The overuse of Tamiflu in treating avian flu

Iwan Darmansjah, Jakarta

Tamiflu -- generic name oseltamivir -- has been promoted too far to be used in avian flu because it is neither proven nor significantly effective even in the treatment of common flu. In influenza Type A and B (common flu) the efficacy in clinical studies is manifested as an improvement of one day (FDA website -- updated on Nov. 17, 2005 -- states one to one-and-a-half days) in the recovery-time of influenza that normally heals within a week. A one-day improvement should be considered an insignificant result in a clinical trial.

Moreover, in the U.S. most flu is that of Type A. Such results, measured by days of recovery for a sickness that lasts normally a week, should be classified as "modestly effective" at the most to justify marketing of the product. Adverse reactions are not yet completely known.

On the current FDA website, adverse event reports from Japan in children documented "primarily unusual neurologic and psychiatric events such as delirium, hallucinations, confusion, abnormal behavior, convulsions and encephalitis... 12 deaths in pediatric patients were documented since Tamiflu's approval".

The FDA is currently investigating these reported events, all of them from Japan. Although adverse events are different from (established) adverse reactions, this phenomenon should not be taken for granted, since this is not the first time that Japan has produced such reports on other drugs that have later been proven to be true.

Against avian flu Tamiflu has never been studied before, while avian flu is quite a different disease than ordinary flu that occurs in countries with a cold climate nearer the Antarctic. Influenza in tropical climates, such as Indonesia, is again not the same as the cold winter flu in the above countries.

Thus the "fever" of stocking Tamiflu is rather surprising; it even is supported by a trusted agency, the WHO. The efficacy criteria adopted by the FDA (1999), which was the data accepted before marketing, is indeed debatable if applied to the avian flu syndrome.

The case of avian flu has a different dimension, and surrogate endpoints such as the use of days in improving a self-limited illness in a clinical trial cannot be extrapolated into the deadly avian flu case. Here, "real endpoints", such as survival rate or total deaths should be measured in controlled clinical trials.

Such data could then guide clinicians to use the drug against avian flu. What then should the control treatment be in such a study, and what causes death in avian flu? In my opinion, clinical studies should be done using an effective antibiotic as an active comparator, because a true placebo may be regarded as unethical.

What about today's treatment of avian flu in humans?

Some of the important causes of fatality in avian flu are caused by its complications, often pneumonia or bronchopneumonia, the cause of which is most likely Streptococcus pneumoniae, and thus bacterial.

Much has been written about the resistance of these microbes against the then commonly-used antibiotic of choice, penicillin G. No other antibiotic, in fact, is as potent as penicillin G in killing properties, unless severe resistance hinders its use; but even this can be overcome by using very high doses of penicillin. For strains of S. pneumoniae with a very high minimum inhibitory concentration (MIC) of

High-dose penicillin G injection (either intramuscular procaine penicillin G or benzyl-penicillin G intravenously administered) is currently still advocated for streptococcus pneumonia or bronchopneumonia (Antibiotic Therapeutic Guidelines and Pharmaceutical Benefits Scheme (PBS), Australia) in many parts of the world.

I have been using penicillin procaine for pneumonia with outstanding results -- remarkable clinical improvement and disappearance of fever within one day. Cephalosporins and ciprofloxacin will be less effective as a first-line choice and may even be dangerous. Penicillin G should be used in high doses; this does not affect the seriousness of a possible anaphylactic reaction.

Where does Tamiflu stand?

The first antiviral drug against flu began with amantadine; the results should be classified as modest. Since flu is widespread, research has focused on this illness. Occasional fatalities and large market prospects have stimulated the continued search for new antivirals against flu.

As a class in itself, antivirals have made relatively slow progress in the development of effective drugs. Clinical trials are also difficult to assess. With a high rate of spontaneous cure of the illness, such studies need a large number of subjects to balance the expected high placebo reactors.

But the development of antivirals to combat viral disease is not like antibiotics against bacteria, which have made a large contribution to our medical armamentarium. From all the antivirals on the market, only a few can be classified as significantly effective.

From a list of antivirals, only acyclovir, lamivudine, ribavirin, and interferon (pegylated included) have reached clinically satisfactory results. Three or four antiretrovirals (a.o. zidovudine, didanosine, nefirapine, and indinavir) have been contributing modestly to the treatment of HIV-positive patients.

Many vaccines against viral diseases are under development and their long term efficacy and safety will further be evaluated.

The writer was a WHO expert panel member on drug evaluation and pharmaceuticals (1975-2003).