.
Selection of H5N1Tamiflu-Resistant Mutants in Vietnam
Recombinomics Commentary
December 22, 2005
The report on the two fatalities linked to Tamiflu resistant H5N1 in Vietnam raise additional concerns about the creation of resistant strains through sub-optimal dosing of H5N1 infected patients.
The causes for this concern are manifold. Initial in vitro experiments using Tamiflu(oseltamivir) and Relenza (zanamivir) against all 9 N serotypes indicated that Tamiflu was about 20 times more effective against N2 than N1. A more recent in vitro study targeting H5N1 from Vietnam showed that an additional 3 fold increase in Tamiflu was required to achieve 50% inhibition of the virus. Thus, H5N1 from Vietnam requires over 50X as much Tamiflu as N2 serotypes.
This difference in in vitro activity is not reflected in FDA approved treatment or prophylactic use of Tamiflu against H1N1 or H3N2. The same dosage is recommended for both human serotypes. This same dosage has been used to treat H5N1 infected patients, but additional animal studies indicate this concentration may be sub-optimal.
Tamiflu was used in vivo at a tiger zoo in Thailand last year. Initially a small number of tigers were infected with H5N1. Tamiflu was used to treat the symptomatic tigers and was used prophylactically to prevent spread. These efforts were not successful and all exposed tigers (147) died or were euthanized.
Mice were used in additional in vivo tests of Tamiflu. Although mice were treated with 20X the prophylactic dose prior to infection, 50% of the mice died after 5 days of treatment (the FDA approved protocol). Extending treatment to 8 days reduced the death rate to 20%. The higher dose was used because of differences in drug metabolism, but the above data indicated that even 20X the approved dose was sub-optimal.
Thus, the above data indicate the FDA approved prophylactic or treatment dose for Tamiflu is sub-optimal for H5N1 infections. Use of Tamiflu at sub-optimal levels can give rise to resistant H5N1.
Resistance has now been reported in three individuals. The first case was initially treated with a prophylactic dose. The level was increased to a treatment dose and the patient recovered. However, 50% of the virus isolated from the patient had the Tamiflu resistance polymorphism H274Y.
The more recent report described two additional patients with the same H274Y polymorphism. However, in these two cases H274Y had become dominant and both patients died. Although wild type 274H could be seen in sequencing data, the wild type H5N1 could not be isolated suggesting 274Y could overgrow the wild type version even in the absence of Tamiflu. These data suggest that under some conditions, the 274Y version was as fit as the wild type.
The number of H5N1 isolates from patients in Vietnam has been limited, so reports of H274Y in three patients is alarming. Tamiflu resistance had been reported in Japan in patients infe3cted with H3N2, so the reduced effectiveness of Tamiflu for inhibition of H5N1 suggests resistance may develop easily.
Resistance is one reason why amantadine usage for treatment of influenza has declined. Initial human H5N1 were limited to Vietnam and Thailand where all isolates reported thus far have two amantadine resistance mutations in M2 that preclude use on those patients. However, now H5N1 has been reported in patients in Indonesia and China, and in all cases those isolates are sensitive to the amantadines.
The first reported human H5N1 case in mainland China was successfully treated with amantadine, suggesting combination therapy may be useful in H5N1 cases that do not have resistance polymorphisms.
The high frequency of H274Y in Tamiflu treated patients raises serious questions about sub-optimal use of the drug and selection of H5N1 with resistance polymorphisms.
Map
Media Resources