Fifty sudden deaths may be related to central suppression
BMJ 2007; 335 doi: https://doi.org/10.1136/bmj.39262.448877.BE (Published 12 July 2007) Cite this as: BMJ 2007;335:59
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
EDITOR-
Japanese authorities set up two working groups on oseltamivir (Tamiflu,
Chugai, Roche), one examining clinical and the other non-clinical
causality issues related to neuropsychiatric events, after they advised
against prescribing Tamiflu to adolescents aged 10-19 years in March
2007[1].
After discussing a series of studies mainly by pharmaceutical
companies (Roche and Chugai) as well as its own, the clinical working
group initially announced at the last (7th) meeting on 10th July 2008 [2]
that it had not found any causal relationship between the drug and
neuropsychiatric events including sudden death and abnormal behavior.
With these results, the Pharmaceutical Affairs and Food Sanitation
Council was scheduled to hold a session on Friday (8th August) to lift a
ban on the use of Tamiflu by teens. Surprising enough, the Ministry of
Health, Labor and Welfare (MHLW) announced on 5th August it will review
its opinion about causality issues, as errors in the processing of data of
an epidemiological study had been identified [3].
This epidemiological study analyzed 10,316 children with flu in the
winter of 2006/07. The study was done by the MHLW's task force headed by
Prof. Yoshio Hirota of Osaka City University, who also led the research.
The preliminary report was first issued on December 25th 2007 [4].
It was a short course cohort study with two groups: one was
prescribed oseltamivir phosphate (Tamiflu) (7,677 children) while the
other is the group of children who were prescribed drugs other than
Tamiflu (2,192). They were followed for four days from onset of fever and
the proportions of children with neuropsychiatric events were compared
[2,4].
The odds ratio in the Tamiflu group was 0.382 (95%CI: 0.338-0.432,
p<_0.0001 in="in" the="the" preliminary="preliminary" report="report" _4.="_4." p="p"/> But there was serious misclassification of events among groups. For
example they withdrew cases with the above-mentioned events occurring
before taking drug(s) only in the Tamiflu-prescribed group (T-pr group)
and they added these withdrawn cases to non-Tamiflu-prescribed group (non-
T-pr group) [5].
Correct odds ratio calculated on intention-to-treat (ITT) analysis is
estimated to be at least 1.37 (95%CI: 1.18-1.58) [5].
I warned the task force head about this misclassification on Feb.
8th 2008 and I requested that results be corrected to show significant
relationship between abnormal behavior and Tamiflu prescribing.
I also suggested that they should withdraw cases with early
neuropsychiatric events before taking prescribed drugs from both groups if
they want to calculate the proportion of patients with events only among
those who took drugs.
At the 7th meeting of the clinical working group for Tamiflu held
10th July 2008, an interim report analyzing a total of 10,017 children
under 18 year old was published. The children had confirmed flu and the
report showed that the proportions of events were 11.9 % (889 in 7,487) in
Tamiflu-treated group and 12.8 % (286 in 2,228) in non-Tamiflu-treated
group, with an odds balance of 0.91 (95%CI: 0.79-1.06) [2].
In this interim report, it is stated that 1,215 children with events
in the T-pr group and 262 in the non-T-pr group were identified. Early
neuropsychiatric events that occurred before visiting physicians were
deleted from both groups (227 from T-pr group and 47 from non-T-pr group),
which was reasonable.
However, cases with events before taking drugs were again removed
only from the Tamiflu-pr group (99 cases) and added to the non-T-pr group.
Correct proportions of children with events (including early events
before taking drugs) calculated on ITT analysis were 13.0 % (988 in 7,586)
and 8.8 % (187 in 2,129) respectively. Odds ratio was estimated at 1.56
(95%C.I. 1.32-1.84) [6].
The task force's analysis is wrong: Suppose Tamiflu does not affect
the neuropsychiatric system and there are two groups (T-pr and non-T-pr
groups) with the same population (n), the same number of patients with
early events before taking drugs (b), and the same number of patients with
events after taking drugs (a).
The proportion of total events calculated on ITT-analysis is the
same: (a+b)/n and the odds ratio=1. If we follow the task force's
analysis the proportion would be a/(n-b) in the T group and (b+a+b)/(n+b)
in the non-T group, with an odds ratio of a/(a+2b), i.e. less than 1. Thus
the results obtained through the task force's analysis contradict the
initial assumption.
Reference
[1] Japan issued Tamiflu warning after child deaths. Times 21 March
2007. www.timesonline.co.uk/tol/news/world/asia/article1549260.ece
[2] MHLW. �gThe seventh meeting of working group for clinical issue
of causality of Tamiflu and neuropsychiatric events held on 10th July 2008
(in Japanese):
http://www.mhlw.go.jp/shingi/2008/07/s0710-6.html
document-1: Epidemiological study on the symptoms during influenza
(Interim report: in Japanese)
http://www.mhlw.go.jp/shingi/2007/12/dl/s1225-7y.pdf
[3] Ministry to review links between Tamiflu, behavior
http://www.yomiuri.co.jp/dy/national/20080807TDY03302.htm
[4] MHLW (2007). �gThe fifth working panel for drug safety in 2007�h
held on 25th Dec. 2007: document-3-4: Epidemiological study on the
symptoms during influenza (in Japanese)
http://www.mhlw.go.jp/shingi/2007/12/dl/s1225-7y.pdf
[5] Hama R, Fatal neuropsychiatric adverse reactions to oseltamivir:
case series and overview of causal relationships. The International
Journal of Risk & Safety in Medicine (2008) 20: 5-36. available at :
http://npojip.org/sokuho/published-paperJRS431.pdf
[6] Hama R. Epidemiological study showed the Tamiflu increase
abnormal behavior by 1.56 in children. �gWeb-Kusuri-no-Check�g No108 (in
Japanese):
http://npojip.org/sokuho/080711.html
and No 109 (in Japanese) http://npojip.org/sokuho/080728.html
Competing interests:
None declared
Competing interests: No competing interests
Before discussing the denominator (the number of patients who have
taken oseltamivir), I would like to mention some important points when
considering benefit/harm balance in using Tamiflu.
First of all, I would like to emphasize the self-limiting manner of
influenza in otherwise healthy people. It is also true even in children
with chronic asthma, because Kaplan-Meier curves for proportion of
patients without freedom from illness in placebo treated patients are
almost the same both in the randomized controlled trial with previously
healthy children (study WV15758[1]) and those with chronic asthma (study
WV15795/15871[1]). Thus, for most of the people with sudden deaths or with
accidental deaths from abnormal behaviour, Tamiflu was unnecessary.
Secondly as mentioned previously [2], percent case mortality among
patients with encephalopathy (Reye's syndrome or so called influenza-
associated encephalopathy) fell from approximately thirty percent to below
fifteen percent before the introduction of Tamiflu for children in Japan
in late 2002. This is due to the restriction of use of non-steroidal anti-
inflammatory drugs (NSAIDs) as antipyretics in 2000, which caused a
percentage of NSAIDs users in encephalopathy children to fall from about
thirty percent to below ten percent in the same period.
Thirdly, one of the most important things I tried to tell in my
letter [3] was that the classifications of 58 deaths (50 sudden deaths and
eight accidental deaths from abnormal behaviour) differ completely between
that by Japanese MHLW (practically causality denied) and that by many
doctors(probable, possible or at least causality could not be ruled out.
For the denominator, no precise data are available except the number
of prescription (IMS data) released by FDA in November 2006 [4] and the
overall number of Tamiflu users estimated by Chugai which Japanese MHLW
released in June 2007 [5]. The number of Tamiflu prescriptions in all ages
from 2001 to 2005 is 6.5 million and 24.5 million in US and Japan
respectively, while that in ages 0-16 is 0.9 million and 11.6 million
respectively.
The number of Tamiflu users is estimated as 35 million in Japan from 2001
to 2007 (estimated from a shipment by Chugai) and 45 million in the world
in the same period (from the press release issued by Roche in March 2007)
[5].
However, it should be noted that spontaneous reporting systems such
as the Yellow card system in UK estimate that physicians report between
only one and ten percent of adverse effects on treatment [6], though no
such estimation is available in Japan.
Erratum:
In the products (capsules and powder for oral suspension),
oseltamivir is present as oseltamivir-phosphate (oseltamivir-P), from
which oseltamivir is easily dissociated in the GI tracts, absorbed,
extensively metabolized to oseltamivir-carboxylate by esterase in the
liver, otherwise up to one fourth of oseltamivir is distributed via
circulation and enter into brain tissue through blood-brain barrier.
Therefore, the agent which has central suppressive action is oseltamivir
(ethyl ester prodrug).
So "oseltamivir-P" should be changed to "oseltamivir" as follows:
Thus adverse reactions to oseltamivir may be roughly classified into
three groups: (a) sudden onset reactions related to central suppressive
action of oseltamivir during cytokine storm, including sudden death,
abnormal behaviours, and other sudden neuropsychiatric disorders; (b) late
onset reactions such as pneumonia, sepsis, hyperglycaemia, and late onset
neuropsychiatric disorders possibly related to inhibition of human
cytosolic neuraminidase (sialidase) activity by oseltamivir carboxylate5;
and (c) allergic reactions and others.
References
1.New drug approval package (NAP) of oseltmivir (in Japanese);
Tamiflu dry syrup (2002):
http://211.132.8.246/shinyaku/g0201/11/5303990_21400AMY00010.html?
2.Hama R. Limited benefit and potential harm of oseltamivir including
sudden death and death from abnormal behaviour.
www.bmj.com/cgi/eletters/331/7526/1203-b#122513
3.Hama R. Oseltamivir's adverse reactions: Fifty sudden deaths may be
related to central suppression. BMJ. 2007 Jul 14;335(7610):59.
http://www.bmj.com/cgi/content/full/335/7610/59
4.FDA, Tamiflu AE Review 2006
http://www.fda.gov/ohrms/dockets/ac/06/briefing/2006-
4254b_09_01_Tamiflu%20AE%20Review%202006%20Redacted_D060309_092.pdf
5.Document No 3 for Advisory Committee on Drug and Food. Second
annual meeting of the Sub-Committee on Safety of Medicine for 2007 held on
4 April 2007 (in Japanese) http://www.mhlw.go.jp/shingi/2007/04/dl/s0404-
2c.pdf
6.Pirmohamed M, Breckenridge AM, Kitteringham NR, Park BK. Adverse
drug reactions. BMJ. 1998;316:1295�E298.
Competing interests:
None declared
Competing interests: No competing interests
The report by Hama gives us a clear indication of the numerator.
However, without any indication of the denominator the usefulness of this
report is severely curtailed.
Can we have an indication of the number of patients who have taken
Oseltamivir in the relevant age groups in Japan, over the same period,
please?
Competing interests:
None declared
Competing interests: No competing interests
Mechanisms of neuropsychiatric reactions to oseltamivir become clearer.
EDITOR-
In spite of the Pharmaceutical Affairs and Food Sanitation Council
repeated announcements that it had not found any causal relationship
between oseltamivir (Tamiflu, Chugai, Roche) and neuropsychiatric adverse
events [1], clinical and non-clinical evidence have been accumulating and
the mechanisms of neuropsychiatric reactions to Tamiflu has now become
clearer [2-8].
It includes a dose-dependent increase in deaths and central
suppressive symptoms in 7-day-old experimental rats, together with the
similarity of symptoms and findings in animals and humans [2]. Symptoms
occurring after taking Tamiflu may be induced by unchanged oseltamivir
(OT), because its level in the brain is 64 times higher in 7-day-old rats
than in mature rats. Such higher level of OT in brain may be due to the
immature activity of liver microsomal carboxyesterase (CE-1) that
metabolize OT to oseltamivir carboxylate (OC) [3] and immature P-
glycoprotein (P-gp) that is the efflux transporter of OT through the blood
-brain barrier (BBB) [4].
It is not difficult to presume that not only the immaturity but also
the impairment of CE-1 and P-gp due to higher level of cytokines at the
peak of flu infection in older infants and adults in human may increase
level of OT in the brain.
On the other hand the mechanism underlying the abnormal behavior on
oseltamivir may be related to dyscontrol of central nervous suppressants
such as barbiturates, benzodiazepines and alcohol [2]. But the precise
mechanisms were not known.
Recently, Yoshino et al [5] found that systemic administration of
oseltamivir (25mg/kg or 100mg/kg; intraperitoneally (i.p.)) significantly
increased extracellular dopamine in the medial prefrontal cortex (mPFC) of
rats on microdialysis by 156 % or 223 % compared to the control values.
Serotonin level was unchanged and it was suggested that the increase of
dopamine (DA) was due to increased DA release in the mPFC [5].
The mechanisms of increased release of DA are clearly different from
those caused by psychostimulants such as amphetamine and narcotics such as
cocaine, and are similar to those induced by benzodiazepines because of
the following reasons:
(1) The extent of DA increase was far less than with
psychostimulants, including phencyclidine and methamphetamine that led to
more than 10-fold increase versus preadministration levels [6].
(2) Neither OT nor OC inhibited the reuptake of DA in presynapses and
promoted its release in postsynapses in the in vitro assay system using
rat brain synaptosomes, while metanphetamine inhibited the reuptake of DA
and promoted its release, and cocaine only inhibited the reuptake of DA
[7].
(3) In rats, ataxia was observed within 10 min after 100mg/kg of OT
was administered continuing for a few minutes [5].
(4) Diazepam was found to have a biphasic effect in mice; increasing
locomotor activity (about 1.5 times more than pretreatment level) at a low
dose (0.25 mg/kg or less), while decreasing it at higher doses (>0.5
mg/kg) [8]. The locomotor stimulating effect of diazepam was effectively
blocked by pretreatment with the benzodiazepine receptor antagonist
flumazenil, as well as with the catecholamine synthesis inhibitor e-
methyltyrosine and the dopamine receptor antagonists haloperidol,
spiperone and SCH 23390 [8]. Taken together, these data indicate that the
locomotor stimulating effect observed after low doses of diazepam is due
to activation of brain dopaminergic systems involved in locomotor activity
[8].
References
[1] MHLW. A series of meetings of the working group for clinical issue of
causality of Tamiflu and neuropsychiatric events�Eand that for non-
clinical issue�E The last meeting was held on 10th July 2008 (in
Japanese): http://www.mhlw.go.jp/shingi/2008/07/s0710-6.html
[2] Hama R, Fatal neuropsychiatric adverse reactions to oseltamivir:
case series and overview of causal relationships. The International
Journal of Risk & Safety in Medicine (2008) 20: 5-36. available at:
http://npojip.org/sokuho/published-paperJRS431.pdf
[3] Chugai Pharm Co (2002) New drug approval package (NAP) of
oseltmivir (in Japanese); Tamiflu dry syrup (in Japanese): available at
http://164.46.226.166/shinyaku/g0201/11/5303990_21400AMY00010.html?
[4] Morimoto K, Nakakariya M, Shirasaka Y, Kakinuma C, Fujita T,
Tamai I, Ogihara T. Oseltamivir (TamifluTM) efflux transport at the blood-
brain barrier via P-glycoprotein. Drug Metab Dispos. 36(1) (2008):6-9.
Epub 2007 Oct 16
[5] Yoshino T, Nisijima K, Shioda K, Yui K, Kato S. Oseltamivir
(Tamiflu) increases dopamine levels in the rat medial prefrontal cortex.
Neurosci Lett. 2008 Jun 13;438(1):67-9. Epub 2008 Apr 9.
[6] Nishijima K, Kashiwa A, Hashimoto A, Iwama H, Umino A, Nishikawa
T. Differential effects of phencyclidine and methamphetamine on dopamine
metabolism in rat frontal cortex and striatum as revealed by in vivo
dialysis. Synapse. 1996 Apr;22(4):304-12.
[7] Satoh K, Nonaka R, Ogata A, Nakae D, Uehara S. Effects of
oseltamivir phosphate (Tamiflu) and its metabolite (GS4071) on monoamine
neurotransmission in the rat brain. Biol Pharm Bull. 2007 Sep;30(9):1816-
8. available at: http://www.jstage.jst.go.jp/article/bpb/30/9/1816/_pdf
[8] Söderpalm B, Svensson L, Hulthe P, Johannessen K, Engel JA.
Evidence for a role for dopamine in the diazepam locomotor stimulating
effect. Psychopharmacology (Berl). 1991;104(1):97-102.
Competing interests:
None declared
Competing interests: No competing interests