Review of Japan's Evidence
Table 1 summarizes the trend of both spontaneous ADR reporting and solicited reporting of NPAEs. What is noteworthy about oseltamivir is the sharp increase of NPAEs reported by the manufacturer (Chugai Pharmaceuticals) after the 2005/2006 season. This coincides with the timing when NPAEs of oseltamivir caught media and public attention. It is also noteworthy that NPAEs of zanamivir also started to rise in the 2006/2007 season. After the MHLW warning against the use of oseltamivir for teenagers in March 2007, the share of oseltamivir declined as shown by the JMDC database (the share of oseltamivir was 45.8% in the 2004/2005 season but declined to 26.7% in the 2007/2008 season). Oseltamivir was increasingly replaced by zanamivir and the change was reflected in the numbers of spontaneous reporting of NPAEs from manufacturers. Spontaneous ADR reporting from doctors/pharmacists was mandated in July 2003 but the reporting of influenza-related ADRs from doctors/pharmacists was scant until the 2005/2006 season. However, it increased sharply in the 2006/2007 season to 20, when the MHLW issued a warning attracting a lot of attention. The number of spontaneous reporting from doctors/pharmacists declined to zero from the next season but the decline could be attributable to the new solicited NPAE reporting. Doctors/pharmacists chose to report NPAEs via the new route.
When combined, there were 92, 41 and 87 reports from doctors/pharmacists and 204, 84 and 124 reports from manufacturers (the number was tallied to avoid duplicates) in the 2006/2007, 2007/2008 and 2008/2009 season, respectively. This trend appears to be in proportion to the severity of the epidemic (the number per sentinel was 225.2, 139.3 and 274.8, respectively), although the number of reports in the 2006/2007 season was somewhat inflated owing to active media coverage. It is also noteworthy that nearly half (96 out of 200) of NPAEs reported by doctors/pharmacists were not related with any antiviral drugs. This lends support to a claim that NPAEs are caused by influenza per se and are not attributable to certain drugs. However, no definite conclusion was drawn from these spontaneous reports alone.
The consequences of two research projects (Yokota and Hirota) were unsatisfactory; they ended up inconclusive after investing time, money and efforts. This may be a reason why neither study has been published in major international journals, despite their large-scale data and importance. Hirota advocated conducting a case–control study in conclusion but the project was terminated and no such additional study had been conducted to date. Sure enough, distributing thousands of questionnaires is not the best choice for pharmacoepidemiological studies.
To detect rare and ill-defined events, such as NPAEs, an alternative to case–control study is to use administrative data, such as electronic health records or health insurance claims. The three US studies using HMO data and health insurance claims data did reach definite conclusions in favor of oseltamivir. My study, although the outcome measures are different (using injuries as outcome and not NPAEs themselves), used a similar approach (propensity score matching) with the three US studies. Table 2 compares the incidence of injuries within 3 days of diagnoses of influenza among five subgroups of the similar propensity score to receive oseltamivir. The incidence of injuries was significantly lower in oseltamivir-treated patients than those otherwise. The tendency held true over five subgroups divided by propensity scores predicted by past history of diagnoses within 6 months. The conclusion was also consistent with the three US studies.
Japan's MHLW also set up a pharmacological working group shortly after they had issued a warning against use of oseltamivir for teenagers in March 2007. The working group conducted a series of animal experiments and received a report on penetration of oseltamivir and its carboxylate into cerebrospinal fluid involving four Caucasian and four Japanese healthy volunteers from the manufacturer (Hoffmann-La Roche, Ltd) in its fourth meeting held in December 2007. The report concluded that oseltamivir and oseltamivir and its carboxylate had limited potential to induce or exacerbate CNS adverse events in individuals with influenza.
Overall, there has been no evidence to support that oseltamivir increased the risk of NPAEs in influenza patients compared with other drugs.
However, there is one problem yet to be answered: the risk of sudden death due to oseltamivir, a hypothesis raised by Hama. In support of his hypothesis, a considerable number of sudden deaths were reported in relation to oseltamivir by the manufacturer as demonstrated in Table 1, a finding not seen in other drugs. Also, it appears that the number of sudden deaths related to oseltamivir increased dramatically in three season (2005/2006, 2006/2007 and 2007/2008): an unnatural phenomenon that cannot be explained by the increase of influenza patients.
Causality between oseltamivir and sudden deaths may not easily be proved with a small number of spontaneous reporting of deaths but, fortunately, Japan may be able to draw a definite conclusion in the near future because Japan has developed a national database collecting all medical and pharmaceutical claims. Since health insurance claims contain prognoses (i.e., death or recover) it may be possible to conduct a population survey by linking a medical and a pharmaceutical claim on an individual basis, thereby elucidating whether increased deaths occur shortly after taking certain drugs. Such a study should be conducted as soon as possible.
Expert Rev Pharmacoeconomics Outcomes Res. 2010;10(1):17-24. © 2010 Expert Reviews Ltd.
Cite this: Is Oseltamivir (Tamiflu®) Safe? Re-examining the Tamiflu 'Ado' from Japan - Medscape - Feb 01, 2010.