting antiretroviral therapy.

Fig. 41.3 Opportunistic complications of HIV infection and the CD4 count ranges at which they commonly occur.

Viral load

The measurement of plasma HIV RNA (viral load) estimates the amount of circulating virus in the blood. This has been proven to correlate with prognosis, with a high viral load predicting faster disease progression (Mellors et al., 1997). Conversely, a reduction in viral load after commencement of antiviral therapy is associated with clinical benefit. This measure, in combination with the CD4 count, allows patients and clinicians to make informed decisions regarding when to start and when to change antiviral therapies, enabling the more effective use of such agents. There are on-line calculators utilising viral load and CD4 count to model risk of disease progression or death based on large cohort studies.

Resistance testing

Due to the implications of transmitted (primary) resistance, it is recommended that all patients have a genotypic HIV resistance test performed soon after diagnosis; this will ensure that appropriate initial therapy is selected. Further resistance tests should be performed at any subsequent virological failure to direct therapy choice.

Tropism testing

Viruses may enter the cell using the CCR5 co-receptor, the CXCR4 co-receptor or both co-receptors. Those which just use one co-receptor are known as CCR5-tropic or CXCR4-tropic viruses; those which can use both receptor types are called dual-tropic. Where a mixture of virus populations is present, the term mixed-tropic is used. Different methods of determining tropism are currently under evaluation. The tests must be performed in real time as viral tropism changes as the disease progresses. If CCR5 inhibitors are to be used, it is essential to determine that the virus is CCR5-tropic, that is, that there is no significant use of the CXCR4 receptor.

Drug treatment

The drug treatment of HIV disease can be classified as antiretroviral therapy, the management of opportunistic infections or malignancies, the management of ‘non-HIV-related’ co-morbidities, and symptom control. For the first decade of the epidemic, most of the available drugs and therapeutic strategies were aimed at treating or preventing opportunistic complications and alleviating HIV-related symptoms. Whilst these are still important, there has been a shift in emphasis towards treatment aimed at reducing the HIV viral load, restoring immune function and reducing the potential consequences of co-morbidities.

Due to the speed at which new antiretroviral agents are being developed, comprehensive data on drug interactions, side effects, etc., are often lacking. Thus, the ability to apply general pharmacological and pharmacokinetic principles, together with common sense, is required.

The treatment of many of the opportunistic complications of HIV comprises an induction phase of high-dose therapy, followed by maintenance and/or secondary prophylaxis using lower doses. This is due to the high rate of relapse or progression after a first episode of diseases such as P. jiroveci pneumonia, cerebral toxoplasmosis (toxoplasmic encephalitis), systemic cryptococcosis and CMV retinitis. Where a cost-effective agent with an acceptable risk/benefit ratio exists, primary prophylaxis may be offered to individuals who are deemed to be at high risk of developing a particular opportunistic infection, for example, P. jiroveci pneumonia prophylaxis. Discontinuation of prophylaxis, both primary and secondary, is now usually possible in individuals who demonstrate immunological restoration on Highly Active Antiretroviral Therapy (HAART).

Paradoxically, this immunological restoration may result in apparent clinical deterioration with opportunistic infections during the first few weeks after initiation of HAART. This is known as immune reconstitution inflammatory syndrome (IRIS).

The goals of therapy in HIV-positive individuals are to:

• improve the quality and duration of life;

• prevent deterioration of immune function and/or restore immune status;

• treat and/or prevent opportunistic infections;

• relieve symptoms.

Antiretroviral therapy

Antiretroviral therapy is currently one of the fastest evolving areas of medicine. The specific details of treatment will therefore continue to change as new drugs emerge, although it is likely that the following general principles will remain:

• A combination of three antiretroviral agents, selected on the basis of treatment history and resistance tests, should usually be prescribed to increase efficacy and reduce the development of drug-resistant virus

• Wherever possible, a regimen should contain at least one drug that penetrates the central nervous system and confers protection against HIV-related encephalopathy/dementia

• Treatment strategies should be adopted that sequence drug combinations, being mindful of potential cross-resistance and future therapy options

• Given the crucial importance of a high level of adherence to these therapies, the regimen adopted for a particular individual should, wherever possible, be tailored to suit the daily lifestyle.

Many organisations, such as the British HIV Association (BHIVA), the European AIDS Clinical Society (EACS) and the International AIDS Society (IAS), produce regularly updated guidelines on the use of antiretroviral therapy, for example, Gazzard et al. (2008). These guidelines include the most up-to-date considerations of:

• when to start therapy;

• what to start with;

• how to monitor, including use of therapeutic drug monitoring and resistance testing;

• when to switch therapy;

• what therapy to switch to;

• treating individuals who have been highly exposed to multiple agents;

• managing individuals with significant co-morbidities, for example, tuberculosis or hepatitis B/C.

Most studies evaluating triple combinations of antiretrovirals have been designed with so-called surrogate marker endpoints, measuring the effect on laboratory parameters such as CD4 count and HIV viral load. These trials are generally smaller and shorter in duration than clinical endpoint studies that are powered to measure the impact on survival and disease progression. The first large clinical endpoint trial that demonstrated the superiority of a triple combination over dual therapy was undertaken by Hammer et al. (1997). Following the results of this trial, the standard approach, where treatment is indicated, has been to use a combination of at least three agents. The reduction in morbidity and mortality associated with HAART has been confirmed in routine clinical practice, as well as in other trials (e.g. Palella et al., 1998; Smit et al., 2006). Subsequent clinical trials have largely been for licensing purposes and/or have served to refine therapeutic choices rather than to change the paradigm of treatment. The concept of intermittent rather than continuous therapy was evaluated in the SMART study but shown to be linked with an increased risk of co-morbidities not previously thought to be associated with HIV (such as cardiovascular disease, hepatic and renal failure) as well as HIV disease progression (El-Sadr et al., 2006). The use of protease inhibitor (PI) ‘monotherapy’ compared to conventional triple therapy has been evaluated in a number of small studies, for example, Arribas et al. (2009), and is being investigated in longer-term strategic studies. A large international study of early versus deferred treatment, to attempt to address the question of when to initiate treatment, is ongoing.

When to start therapy

Current UK guidelines recommend starting antiretrovirals when the CD4 count drops below 350 cells/mm³. Therapy should be considered at a higher CD4 count in specific situations, for example, in the presence of an AIDS-defining illness or HIV-related morbidity.

Choosing and monitoring therapy

The majority of individuals are currently commenced on a combination of two nucleoside/nucleotide reverse transcriptase inhibitors (NRTIs) and a non-nucleoside reverse transcriptase inhibitor (NNRTI) or two NRTIs and a boosted PI. The term ‘boosted PI’ refers to a combination of one PI combined with a low dose (usually 100–200 mg once or twice daily) of ritonavir, another PI. The ritonavir does not directly add to the antiretroviral activity of the regimen; it is used purely as a pharmacokinetic enhancer of the other PI, by increasing the maximum plasma concentration, C_max, due to inhibition of cytochrome P450 enzymes in the gut wall and/or extending the half-life, t_1/2, by inhibition of hepatic cytochrome P450 enzymes. Triple NRTI therapy is no longer recommended, as it is associated with unacceptable rates of virological failure. Alternative strategies, such as NRTI-sparing regimens and boosted PI monotherapy, are currently only routinely recommended in a research setting. More recently, integrase inhibitors have been approved for initial therapy and may be used as an alternative to NNRTIs or PIs.

The aim of initial therapy is to achieve viral load suppression in the plasma to levels below the detection limits of available assays (40 or 50 copies/mL). Such virological suppression is almost invariably accompanied by an elevation in CD4 count and clinical evidence of immune reconstitution. Whilst sustained suppression over many years is usually possible, viral rebound may occur and is often accompanied by the development of resistance to one or more agents in the combination. Upon confirmed virological failure, a resistance test is performed which will help to identify to which agents the virus may have adapted and the extent to which any such resistance mutations may confer cross-resistance to other available drugs. A second-line regimen is then constructed, wherever possible utilising a new class of drug to which the individual has not previously been exposed. Upon virological failure of subsequent regimens, the therapeutic options available become increasingly complex, but with the availability of more agents targeting different parts of the virus life cycle, virological suppression is still usually possible and should remain the goal of treatment.

Many of the antiretrovirals, particularly the PIs and NNRTIs, exhibit a wide range of interactions, especially with other drugs that are metabolised by the cytochrome P450 enzyme system, including prescribed, ‘over-the-counter’, herbal and recreational drugs. HAART failure (detectable viral load and drug resistance) has been documented following co-administration of hepatic enzyme inducers, including non-prescribed agents such as St John’s Wort. Conversely, serious and even fatal toxicities due to enzyme inhibition by the PIs continue to be reported. These include Cushing’s syndrome following concomitant use of fluticasone or budesonide inhaler or nasal spray with a PI. This highlights the necessity of taking a comprehensive drug history prior to starting or switching HAART and ensuring patients and prescribers are aware of the need to check the interaction potential of new medicines. General prescribing guidelines for antiretrovirals are presented in Box 41.1 whilst details of common side effects and interactions of the currently available agents are summarised in Tables 41.1 to 41.5.

Box 41.1 General prescribing and monitoring information for antiretroviral agents

• The Summary of Product Characteristics, current British National Formulary (BNF) and national guidelines should be consulted when managing the treatment of an HIV-positive patient.

• Adverse events should be reported. With greater exposure and long-term use, it is likely that more adverse events will be recognised. Some of these may be class effects associated with particular groups of antiretrovirals and should therefore be monitored even in new agents within a class. Examples include dyslipidaemia and diabetes mellitus with protease inhibitors, mitochondrial myopathy and lipoatrophy with NRTIs, rash with NNRTIs.

• There are limited data on the safety of many antiretroviral drugs when taken during pregnancy, or their long-term effects on babies/children. Updated safety information from the Antiretroviral Pregnancy Register is published twice yearly. Efavirenz has been shown to be teratogenic in animal studies, but comparable data are not available for the other antiretrovirals. Caution should therefore be exercised with all agents. Treatment of women who are planning conception or who are pregnant must be discussed with the relevant experts. All pregnant women who are treated with antiretrovirals should be reported prospectively to the Antiretroviral Pregnancy Register.

Table 41.1 General prescribing points for nucleoside/nucleotide analogue reverse transcriptase inhibitors (NRTIs)

Table 41.2 General prescribing points for non-nucleoside reverse transcriptase inhibitors (NNRTIs)

Table 41.3 General prescribing points for protease inhibitors

Table 41.4 General prescribing points for Integrase Inhibitors

Table 41.5 General prescribing points for entry inhibitors

The routine use of therapeutic drug monitoring is not recommended but blood levels of PIs and NNRTIs should be measured in selected patients, for example, during pregnancy, where there is liver impairment and where there are concerns regarding potentially interacting drugs (Gazzard et al., 2008).

HIV mutates readily and resistance to some antiretrovirals, particularly reverse transcriptase inhibitors and integrase inhibitors, develops rapidly in the face of suboptimal treatment, for example, monotherapy or subtherapeutic blood levels. A high level of adherence to treatment is crucial to the sustained, successful outcome of antiretroviral regimens and has been the subject of much research. For example, in one study of people taking their first regimen containing nelfinavir, it was found that at least 95% adherence was required to achieve a sustained response in the majority (78%) of patients. The chances of treatment success declined as the level of adherence dropped, such that 80% of patients whose adherence was below 80% experienced virological failure. Virological success was also found to correlate with a better clinical outcome in terms of fewer hospitalisations, opportunistic infections and deaths (Paterson et al., 2000). Such clinical trial data have also been supported by clinical experience in the UK and elsewhere, although it has yet to be established if the level of adherence required is the same for all regimens and every patient. In view of this, patients should be advised to take HAART as close as possible to the same time every day and certainly within 1 hour of the agreed time each day. If they forget a dose, it should be taken as soon as they remember and then return to the original schedule.

There has been significant progress over recent years in reducing some of the physical burden of therapy, through the development of combination tablets and the use of strategies such as ritonavir boosting to reduce dietary restrictions and dosing frequency. Adherence aids such as pill boxes, medication record cards and alarms (e.g. on mobile phone) can also help to support adherence. However, practical issues are not the only barriers to adherence and the individual’s health beliefs and motivation, particularly around HIV and antiretroviral therapy, should also be addressed before treatment is commenced, as these are likely to have a significant impact on outcome (Horne et al., 2004). Although there is little evidence to demonstrate what the optimal interventions to improve adherence are, multidisciplinary and multiagency approaches appear to be most useful (Poppa et al., 2003).

Treatment interruptions

For many reasons, including toxicity, cost and adherence, patients and clinicians have been interested in considering ‘drug holidays’ or treatment interruptions. However, this strategy is no longer recommended in routine practice (El-Sadr et al., 2006). It is now recognised that there are dangers associated with this approach because of CD4 decline, disease progression, mortality related to co-morbidities, for example, cardiovascular disease, and viral load rebound associated with increased transmission risk and a seroconversion-like syndrome. Further, as different anti-HIV medications have different half-lives, there may be a risk of functional monotherapy, particularly with NNRTIs, and the development of resistance if combinations are stopped abruptly in an unplanned fashion.

Post-exposure prophylaxis

Post-exposure prophylaxis (PEP) involves the use of antiretroviral drugs to prevent infection with HIV after possible exposure, which may be recommended after occupational injuries (DH, 2008) or sexual exposure (Fisher et al., 2006). Whilst PEP is a largely unproven and unlicensed indication for the drugs used, it is supported by animal model data and case–control studies. Where recommended in guidelines, PEP is usually commenced as a 3–5-day starter regimen of two NRTIs and a boosted PI, followed by an ongoing course for a total of 4-week post-exposure. It is believed this will reduce the likelihood of infection by at least 80%, although toxicity issues are not insignificant. Therefore, the decision to prescribe or take PEP must reflect a careful risk/benefit evaluation. Studies of pre-exposure prophylaxis (PREP), using one or two antiretrovirals (orally or topically) before potential exposure to HIV, have so far yielded mixed results, but may offer additional options to reduce transmission.

Cardiovascular disease

Cohort studies have suggested an increased risk of cardiovascular disease with some PIs (Kaletra® and indinavir) and with the NRTI, abacavir (Sabin et al., 2008; Worm et al., 2010). This risk is independent of the effect on lipids and the mechanism has yet to be determined. As with lipid disturbances, the decision to start or continue these agents needs to be considered as part of a holistic approach to cardiovascular risk.

Opportunistic infections and malignancies

Detailed information on treatment guidelines can be obtained by consulting the relevant national guidelines, for example, British HIV Association guidelines for the management of opportunistic infections (Nelson et al., 2010), tuberculosis (Pozniak et al., 2010), malignancies (Bower et al., 2008). Dosing schedules for the most commonly used drugs used to treat opportunistic infections are summarised in Table 41.6.

Table 41.6 Selected drugs to treat opportunistic infections in HIV disease

Fungal infections

P. jiroveci pneumonia

This remains one of the most common causes of morbidity and mortality in HIV-positive individuals. Classically, patients present with an insidious onset of a non-productive cough, shortness of breath on exertion and an inability to take a deep breath. Fever, anorexia and weight loss are common accompanying symptoms. Patients are usually markedly immunosuppressed, with CD4 counts less than 200 cells/mm³. Diagnosis is supported by the presence of exercise-induced oxygen desaturation and the typical chest radiographic appearance of bilateral interstitial shadowing, though in mild cases, the chest X-ray may be normal. P. jiroveci cannot be cultured in vitro, so the diagnosis is confirmed by demonstration of the organism by immunofluorescence or silver staining, or by nucleic acid amplification techniques (NAAT), of samples obtained by bronchoalveolar lavage. Sputum induction, by nebulisation of hypertonic sodium chloride, is not recommended for obtaining samples unless the procedure is being performed in the treatment centre sufficiently regularly to enable staff to maintain competence, as diagnostic sensitivity of the test is significantly affected by technique.

Treatment is instigated in patients with a proven diagnosis, or empirically where there is a suspicion prior to confirmation. Adequate samples may be obtained to make a diagnosis up to 7–10 days after starting treatment. Oxygen is essential for patients with compromised respiratory function. First-line therapy is high-dose co-trimoxazole for 21 days: oral for mild cases (1920 mg three times daily), and intravenous for moderate-to-severe disease (120 mg/kg/day in divided doses for 3 days, then 90 mg/kg/day for 18 days). Nausea and vomiting commonly occur and may be best managed pre-emptively by administration of a prophylactic antiemetic.

In cases of co-trimoxazole intolerance, several alternative therapies are available. For mild P. jiroveci pneumonia, a combination of oral trimethoprim (10–15 mg/kg/day in two divided doses) with dapsone (100 mg daily) may be effective. For moderate-to-severe disease, a combination of clindamycin (intravenous or oral) and primaquine is often used. Intravenous pentamidine is another alternative, though its use may be associated with more significant adverse reactions. Pentamidine should also be given in nebulised form (600 mg) via a suitable nebuliser, for example, Respirgard II, for the first 3 days, to ensure prompt attainment of adequate lung tissue levels. Oral atovaquone suspension can be used for mild-to-moderate P. jiroveci pneumonia but must be taken with food, particularly fatty food, to be effective. Glucose-6-phosphate dehydrogenase (G6PD) levels should be checked prior to (or as soon as possible after starting) treatment with co-trimoxazole, dapsone or primaquine.

For cases of moderate-to-severe P. jiroveci pneumonia (PaO₂ < 9.3 kPa or SpO₂ < 92%), adjunctive corticosteroid therapy is recommended, for example, prednisolone 75 mg daily for 5 days, 50 mg for 5 days, and then 25 mg for 5 days. Ventilatory support should be considered for patients in whom the underlying prognosis is good, for example, those presenting for the first time with P. jiroveci pneumonia, those without severe co-existing medical complications and those with remaining effective antiretroviral options.

It has been clearly demonstrated that prophylactic therapy reduces both the incidence and severity of P. jiroveci pneumonia in patients with either prior disease or those at risk of a first episode, and that this intervention significantly improves survival. Primary prophylaxis is recommended for those individuals with a previous AIDS-defining illness, markedly symptomatic disease (including oral candidiasis), a CD4 count < 200 cells/mm³ or CD4 percentage <14%. Prophylaxis can be discontinued when CD4 is sustained above 200 cells/mm³ for over 3 months on HAART.

Co-trimoxazole is the gold standard and also confers protection against toxoplasmosis and some other bacterial infections. Commonly used regimens with proven efficacy are 960 mg daily or three times a week or 480 mg daily. The incidence of adverse reactions to co-trimoxazole in HIV-positive individuals is higher than in the general population, although many patients who are intolerant of high-dose treatment do not experience problems at prophylactic doses. In cases of intolerance, several alternative approaches may be adopted. Desensitisation may be attempted or other agents may be used. Dapsone 100 mg daily, with or without pyrimethamine, according to Toxoplasma status, is effective but does not offer such broad protection against bacterial infections. Nebulised pentamidine at a dose of 300 mg every month via an appropriate nebuliser (with prior nebulised or inhaled β₂-agonist to prevent bronchospasm) can also be used but does not protect against extra-pulmonary P. jiroveci pneumonia.

Oropharyngeal candidiasis

Candidiasis is a frequent manifestation of HIV infection and may occur early in the disease. Clinically, it is usually characterised by white plaques on the oral mucosa, but may present as erythematous patches or as angular cheilitis. If swallowing is difficult (dysphagia) or painful (odynophagia), oesophageal involvement may be suspected.

First-line therapy for oral candidiasis is a systemic agent, usually fluconazole (50 mg daily for 7 days). An alternative is itraconazole (200 mg once daily); the solution has higher bioavailability than the capsule formulation and does not require an acid pH for absorption. Topical therapies are available where there is a clinical decision to avoid systemic antifungals (e.g. in patients with hepatic failure). They are as effective as oral azoles for oropharyngeal disease, but time to yeast clearance is longer and relapse rate is higher.

In cases of oesophageal candida, which is an AIDS-defining illness, systemic therapy is necessary using higher doses of the above agents for a longer duration, for example, fluconazole 100 mg once daily for 2 weeks. Continuous azole therapy or frequent courses of these drugs predispose to the development of azole resistance. In such instances, an alternative azole or an echinocandin may be used, or occasionally higher than usual doses of the original agent, for example, fluconazole 400 mg daily. In intractable cases, intravenous amphotericin may be required. With the use of HAART, such complications are rarely seen.

Cryptococcus neoformans

This causes a disseminated infection, usually with meningeal involvement, in individuals with HIV infection. Patients present with fever and headaches, often without the characteristic symptoms of meningism such as photophobia and neck stiffness. Diagnosis is normally made on the basis of CSF analysis, though serum cryptococcal antigen and blood cultures may also be indicative.

For patients who are moderately or severely unwell, intravenous amphotericin B deoxycholate (0.7–1 mg/kg/day) or a lipid complex/liposomal formulation, with or without flucytosine (100 mg/kg/day in divided doses, oral or intravenous) is the first-line therapy. The place of flucytosine in therapy remains uncertain, with the only clear benefit seen in patients not on HAART, in whom it speeds the rate of CSF sterilisation and reduces the relapse rate. However, it has not been shown to reduce mortality and can be associated with additional toxicity. Conventional amphotericin is associated with high rates of nephrotoxicity, leading in some cases to acute renal failure and death, although strategies can be employed to minimise adverse effects. However, despite the higher cost, the liposomal form (AmBisome®, 4 mg/kg/day) is now usually the agent of choice due to its more favourable toxicity profile. Many UK hospitals now only stock one intravenous amphotericin product, due to reports of errors following confusion between different formulations because doses, reconstitution procedures and administration vary. The usual duration of amphotericin (± flucytosine) treatment is 2 weeks, after which high-dose fluconazole (400 mg daily orally) should be continued for further 10 weeks. Subsequent maintenance therapy with fluconazole 200 mg daily has been shown to be effective in reducing the incidence of relapse and should be continued for life, or until immune function is restored (for example, CD4 > 100 cells/mm³ in the presence of an undetectable viral load for at least 3 months). In milder cases fluconazole may be given for the entire duration of treatment. Itraconazole (400 mg/day) has been used but is less effective than fluconazole and should be used only if other agents are contraindicated.

Protozoal infections

Toxoplasmosis

T. gondii is a frequent cause of central nervous system disease in patients with AIDS. Individuals may present with headaches, fever, confusion, seizures or focal neurological symptoms and signs. Diagnosis is usually based on the appearance of ring-enhancing lesion(s) on computed tomography (CT scan). Definitive diagnosis is based on brain biopsy, which is rarely performed, but is generally made presumptively after response to therapy.

First-line treatment is sulphadiazine and pyrimethamine, with folinic acid to prevent myelosuppression, for 6 weeks, followed by maintenance therapy with lower doses of the same agents until CD4 count is consistently maintained above 200 cells/mm³. The preferred alternative is clindamycin and pyrimethamine with folinic acid, with limited data for atovaquone, co-trimoxazole, clarithromycin and doxycycline. Adjunctive therapy with corticosteroids or anticonvulsants may be used in cases of severe oedema or seizures, respectively.

Cryptosporidiosis

Cryptosporidium parvum is a ubiquitous organism and a common cause of diarrhoea in immunocompetent individuals. In patients who are immunocompromised, persistent infection may occur characterised by abdominal pain, weight loss and severe diarrhoea. Diagnosis is generally based on stool analysis.

Although many agents have been investigated for the treatment of cryptosporidiosis, the majority of results have been disappointing. Nitazoxanide is possibly the most promising agent, but failed to demonstrate superiority over placebo in severely immunocompromised patients. The optimal treatment for cryptosporidiosis (and indeed the majority of chronic opportunistic infections) is to increase immunological function with HAART. The mainstay of management in patients who are not able or willing to take HAART remains symptomatic control with nutritional supplementation, adequate hydration and antidiarrhoeal agents.

Bacterial infections

Bacterial infections are common in the context of HIV infection. Recurrent bacterial pneumonia, particularly Streptococcus pneumoniae, and diarrhoeal illnesses associated with Salmonella, Shigella or Campylobacter are particularly common. In general, these are treated the same as in immunocompetent individuals, although recurrent infections and/or septicaemia occur more frequently.

Mycobacteria

In HIV-positive individuals, M. tuberculosis (TB) is characterised by increased likelihood of reactivation of latent disease; more rapid progression to clinical disease following acquisition; more frequent extrapulmonary manifestations of tuberculosis; and more rapid progression of HIV disease (if the individual is not receiving HAART). Overall, there is no increase in infectivity of tuberculosis compared with HIV-negative patients.

Definitive diagnosis is reliant on culture of the organism from biological specimens but may be complicated by atypical clinical features and reduced response to tuberculin testing; it is often necessary to initiate treatment empirically.

Treatment for pulmonary and extrapulmonary tuberculosis should follow conventional guidelines for immunocompetent individuals. Meningitis is an unusual but significant complication of tuberculosis. Treatment for all forms of CNS tuberculosis should consist of four drugs (isoniazid, rifampicin, pyrazinamide, ethambutol) for 2 months followed by two drugs (isoniazid, rifampicin) for at least 10 months. Adjunctive corticosteroids (either dexamethasone or prednisolone) should be given to all patients with tuberculous meningitis, regardless of disease severity (Thwaites et al., 2009).

The use of primary and secondary prophylaxis remains controversial but may be appropriate in high-incidence groups. The increased incidence of multidrug-resistant tuberculosis (MDRTB) and the emergence of extremely drug-resistant disease (XDRTB) is a cause for concern and raises many infection control issues. It also highlights the need for antibiotic therapy driven by bacteriological sensitivities.

Managing tuberculosis and HIV co-infection is further complicated by drug–drug interactions between anti-tuberculous and antiretroviral agents, overlapping toxicities and the risk of development of immune reconstitution disease (see later). This is a complex area and treatment should be guided by clinicians with the relevant specialist expertise (Pozniak et al., 2010). One of the most frequent concerns is when to initiate therapy for HIV in an individual receiving TB treatment. This decision is based upon the risk of developing other opportunistic infections in the medium term. HAART is usually started after 2 weeks of tuberculosis treatment in those with severe immunosuppression (CD4 <100 cells/mm³), after 2 months if the CD4 is between 100 and 200 cells/mm³, and on completion of tuberculosis treatment or at 6 months if the CD4 is greater than 200 cells/mm³.

M. avium intracellulare (MAI) or M. avium complex (MAC) infection was historically a frequent manifestation of late-stage HIV disease, but is rarely seen now that HAART is widely used. Patients with disseminated infection classically present with fevers, weight loss, diarrhoea and hepatosplenomegaly. Diagnosis is sometimes made presumptively but is usually based on culture of the organism(s). In vitro sensitivities may not be good predictors of response to therapy. Therapy may need to be tailored to account for drug interactions with concomitant antiretrovirals, but usually includes azithromycin and ethambutol (ideally with rifabutin). Alternative agents include the quinolones and amikacin. Corticosteroids may be useful for symptomatic control. Although rifabutin, clarithromycin and azithromycin have all been demonstrated to be effective agents for primary prophylaxis against MAI, their cost–benefit remains controversial and use is not widespread in the UK. The strongest indications are for those patients with CD4 <50 cells/mm³ who decline HAART, or who experience HAART failure (without further effective options) or who are receiving chemotherapy.

Viral infections

Cytomegalovirus

CMV is a herpes virus that is acquired by approximately 50% of the general population and over 90% of homosexual men. Like other herpes viruses, once infection has occurred, the virus remains dormant thereafter, but in individuals with advanced immunosuppression, reactivation may occur and cause disease. In the context of HIV infection, the most common sites of disease are the retina and gastro-intestinal tract, though neurological involvement and pneumonitis are well reported.

Diagnosis of CMV retinitis is based on clinical appearance; it may be detected in asymptomatic individuals but usually presents with symptoms of blurred vision, visual field defects or ‘floaters’. Untreated CMV retinitis progresses rapidly to blindness and treatment substantially reduces the morbidity associated with this condition. Although previously lifelong treatment had been recommended, where immunological restoration occurs discontinuation may be possible. Conventional therapeutic approaches are based upon an initial induction period of high-dose therapy for 2–3 weeks, until the retinitis is quiescent, followed by lower dose maintenance treatment, with reinduction if disease progression occurs.

The most commonly used agent for induction therapy in the UK is ganciclovir, which can be given intravenously or orally, as the pro-drug valganciclovir. It is usually administered (5 mg/kg) via a central line over 1 h and should be handled as a cytotoxic agent. Valganciclovir is well absorbed and a dose of 900 mg twice daily has been shown to be as effective as intravenous ganciclovir for induction therapy. Significant side effects encountered with these agents include neutropenia, which may require colony-stimulating factor support, and thrombocytopenia. Maintenance treatment may also be given either intravenously (6 mg/kg on 5 days a week) or orally (valganciclovir 900 mg once daily).

Intravenous maintenance therapy requires the insertion of a permanent indwelling catheter and is therefore usually used only when oral administration is not possible. Intravitreal administration of ganciclovir is possible, but rarely used, as this does not confer any systemic protection.

An alternative agent to ganciclovir is foscarnet. It has a less favourable toxicity profile and is thus usually reserved for cases of therapeutic failure with ganciclovir. Its main adverse effects are electrolyte abnormalities, nephrotoxicity that requires dose adjustment or cessation of therapy, and ulceration, particularly of the genitals, which may be prevented by assiduous attention to personal hygiene after micturition. No effective oral formulation is currently available.

Cidofovir requires less frequent administration: two doses at weekly intervals for induction and fortnightly for maintenance, though intravenous administration is again required and nephrotoxicity and other metabolic disturbances are well recognised. A strict regimen of intravenous hydration and oral probenecid (2 g given 3 h prior to infusion and 1 g 2 and 8 h post-cidofovir) must be followed. The risk of nephrotoxicity is increased if cidofovir is co-administered with agents such as tenofovir that are excreted via the same renal tubular anion transporter.

CMV disease of the gastro-intestinal tract usually affects the oesophagus or colon, causing dysphagia and abdominal pain with diarrhoea, respectively. Diagnosis is based upon histological analysis of biopsy specimens. Treatment is as for CMV retinitis induction therapy; maintenance therapy is not usually given unless relapses occur. Neurological disease may present in a variety of ways, is difficult to diagnose and frequently carries a poor prognosis, even with treatment. The optimal agent, dosage and duration of therapy remain undetermined.

Impact of HAART on opportunistic infections

The widespread use of HAART has had a dramatic effect on the incidence, prognosis and clinical aspects of opportunistic infections.

Decreased incidence of opportunistic infections

HAART has resulted in a major reduction in the vast majority of opportunistic infections and a consequent reduction in mortality rates and requirement for hospital admissions.

Withdrawal of prophylaxis

The rise in CD4 count associated with HAART has led to an improvement in functional immunity. Clinical trials have suggested that the withdrawal of both primary and secondary prophylaxis against P. jiroveci pneumonia is safe, and similar results are thought to be likely for cryptococcosis, toxoplasmosis, MAI and CMV. Most would consider withdrawing prophylaxis in individuals on successful HAART if the CD4 count is consistently greater than 200 cells/mm³ (P. jiroveci pneumonia, toxoplasmosis, cryptococcosis) or 100 cells/mm³ (MAI and CMV).

Successful treatment of opportunistic infections

Some previously difficult to treat infections, notably cryptosporidiosis and microsporidiosis, appear to resolve with significant CD4 count improvements associated with HAART. However, in up to 30% of individuals, initiation of HAART can be associated with a clinical deterioration known as immune reconstitution inflammatory syndrome or IRIS. This typically occurs in the first 2–6 weeks after starting HAART and presents as fevers and localised symptoms pertaining to a recently treated or previously undiagnosed opportunistic pathogen. It is most frequently seen with mycobacteria (lymphadenitis with both M. tuberculosis and MAI) and CMV (vitritis, iritis and retinal oedema), but has been reported for most infections. Diagnosis is difficult as the signs and symptoms mimic those of resistant disease, non-adherence and co-morbidity. No management strategy is proven but corticosteroids, NSAIDs and immunomodulatory therapies (IL-2 and GM-CSF) have all been used.

Cancers

Although there are a number of malignancies associated with HIV infection, the most common are Kaposi’s sarcoma and lymphoma.

Kaposi’s sarcoma

This is the most common malignancy in people with HIV infection and may be triggered by infection with human herpes virus 8 (HHV-8). The majority of lesions affect the skin and appear as raised purple papules. These may be single or multiple and in severe cases may result in oedema, ulceration and infection. Visceral involvement is not uncommon but rarely causes clinically significant disease. In some cases, no therapeutic intervention is necessary and cosmetic camouflage may be sufficient. Indeed, treatment of HIV with antiretroviral therapy usually results in improvement, and in most cases, complete resolution, of Kaposi’s sarcoma. When individual lesions are troublesome, local radiotherapy or intralesional vinblastine can be beneficial. Alitretinoin gel (0.1%) (9-cis-retinoic acid) is a topical, self-administered therapy approved for the treatment of KS in the USA but not licensed in Europe. In cases of widespread cutaneous disease or significant visceral involvement, systemic chemotherapy is used, though this is often withheld until the potential benefits of HAART have been established. The liposomal anthracyclines and taxanes are now the mainstay of systemic chemotherapy for Kaposi’s sarcoma (Bower et al., 2008).

Lymphomas

The most common lymphomas in patients with HIV infection are high-grade B-cell (non-Hodgkin’s) types. Primary central nervous system lymphomas, which are extremely rare in the general population, are more common in individuals with HIV infection but tend to occur only in those with severe immunosuppression. Diagnosis of lymphoma is usually based upon histological confirmation from biopsy specimens. This may not be possible for primary central nervous system disease. The advent of HAART has dramatically reduced the incidence of all lymphomas. However, whilst it has similarly improved the outcome for most cases, this is unfortunately not true for primary central nervous system lymphomas. Lymphoma of the central nervous system is associated with an extremely poor outcome, and in many cases, even palliative radiotherapy or corticosteroids confer little benefit.

The optimal therapy for HIV-associated lymphomas has yet to be determined, but the management plan should either be supervised by, or have input from, an oncologist with relevant specialist experience. The outcome for patients with relatively preserved immune function or those receiving HAART is comparable to that in the general population. However, many individuals are unable to tolerate treatment without dose modification. Drug interactions between antiretrovirals (particularly PIs) and chemotherapeutic agents need to be considered to minimise the risk of treatment-limiting toxicities, whilst the proactive use of colony-stimulating factors may enable optimal dosing.

Cervical intraepithelial neoplasia and anal intraepithelial neoplasia

These are both associated with human papillomavirus infection, are more common in individuals with HIV infection and may progress to cervical cancer and anal cancer, respectively. It is currently believed that HAART reduces the progression of cervical intraepithelial neoplasia (CIN) but this does not appear to be true of anal intraepithelial neoplasia (AIN), and there are concerns that the incidence of this malignancy may increase with improvements in HIV survival.

The optimal management is a combination of early diagnosis, surgery, chemotherapy and/or radiotherapy. It is possible that screening by smear tests and early treatment, for example, with imiquimod, may reduce the need for such aggressive treatment approaches.

Neurological manifestations

Neurological symptoms may be due to opportunistic infections, tumours or the primary neurological effects of HIV.

HIV encephalopathy or AIDS dementia complex (ADC) is believed to result from direct infection of the central nervous system with HIV itself. Individuals who may otherwise be physically well can be debilitated by profound cognitive dysfunction and amnesia. Although psychometric test results are usually suggestive of the underlying aetiology, it is wise to rule out any other cause with brain scanning and CSF analysis.

The incidence of ADC has reduced dramatically with the use of HAART, and similarly, the use of HAART has been anecdotally associated with an improvement in outcome in many cases. Whilst it is known that the central nervous system penetration of some antiretroviral agents is better than others, the beneficial effects on ADC do not appear to be limited to those agents which penetrate well. Nonetheless, many clinicians would choose to include at least one agent with good penetration of the central nervous system in most HAART regimens, particularly in individuals with cognitive impairment. More recently, there have been increasing reports of more subtle cognitive impairment in patients with HIV receiving effective HAART therapy. The role of varying neuropenetrative agents in either the aetiology or the management of such patients is currently under investigation.

Progressive multifocal leucoencephalopathy (PML) is caused by JC virus and may, at presentation, appear similar to a cerebrovascular accident but will have characteristic white matter lesions on an MRI scan, with or without the presence of JC virus in the CSF. In many cases, the introduction of HAART prevents progression of disease, but it is unlikely to reverse the functional deficit at presentation. The role of adjunctive cidofovir in treatment remains controversial.

Hepatitis B co-infection

There are a number of ways in which HIV can impact on hepatitis B (HBV) infection:

• Hepatitis B vaccination is less successful

• Hepatitis B is less likely to be cleared and hence more likely to become chronic

• Hepatitis B infection is likely to be associated with higher hepatitis B DNA levels

• Progression to cirrhosis is more rapid

• Hepatocellular carcinoma is more common.

Although many individuals with HIV will have hepatitis B serological markers suggestive of previous infection, only 6–10% in most series have active hepatitis B infection.

Management requires an understanding of both viruses and is facilitated by the availability of drugs with dual HIV and hepatitis B activity (Brook et al., 2010). The HAART regimen should include two agents with anti-hepatitis B activity, usually tenofovir with either emtricitabine or lamivudine. If HIV develops resistance to these agents, they can still be continued for their anti-hepatitis B activity.

Hepatitis C co-infection

HIV also impacts on hepatitis C infection in a number of ways:

• Hepatitis C is less likely to be cleared spontaneously

• Higher levels of hepatitis C RNA are seen

• Hepatitis C progresses to cirrhosis more rapidly

• Hepatocellular carcinoma occurs more frequently

• Response to hepatitis C therapy is poor.

In the UK, approximately 5–10% of individuals are co-infected with hepatitis C; there are increasing reports of acute hepatitis C infection in gay men with HIV. In Eastern Europe, where the predominant route of HIV transmission is needle sharing, co-infection rates of over 50% are reported.

The management of hepatitis C/HIV co-infection is complicated (Brook et al., 2010). If treatment for hepatitis C is needed for someone on HAART, the antiretrovirals used must be compatible with hepatitis C therapy (pegylated α interferon and ribavirin). Didanosine, abacavir, stavudine and zidovudine should be avoided. It is commonplace to treat individuals with co-infection for a longer duration, for example, 48 weeks rather than 24 weeks for individuals with genotype 2 and 3 hepatitis C, and possibly 72 weeks for genotypes 1 and 4 – dependent upon early virological responses. Side effects of hepatitis C therapy tend to be more frequent and severe in the co-infected population. The proactive use of the colony-stimulating factors erythropoietin and G-CSF may enable optimal dosing of hepatitis C therapy and thereby improve outcome. The management of HCV is likely to be revolutionised by the availability of increasing numbers of Directly Acting Antivirals (DAAs) of which a large number are currently in development. The optimal way to use these and the potential drug-drug interactions with antiretrovirals have yet to be fully elucidated.

Women with HIV

In addition to the general points covered elsewhere in this chapter, specific issues for women with HIV include:

• Cervical screening should be carried out at least annually, to check for gynaecological manifestations of HIV

• Drug toxicity may manifest in different ways or occur with different frequencies, for example, nevirapine hypersensitivity reaction occurs at a lower CD4 count in women whilst lipodystrophy syndrome may also present differently, with breast hypertrophy commonly seen.

Pregnancy and contraception impact on the medicines prescribed for women of childbearing potential. The factors that need to be taken into account (where appropriate) may include:

• Interactions between many antiretrovirals and oral and injectable contraceptive agents (see Table 41.7). There are no interactions with the levonorgestrel-releasing intrauterine system, Mirena. (Note: barrier methods should also be recommended in addition to hormonal contraception, to prevent transmission of HIV and other sexually transmitted infections)

• Potential teratogenicity of the drugs prescribed

• Possible increased toxicity of the drugs prescribed to both mother and child

• The use of antiretroviral agents and other strategies to reduce vertical (mother-to-child) transmission of HIV.

Table 41.7 Interactions between antiretroviral (ARVs) and contraceptive agents

The use of combination antiretroviral therapy, with or without caesarean section, with zidovudine as PEP to the neonate, together with a non-breast feeding strategy has reduced vertical transmission rates from 35% to less than 1% where HIV status is known and where antiretroviral therapies are widely available.

Nonetheless, questions remain regarding the optimal therapies to use, the risk of transmission of resistant virus to the neonate and the risk of toxicity from antiretroviral agents administered during pregnancy, particularly in the first trimester. Ideally, all HIV-positive women should be counselled regarding these issues before they become pregnant, so they can make informed decisions regarding both therapy and timing of pregnancy.

Guidelines are available that set out the management of HIV infection in pregnant women and the prevention of mother-to-child transmission (de Ruiter et al., 2008). In general, intervention reflects a risk/benefit evaluation between the efficacy of reducing transmission and the potential harmful effects to the mother and fetus. Where HAART is clinically indicated for the mother herself, this should utilise a regimen of optimal efficacy with a favourable safety profile. Where therapy is initiated to reduce transmission, this is usually a HAART regimen but could be zidovudine monotherapy if the viral load is low (consistently <10,000 copies/mL). In the latter situation, the time of therapy initiation is guided by the viral load (Read et al., 2010):

if <10,000 copies/mL – start by 26 weeks;

if >10,000 copies/mL – start by 20 weeks;

if >32,000 copies/mL – start without delay.

Adjunctive caesarean section is recommended where monotherapy is used or there is a detectable viral load prior to delivery, but may not be essential if the mother is on fully suppressive HAART. Breastfeeding should be avoided and antiretroviral therapy is usually administered to the newborn for 4 weeks after birth. In the developing world, different strategies may be adopted. For example, single-dose nevirapine monotherapy is effective in reducing transmission but may be associated with the development of resistance, and exclusive breast feeding, with or without antiretrovirals for the baby, may be recommended where water safety is poor.