Antimicrobial Chemotherapy, 4th Edition
The therapeutic use of antimicrobial agents
- L. Irving
Diseases caused by viruses are extremely common and range from trivial coughs and colds to life-threatening infections. Most viral infections are self-limiting, although some that normally cause relatively minor problems can be serious in immunosuppressed patients or neonates. Some viruses nearly always cause severe disease regardless of the initial status of the host. Rabies, human immunodeficiency virus (HIV), and several of the viruses causing haemorrhagic fevers fall into this category.
Most success in the control of serious virus infections has been achieved by immunoprophylaxis, and to a lesser extent immunotherapy, rather than by the use of specific antiviral chemotherapy. However, there is an ever-increasing list of useful antiviral drugs (see Chapter 6and Chapter 7) and, as knowledge of the molecular biology of viral replication increases, more potential targets are being identified.
All eight herpesviruses that infect man (Table 28.1) exhibit the phenomenon of latency. Following acute infection, virus is not eliminated from the body, but viral DNA remains quiescent without causing any apparent damage to the latently infected cells. Later, virus may be reactivated, the virus replicates, and disease may re-emerge. Herpesvirus infections can therefore be classified as primary, the host's first encounter with the virus; or secondary, due to reactivation of latent virus, or occasionally to re-infection with exogenous virus. The clinical features of primary and secondary herpesvirus infections are often very different.
Table 28.1 Human herpesviruses
Aciclovir is very active against certain herpesviruses and was the first truly selective antiviral agent. Two more oral antiherpetic drugs have been developed since the introduction of aciclovir into clinical practice: valaciclovir (an oral prodrug of aciclovir) and famciclovir (a pro-drug of penciclovir) (p. 85). These
newer agents exhibit greater oral bioavailability, which allows an easier dosing schedule of 1–3 tablets per day, as opposed to the 5 necessary for aciclovir.
Not all herpesviruses are equally sensitive to aciclovir and penciclovir (Table 28.1). Cytomegalovirus (CMV) and human herpesvirus type 6 (HHV6) do not possess thymidine kinase, and cannot activate the drugs efficiently. In practice, these drugs are useful only in the treatment of disease due to the two types of herpes simplex virus (HSV), in varicella-zoster virus (VZV) infection, and in certain unusual manifestations of Epstein–Barr virus (EBV) infection. Clinical trials in various manifestations of HSV and VZV infection have shown the two compounds to be equipotent. Which drug should be prescribed—aciclovir, valaciclovir, or famciclovir—depends on the circumstances. Aciclovir is the only agent licensed for intravenous use, and thus must be used in life-threatening diseases necessitating intravenous therapy. For oral therapy, the choice lies between valaciclovir and famciclovir. As there are no clinical trials evaluating the newer drugs in the treatment of ocular herpes simplex or zoster, aciclovir is still the drug of choice for these conditions.
Herpes simplex viruses
Although most HSV infections are asymptomatic, there can be a wide range of clinical manifestations.
Primary orolabial HSV infection
Symptomatic primary infection, which usually occurs in children, presents with extensive painful blistering and ulceration of the lips, tongue, and mucous membranes of the mouth, sometimes with marked cervical lymphadenopathy and fever. Left untreated, the infection is self-limiting with complete healing in 2–3 weeks. Oral valaciclovir or famciclovir therapy reduces the manifestations of systemic disease, the formation of new lesions, the period of virus excretion, and the time to healing; they are of unquestioned therapeutic benefit.
Recurrent orolabial HSV infection
Resolution of primary disease is followed by the establishment of latency of HSV within the trigeminal nerve ganglion. This virus can be reactivated in later life and track back down the nerve to reach the mouth. The manifestations of reactivated disease in an immunocompetent host are much milder: there is no systemic upset, only a few lesions (cold sores), localized to a small area on the lips; the inflammatory response is less, with less pain and swelling, and the lesions heal in about 5–6 days. Aciclovir therapy enhances resolution of lesions by about 24 h. Given the mild nature of recurrent disease (as compared to primary infection), and the marginal benefit of aciclovir, routine systemic treatment of cold sores is not recommended. However, aciclovir cream is now available over the counter in the UK, and many patients with recurrent cold sores give eloquent testimony to the value of this therapy, whatever the results of double-blind placebo-controlled trials may say!
Intact skin is a very efficient barrier to the spread of viruses. However, in patients with chronic dermatitis or eczema, this barrier is not intact, and virus can spread freely. Eczema herpeticum is a severe complication of primary or secondary orolabial HSV infection in patients with eczema. Lesions extend to large areas on the face, neck and upper chest. Virus may also gain access to the bloodstream through cracks in the skin, putting the patient at risk of death from disseminated HSV infection. Antiviral therapy is mandatory. Patients should be given supplies of valaciclovir or famciclovir so that they can initiate therapy themselves in the early stages of recurrent disease.
Ocular HSV infection
HSV is the commonest infectious cause of blindness in the UK. Children are particularly adept at self-inoculating the eye with virus-infected oral secretions. Spread of virus onto the cornea results in keratitis and may lead to chronic ulceration and inflammatory changes. Management should include the early involvement of specialist ophthalmological advice, and systemic and topical aciclovir.
The clinical manifestations and management of primary and secondary genital herpes are analogous to those of orolabial disease. Thus primary infection results in extensive bilateral, painful ulceration with spread to adjacent skin, inguinal adenopathy, and fever, lasting 2–3 weeks. Patients may experience difficulty in passing urine, and cervicitis with profuse discharge is common in women. Oral valaciclovir or famciclovir is of proven benefit. Recurrent disease is much more localized, with no systemic upset, and shorter-lasting; oral antiviral therapy is of marginal value.
Some unfortunate individuals may suffer severe recurrent attacks of genital herpes as frequently as once a month. The immunological basis for this debilitating and depressing condition is not understood. One approach to the management of these patients is to give them continuous prophylactic antiviral therapy. This results in a considerable decrease in the frequency of attacks, and many patients have now been on prophylaxis for many years, without apparent side-effects.
Pregnant women with genital herpes are at risk of passing on the infection during childbirth. Neonatal HSV infection is a potentially devastating disease, as virus may disseminate to internal organs very easily, and such babies may die of HSV hepatitis, pneumonitis, or encephalitis. Survivors are almost invariably left with severe neurological sequelae. Although mortality from neonatal herpes has improved with the use of intravenous aciclovir, overall morbidity has scarcely been affected, presumably because by the time the diagnosis is made, and therapy initiated, the damage has already been done.
Herpes simplex encephalitis
Herpes simplex encephalitis is the commonest form of sporadic viral encephalitis, with an annual incidence of about 1 case per million population. Most cases occur in individuals with evidence of prior infection with HSV. Thus the disease is usually a manifestation of secondary infection, although the route by which virus gains access to the brain is not clear. Untreated, mortality is high, and survivors suffer severe long-term damage. Diagnosis is difficult. The cerebro-spinal fluid (CSF) usually shows a pleocytosis, but it is very rare to succeed in isolating the virus from CSF. HSV DNA may be detected in CSF at the acute stage by the polymerase chain reaction (PCR), but this test is not universally available. Intravenous aciclovir, in high dosage, improves the prognosis dramatically, provided therapy is started as soon as the diagnosis is suspected, without waiting for laboratory confirmation. Vidarabine was formerly used successfully in herpes encephalitis but aciclovir is at least as beneficial, considerably less toxic, easier to administer and more readily available.
Patients who are immunocompromised for whatever reason are at risk of severe primary or secondary HSV disease. All HSV infections in this group of patients should be treated promptly with aciclovir, if necessary by the intravenous route. In bone marrow transplant recipients, prophylactic aciclovir is recommended.
The primary manifestation of infection with VZV is chickenpox. This is usually a trivial disease of children, despite the dramatic rash. The most important complication is varicella pneumonia, which can be life-threatening. This is considerably more common in adults than in children, and seems to be more common in pregnant women who develop chickenpox. Adults with chickenpox should be referred to hospital for intravenous aciclovir therapy if evidence of respiratory involvement occurs.
It has been suggested that all adults, and even all children, with chickenpox should be given aciclovir, and, indeed, the drug is now licensed in the UK for these purposes. The arguments in favour of this blanket approach are economic rather than medical. Aciclovir allows resolution of the disease 24–48 h earlier than would otherwise be the case, enabling adults, or carers of sick children to return to work that much sooner. The frequency of complications in adults is so small that even large-scale clinical trials have failed to prove that aciclovir therapy reduces their incidence.
Pregnant women who develop chickenpox towards the end of their pregnancy may pass infection on to their baby if it is born within a week of onset of the maternal rash; i.e. before sufficient time has elapsed to allow the mother to generate protective antibodies which can be transferred across the placenta. Neonatal chickenpox is a feared disease, and babies at risk should be given passive immunization with hyperimmune zoster immune globulin. It is also reasonable to give such babies prophylactic aciclovir by mouth, although some would prefer to withhold the drug until the baby developed signs of disease.
The site of latency of VZV is the dorsal root ganglion. The systemic nature of the rash in chickenpox means that dorsal root ganglia up and down the spinal cord become latently infected. Reactivation of infection results in virus travelling down the nerve in question, to reach the dermatome supplied by that nerve. This accounts for the characteristic rash of shingles, or herpes zoster, the clinical manifestation of reactivated VZV infection. The onset of rash is often preceded by pain or abnormal sensation in the distribution of the dermatome. The rash usually heals uneventfully, but pain in the area of the rash may persist long after the rash itself has resolved. This post-herpetic neuralgia is more likely to occur in elderly patients, and can be extremely debilitating.
Other complications of secondary VZV infection occasionally occur: reactivation of virus in the ophthalmic branch of the trigeminal nerve may result in keratitis and damage to the cornea; facial nerve involvement may cause a form
of Bell's palsy; motor nerve damage may also be evident in zoster affecting the limbs; involvement of sacral ganglia may result in urinary and anal retention.
A prolonged or atypical attack of herpes zoster may be the presenting feature of a number of diseases in which host immune responses are impaired. These include malignancy of the reticuloendothelial system and HIV infection. In any immunocompromised patient with zoster, dissemination of the virus in the bloodstream may occur, with the appearance of lesions outside the initial dermatome. This spread of virus also puts the patient at risk of life-threatening internal organ infection.
The aims of antiviral therapy of reactivated VZV infection include alleviation of the pain and discomfort of the rash, and the prevention of complications, including post-herpetic neuralgia and dissemination. Antiviral therapy given early in the course of zoster reduces the incidence of post-herpetic neuralgia and accelerates the resolution of pain. It is therefore logical to use antiviral therapy (aciclovir or its con-geners) in the following types of zoster, provided this is started within 72 h of onset of the rash (or longer for immunocompromised patients): any form of zoster in an immunocompromised patient, no matter how mild the attack appears to be on first presentation; ophthalmic zoster; zoster involving motor nerves, including the facial nerve; sacral zoster; zoster occurring in patients above about 50 who are therefore at increased risk of post-herpetic neuralgia. The exact cut-off age may vary, perhaps in relation to the age of the clinician drawing up the guidelines!
Considerably larger doses of aciclovir are necessary to inhibit EBV replication than HSV or VZV in vitro. This presumably reflects the efficiency of the phosphorylation of the drug by the different viruses. The results of clinical trials of aciclovir in patients with infectious mononucleosis have been disappointing. The only manifestation of EBV infection in which aciclovir is useful is that of oral hairy leukoplakia. This unusual disease arises only in immunocompromised patients, most commonly those with HIV infection. It presents as a whitish coating on the tongue and/or buccal mucosa, which may resemble candidiasis. The lesions are packed with replicating virus. The lesions may respond to aciclovir therapy, but reappear when treatment is stopped, so that long-term use of the drug may be necessary.
Primary CMV infections of immunocompetent individuals are usually asymptomatic, although a small minority result in the infectious mononucleosis syndrome (glandular fever). The site of virus latency is currently unknown, and reactivation fails to induce any recognizable disease in the individual concerned. There are two groups of patients, however, in whom CMV is a significant pathogen: babies infected in utero, and immunocompromised patients.
Transfer of CMV across the placenta may arise from both primary and reactivated maternal infection. Congenital infection affects about 1 in 300 live births in the UK. Most of these babies develop normally, but about 5–10 per cent are born with so-called cytomegalic inclusion disease, which is often fatal. A further 5–10 per cent are normal at birth, but later develop abnormalities such as deafness, impaired neurodevelopment, or learning difficulties. There is presently no effective strategy for the prevention of congenital CMV infection. In theory, antiviral therapy may be beneficial for normal babies who are destined to develop abnormalities, but there is currently no way of identifying those babies at risk, and the toxicity of the available anti-CMV drugs precludes their widespread use for this purpose.
Symptomatic CMV disease may arise from primary or secondary infection in immunocompromised patients, including transplant recipients and those infected with HIV. In these patients, active CMV infection is a multi-system disease. In transplant recipients, it often presents with fever and leucopenia; other complications include hepatitis and pneumonitis. In HIV-infected patients, retinitis is the commonest manifestation, accounting for 85 per cent of all CMV disease in patients with AIDS, but infection may also involve any part of the gastrointestinal tract, the brain, lungs, liver, adrenals and peripheral nerves.
Treatment of CMV infection
Although several anti-CMV drugs are now available, the goal of a safe, easily administered, and effective therapy for CMV infection has not yet been reached. Ganciclovir, foscarnet, and cidofovir must all be given by intravenous infusion, but fomivirsen is administered by intraocular injection. Side-effects may be serious: bone marrow suppression (ganciclovir), nephrotoxicity (foscarnet and cidofovir), and painful penile ulceration (foscarnet) are the most notable.
Ganciclovir is the drug of first choice, with foscarnet as an alternative. Experience with cidofovir and fomivirsen is so far restricted to the management of CMV retinitis. Not all complications of CMV infection respond equally well to therapy. Drugs can save the sight of patients with retinitis, and infections of the bowel and liver also respond well, but treatment of CMV pneumonitis in bone marrow transplant recipients is often unsuccessful because immune responses to the virus may also contribute to the disease process.
Therapy does not eliminate the virus, and in patients who remain immunosuppressed disease often recurs when therapy is stopped. This is a particular problem for AIDS patients with retinitis in whom relapse is almost universal and some form of maintenance therapy is needed. Regular intravenous injections of ganciclovir or foscarnet are effective, but far from optimal, since the drugs penetrate the eye poorly. This has led to the use of intravitreous injections, and
the development of intraocular devices designed to deliver drug continuously at the site of disease. A further possibility about which there is still controversy is to give oral ganciclovir as maintenance treatment, and there are hopes that valgan-ciclovir, an oral prodrug of ganciclovir, may prove useful. The restoration of immune function in HIV-infected patients treated with antiretroviral regimens (p. 330) has resulted in a marked decrease in the incidence of CMV-associated disease in these individuals.
Prophylaxis of CMV disease
Prevention of serious CMV disease in solid organ transplant recipients can be achieved by avoiding transplantation of material from latently infected donors to uninfected recipients. In contrast, bone marrow recipients most at risk of life-threatening CMV disease are those who are seropositive themselves, but receive marrow from a seronegative donor. However, matching of CMV serostatus between donor and recipient in order to prevent the risk of disease is not a practicable solution to the problem. If recipients are deemed to be at risk, administration of some form of CMV prophylaxis is appropriate. The exact regimen is a matter of debate, but trials in different transplant settings have demonstrated the benefit of CMV immunoglobulin, oral ganciclovir, and, surprisingly, oral aciclovir or valaciclovir. Serum levels of aciclovir in these patients are well below the concentrations necessary to inhibit CMV replication, but sufficiently high levels may be achieved inside infected cells.
Human herpesviruses 6, 7, and 8
HHV6 is the causative agent of roseola infantum, one of the many rashes of childhood. Occasional cases are associated with a hepatitis or encephalitis. The virus is sensitive in vitro to ganciclovir, but not aciclovir; however, the value of treatment is unknown. No disease has yet been associated with either primary or secondary infection with HHV7.
HHV8, also known as Kaposi's sarcoma-associated herpesvirus (KSHV), is found in almost all forms of Kaposi's sarcoma, although strict proof that the virus is the cause of this malignant disease has not yet been adduced—the virus could be an innocent passenger within the tumour tissue. The weight of evidence does suggest an aetiological role in tumour formation, though any role for antiviral drugs remains to be defined.
Respiratory tract infections
Upper respiratory tract infections
Infections of the upper respiratory tract with rhinoviruses (of which there are over 100 serotypes), corona-, entero-, adeno-, respiratory syncytial and parainfluenza
viruses are extremely common, but, fortunately, result in very little serious morbidity or mortality. Despite considerable effort, no specific antiviral therapy has been successfully developed for treatment or prevention of these infections. Intranasal interferon spray is effective in the treatment and prophylaxis of common colds that are caused by rhinoviruses, but patients experience nasal stuffiness, irritation, and bleeding, which negates any possible therapeutic benefit. This approach to the management of the common cold has now been abandoned.
Lower respiratory tract infections
Viral infection of the lower respiratory tract is potentially much more dangerous. The commonest viral infections giving rise to bronchiolitis and pneumonia are those due to respiratory syncytial and influenza viruses. Varicella pneumonia and CMV pneumonitis have been referred to above. In addition, giant cell pneumonia is a rare and fatal complication of measles, usually in leukaemic children who escaped vaccination.
Respiratory syncytial virus infection of infants
Bronchiolitis and pneumonia due to respiratory syncytial virus (RSV) infection are relatively common in infants under 1 year of age. Ribavirin reduces viral shedding, hastens resolution of fever, improves respiration, and shortens stay in hospital. The main difficulty is that the drug needs to be administered by use of a small particle aerosol generator. Moreover, for maximum benefit, the infant must breathe nebulized drug for at least 12 h each day. Thus, ribavirin is not usually used in otherwise uncomplicated RSV infection in an immunocompetent host. In babies with congenital immunodeficiency or heart or lung defects, ribavirin can be life saving.
Influenza A (but not B) virus is susceptible to amantadine. Clinical trials in boarding schools and other places in which large numbers of individuals are crowded together have shown that the drug is effective in the treatment and prophylaxis of influenza A. Those treated with the drug have a milder, shorter-lasting infection. About 70 per cent of contacts given the drug are protected from infection. However, amantadine is unpopular in the elderly, who suffer the brunt of serious influenza, because of its stimulatory actions on the central nervous system. Patients become confused and agitated, and may be unable to sleep.
Amantadine may be of more use in the prevention of spread of infection in an institution in which influenza has already occurred. Vaccination of non-infected individuals may take 2 weeks to induce protection, whereas protection with amantadine is demonstrable after a few hours. The drug is also an alternative for those in whom vaccination is contra-indicated because of egg allergy or for other
reasons. Rimantadine, a derivative of amantadine, is said to be equipotent in its anti-influenzal properties but less prone to side-effects. However, it is not currently licensed in the UK. Viral resistance emerges rapidly in patients treated with either of these drugs.
Influenza A and B viruses, as well as the parainfluenza viruses, are sensitive to ribavirin in vitro. The drug has been successfully used in severe lower respiratory tract infection with these viruses, but experience is limited. Much more promising are the neuraminidase inhibitors, zanamivir and oseltamivir (p. 90). Controlled trials have shown a shorter time to the resolution of symptoms and lower usage of antibiotics among patients treated with these drugs. However, they must be given within 30 h of onset of symptoms for maximum benefit—a time frame not often met by patients suffering influenzal symptoms. Moveover, zanamivir has poor oral bioavailability and has to be given by oral or nasal inhalation. Oseltamivir, in contrast, can be administered by mouth in capsules.
It is difficult to predict how valuable these drugs will turn out to be in clinical practice. There is as yet no evidence that they reduce complications, hospital admissions, and deaths in patient groups who are at most risk of serious influenza infection; i.e. those with chronic pulmonary, cardiac, renal or endocrine diseases, the immunodeficient, or the very elderly (>75 years). Their use in a younger, fitter population will depend on any economic benefits to be gained from earlier return to work, as much as on medical grounds.
Chronic viral hepatitis
Patients infected with hepatitis B or C viruses (HBV or HCV) may become chronic carriers. These patients are at risk of continuing inflammation in the liver, leading to chronic hepatitis, cirrhosis, and ultimately, hepatocellular carcinoma.
All chronic carriers of HBV have detectable hepatitis B surface antigen (HBsAg) in their peripheral blood. However, HBsAg-positive individuals can be subdivided further into those in whom HBeAg (a breakdown product of the viral core antigen) can be detected, and those with antibodies to this antigen, anti-HBe. HBeAg is a surrogate marker of active viral replication in hepatocytes. HBeAg positive patients are considerably more infectious, and are at much greater risk of the long-term deleterious consequences of HBV carriage, than are anti-HBe positive patients. For these reasons, therapy for chronic HBV infection is targeted at those HBsAg carriers who are HBeAg positive.
Interferon-αis beneficial in the treatment of chronic hepatitis B infection, but its action is immunomodulatory, rather than antiviral. Interferon enhances the expression of human leucocyte antigen (HLA) class I molecules on the surface of HBV-infected hepatocytes resulting in the efficient elimination of these cells
by circulating cytotoxic T lymphocytes. This in turn reduces the amount of virus in the liver, and the production of infectious virions. HBeAg in the peripheral blood is replaced by anti-HBe.
A typical response of an HBV carrier to interferon is shown in Fig. 28.1. At the onset of therapy, the patient is HBsAg and HBeAg positive, and has a raised alanine aminotransferase level, which indicates continuing liver cell damage. Some weeks after the initiation of interferon therapy, there is a marked rise in alanine aminotransferase. Paradoxically, this indicates a good therapeutic response, since it shows that hepatocytes now express sufficient HLA Class I molecules to allow recognition and destruction by cytotoxic lymphocytes. At the end of therapy, liver function returns to normal and the patient is now anti-HBe positive. Note that the patient remains HBsAg positive. Thus virus has not been completely eliminated, but the risk of ongoing liver damage is much reduced, and the patient is much less of an infection risk to sexual partners and family members. After a number of years, many patients do in fact eventually lose HBsAg.
Fig. 28.1 Treatment of chronic hepatitis B infection with interferon showing the change in alanine aminotransferase (ALT) levels, loss of hepatitis B e antigen (HBeAg), and the appearance of antibody to HBe (anti-HBe). Hepatitis B surface antigen (HbsAg) is not eliminated.
Not all HBV carriers respond to interferon. Those who have been carriers of the virus from birth (the vast majority world-wide) and patients who are immunocompromised (including those with HIV infection, in whom HBV carriage is common, as the routes of transmission of the two viruses are similar) respond poorly. In other patients, response rates are >30 per cent.
Other forms of therapy of chronic HBV infection are being investigated. Most progress has been made with lamivudine (p. 88). Prolonged (>1 year) oral therapy with this drug results in at least a 100-fold reduction in viral DNA titre, and his-tological improvement of liver biopsy appearance in >80 per cent of patients. However, seroconversion rates to anti-HBe are much lower, at best 20 per cent. Cessation of therapy leads to recurrence of HBV replication. More prolonged therapy may improve the seroconversion rate, but this benefit must be weighed against the increased risk of the emergence of resistant virus. Famciclovir therapy similarly reduces HBV DNA levels, but also results in low seroconversion rates. Furthermore, mutation to famciclovir resistance is associated with cross-resistance to lamivudine. Adefovir (p. 87) is also under clinical investigation.
One area in which these newer agents have shown promise is in the prophylaxis and treatment of hepatitis B positive patients undergoing liver transplantation, with reduction in risk of severe disease in the grafted liver.
The relative success of interferon therapy in chronic HBV infection led to trials of its use in chronic HCV infection. About 40–50 per cent of such patients respond to a year's course of interferon, as judged by improvement in liver function and loss of HCV RNA from serum, but about half relapse when therapy is stopped, resulting in an overall cure rate of less than 20 per cent. The mechanism of action of interferon in HCV infection appears to be different from that in HBV, as response is not accompanied by a rise in alanine aminotransferase levels.
Ribavirin therapy provides modest benefit in liver function in chronic HCV disease. When combined with interferon, response rates are much better than with interferon alone, so that interferon monotherapy is no longer justifiable. The mechanism of action of ribavirin in HCV infection and the basis of the synergy with interferon are not clear. Large-scale trials of combination therapy have identified five unfavourable indicators of response: infection with genotype 1 virus (HCV exists world-wide as at least six distinct genotypes); a high viral load; age >40; pre-existing fibrosis; and male sex. Patients with three or more of these factors should be treated for at least a year. About 50 per cent of patients with a more favourable profile show a sustained response (loss of HCV RNA from serum, maintained for at least 6 months after cessation of therapy) after 6 months of therapy.
The management of HIV-infected patients has undergone enormous changes in the past few years. Advances in drug design and in understanding the replication cycle of HIV have led to the development of an increasing number of antiretroviral agents (see Table 7.1, p. 94). Moreover, application of molecular biological techniques allows accurate monitoring of the amount of HIV RNA in
peripheral blood (viral load) in individual patients. Treatment regimens and management protocols are becoming ever more complex. In order to achieve some sort of standardization of clinical practice, bodies such as the British HIV Association publish consensus guidelines for antiretroviral treatment of HIV seropositive individuals. Such is the pace of change that these guidelines need almost annual revision. It is nevertheless possible to discern some important underlying principles governing the appropriate use of antiretroviral therapy, although their detailed application may vary from place to place and over time.
For many years, monotherapy with zidovudine (AZT) was the only option available. However, resistance invariably develops through the selection of mutations in the error-prone reverse transcriptase gene, and AZT monotherapy loses its clinical benefit after about 6 months.
With the advent of more antiretroviral drugs, it became possible to model treatment of HIV infection on that used for the treatment of tuberculosis: use of multiple agents to reduce the chances that the virus will become resistant to all of the drugs simultaneously. This principle has now been validated for a variety of combination therapies and there is no further justification for treating an HIV-infected individual with a single antiretroviral drug.
By necessity, initial trials of combination therapy involved use of two nucleoside analogues. However, triple drug combinations are even more effective in reducing viral load, both in degree and in duration of the effect. With six nucleoside analogues, three non-nucleoside reverse transcriptase inhibitors and five protease inhibitors presently available (Table 7.1; p. 94), there are numerous possible triple drug combinations, and the continuing emergence of new agents will complicate this even further. It is clearly not possible to test each one of these combinations in full scale clinical trials, but several regimens have proved to be highly effective. Such combination therapies are now collectively referred to, somewhat euphemistically, as ‘highly active antiretroviral therapy’ (HAART).
Regimens recommended in the UK include two nucleoside analogues plus either a non-nucleoside reverse transcriptase inhibitor or a protease inhibitor, and two protease inhibitors with or without a nucleoside analogue. The reason for not selecting one drug of each category, which might appear intuitively to be the best idea, is to reserve one class of drugs for the time when initial therapy begins to fail. At this time, the virus present within the patient should at least be fully sensitive to all drugs of the unused class. In the absence of such ‘class-sparing’, the options for salvage therapy in the event of treatment failure are much reduced, as cross-resistance between drugs of the same class is common.
The exact choice of drugs for an individual patient depends on a number of variables. It cannot be assumed that virus in a patient who has received no previous antiretroviral therapy will be sensitive to all drugs, as the patient may have acquired a resistant strain. Moreover, adverse events are often unpredictable and regimens
may need tailoring to individual patients. Clearly, the more drugs a patient takes, the greater the risk of adverse reactions. There is particular concern about the long-term safety of protease inhibitors, owing to their effect on lipid metabolism.
Monitoring of therapy
For many years, the antiviral effect of drugs used in HIV-infected individuals could be determined only by measuring surrogate markers, such as the level of circulating CD4-positive T cells. The advent of viral load testing provides a much more direct measure of the efficacy of an antiviral drug. Viral load is of prognostic importance: the higher the load in an individual patient, the sooner that individual will develop AIDS and die. Thus a new goal of therapy has been defined—to reduce viral load to undetectable levels. It is now possible routinely to detect as few as 50 copies of viral RNA per ml of plasma. Triple drug regimens can achieve and maintain viral load below this level.
Failure of treatment for whatever reason can be detected by a rise in viral titres. Thus, long-term antiretroviral therapy should be monitored by serial viral load measurements, and action taken if these levels rise. If failure is due to the emergence of resistance to one or more of the agents being taken, then appropriate changes in regimen should be instituted. It is possible to analyse the genome of the HIV present in a patient who is failing therapy, and identify which resistance mutations are actually present. Comparative trials have shown that genotypic resistance testing confers a significant benefit on the virological response when choosing therapeutic alternatives. At present, drug-resistance genotyping is expensive and not widely available, and clinicians must make an educated guess as to which of the combination of drugs should be changed.
In parallel with the drop in viral load, CD4 counts rise, and there is at least a partial restoration of the previously damaged host immune system. As the immunodeficiency recedes, then the risk of those complications of HIV infection which define the acquired immune deficiency syndrome also recede. For example, it may be possible to discontinue pneumocystis prophylaxis in patients responding successfully to HAART.
When to treat?
Given that there are now drugs available which can substantially decrease the viral load within an individual patient, and that this can be monitored, the question arises as to when antiretroviral therapy should be started. This decision is an important one. It requires commitment from the patient to take the medication for a prolonged period, possibly for ever. This may incur both physical and psychological morbidity, disadvantages which have to be weighed against the potential therapeutic benefits to be gained.
UK and US guidelines suggest that antiviral therapy is appropriate for patients with a viral load above a certain threshold level, e.g. 10 000 copies/ml, a value
associated with a risk of disease progression. A falling CD4 count is another poor prognostic indicator that should encourage therapeutic intervention. Ideally, therapy should start before the CD4 count drops too low, with irrevocable damage to the immune system, and before the patient suffers an AIDS-defining illness. Ultimately, the decision can only be made after informed discussion between patient and physician.
A special consideration is the use of antiretroviral drugs in pregnancy. There is now compelling evidence that the risk of vertical transmission of HIV from mother to baby is proportional to the maternal viral load during pregnancy. Reduction of maternal viral load by use of antiretroviral agents significantly reduces the risk of transmission. Thus, all pregnant HIV carriers should be offered antiretroviral therapy regardless of whether or not they would otherwise qualify for such treatment. The choice of regimen will depend on the patient's past history of antiretroviral usage, but the optimum protocol has not yet been established. A short course at the end of pregnancy and during labour may offer the best balance between therapeutic effect on the one hand, and cost and adverse toxicity to the mother and fetus on the other.
Importance of compliance
Drugs are only effective for as long as their concentration in the tissues is high enough. Failure to take the requisite regular doses of an antiviral drug leads to a fall in tissue levels, with a consequent risk of escape of the virus from inhibition of replication, and an increased likelihood of mutation to resistance. It is important, therefore, that patients adhere to the prescribed regimens. The more doses that are missed, the faster resistant virus will emerge.
Compliance is a significant problem, especially with multi-drug regimens, in which some tablets should be taken with food and others on an empty stomach; some twice a day, others three or four times a day. These patients may also be on a variety of other agents; e.g. for prophylaxis against pneumocystis pneumonia or recurrent herpes simplex disease. Moreover, in any disease, and HIV infection is no exception, adherence to treatment is poorer in patients with symptom-free disease. Failure to take the tablets must always be considered as an explanation for a sudden rise in viral titre.
Overall strategy of antiretroviral therapy
Once the decision to begin therapy has been taken, the aim must be to ‘hit the virus hard’—if possible to eliminate completely any viral replication. Continuing viral replication, even at very low levels, will eventually lead to drug resistance, and a rise in viral titre. The nadir of viral load is predictive of the duration of the response. Initial treatment regimens should involve combination therapy with three agents. Effort must be made to ensure maximum compliance. Response to treatment must be monitored by regular viral load testing.
By adhering to the above principles, it is possible to render HIV undetectable in most patients within a few weeks of initiation of therapy. Indeed, many patients maintain this status for many months and even years. This led to the hope that if virus could be prevented from replicating until all the resting lymphocytes carrying the viral genome in a latent state reached the end of their life span and died, then the infection might ‘die out’. In this way it may be possible to cure a patient of HIV infection. This optimistic view has been tempered by the knowledge that inhibition of virus replication may need to be maintained for 20 years or more before the latently infected cells disappear. Attention is now turning towards developing ways of encouraging the host immune system to eliminate this reservoir of HIV-infected cells, but this strategy is at a theoretical rather than practical stage.
Another hope which has been dashed was that, after initial induction of potent combination therapy, it might be possible to relax the therapeutic regimen to a maintenance level. This would reduce the risk of long-term adverse effects and of the development of resistance to multiple classes of drugs. Sadly, relaxation of the antiviral regimen merely allows re-emergence of detectable virus.
Despite these gloomy comments, there is no doubt that the outlook for patients infected with HIV has been transformed in the past 5 years, at least for patients in the developed world with access to the latest therapeutic advances. The advent of new classes of antiretroviral drugs, and the application of triple therapy with viral load monitoring has led to a dramatic reduction in HIV-associated hospital admissions, morbidity, and mortality. However, there is no room for complacency. HAART as presently constituted is likely to prove only a temporary respite for most patients. There is still a great need for newer and better anti-retroviral agents, with novel modes of action, and few or no adverse side-effects.
HIV-positive patients are at risk of a wide range of serious infections. These are discussed in Chapter 29.
Papillomavirus infections (warts) are amenable to physical (cryotherapy; surgery) and chemical (podophyllin and derivatives; salicylic acid) therapies. Interferons may have a place in the management of recalcitrant warts. About 50 per cent of warts disappear following intralesional or intramuscular interferon, but they often recur. Topical application of imiquimod, a new imidazoquinoline that stimulates interferon and other cytokines, appears to be modestly effective.
Polyoma virus infections
Progressive multifocal leukoencephalopathy, a disease which occurs almost exclusively in HIV-positive and other immunosuppressed patients, is a manifestation of
reactivation of a polyoma virus in the brain. Although it is of limited availability, the cytotoxic drug, cytosine arabinoside (Ara C) has been successfully used. Cidofovir (p. 86) may prove to be more potent and reliable.
This large group of viruses causes a wide range of clinical disease, from skin rashes in children to paralytic polio. The more serious infections include meningitis and, in immunodeficient patients, chronic enteroviral encephalitis, a rare but extremely debilitating disease. Antiviral therapy presently has little to offer these patients, but this may change. The investigational drug, pleconaril, prevents enteroviral entry into target cells by interfering with the binding of virus to its cellular receptor.
Ribavirin, taken orally, is very effective in the treatment and prophylaxis of lassa fever, a haemorrhagic virus infection with a high mortality rate encountered in certain parts of West Africa.
The treatment of viral gastroenteritis is considered in Chapter 22.