Earlier Tamiflu May Cut Death Risk in Some Severe Cases

Earlier administration of oseltamivir (Tamiflu) to severely ill patients with the flu was not linked to a decrease in overall influenza-associated mortality, though there was an observed effect on patients with influenza A/H3N2, researchers found.

After adjustment, early oseltamivir did not affect overall mortality (RR 0.94, 95% CI 0.78-1.14) compared with receiving the drug late, but early oseltamivir did have a significant effect in patients with influenza A/H3N2 (RR 0.69, 95% CI 0.49-0.94), reported Theodore Lytras, MD, of the Hellenic Centers for Disease Control and Prevention in Athens, Greece, and colleagues, in Clinical Infectious Diseases.

Use of neuraminidase inhibitors, such as oseltamivir, are recommended for severely ill patients or those at high risk of complications, the authors said, but added that “a debate persists around the evidence supporting these recommendations.” They cited lack of evidence from randomized trials about the effects of neuraminidase inhibitors on mortality, and the ethics surrounding randomizing severely ill patients.

As a result, they said that recommendations have been forced to rely on observational studies, which are often low-quality, with a high risk of bias and “incompletely adjusted for confounding” with a patient population that is “relatively young and healthy” with a low mortality rate. They added that “practically all available evidence” on the topic pertains to influenza A/H1N1pdm09.

Researchers examined data from an active surveillance system for severe influenza in Greece. The study population was comprised of adults ages ≥18 with laboratory-confirmed influenza over eight winter seasons (2010-2011 to 2017-2018). Patients were treated in an ICU, mechanically ventilated, and received oseltamivir alone. Primary outcome was death in the ICU (vs live discharge from the ICU), the authors said.

Overall, the study included 1,330 ICU patients, where 362 received early oseltamivir. About 40% of patients were women, median age was 62, and about three-quarters of patients were age >50. Over half of patients had influenza A/H1N1pdm09, while about 20% had influenza A(H3N2). Only 10% had been vaccinated against seasonal influenza.

Researchers noted that those receiving early oseltamivir were slightly older, were more likely to have comorbidities and be vaccinated for seasonal influenza, and had a shorter time between symptom onset and ICU admission.

There were 622 deaths out of 1,330 patients, with no significant difference between the early and late oseltamivir groups, the authors said. However, they noted that fewer patients with influenza A/H3N2 died in the early group versus the late group (33.7% vs 48.4%, P=0.029). In addition, median length of stay in the ICU was shorter in the early versus late group (12 days vs 15 days, P=0.003).

The authors characterized this “observed 30% decrease in mortality” among A/H3N2 patients as “especially significant.”

“This subtype is often poorly covered by the seasonal influenza vaccine, and frequently affects elderly patients causing significant morbidity and mortality, as recently seen in the United States,” the authors wrote. “Such a large decrease in mortality among the most critically ill influenza patients indicates a considerable potential for prevention in the population.”

But they also found that there was no mortality benefit of early oseltamivir in influenza A/H1N1pdm09 or influenza B patients, and called for more research on the effects of oseltamivir against all seasonal influenza types.

“This … changes current thinking about antivirals, and is particularly relevant for pandemic planning; the effectiveness (or lack thereof) of antivirals against one strain of influenza does not guarantee (or preclude) effectiveness against a future pandemic strain,” the authors wrote.

The observational nature of the study was a limitation, making it potentially subject to unmeasured confounding. Other limitations included the fact that dosage and duration of oseltamivir treatment was not available and that safety outcomes were not evaluated.