Oseltamivir Resistance: What Does It Mean Clinically?

Public health officials, infectious disease specialists, and the world's population at large are girding their loins for the next influenza season. The winter months portend outbreaks, if not pandemics, of infections caused by both the seasonal influenza A virus subtype H3N2 and the novel influenza A virus subtype H1N1 or swine-origin virus. The swine-origin virus has never entirely gone away since its emergence last season and has manifested itself as a smoldering pandemic that refuses to die down but which has thankfully also not yet flared up, as the most virulent of influenza pandemics do.

Accelerated efforts are under way to produce, test, and distribute vaccines against the “usual suspects”, as well as a vaccine directed at the novel H1N1 variant. However, in the absence of a vaccine, there has been universal reliance on the neuraminidase inhibitors for prophylaxis and treatment of presumed and proven novel H1N1 virus infection. Physicians and occupational health clinics in medical facilities have been faced with a number of conundrums related to the administration of neuraminidase inhibitors, primarily oseltamivir, to high-risk patients with influenza-like illness or to health care workers exposed to patients with influenza-like illness. Only now, after some health care workers have received multiple courses of oseltamivir because of repeated influenza-like illness exposures, has the futility of this approach, to say nothing of its propensity to select for resistance, become clear.

Worldwide, children seem to have borne an excessive burden of influenza-like illness and influenza-related mortality. Oseltamivir has been the mainstay antiviral agent for novel H1N1 virus and seasonal flu therapy because recent circulating strains have shown resistance to the adamantanes. In addition, the other available neuraminidase inhibitor, zanamivir, is contraindicated in people with underlying respiratory conditions and is difficult to self-administer.

Resistance to oseltamivir has emerged in many countries during the past few years. In Japan, resistance rates have skyrocketed from <3% during the 2008–2009 influenza season to 98%–100% during the 2008–2009 season.

In this issue of Clinical Infectious Diseases , Kawai et al [1] report the results of a study examining the clinical efficacy of oseltamivir in the era of high resistance among novel H1N1 virus isolates in Japan. They measured the duration of fever in oseltamivir-treated patients with H1N1 infection during the 2008–2009 influenza season. Time to afebrility in these patients was compared with the duration of fever among oseltamivir-treated patients who had H3N2 virus infection during the same season and among those who had novel H1N1 virus infection during the 2007–2008 influenza season. The authors also compared time to afebrility in 2008–2009 novel H1N1 virus-infected patients treated with oseltamivir or zanamivir, an agent to which the virus has shown little or no in vitro resistance. All of the 2008–2009 H1N1 virus isolates, but none of the 2007–2008 isolates, contained the H274Y oseltamivir resistance mutation.

The duration of fever was significantly longer in the 2008–2009 H1N1 virus-infected patients of all age groups who were treated with oseltamivir than it was in any of the other treated control groups. There were no untreated control patients. All patients were seen as outpatients. The delay in defervescence was longest in oseltamivir-treated children ⩽15 years of age, and these children had higher virus re-isolation rates. Oseltamivir-resistant isolates remained susceptible to zanamivir. No differences were found with respect to times to defervescence in any of the zanamivir-treated groups. There were no deaths.

The study has several significant limitations. The number of patients studied is relatively small, and we are not told much about their characteristics and differences, if any. There is no nontreated group of infected patients. Controls were either historical from the previous season or were isolates and patients representing another influenza A virus serotype. The decision whether to administer zanamivir or oseltamivir was left to the treating clinician. Data on the onset of illness and time to resolution of fever were reported by patients or their family members rather than being directly assessed by the study clinicians. Although antipyretics were not routinely administered, some patients received acetaminophen for variable indeterminate periods. The authors state that influenza vaccination status did not affect the results, but these children with H1N1 virus infection would not have been vaccinated against this variant in any case, so this is not meaningful. It is surprising that so many (>50%) of the children who were infected with H3N2 virus had been vaccinated against this virus and that this had little or no impact on fever duration.

I would consider one of the authors' self-professed limitations to actually be a strength of the study: that is, the use of widely dispersed outpatient clinics linked by an internet-based computer system. Japanese researchers have proven the usefulness of this approach in previous studies, and it is a model that can and should be emulated elsewhere.

Oseltamivir is used much more commonly than zanamivir in both adults and children. Resistance to oseltamivir is also much more common than resistance to zanamivir. Although some of the discrepancy in resistance prevalence may represent differences in the ease with which drug-specific resistance mutations can occur, it seems difficult to escape the notion that, as is true for all other antimicrobial agents, increased use provides the selection pressure for resistance selection. Although the authors suggest that the mutation found in their isolates may have surfaced earlier in Europe, widespread oseltamivir use in Japan certainly has played a major role in the increase from 3% to 99% resistance that occurred in only a single season.

The authors state that oseltamivir remains clinically effective despite the ubiquitous prevalence of the resistance mutation. This statement is presumably based on the fact that all patients survived and eventually became afebrile and the finding that serum oseltamivir concentrations far exceeded the mean inhibitory concentration of the drug. However, the numbers of patients are too small to accurately predict the detrimental effects of this resistance in an outbreak involving compromised patients with the potential for high morbidity and mortality rates. In assessing the clinical effects of drug resistance, it would have been helpful and important to measure duration of viral shedding in the various groups.

Plans for dealing with major outbreaks of both seasonal and H1N1 influenza infections in Europe and the United States include stockpiling and distributing oseltamivir. It seems likely that resistance to neuraminidase inhibitors will increase over time in all parts of the world where these drugs are used extensively. In fact, this process of emerging resistance may have already begun in the United States [2]. A slight prolongation of fever as a result of drug resistance may not be significant, but deleterious effects on shedding duration or morbidity or mortality may lessen the utility of these distribution plans.

The most significant future challenge facing infectious diseases clinicians and investigators is not human immunodeficiency virus infection, mycobacterial infection, or malarial infection per se. It is the development and persistence of multidrug resistance in any and all infectious diseases. There will be a continued need for antiviral agents to treat patients who are not vaccinated, and the development of new effective agents should be pursued vigorously. However, in the face of increasing threats of influenza pandemics caused by newly emerging virus variants, the most important advances we can make will be in the direction of the development of rapid and safe vaccines through the use of modern molecular and biological techniques, not in the development and stockpiling of antiviral drugs—whether new or old.

• Received August 31, 2009.
• Accepted September 10, 2009.

• © 2009 by the Infectious Diseases Society of America