As the potential for an influenza pandemic has galvanized the medical community and the public into action, physicians and patients alike have been heartened by the availability of effective antiviral drugs. The neuraminidase inhibitors provide valuable defenses against pandemic and seasonal influenza, and physicians have been flooded with requests from patients for personal supplies of oseltamivir (Tamiflu). A benefit of having oseltamivir at home is that the sooner the drug is taken after the onset of symptoms, the better its clinical efficacy.¹ And certainly, enabling ill people to stay home and out of waiting rooms and pharmacies should limit the spread of influenza. So it is not surprising that many believe there should be a supply of oseltamivir in every medicine cabinet. This scenario, however, is potentially dangerous. Misuse of the drug could rob us of the advantages that neuraminidase inhibitors provide, by favoring the emergence of oseltamivir-resistant influenza virus. The potentially serious consequences of oseltamivir resistance in patients with influenza A (H5N1) virus infections is alarmingly underscored by the report by de Jong and colleagues in this issue of the Journal (pages 2667–2672).

One strength of the neuraminidase inhibitors oseltamivir and zanamivir (Relenza) over the older adamantanes is that they are less prone to selecting for resistant influenza viruses.² Indeed, no virus resistant to zanamivir, which is currently available only in an inhaled form, has yet been isolated from immunocompetent patients after treatment. The recent emergence of oseltamivir-resistant variants is therefore a matter of immediate concern.

Why is resistance developing to oseltamivir? Several years ago,structural analysis³ predicted that aspects of the chemical structure of oseltamivir (not present in zanamivir) could facilitate the development of resistance mutations that would permit neuraminidase to function, allowing drug-resistant virus to survive and propagate. This prediction is now being validated by clinical data.

The mechanism of the development of resistance is illustrated in the diagram. The influenza neuraminidase releases newly formed viruses from infected cells, allowing them to spread from cell to cell. The inhibitor molecules mimic the natural substrate of the influenza neuraminidase (the sialic acid receptors) and bind to the active site, preventing neuraminidase from cleaving host-cell receptors and releasing new virus. All the resistant variants thus far have contained specific mutations in the neuraminidase molecule; but since neuraminidase serves an essential purpose, mutations that allow the virus to survive must not inactivate the enzyme.

View larger version (42K): [in this window] [in a new window]   Mechanism of Resistance to Oseltamivir. The neuraminidase active site changes shape to create a pocket for oseltamivir, whereas it accommodates zanamivir without such a change (Panel A). Any of several mutations may prevent the binding of oseltamivir by preventing the formation of this pocket (Panel B); the oseltamivir-resistant virus can nonetheless bind to the host-cell sialic acid receptor and to zanamivir. The pocket for oseltamivir, illustrated by key amino acids in Panel C, is created by the rotation of E276 and bonding of the amino acid to R224 — events that are prevented by the mutations R292K, N294S, and H274Y and therefore result in resistance to oseltamivir. An E119V mutation may permit the binding of a water molecule in the space created by the smaller valine, also interfering with oseltamivir binding. None of these mutations prevent the binding of zanamivir or of the natural sialic acid substrate. An animated version of this figure is available with the full text of the article at www.nejm.org.

 
To accommodate the bulky side chain of oseltamivir in the active site, the neuraminidase molecule must undergo rearrangement to create a pocket (Panel A). Zanamivir, by contrast, binds to the active site without any rearrangement of the molecule. Several mutations that limit the necessary molecular rearrangement may diminish the binding of oseltamivir (Panel B). Molecular-level analysis (Panel C) shows that the amino acid termed E276 must rotate and bond with R224 to form a pocket for the side chain of oseltamivir. The mutations R292K, N294S, and H274Y inhibit this rotation and prevent the pocket from forming, resulting in resistance to oseltamivir. The mutations nonetheless allow the binding of natural sialic acid substrate, so mutated virus can survive and propagate. In contrast, the binding of zanamivir does not require any reorientation of amino acids, so these mutated viruses remain sensitive to that drug. An E119V mutation also interferes only with oseltamivir binding, possibly because a water molecule can fit between oseltamivir and valine at the active site but cannot insinuate itself between zanamivir and valine at residue 119.

These mechanisms have clinical implications. The mutations identified in the resistant viruses have thus far all been in the amino acids mentioned above. A 2004 study in Japan found that 9 of 50 children with influenza A (H3N2) virus infection who had been treated with oseltamivir (18 percent) had a virus with a drug-resistance mutation in the neuraminidase gene (R292K, N294S, or E119V).¹ A 2000–2001 Japanese study also revealed resistant influenza A (H1N1) viruses with the H274Y mutation in 7 of 43 oseltamivir-treated children (16 percent).⁴

The surprisingly high rate of emerging resistance in the Japanese studies may have been due to the use of insufficient doses of the drug and resultant failure to eradicate the virus. In both studies, the children were given 2 mg of oseltamivir per kilogram of body weight, and many were very young (75 percent were one to three years of age in the 2004 study, as were 43 percent of those in the 2000–2001 study). Of 147 children in a U.S. trial (including 26 younger than five years) who received the age- and weight-tailored (and therefore sometimes substantially higher) doses that have been approved outside of Japan, none shed resistant virus.¹

It is therefore worrisome that personal stockpiling of oseltamivir is likely to lead to the use of insufficient doses or inadequate courses of therapy. Shortages during a pandemic would inspire sharing of personal supplies, resulting in inadequate treatment. Such undertreatment is of particular concern in children — the main source for the dissemination of influenza within the community, since they usually have higher viral loads than adults and excrete infectious virus for longer periods. The habit of stopping treatment prematurely when symptoms resolve (a well-established tendency with antibiotic therapy) could also lead to suboptimal treatment of influenza and promote the development of drug resistance.

Could drug-resistant viruses then spread? Although many oseltamivir-resistant (non-H5N1) viruses that have been studied in animals have compromised biologic fitness, some resistant variants have been transmitted among ferrets, arousing concern about transmissibility among humans.⁵ In fact, according to recent data collected in Japan by the Neuraminidase Inhibitor Susceptibility Network, 3 of 1200 isolates from ill patients without known exposure to neuraminidase inhibitors contained resistance mutations, suggesting that these resistant viruses are transmitted at least at a low level in humans and are not severely biologically compromised.

There have now been several reports that oseltamivir-resistant influenza A (H5N1) viruses with the H274Y mutation have been isolated from humans with avian influenza infection who were treated with oseltamivir.⁴ The cases described by de Jong et al. raise the worrisome prospect that even with therapeutic doses, oseltamivir resistance may develop during the course of illness and may affect clinical outcomes. Nothing is yet known about the transmissibility of oseltamivir-resistant influenza A (H5N1) viruses in humans, and it will be important to study these isolates in animals to determine how the H274Y mutation affects virulence, pathogenicity, and transmissibility.

There is much to be learned about the clinical and virologic course of H5N1 infection in humans, as well as the response to therapy and the development of resistance. We know that the virus may have a longer incubation period than other influenza viruses, potentially increasing the period of transmissibility before symptoms appear, and that the virus frequently leads to fulminant lower respiratory tract infection.⁴ Interventions of any kind have failed when initiated late in the course of illness, but early therapy with neuraminidase inhibitors is probably beneficial; the cases reported by de Jong et al. suggest that even therapy initiated later in the illness may limit ongoing viral replication. H5N1 virus infections may require higher doses of oseltamivir for longer periods than do other types of influenza.⁵ Indeed, it is becoming clear that more medication than the currently recommended doses may be required for adequate treatment. If so, treatment with the current doses could not only fail but also select for resistant influenza A (H5N1) viruses.

Like any successful infectious agent, influenza virus will most likely evolve to evade any single drug. By targeting several points in the viral life cycle simultaneously with different drugs, we are more likely to discourage the emergence of viruses that can resist all drugs at once. But we currently rely solely on the neuraminidase inhibitors — and solely on oseltamivir in many situations, such as in patients who cannot use inhaled medication or in patients infected with H5N1 virus, in whom systemic drug levels may be important. We must not abrogate the usefulness of these drugs by exposing circulating influenza to them in such a way as to facilitate the selection of resistant viruses. The study by de Jong et al. confirms that oseltamivir-resistant H5N1 virus is now a reality. The need to learn more about how and when resistance to the neuraminidase inhibitors develops, while we focus on the development of new antiviral drugs, is pressing. This frightening report should inspire us to devise pandemic strategies that do not favor the development of oseltamivir-resistant strains. Improper use of personal stockpiles of oseltamivir may promote resistance, thereby lessening the usefulness of our frontline defense against influenza, and should be strongly discouraged.

Source Information

Dr. Moscona is a professor in the Departments of Pediatrics and Microbiology and Immunology at Weill Medical College of Cornell University, New York.

An interview with Dr. Moscona can be heard at www.nejm.org.

References

1. Moscona A. Neuraminidase inhibitors for influenza. N Engl J Med 2005;353:1363-1373. [Full Text]
2. Bright RA, Medina MJ, Xu X, et al. Incidence of adamantane resistance among influenza A (H3N2) viruses isolated worldwide from 1994 to 2005: a cause for concern. Lancet 2005;366:1175-1181. [CrossRef][Medline]
3. Varghese JN, Smith PW, Sollis SL, et al. Drug design against a shifting target: a structural basis for resistance to inhibitors in a variant of influenza virus neuraminidase. Structure 1998;6:735-746. [CrossRef][ISI][Medline]
4. The Writing Committee of the World Health Organization (WHO) Consultation on Human Influenza A/H5. Avian influenza A (H5N1) infection in humans. N Engl J Med 2005;353:1374-1385. [Full Text]
5. McKimm-Breschkin JL. Management of influenza virus infections with neuraminidase inhibitors: detection, incidence, and implications of drug resistance. Treat Respir Med 2005;4:107-116. [Medline]