Thursday, June 02, 2005

Tamiflu (oseltamivir) and some other antivirals

We have previously posted (here and here) about Tamiflu (oseltamivir), the oral antiviral that so far is effective for H5N1 influenza, both prophylactically and if given within the first 48 hours of symptoms. This seems like a good time to recap some of it again (see earlier posts for more and different details).

Tamiflu is in fairly short supply and there isn't likely to be enough or in the right places to stem a pandemic, but it does seem to work, which is something at least. Unfortunately it has a very long production cycle (from raw material to end product, currently 12 months) and only one producer (Swiss-based Roche). Another antiviral, of the same mechanistic class (a neuraminidase inhibitor) is Relenza, which must be inhaled. There are reports Relenza is also effective intravenously, but for now is only approved for inhalation. There is less experience with Relenza and its effectiveness as a prophylactic agent hasn't been demonstrated to my knowledge, but there is no reason why it shouldn't also work. The mode of administration is problematic for those having respiratory distress and possibly those with asthma and for young children. Another drug of the same class, peramivir, has been developed but is not in commerical production. Two other anti-virals used for treating influenza, amantadine and rimantadine, work by different mechanisms (ion channel inhibitors) and they tend to provoke resistance from influenza viruses. The currently circulating H5N1 bird flu strain is reportedly resistant to both of these drugs (possibly a result of their use by the Chinese in chicken feed).

WHO recently reported that a single patient in Vietnam showed resistance to Tamiflu as well. This is less common than for the adamantine class of drugs, however, and may also be less of a problem because Tamiflu works by inhibiting the action of one of the virus's two main surface proteins, neuraminidase (the N part of the H5N1 name). After the virus has replicated and reassembled inside a host cell (for example, a cell in your respiratory tract) it "buds" at the surface of the cell but gets stuck there by attachments of the H protein (hemagglutinin) to the cell surface. The neuraminidase on the virus breaks this sticky attachment, releasing the virus to infect another cell. If you inhibit neuraminidase with Tamiflu or Relenza, the virus remains stuck to the surface of the cell. It also remains visible to the immune system, so the drug doesn't interfere with the development of natural immunity. It is possible for there to be a change in the neuraminidase protein that is impervious to the inhibitor, but the protein seems to be sufficiently specialized and precise that the mutated neuraminidase results in a virus that is much less contagious. At least that is the evidence so far. Fred Hayden at the University of Virginia is quoted in New Scientist to the effect that in normal flu strains the mutated resistant virus is a hundred times less contagious. In Japan where Tamiflu is used routinely for ordinary influenza, about one in six children develops a resistant virus but it doesn't spread. Hayden believes the same will be true of a resistant H5N1 strain, but at this point we don't know this for sure as only one case has been reported.

On the basis of mathematical modeling, Ira Longini at Emory has suggested a pandemic might be stopped by flooding a local outbreak with Tamiflu, thus reducing the basic reproductive number (R "naught") to a point where the spread would die out. Even assuming the model form were correct and the parameters sufficiently accurate, it is unlikely we could identify an infective focus of this very contagious disease rapidly enough and get the drug there quickly enough to pull this off.

The bottom line here is that we currently have some potentially effective pharmaceuticals but probably neither a sufficient supply nor the public health infrastructure to make use of them to stem a pandemic or even make anything but a marginal difference. Much will depend on timing. If we have two or three years, we might be able to build up a sufficient stockpile of antivirals and produce a vaccine with cell culture methods to have an effect, but even then we would need good public health infrastructures. Not likely for the US and impossible for much of the world.

Anyway, at this point two or three years seems like a very, very long time.