New treatments for viral respiratory tract infections—opportunities and problems

Introduction

Viruses are the most common cause of respiratory tract infections (RTIs), yet in contrast to the plethora of antibiotics available for the treatment of bacterial RTI, until very recently only three agents were widely approved for the treatment of viral RTIs: amantadine and rimantadine for influenza A, and ribavirin for respiratory syncytial virus (RSV) infection—and amantadine was first marketed in 1966.

In 1999 the novel anti-influenza agents, zanamivir and oseltamivir, were launched. Controversy has attended the failure to reimburse zanamivir therapy in some countries, and the initial advice from the new National Institute for Clinical Excellence (NICE) in the UK that it should not be generally prescribed.¹ With several new therapies for viral RTIs in late stages of clinical development, now is perhaps an appropriate time to take stock of the situation.

Disorders due to viral RTI

The most frequent symptom complex due to viral infection of the respiratory tract is the common cold (coryza). This is most often caused by rhinoviruses (of which there are more than 100 serotypes). Some 20% of colds are caused by coronavirus infection; identical symptoms can be caused by influenza and parainfluenza viruses, adenoviruses, RSV and various enteroviruses.² The same viruses cause descending infections of the respiratory tract, but the aetiology differs in frequency; adenoviruses are common causes of viral pharyngitis and tonsillitis, while croup (subglottic oedema due to laryngotracheitis) is particularly associated with parainfluenza virus infection.³ All these viruses can cause acute bronchitis in healthy people, and infective exacerbations in patients with chronic obstructive pulmonary disease (COPD).⁴ Viral RTIs are also implicated in exacerbations of asthma⁵ and cystic fibrosis.⁶ Acute bronchiolitis in infants is most commonly the result of RSV infection, but this virus is also an under-recognized cause of lower RTI in the elderly.⁷ Viral infection (particularly with influenza, parainfluenza and RSV) is a common cause of acute respiratory disorders leading to hospitalization in patients with chronic underlying conditions,⁸ particularly COPD, asthma, cardiac disease and diabetes.

Viral pneumonias are uncommon in the previously healthy and are most commonly associated with influenza, but also occur rarely owing to measles and varicella-zoster infection.⁹ Life-threatening viral pneumonias may occur in immunocompromised patients; this increasing group includes subjects with advanced HIV infection, patients receiving anti-cancer treatment, transplant patients on immunosuppressive therapy and the very young (premature neonates) and very old. Pneumonia may be caused by any of the ‘conventional’ respiratory viruses described above, but is commonly due to herpesviruses, particularly cytomegalovirus (CMV) and herpes simplex virus (HSV).¹⁰

In addition to predisposing to opportunistic lung infections, HIV infection may be associated with lymphocytic and non-specific interstitial pneumonitis, and also appears to increase susceptibility to smoking-induced emphysema.¹¹

Rare specific viral RTIs include recurrent respiratory papillomatosis (caused by the human papillomavirus) and the serious pneumonia caused by a hantavirus that has recently been described in parts of North and South America.¹² Chronic or latent viral infection has been implicated in the aetiology of idiopathic pulmonary fibrosis (Epstein–Barr and hepatitis C viruses)¹³ and COPD (adenovirus),¹⁴ but these links remain unproven. Empirical therapy of idiopathic pulmonary fibrosis with nebulized ribavirin (a broad-spectrum antiviral) produced no clinical benefit¹⁵ (and N. J. C. Snell, unpublished observations). It has also been suggested that viral RTIs could be co-factors in the development of hypersensitivity pneumonitis.¹⁶

Therapy for rhinovirus and coronavirus infection

The British Medical Research Council's Common Cold Research Unit laboured for 43 years (1946–1989) without discovering a cure. Several compounds with promising anti-rhinovirus activity in vitro failed to demonstrate useful clinical activity. Both α- and β-interferon, given intranasally before viral challenge, were shown to be effective in protecting against rhinovirus, coronavirus, influenza and RSV infection; however, local side-effects and the fact that they were most effective given prophylactically inhibited their utility in these indications.¹⁷ Recently, better understanding of the morphology and mode of action of the rhinovirus has permitted the development of more specific antiviral agents.

The discovery that a deep surface depression or ‘canyon’ on the surface of the rhinovirus was the site of receptor attachment¹⁸ led to the development of drugs that bind to this site. A series of compounds synthesized by Sterling-Winthrop bind to a hydrophobic pocket in the canyon floor and are thought to act by preventing viral uncoating, through stabilization of the viral capsid.¹⁹ The most promising of this series is pleconaril (WIN 63843), which has been licensed out to Viropharma for clinical development in rhinovirus and other picornavirus infections. It has good oral bioavailability and is widely distributed in the tissues, attaining high concentrations in nasal epithelium, and is well tolerated in animal models and in man.²⁰ Prophylactic administration of pleconaril to healthy volunteers before intranasal inoculation of coxsackievirus A21 (a model of picornavirus RTI) led to significant reductions in viral shedding, nasal secretions and symptom scores compared with placebo.²¹ Phase II clinical trials in approximately 1500 patients presenting with a viral respiratory illness showed that pleconaril reduced the duration of symptomatic illness by about 3 days compared with placebo, and had a similar adverse event profile.²² However, a more detailed abstract of a single large (875 subject) placebo-controlled study found a 1–2.25 day reduction in symptoms that was statistically significant only in patients positive for picornavirus infection by culture or polymerase chain reaction (PCR), who were not taking concomitant cold medications.²³

Agouron Pharmaceuticals has elucidated the structure of the rhinoviral enzyme 3C protease, which splits viral precursor proteins into structural and enzymic proteins essential for replication, and has synthesized a specific 3C protease inhibitor, AG-7088. This has similar in vitro antirhinoviral activity to pleconaril, and was well tolerated when given intranasally up to six times daily in volunteer studies. Phase II clinical trials in experimentally infected healthy volunteers showed significant reductions in symptoms, nasal secretions and viral concentrations compared with placebo.²⁴ Other companies developing rhinovirus 3C protease inhibitors include Lilly, Cytoclonal Therapeutics and Apotex. Lilly is also studying a 2A protease inhibitor. All these projects are currently at the preclinical stage.²⁵

Rhinoviruses attach to specific cellular receptors in order to initiate infection. It has been shown that 90% of rhinovirus serotypes attach to a cell surface glycoprotein known as intercellular adhesion molecule-1 (ICAM-1). The remaining 10% probably all utilize members of the low-density lipoprotein receptor (LDLR) family.²⁶ Two pharmaceutical companies (Boehringer Ingelheim and Bayer) have managed to produce truncated, soluble forms of ICAM-1 by genetic engineering; since free ICAM-1 will attach to the rhinovirus binding sites, the virus should effectively be ‘neutralized’ and prevented from attaching to the cell-surface receptors and initiating replication. Boehringer's product, tremacamra, has demonstrated a potent inhibitory effect on human rhinovirus in vitro.²⁷ In phase II clinical trials, when given intranasally, tremacamra significantly reduced the symptoms of experimental rhinovirus colds, regardless of whether it was administered before or after viral challenge. There were no adverse effects, and drug administration was not associated with the development of neutralizing antibodies.²⁸ Bayer's product, given intranasally to chimpanzees, has been shown to protect against infection on subsequent challenge with virus.²⁹ Despite these promising findings clinical development of both projects is apparently on hold. One problem is the short residence time of intranasal medication, owing to mucociliary clearance and evacuation by sneezing and nose-blowing;³⁰ another is the fact that protection is specific for only 90% of rhinovirus strains and not for other causes of the common cold.

The difficulty in attaining and maintaining effective concentrations of drug in the nasopharynx is probably the reason for the failure of a number of compounds that showed promise in preclinical studies—for example, the compound 7-thia-8-oxoguanosine (NARI 10146) was highly effective as prophylaxis against an otherwise lethal coronavirus infection in a rat model, but had no influence on the course of an experimental coronavirus infection in human volunteers.³¹ On the other hand, the anti-influenza agent zanamavir has been proved to be efficacious when administered intranasally; positron emission tomography has shown that 50% of the deposited drug remains in situ for at least 1.5 h.³² The reason for this long residence time is unclear, although in subjects with active influenzal infection mucociliary clearance may be impaired. It may prove possible to prolong the intranasal residence time of inhaled drugs by utilizing thixotropic formulations.³³

Anti-influenza agents

The anti-influenza activity of amantadine was first described in 1964. It and its analogue rimantadine have been shown to be effective both in the prophylaxis and treatment of influenza A infection.³⁴ Their activity appears to be because of interaction with the M₂ membrane protein of the influenza A virus, which acts as an ion channel with a pH regulatory function.³⁵ However, their usefulness is limited by their specificity for influenza A (they have no activity against influenza B, which does not possess the M₂ membrane protein), and by the occurrence of central nervous system side-effects (less frequent with rimantadine).³⁶ Both compounds are fetotoxic in rodent studies.³⁴ Administration by nebulization has been tried in an attempt to improve the risk:benefit ratio, but has not become an accepted route of treatment.³⁷ Resistance to both agents develops readily during treatment (see below). Until recently the only other agent available for the treatment of influenza was the broad-spectrum antiviral nucleoside, ribavirin. This is active against both influenza A and B in vitro; it showed low clinical efficacy in a series of trials when given orally, but by the nebulized route was shown to accelerate recovery from influenza A and B infections in infants and young children,³⁸ although it has not been widely approved for this indication.

Advances in our knowledge of the biology of the influenza virus have led to the recent development of a new generation of anti-influenza agents. The major influenza virus surface proteins are haemagglutinin and neuraminidase (sialidase). Haemagglutinin mediates binding and fusion of the viral and host cell membranes; it is the major target for neutralizing antibodies (and hence influenza vaccines) but is highly variable. Neuraminidase is essential for facilitating release of new virions from the host cell surface and preventing their aggregation. Although there is also variability in this protein, the amino acid sequence and three-dimensional structure of the enzyme's active site are highly conserved, making it a suitable therapeutic target.³⁹ Several neuraminidase inhibitors are now in preclinical or clinical development,⁴⁰ and two (GlaxoWellcome's zanamivir and Roche/Gilead's oseltamivir) have received marketing authorization. Both have been the subject of recent comprehensive reviews.^41–43

Zanamivir has poor oral bioavailability and is administered by po inhalation from a dry-powder inhaler (the Diskhaler), although in early volunteer studies it was given intranasally. Clinical trials have shown that it shortens the symptomatic duration of the disease and the time taken to resume normal activities by at least a day, and its use is associated with a reduced consumption of over-the-counter medications and prescribed antibiotics. It has also given promising results when used for the prophylaxis of influenza,⁴⁴ but this is not yet an approved indication. The fact that the nasopharynx is bypassed when administering medication is a potential drawback, and may relate to the fact that treatment has no effect on the incidence of complicating sinusitis and otitis media,⁴⁵ although concentrations in excess of the viral neuraminidase MIC₅₀ have been demonstrated to persist in nasal washings for at least 12 h after oral inhalation.⁴⁶ On the other hand, po inhalation does limit systemic exposure and hence potential systemic side-effects.

In contrast, oseltamivir is well absorbed orally and widely distributed throughout the tissues, attaining high concentrations throughout the respiratory tract. A plasma half-life of approximately 8 h allows bd dosing (zanamivir is also recommended for bd administration, although the serum half-life is shorter at 4–5 h). The zanamivir and oseltamivir clinical trials are not strictly comparable, but overall the oseltamivir studies show a shortening of symptom duration of about 1.5 days, and a greater reduction than for zanamivir in associated antibiotic prescribing, for secondary infections of both the upper and lower respiratory tract.^43,47 Both drugs are well tolerated, although a few cases of bronchospasm following inhalation of zanamivir have been reported;⁴⁸ a small study in mild to moderate asthmatics found no significant effects of zanamivir inhalation on lung function or airway responsiveness.⁴⁹

Several more neuraminidase inhibitors are under development, including analogues of zanamivir and oseltamavir, in addition to novel compounds.⁴⁰ The most advanced of these is a cyclopentane derivative, RWJ-270201 (BCX-1812),⁵⁰ which is under joint development by Johnson and Johnson and BioCryst Pharmaceuticals. Given orally in mouse models of influenza A and B infection it appears to be at least as effective as oseltamivir, with no observed toxicity. Phase II clinical data were promising, and it is currently in phase III clinical trials.⁵¹ Abbott Laboratories is studying a series of novel substituted pyrrolidine neuraminidase inhibitors.⁵²

Other recently described agents with potentially useful activity include a series of compounds derived from podocarpic acid⁵³ and antisense phosphorothioate oligonucleotides,⁵⁴ both of which exhibit in vivo activity against influenza A. Provir (SP303), a naturally occurring phenolic polymer derived from a plant of the Euphorbiaceae family, showed promising activity against influenza viruses in vitro and in vivo (and also against RSV and parainfluenza type 1),⁵⁵ but proved to have very poor oral bioavailability, and clinical development in these indications has been discontinued.

Therapy for RSV infection

There is currently no effective vaccine against RSV infection; clinical trials with a formalin-inactivated vaccine in the 1960s not only demonstrated a lack of protective effect, but infected infants actually experienced more severe disease, probably because of an immune-mediated response. Efforts to develop an effective vaccine continue.⁵⁶ Meanwhile it has been shown that both hyperimmune globulin^56,57 and a humanized RSV-specific monoclonal antibody (palivizumab)⁵⁸ can reduce the incidence of RSV bronchiolitis requiring hospitalization when given prophylactically. Palivizumab significantly decreased the replication of RSV in tracheal secretions of ventilated infants,⁵⁹ but is not currently licensed for treatment of established disease. It is expensive, and it has been suggested that its use is not cost-effective in the majority of premature infants, with or without chronic lung disease.⁶⁰

The only approved therapy for established RSV infection is the nucleoside antiviral, ribavirin. A series of clinical trials and ancillary studies established the safety and tolerability of this agent when administered by nebulization, and suggested efficacy in improving clinical signs, oxygen saturation, and duration and severity of the illness.⁶¹ However, most of these studies were small, and their design and the relevance of the clinical effects seen have been questioned.⁶² Subsequent studies looking at long-term effects of therapy on lung function and bronchial hyperactivity have given varying results, some showing marked benefit, some none.⁶³ Ribavirin therapy is expensive and inconvenient (current recommendations are to nebulize for 12–18 h daily, although studies have shown equivalent effects from a higher-dose, shorter-duration administration).⁶⁴ Its use should probably be restricted to high-risk infants with pre-existing cardiopulmonary disorders or immune deficiencies. Several ribavirin analogues have been shown to inhibit RSV replication in vitro, but in vivo studies have not been reported.⁶⁵

α-Interferon has been shown to have modest beneficial effects on the course of RSV infection when given intramuscularly⁶⁶ or as an aerosol,⁶⁷ but this mode of therapy does not seem to have been pursued. Several other novel agents are at an early stage of development,^65,68 including a nebulized antisense oligonucleotide (NIH 315) and a number of compounds that interact with the viral F protein and inhibit fusion with the host cell, including the biphenyl derivative CL-387626 (Wyeth-Ayerst), VP14637 (Viropharma) and a RhoA-derived peptide.⁶⁹

Rare or novel viral RTI

Juvenile laryngeal papillomatosis (recurrent respiratory papillomatosis) is a rare condition (prevalence approximately 2 per 100 000 population in the USA)⁷⁰ caused by the human papillomavirus, and characterized by recurrent papillomas of the aerodigestive tract, most frequently affecting the glottis where they can cause hoarseness, stridor and respiratory distress; rarely, spread to the airways and lung parenchyma can occur, and the condition can progress to malignancy. The condition also affects adults, when it is generally less aggressive.⁷¹

Conventional management involves frequent endoscopic surgery. Treatment with several antiviral agents has been studied; intramuscular α-interferon has been associated with sustained or repeated responses in a substantial proportion of patients.⁷² There have been anecdotal reports of benefit from ribavirin⁷³ and from acyclovir therapy.⁷⁴ A promising new approach appears to be the use of cidofovir, either intralesionally⁷⁵ or intravenously.⁷⁶

Hantaviruses have been known for many years as the cause of haemorrhagic fever with renal syndrome (HFRS), which varies from a mild condition (nephropathia epidemica) in Scandinavia, to a more serious form in Asia (Korean haemorrhagic fever). Respiratory symptoms are not a feature. However, in 1993 a cluster of cases of severe respiratory disease occurred in the South Western USA, characterized by non-cardiogenic pulmonary oedema and a high mortality rate. These were eventually shown to be due to a hantavirus (Sin Nombre virus), and subsequently further cases due to the same or closely related viruses have been reported in both North and South America.¹² Intravenous ribavirin has been shown to reduce mortality in HFRS,⁷⁷ but an open-label trial in hantavirus pulmonary syndrome was not associated with a dramatic effect on mortality.⁷⁸ Treatment with corticosteroids has given promising results in South America.¹²

Viral pneumonitis in the immunosuppressed

Immunosuppressed subjects, particularly bone-marrow and solid-organ transplant patients, are prone to develop severe viral RTI. Normal community viruses (adenovirus, influenza, parainfluenza and RSV) are commonly implicated, but a specific problem in these patients is severe herpesvirus infections, most importantly due to CMV and HSV. The relative scarcity of such patients and the fact that they are usually already the subjects of polypharmacy have made formal clinical trials of antiviral agents difficult.⁷⁹ There are no therapeutic agents approved specifically for the treatment of parainfluenza or adenovirus infections; because of its broad spectrum of antiviral activity, ribavirin has been tried empirically in RSV infection,⁸⁰ influenza,⁸¹ parainfluenza⁸² and adenovirus infection,⁸³ with varying degrees of success.

The management of herpesvirus infections in transplant patients has been the subject of a recent report from a working party of the British Society for Antimicrobial Chemotherapy.⁸⁴ The antiviral agent ganciclovir has been shown to be of value in the prophylaxis of CMV pneumonitis associated with bone-marrow⁸⁵ and solid-organ⁸⁶ transplants, while acyclovir is inferior to ganciclovir in the prophylaxis of CMV infection and obliterative bronchiolitis in lung transplant patients.⁸⁷ Current therapy for established CMV pneumonitis is a combination of ganciclovir with immune globulin, based on open-label studies that showed, respectively, a significantly improved survival on this regimen compared with historical controls⁸⁸ and significantly better survival with the combination therapy than with ganciclovir or immune globulin alone.⁸⁹ Foscarnet is also active against CMV but its use is limited by renal toxicity. Cidofovir is undergoing clinical trials. Novel agents (e.g. BAY 38-4766)⁹⁰ are under investigation; an interesting new approach is the use of a bioengineered proteinase inhibitor (α1-PDX), a selective and potent furin inhibitor, which is highly active against herpesviruses (including CMV) in cell-culture models.⁹¹ High-dose acyclovir (plus α-interferon) was ineffective in the treatment of CMV pneumonitis following bone-marrow transplantation,⁹² but acyclovir is currently the recommended agent for prophylaxis and treatment of HSV pneumonitis^84,93 and also for varicella-zoster virus infection,⁸⁴ although both valaciclovir and famciclovir exhibit better oral availability and may well become the treatments of choice in these two infections.

The problem of resistance

The problem of increasing resistance to antiviral agents was highlighted in an editorial in the New England Journal of Medicine in 1989.⁹⁴ However, selection of a strain of influenza A virus in vivo with increased resistance to amantadine was demonstrated as early as 1970.⁹⁵ The emergence of resistance to rimantadine during therapy was first noted in 1987.⁹⁶ Numerous subsequent reports have documented the development of resistance to both amantadine and rimantadine during treatment.⁹⁷ The genetic basis for the development of resistance appears to be a single amino acid change in the transmembrane portion of the M₂ protein, which is the target area for these two antiviral agents (leading to blockade of viral fusion or release).⁹⁷

Although transmission of resistant influenza virus has been demonstrated in the family setting and probably during nursing-home outbreaks,⁹⁸ and prolonged shedding may occur in immunosuppressed patients,⁹⁹ persistence of resistant strains in the community does not seem to be a problem, perhaps owing to the regular emergence of new strains.⁹⁸ Interestingly, the development of resistance during therapy appears to have little effect on the clinical outcome in treated patients,^97,98 although it does limit the use of the adamantanes for concurrent treatment of an index case and prophylaxis of contacts. A survey of rimantadine resistance in field isolates of influenza A virus in 43 countries and territories, in some of which rimantadine has been widely used for more than 20 years, found only 0.8% of 2017 isolates to be drug resistant,¹⁰⁰ which is quite reassuring.

The nucleoside antiviral agent, ribavirin, has a broad spectrum of activity against respiratory tract and other viruses. In contrast to amantadine and rimantadine, viral resistance to ribavirin has not been demonstrated. This may be because at least three mechanisms of antiviral activity have been described:¹⁰¹ depletion of cellular nucleotide pools, inhibition of cap formation of mRNA and inhibition of influenza virus RNA polymerase.

Strains of both influenza A and B virus resistant to the neuraminidase inhibitor zanamivir have been produced by repeated passage in vitro, owing to mutations in the haemagglutinin or neuraminidase. However, resistance to zanamivir develops much less readily than it does to amantadine and rimantadine in similar experiments, and has not been shown in animal models under conditions that readily select for resistance to amantadine and rimantadine.¹⁰² The zanamivir-resistant strains appear to replicate less well in culture and are very much less infectious in mice than the wild-type virus.¹⁰³ In clinical practice only one case of resistance developing to drug during therapy has been reported, in an immunocompromised child receiving prolonged treatment with zanamivir.¹⁰⁴ Oseltamivir-resistant influenza virus has been detected in in vitro experiments and in <1% of isolates from treated patients; this is because of mutations at the active site of neuraminidase, which are associated with greatly impaired replication in vivo, suggesting that the development of resistance is unlikely to limit the clinical usefulness of this class of drugs in immunocompetent subjects,⁴¹ for therapy or prophylaxis.

The development of resistance by herpesviruses to specific antiviral agents has been well studied. At least three mechanisms have been described in HSV, two involving mutations in the viral enzyme thymidine kinase, the other involving alterations in the viral DNA polymerase.¹⁰⁵ Although acyclovir-resistant HSV is generally considered to have reduced virulence, cases of severe HSV pneumonia due to resistant strains of virus have been observed.¹⁰⁶ Variants of CMV conferring multidrug resistance have also been reported.¹⁰⁷ In general, the development of resistance has been associated with prolonged monotherapy, usually in immunocompromised patients.

The use of soluble ICAM receptor therapy as prophylaxis against rhinovirus over several winter months could put a selection pressure on those viruses that do not use ICAM-1 as their attachment receptor and they might become dominant in the community. Strains of rhinovirus resistant to several novel anti-rhinoviral agents including dichloroflavan and chalcone, neither of which is still in development, have been described.¹⁰⁸

What can be done to limit the development of viral resistance? By analogy with bacterial infections one can propose a general rule that monotherapy of acute viral infections in patients with normal host defences seldom leads to the development of clinically significant resistance, whereas monotherapy of chronic infections is much more likely to, particularly in the immunosuppressed (this was discovered half a century ago with tuberculosis, and the same lesson had to be relearned in the treatment of HIV infection). Additive or synergistic effects of combinations of agents active against influenza viruses have been reported^109,110 and it may be prudent to consider whether we should be prescribing combination chemotherapy for viral RTI, where the patient is immunosuppressed or treatment is likely to be prolonged. Combination therapy with ganciclovir and foscarnet has been proposed for the treatment of CMV patients with a high viral load.⁸⁴ Now that we have reached a point where effective chemotherapy for most viral RTIs is a real possibility, it would be sensible to try to avoid a similar situation to that with antibacterial agents, where widespread resistance has developed largely because of inappropriate prescribing.

Table.

Selected viral pathogens of the respiratory tract, approved therapy and key antiviral agents undergoing clinical trials or with evidence of clinical activity

Causative virus	Currently approved therapya	Agents under clinical investigation or used empirically
aLicensed in North America and/or EU countries.
bProphylaxis only.
Rhinovirus	–	t-ICAMs (tremacamra, BAY z 9700)
		pleconaril
		AG-7088
Influenza A virus	amantadine, rimantadine
Influenza A + B virus	zanamivir, oseltamivir, ribavirin	RWJ-270201
Respiratory syncytial virus	ribavirin
	palivizumabb
Parainfluenza virus	–	ribavirin
Adenovirus	–	ribavirin
Papillomavirus (recurrent respiratory papillomatosis)	–	α-interferon
		cidofovir
Hantavirus	–	ribavirin
Cytomegalovirus	ganciclovir	cidofovir
	immune globulin	foscarnet
Herpes simplex virus	acyclovir	famciclovir
		valaciclovir
Varicella-zoster virus	acyclovir	famciclovir
		valaciclovir