Antivirals in the management of an influenza pandemic

Akey component of the Australian health management plan for pandemic influenza is the targeted use of stockpiled antivirals to contain infection in early pandemic stages by treating clinical cases, providing post-exposure prophylaxis to close contacts of ill patients, and providing longer term prophylaxis to health care workers caring for ill patients.1 During the pandemic, maintenance of essential health services will include use of antiviral drugs for treatment and longer term prophylaxis for health care workers. The use of neuraminidase inhibitors (NIs) is based on outcomes demonstrated in seasonal influenza; NIs should potentially reduce the duration of illness and prevent further infection in an influenza pandemic. The Australian stockpile currently holds 3.8 million courses of oseltamivir, 50 000 courses of oseltamivir suspension for children and 275 000 courses of zanamivir, all available for use within 24 hours. This stockpile is currently one of the highest per capita supplies of NIs held worldwide, and will increase to a total of 8.75 million courses by early 2007.1

Current World Health Organization treatment guidelines for influenza A/H5N1 infection, the strain of concern in current pandemic planning, recommend NIs for treatment and prophylaxis. The guidelines also include a soft recommendation for the use of M2 inhibitors (amantadine and rimantadine) for strains of the virus still susceptible to these drugs. M2 inhibitors have a relatively small role in pandemic planning worldwide because of their adverse effect profile and propensity to produce clinically significant viral resistance.2 The safety and efficacy of antivirals during an influenza pandemic is unknown, and very little is known about their clinical effectiveness against H5N1. Clinical experience with NIs in the treatment of human H5N1 infection is extremely limited and has often been inconsistent with current guidelines.3,4

The uncertainty about the effectiveness of antivirals against H5N1 makes planning for the evaluation of the use of antivirals as part of an overall public health response in an influenza pandemic vital. Given the potentially high pathogenicity of a pandemic strain (H5N1 infection in humans has a mortality rate of 59% in 256 cases worldwide, as of 16 October 20065) and the substantial social upheaval that might arise during a pandemic, it is unlikely that it will be ethically and logistically possible to conduct large randomised clinical trials. This necessitates the use of other research methods, such as observational studies, to assess the safety and effectiveness of NIs early in the course of a pandemic.

There are several concerns with the use of the older class of antivirals, M2 inhibitors (amantadine and rimantadine), for seasonal influenza. These concerns include central nervous system (dizziness, nervousness, and insomnia) and gastrointestinal toxicities, as well as antiviral resistance.11 However, amantadine might be considered as part of combination antiviral therapy by clinicians treating severely ill patients in a pandemic. Current WHO treatment guidelines for pandemic influenza recommend the use of an M2 inhibitor along with an NI only if administered in the context of prospective data collection and if local surveillance data show that the H5N1 variant of concern is known or likely to be susceptible to this drug.12 Amantadine is currently licensed for prophylactic use, while rimantadine is not licensed for use in Australia and will not be available during a pandemic.

Antiviral drug resistance is a major concern when proposing widespread use of antiviral medications. This is particularly evident with the adamantanes, where a single amino acid substitution in the transmembrane portion of the M2 ion channel protein causes amantadine resistance (resistant variants have been detected in up to 80% of patients), leading to treatment failure and transmission of resistant virus to others. Amantadine resistance has been found in human and avian H5N1 strains in China, Thailand, Vietnam and Cambodia, although other variants found in Indonesia, China, Mongolia, Russia and Turkey remain sensitive.13,15 Worldwide seasonal A/H3N2 and A/H1N1 strains are becoming increasingly resistant to the adamantanes.16

Resistance to the NIs is less frequent, occurring in less than 1% of virus isolated from adults and 5%–18% of children treated with these drugs. Resistant strains are not generally associated with clinical deterioration except in rare cases of immunocompromised patients and possibly in patients with H5N1 infection.15,16 Resistance in vivo is usually due to mutations in the neuraminidase region, and in vitro involves the haemagglutinin region, leading to impaired enzyme activity.13 Although oseltamivir resistance in human H5N1 infection has been documented, these mutations occurred in the context of commencement of oseltamivir after 48 hours of illness and suboptimal dosing regimens.17 Circulating strains of avian H5N1 remain susceptible to NIs.13,18 In animal models, oseltamivir resistance mutations lead to reduced viral replication and reduced transmissibility.13 There is little evidence of clinical zanamivir resistance in seasonal influenza or H5N1 strains,19 and zanamivir remains active against most oseltamivir-resistant variants.15

The safety and clinical effectiveness of NIs against the H5N1 strain is unknown, and currently published data include a limited number of observational cases from Thailand and Vietnam. The role of NIs in the management of seasonal influenza, and the safety profile of these drugs, is well known. Although H5N1 is susceptible in vitro to NIs,19 recent studies demonstrate that treatment of human H5N1 infection may require higher doses and longer administration of oseltamivir to be effective.2,20 The pathogenicity of the currently circulating strain of H5N1 is significant (59% compared with 20% in the 1918 pandemic). The NIs were developed and licensed in the setting of seasonal influenza and geared towards identifying the minimal effective dose and regimen duration. NIs may also be used in conjunction with M2 inhibitors in susceptible strains.11 There is little evidence to support these approaches, and research is necessary to evaluate NIs in this “off-label” context.

The paucity of information about the safety and efficacy of NIs against H5N1 is a current cause of concern in pandemic influenza planning. The National Centre in HIV Epidemiology and Clinical Research at the University of New South Wales, in cooperation with infectious diseases physicians from major teaching hospitals in NSW, South Australia, Victoria, Western Australia, Singapore and Hong Kong, has developed research protocols designed to assess the safety and efficacy of antiviral medications, specifically NIs, in the event of an influenza pandemic reaching Australia, Hong Kong or Singapore. These studies will also generate a tissue repository (plasma or sera, nasopharyngeal aspirates, and peripheral blood mononuclear cells) for prospective and retrospective laboratory research.

The three Pandemic Influenza — Prospective Evaluation of Treatment (PIPET) studies will examine NI use in pandemic influenza index cases, as prophylaxis in the contacts of individuals with pandemic influenza infection, and in essential workers at high risk of exposure who receive long-term prophylaxis (Box 2). To enrol patients in the index case study, the diagnosis of influenza infection will be made using the applicable clinical case definition. At the time of analysis of study data, more stringent criteria will be used to identify the evaluable populations for the planned research questions. Currently, the necessary laboratory techniques for confirmatory purposes are not available. The planned studies are purely observational, as we took the view that the conduct of randomised controlled trials during pandemic influenza would be operationally problematic. Furthermore, randomisation may not be ethically possible in a study of a disease about which we know relatively little, as it would be difficult to establish clinical equipoise.

To allow the PIPET studies to be conducted by any doctor and in any location where pandemic cases are diagnosed, web-enabled case report forms will be used and data collection will be kept to a minimum. Web-based data collection will allow the study to proceed in the event of restricted movement during a pandemic, as it will be possible to provide support to study centres remotely. Any doctor or study centre with a computer and Internet connection will be able to participate in the PIPET studies.

The PIPET study group aims to capture data on as many as possible of the clinically diagnosed cases of pandemic influenza and individuals receiving prophylaxis during the early stages following a pandemic being declared. To achieve this goal, we are asking that each clinician likely to be managing influenza patients in the early stages of a pandemic in Australia consider participating in this project by contacting one of the authors to receive protocol documents and support in the ethics submission process. The critical aim of this study is to capture data for all patients from their first diagnosis of pandemic influenza, making widespread participation of hospitals vital. It is hoped that early data on antiviral treatment and prophylaxis will be used to inform further policy decisions should pandemic influenza become more widespread in Australia. The data collected in this study will not be the only measure used to determine policy; however, we will be able to demonstrate overall rates of morbidity and mortality with some accuracy, as well as identify subgroups in whom there is evidence of either increased or decreased effectiveness of NIs.

Pandemic planning in Australia and around the world has required a significant investment, both financial and temporal, to develop national responses to pandemic influenza. A cornerstone of this planning has been stockpiling antivirals for use in both treatment and prophylaxis. It is unknown whether oseltamivir will improve clinical outcomes in infected patients and whether resistance will become a clinically important issue in a pandemic. It is difficult to assess if the average reduction of duration of symptoms of 1 or 2 days will be of overwhelming benefit in a pandemic setting: the effect on mortality and hospitalisation, having not been previously measured in seasonal influenza trials, is entirely unknown. For the more relevant concern of transmission by infected individuals, there is little evidence confirming the relationship between symptom duration and viral shedding, so the benefit in this regard is unknown. It is more likely that a benefit to the community will be derived from the short-term prophylactic use of NIs in preventing transmission and illness, and long-term prophylaxis in protecting health care and other essential workers. It is crucial, given the likely climate of fear that will exist during a pandemic, that prophylactic use is guided by evidence rather than panic or politics. Although Australia has an extensive stockpile, judicious and evidence-based use is required to best protect the whole population. The ability to make evidence-based decisions through research during a pandemic is vital.