PARLIAMENTARY COMMITTEES ACT 2003

INQUIRY INTO VIOLENCE AND SECURITY ARRANGEMENTS IN VICTORIAN HOSPITALS

“If at first, the idea is not absurd, then there is no hope for it” Albert Einstein “The world is not dangerous because of those who do harm but because of those who look at it without doing anything” Albert Einstein

This submission does not address security, it addresses violence. The root causes of violence by persons treated with drugs used in psychiatry. This submission may not fit into the Terms of Reference. The Terms of Reference do not address the root causes of violence in persons treated with psychiatric and non-psychiatric drugs that have akathisia, suicidal and homicidal ideation, suicidal and homicidal acts, aggression worsening of clinical conditions among their listed side effects. I have practiced in open wards since 1964. It is only in the last 20 years that security guards have been needed. Psychiatric patients have become more violent, more suicidal and more homicidal with the addition of narrow metabolic therapeutic window drugs (antidepressants and antipsychotics) that that carry high rates of akathisia in clinical trials and more in the community in Australia where these drugs are prescribed by general practitioners where they are not asked relevant questions. Psychiatric patients started to become dangerous to the extent that security guards and locked doors were needed on busy wards from about 1990 in Australia. I do not blame doctors, as they are being mis-eduacated by a louder voice than mine, Big Pharma’s. 1990 was the year that Prozac (fluoxetine), the first of the SSRIs, was introduced. The suicide rate in 1997, the year that Risperdal (risperidone) and Zyprexa (olanzapine) were introduced and often given on top of by then five new antidepressants produced the highest suicide rate since 1964. 964 was just after tricyclic antidepressants escaped from the offices of into doctors surgeries to be used by general practitioners who did not know, as 1960 textbooks told us, that these drugs had suicide among their listed side effects. In the 1960s, 1970s psychiatrist who knew they induced suicide and aggression (to self and others) in some people. It was in our undergraduate textbook. Following is a statement from a textbook of psychiatry published over 40 years ago that is referring to observations in patients during initial treatment with tricyclic antidepressants [Clinical Psychiatry, by Mayer-Gross, Slater, and Roth, 1960, p. 231]: “With beginning convalescence (following initiation of treatment with tricyclic antidepressants), the risk of suicide once more becomes serious as retardation fades.” 2

Some patients taking psychiatric drugs develop extra pyramidal side effects including akathisia. Some persons who develop akathisia kill themselves and kill others. Most persons who develop akathisia have violent thoughts, homicidal and suicidal ideation. Many people with akathisia have personality changes towards becoming aggressive. In my experience of making visits to see clients in NSW Jails, 9 out of ten persons I see are suffering from drug induced akathisiai and many have come into prison following acts of violence committed in that state and akathisia violence is not being recognised and they get more drugs that make them more violent and suicidal. This is tragic, as they are doubly mistreated. Drugs that cause violence have been written up in PLOS. http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0015337 This was taken up in TIME Magazine and should be common knowledge but psychiatrists just do not believe it as they are trained to diagnose mental illness not side effects. http://healthland.time.com/2011/01/07/top-ten-legal-drugs-linked-to-violence 10. Desvenlafaxine (Pristiq) An antidepressant that affects serotonin and noradrenalin, this drug is 7.9 times more likely to be associated with violence than other drugs. 9. Venlafaxine (Effexor) A drug related to Pristiq in the same class of antidepressants, both are also used to treat anxiety disorders. Effexor is 8.3 times more likely than other drugs to be related to violent behavior. (More on Time.com: Adderall May Not Make You Smarter, But It Makes You Think You Are) 8. Fluvoxamine (Luvox) An antidepressant that affects serotonin (SSRI), Luvox is 8.4 times more likely than other medications to be linked with violence 7. Triazolam (Halcion) A benzodiazepine which can be addictive, used to treat insomnia. 0.Halcion is 8.7 times more likely to be linked with violence than other drugs, according to the study. 6) Atomoxetine (Strattera) Used to treat attention-deficit hyperactivity disorder (ADHD), Strattera affects the neurotransmitter noradrenaline and is 9 times more likely to be linked with violence compared to the average medication. 5) Mefoquine (Lariam) A treatment for malaria, Lariam has long been linked with reports of bizarre behavior. It is 9.5 times more likely to be linked with violence than other drugs. 4) Amphetamines: (Various) Amphetamines are used to treat ADHD and affect the brain's dopamine and noradrenaline systems. They are 9.6 times more likely to be linked to violence, compared to other drugs. 3) Paroxetine (Paxil) An SSRI antidepressant, Paxil is also linked with more severe withdrawal symptoms and a greater risk of birth defects compared to other medications in that class. It is 10.3 times more likely to be linked with violence compared to other drugs. (More on Time.com: Healthland's Guide to Life 2011) 2) Fluoxetine (Prozac) The first well-known SSRI antidepressant, Prozac is 10.9 times more likely to be linked with violence in comparison with other medications. 1) Varenicline (Chantix) The anti-smoking medication Chantix affects the nicotinic acetylcholine receptor, which helps reduce craving for smoking. Unfortunately, it's 18 times more likely to be linked with violence compared to other drugs — by comparison, that number for Xyban is 3.9 and just 1.9 for nicotine replacement. Because Chantix is slightly superior in terms of quit rates in comparison to other drugs, it shouldn't necessarily be ruled out as an option for those trying to quit, however. Akathisia is one of the extra pyramidal side effects. The others are dyskinesia (acute, tardive, chronic), neuroleptic malignant syndrome, and drug induced Parkinson’s symptoms. 3

Product Information (PI) and current practice suggests that extra pyramidal side effects are dose dependent. However, these extra pyramidal side effects are not only dose dependent but idiosyncratic. That is, if you give any person too much they will develop all of these side effects. Some persons develop them on low doses. Since the availability of the discipline of pharmacogenetics, it is known that there is a thousand-fold variation in metabolic rates for some drugs among people and this is caused by genetic factors. We can test for these problems buy buccal swab and $290, half the cost of a day in hospital. There is a huge literature on this. There are textbooks on this published by the American Psychiatric Press. All this knowledge seems to have escaped psychiatrists in Australia. Extra pyramidal side effects (akathisia, dyskinesia) are not only dose dependent but idiosyncratic. These formerly regarded as idiosyncratic are attributable to diminished metabolism at cytochrome P450 and perhaps at the phase two metabolism of UGTs. Where they are not due to overdosing, size of dose, co- prescribed cytochrome inhibitors and inducers, removal of inducers, competition for substrate enzymes they are understood by doing genetic testing and examining drug drug interactions. Australian psychiatrists are poorly educated about akathisia. I have rarely had a reasonable conversation with an Australian trained psychiatrist about the condition. Even though the world expert Professor Perminder Sachdev is at University of NSW. Neuroleptic-induced and SSRI-induced akathisia has been in the DSM at 333.99 since 1994. Key opinion leaders who are favourites of the pharmaceutical industry also educate Australian psychiatrists and much of the literature about these drugs has been ghost written over their well-remunerated signatures. For example, PHaRMAs that commission who publish efficacy studies feel no obligation to report side effects, strokes, deaths, violence. I can provide detailed evidence of this for a number of drugs. There have been homicides and suicides in clinical trials for SSRIs and atypicals. Homicide on psychiatric drugs has been well reported since 1975. Homicide is the tip of an iceberg of aggression and violence and that is what this Inquiry should be about. Akathisia became epidemic after the introduction of Prozac (fluoxetine) in 1990. There was and akathisia rate of 27% for antidepressants and variable rates in atypical antipsychotics, 27% akathisia for Zyprexa at 15 mg daily is reported in US product information, Australian PI is inadequate. Indeed, the Diagnostic and Statistical Manual says that akathisia rates are 25% - 75% of users depending on criteria. Neurotoxic side effects appear to be the cause of the crisis in mental health and the increasing violence, suicide and homicide among mental health patients. This seems to be better known in the United Kingdom and the United States than in Australia. In Australia, the notion that these drugs have any side effects at all has been strongly resisted by peer reviewers in various situations. I enclose some papers from the New York Review of Books. 4

Here are footnoted some questions in New South Wales Parliament which were asked last year. 10218—PSYCHIATRIC DRUGS Mr. Daryl Maguire to the Deputy Premier, and Minister for Health—

Has the Minister and her predecessors been warned by individuals that suicides committed by patients and clients under mental health care could be caused by psychiatric drugs:

That affect persons who have a genetically determined inability to metabolize them;

That such persons should be recognized by their adverse medication responses?

How many persons have committed suicide whilst under mental health care in the years 2003 to 2008?

How many have committed homicide?

Do these figures represent a deterioration or improvement in the numbers of suicides under mental health care:

Before 1990;

Before 2002?

Answer—

I am advised:

NSW Health advises me that there has been correspondence to previous Health Ministers in relation to this issue. I have also received such correspondence. The Chief Psychiatrist in consultation with the NSW Mental Health Clinical Advisory Council is currently considering these issues.

According to the Mental Health Client Incident Information System, there were 937 notifications of suspected suicides of persons under mental health care that were reported to the NSW Health Department between 1 Jan 2003 and 31 Dec 2008.

According to the Mental Health Client Incident Information System, there were 43 notifications of suspected homicides by persons under mental health care that were reported to the NSW Health Department between 1 Jan 2003 and 31 Dec 2008. 5

It is not possible to compare the data over this time period due to the fact that different methodology was used to collate this data.

This is not entirely true, as the Sentinel Events Committee itself charted a rising rate of suicide and homicides under state mental health care; they also occur under private care, where coroners pay no attention.

From Tracking Tragedy reports Reported suicide deaths of patients in contact with mental health services, and all suicide deaths in NSW 1993-2001.ii,iii Year Suicides Suicides in Percent of in NSW mental all NSW health care suicides Suicides include inpatient suicides and suicides within 28 days of contact with mental health care. (Updated numbers in brackets) 1993 676 68 10% 1994 798 72 9% 1995 747 (748) 100 13% 1996 811 (817) 136 17% 1997 946 (950) 166 18% 1998 827 (832) 143 17% 1999 846 (854) 173 20% 2000 738 156 21% 2001 775 (789) 159 21% 2002 643 117 ? 2003 ? ? ? 2004 In 2004, the mode of counting changed to counting mental health and suicides of inpatients or within 7 days of contact with mental health care. 2005 80-110 Information requests for this data have been denied 2006 80-110 2007 80-110 2008 2009

Between 2002 and 2009 the NSW Mental Health Sentinel Events Review Committee recorded well over 9 suicides and 38 homicides committed by persons under public sector mental health care. It did not examine medication. Suicide numbers under mental health care in NSW rose from 68 in 1993 to 173 in 1999 and 159 in 2001, a proportional increase from 10% to 6

21% of all suicides in NSW and one would expect pro rata to population rises in The Review Committee found in that half the sample of patients died by suicide before the tenth day of their stay in mental health facilities, and 30% died within the first three days of their episode of care. By 2008, this Review Committee also examined 38 homicides committed by persons under public sector mental health care and nearly 2000 suicides. A Victorian psychiatric unit had 13 suicides in 13 months in 2002-3 but this did not attract inquiry.iv v Leaving aside suicide and its underpinning presentations and correlates, outcomes other then suicide for treated mental illness have been deteriorating. The number of people under mental health care has also increased. Question asked on 13 May 2010 (session 54-1) and published in Questions & Answers Paper No. 197 .

Further problems include that there is a poor understanding in Australia of either therapeutic window of opportunity or half lives of drugs. These determine dosing. As a result, doses are very large and often cumulative. For instance, Product Information (PI) on Zyprexa (olanzapine) (which causes hallucinosis in 4% of the population who have never hallucinated before and has an akathisia rate of 27% o in clinical trials on 15 mg daily)) PI states the following. The following information comes for both Australian and American product information: Oral Administration Olanzapine is well absorbed and reaches peak concentrations in approximately 6 hours following an oral dose. It is eliminated extensively by first pass metabolism, with approximately 40% of the dose metabolised before reaching the systemic circulation. Food does not affect the rate or extent of olanzapine absorption. Pharmacokinetic studies showed that ZYPREXA tablets and ZYPREXA ZYDIS (olanzapine orally disintegrating tablets) dosage forms of olanzapine are bio-equivalent. Olanzapine displays linear kinetics over the clinical dosing range. Its half-life ranges from 21 to 54 hours (5th to 95th percentile; mean of 30 hr), and apparent plasma clearance ranges from 12 to 47 L/hr (5th to 95th percentile; mean of 25 L/hr). Administration of olanzapine once daily leads to steady-state concentrations in about one week that are approximately twice the concentrations after single doses. Plasma concentrations, Half-life, and clearance of olanzapine may vary between individuals on the basis of smoking status, gender, and age (see Special Populations). 7

The significance is if a drug has a half life of more then 24 hours it accumulates in the body as only half of what is there is excreted in next 24 hours, and it accumulates to toxic levels. Doses should be decreased and sooner or later more than half if its users become toxic, as therapeutic level is low 9-29 ng/mL, this overloading by people who still believe that 100 ngm/ml is a therapeutic level accounts of a lot of aggression, too high blood levels and other forms of idiosyncrasy. The half-life of Clopixol (zuclopenthixol depot) is 19 days. It is almost inevitably given fortnightly with resulting accumulation, increasing violence as akathisia toxicity develops. It was being done in your prison hospital when I visited a patient. He was on another antipsychotic, amisulpride, as well. There is no guideline in the world that would permit that. It is an undisclosed experiment for which consent was being daily refused by his better informed guardian. There is no culture of therapeutic drug monitoring in Australia. Australian psychiatrists show no inclination to learn new material. My experience is that I might as well be talking about star signs when I tell my colleagues that someone is a poor metaboliser. I believe that the Federal Government needs to hold a Royal Commission into pharmaceutical industry fraud misinformation and the practice of psychiatry, and education of psychiatrists and role of coroners in investigating this evidence in suicide inquests. So far they simply refuse to hear it, refuse. Coroners investigate these matters in USA and UK, why no there? In the United States, qui tam whistleblower cases are netting literally billions of dollars for the state. This still does not compensate a country like Australia which funds public health cares while drug companies get privatised profits and socialised costs. I am happy to give evidence at any open hearing.

I am happy to provide peer reviewed evidence for everything that I have said here. Yours sincerely,

Dr Yolande Lucire Consultant Psychiatrist

i ii NSW Mental Health Sentinel Events Review Committee. Tracking Tragedy A systemic look at suicides and homicides amongst mental health inpatients. First Report of the Committee. December 2003. 8

iii NSW Mental Health Sentinel Events Review Committee Tracking Tragedy A systemic look at homicide and non-fatal serious injury by mental health patients, and suicide death of mental health inpatients Fourth Report of the Committee. March 2008 http://www.health.nsw.gov.au/pubs/2009/pdf/tracking_tragedy_2008_fourth_report.PDF iv Greenland H. The Bulletin October 3, 2003. 333.99 Neuroleptic-Induced Acute Akathisia DSM IV Diagnostic Features The essential features of Neuroleptic-Induced Acute Akathisia are subjective complaints of restlessness and at least one of the following observed movements: fidgety movements or swinging of the legs while seated, rocking from foot to foot or "walking on the spot" while standing, pacing to relieve the restlessness, or a inability to sit or stand still for at least several minutes. In its most severe form, the individual may be unable to maintain any position for more than a few seconds. The subjective complaints include a sense of inner restlessness, most often in the legs; a compulsion to move one's legs; distress if one is asked not to move one's legs; and dysphoria and anxiety. The symptoms typically occur within 4 weeks of initiating or increasing the dose of a neuroleptic medication and can occasionally follow the reduction of medication used to treat or prevent acute extrapyramidal symptoms (e.g., anticholinergic agents). The symptoms are not better accounted for by a mental disorder (e.g., Schizophrenia, Substance Withdrawal, agitation from a Major Depressive or Manic Episode, hyperactivity in Attention-Deficit/Hyperactivity Disorder) and are not due to a nonneuroleptic substance or to a neurological or other general medical condition (e.g., Parkinson's disease, iron-deficiency anemia). Associated Features and Disorders The subjective distress resulting from akathisia is significant and can lead to noncompliance with neuroleptic treatment. Akathisia may be associated with dysphoria, irritability, aggression or suicide attempts. Worsening of psychotic symptoms or behavioural dyscontrol may lead to an increase in neuroleptic medication dose, which may exacerbate the problem. Akathisia can develop very rapidly after initiating or increasing neuroleptic medication. The development of akathisia appears to be dose dependent and to be more frequently associated with particular neuroleptic medications. Acute akathisia tends to persist for as long as neuroleptic medications are continued, although the intensity may fluctuate over time. The reported prevalence of akathisia among individuals receiving neuroleptic medication has varied widely (20%-75%). Variations in reported prevalence may be due to a lack of consistency in the definition of caseness, neuroleptic prescribing practices, study design, and the demographics of the population being studied. Differential Diagnosis Neuroleptic-Induced Acute Akathisia may be clinically indistinguishable from syndromes of restlessness due to certain neurological or other general medical conditions, to nonneuroleptic substances, and to agitation presenting as part of a mental disorder (e.g., a Manic Episode). The akathisia of Parkinson's disease and iron-deficiency anemia are phenomenologically similar to Neuroleptic- Induced Acute Akathisia. The frequently abrupt appearance of restlessness soon after initiation or increase in neuroleptic medication usually distinguishes Neuroleptic-Induced Acute Akathisia. Serotonin-specific reuptake inhibitor antidepressant medications may produce akathisia that appears to be identical in phenomenology and treatment response to Neuroleptic-Induced Acute Akathisia. Akathisia due to nonneuroleptic medication can be diagnosed as Medication- Induced Movement Disorder Not Otherwise Specified. Other situations that might be included under Medication-Induced Movement Disorders Not Otherwise Specified are acute akathisia with only subjective or only objective complaints, but not both; and akathisia occurring late in the course of treatment (e.g., 6 months after initiation of, or increase in the dose of, a neuroleptic). Neuroleptic-Induced Tardive Dyskinesia also often has a component of generalized restlessness that may coexist with akathisia in an individual receiving neuroleptic medication. 9

Neuroleptic-Induced Acute Akathisia is differentiated from Neuroleptic-Induced Tardive Dyskinesia by the nature of the movements and their relationship to the initiation of medication. The time course of symptomatic presentation relative to neuroleptic dose changes may aid in this distinction. An increase in neuroleptic medication will often exacerbate akathisia, whereas it often temporarily relieves the symptoms of Tardive Dyskinesia. Neuroleptic-Induced Acute Akathisia should be distinguished from symptoms that are bettered accounted for by a mental disorder. Individuals with Depressive Episodes, Manic Episodes, Generalized Anxiety Disorder, Schizophrenia and other Psychotic Disorders, Attention-Deficit/Hyperactivity Disorder, dementia, delirium, Substance Intoxication (e.g., with cocaine), or Substance Withdrawal (e.g., from an opioid) may also display agitation that is difficult to distinguish from akathisia. Some of these individuals are able to differentiate akathisia from the anxiety, restlessness and agitation characteristic of a mental disorder by their experience of akathisia as being different from previously experienced feelings. Other evidence that restlessness or agitation may be better accounted for by a mental disorder includes the onset of agitation prior to exposure to the neuroleptic medication, absence of increasing restlessness with increasing neuroleptic medication doses, and absence of relief with pharmacological interventions (e.g., no improvement after decreasing the neuroleptic dose or treatment with medication intended to treat the akathisia). iv Adverse reactions to antipsychotic in the cases that have been mistaken for schizophrenia and treated by further neurotoxic medicines such as Zyprexa and Risperdal, through competition for CYP450 enzymes, in persons who should be already considered suspect of being genetically polymorphic or ‘slow metabolizes

International Journal of Risk & Safety in Medicine 16 (2003/2004) 31–49 31 IOS Press

Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs): A review and analysis ∗

Peter R. Breggin 101 East State Street, No. 112, Ithaca, NY 14850, USA

Abstract. Evidence from many sources confirms that selective serotonin reuptake inhibitors (SSRIs) commonly cause or ex- acerbate a wide range of abnormal mental and behavioral conditions. These adverse drug reactions include the following overlapping clinical phenomena: a stimulant profile that ranges from mild agitation to manic psychoses, agitated depression, obsessive preoccupations that are alien or uncharacteristic of the individual, and akathisia. Each of these reactions can worsen the individual’s mental condition and can result in suicidality, violence, and other forms of extreme abnormal behavior. Ev- idence for these reactions is found in clinical reports, controlled clinical trials, and epidemiological studies in children and adults. Recognition of these adverse drug reactions and withdrawal from the offending drugs can prevent misdiagnosis and the worsening of potentially severe iatrogenic disorders. These findings also have forensic application in criminal, malpractice, and product liability cases.

1. Introduction

Soon after the introduction of the first selective serotonin reuptake inhibitor (SSRI), fluoxetine (Prozac) into the United States marketplace in January 1988, reports began to appear describing fluoxetine- induced violence against self and others. In May 1990 the U.S. Food and Drug Administration required the manufacturer of Prozac, Eli Lilly and Company, to add “suicidal ideation” and “violent behaviors” to the Postintroduction Reports section of its label.1 In 2003 the British Committee on the Safety of Medi- cines and the U.S. Food and Drug Administration issued warnings about increased rates of self-harm and suicidal behavior in children and youth under the age of 18 exposed to paroxetine (Paxil) [78]. Most recently, on August 22, 2003, the manufacturer of venlafaxine (Effexor) issued a similar “Dear Doctor” letter warning about the increased risk of “hostility and suicide-related adverse events, such as suicidal ideation and self-harm” in children age 6 to seventeen [79]. In August 11, 1990, an editorial in The Lancet [53] included “the promotion of suicidal thoughts and behaviour” (p. 346) among the adverse effects of fluoxetine. The following year, the British National Formulary, a joint publication of the British Medical Association and the Royal Pharmaceutical Society (1991), listed suicidal ideation and violent behavior as fluoxetine side effects. Subsequently, many books and reports have dealt with the subject of SSRI-induced violence and suicide (e.g., [10,12,14,35,40,72]). This report will provide an extensive review and analyze the literature concerning SSRI-induced sui- cide and violence, and identify several clinical syndromes that can cause the phenomena. It will examine

*The present paper appears simultaneously in Ethical Human Sciences, Journal of the International Center for the Study of Psychiatry and Psychology, and is published with permission of Springer Publishing Company, New York, NY, USA. 1The section that lists violence and suicide as possible adverse drug reactions begins with a caveat that the reported reactions “may have no causal relationship with the drug.”

0924-6479/03/04/$8.00  2003/2004 – IOS Press. All rights reserved 32 P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) the clinical and forensic implications of these findings, and also examine the ethical and scientific contro- versy surrounding the capacity of psychoactive agents to cause “bad behavior.” (The subject of abnormal behavior induced by withdrawal from SSRIs will be considered in a later publication. Also see [15]).

2. The class of SSRIs

These selective serotonin reuptake inhibitors (SSRIs) include fluoxetine (Prozac), sertraline (Zoloft), paroxetine (Paxil), fluvoxamine (Luvox), citalopram (Celexa), and, most recently, escitalopram (Lexapro). These drugs block the removal of the neurotransmitter serotonin from the synaptic cleft. A number of other antidepressants are potent non-selective serotonin reuptake inhibitors (NSRIs). These include the atypicals venlafaxine (Effexor) and nefazodone (Serzone) and the tricyclic clomipramine (Anafranil). When observations are made in clinical practice and in the scientific literature concerning the impact of SSRIs, they are typically treated as a single category or class of pharmacological agents. It is generally recognized that an adverse mental or behavioral reaction, such as agitation or mania, that is observed in regard to one SSRI is likely to be found with all the other SSRIs.2 While usually examined as sepa- rate classes of antidepressants, the NSRIs also share many characteristics with the SSRIs, including the capacity to induce stimulation, anxiety, agitation, and mania.

3. SSRI-induced mania and the continuum of stimulation

All antidepressants cause mania and mania is an acknowledged adverse effect in the FDA-approved label of all antidepressants. Preda et al. [66] carried out a retrospective study of 533 psychiatric hospital admissions over a fourteen month period and found that 43 (8.1%) could be attributed to antidepressant- induced mania and/or psychosis. The percentages for each antidepressant were as follows: the SSRIs (70%), the newer atypicals (venlafaxine, nefazodone, and buproprion) (21%), and the older tricyclic antidepressants (amitryptyline, desipramine, imipramine, nortriptyline) (21%). The total percentage ex- ceeded 100% because of overlapping medications in five cases. Twelve of the cases represented new- onset mania or psychosis. The three illustrative cases were severe, including two with marked suicidal potential. A 52-year-old married woman with a past history of bipolar disorder developed “command auditory hallucinations with suicidal content” while taking desipramine and fluvoxamine, as well as risperidone, zolpidem, and oxazepam (p. 31). A 42-year-old woman with a one-year history of depres- sion “began to experience derogatory and then command auditory hallucinations to kill herself” while on fluoxetine as well as lithium and thioridazine (p. 31). Finally, a 49-year-old woman taking venlafaxine for “low mood and anxiety” developed symptoms of paranoia, feelings of doom, and a delusion that television messages were being directed at her (p. 31). All three patients improved rapidly with treatment that included termination of the antidepressants. Mania with psychosis is the extreme end of a stimulant continuum that often begins with lesser degrees of insomnia, nervousness, anxiety, hyperactivity and irritability and then progresses toward more severe agitation, aggression, and varying degrees of mania. At the lower end of the continuum, an ordinarily shy

2Marangell, Yudofsky and Silver [58] observed, “All SSRIs have a similar spectrum of efficacy and a similar side-effect profile” (p. 1035). In the same vein, Borg and Brodin [6] remarked, “There seems to be little difference between the SSRIs with respect to frequency and severity of adverse effects” (p. 66) and Grimsley and Jann [37] concluded, “Overall, the adverse-effect profiles of the different SSRIs are comparable” (p. 938). P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) 33

Table 1 Mental and behavioral Adverse Drug Reactions (ADRs) in adults caused by paroxetine from the 2001 FDA-approved label for Paxil Frequenta Infrequentb Mania/Hypomania 2.2% of bipolar patients Paranoid reaction Mania/Hypomania 1% of depressed patients Psychosis Insomnia 13% Hostility Nervousness 5% Euphoria Anxiety 5% Delirium Agitation 1% Hallucinations Drugged feeling 2% Abnormal thinking Confusion 1% Depersonalization Central nervous system stimulation Neurosis Emotional lability Lack of emotion Concentration impaired Libido increased Amnesia Depression Tremor 8% Sweating 11% Palpitation 3% aFrequent means at a rate of 1% or greater. bInfrequent means at a rate between 1% and 0.1%. All ADRs with percent- ages (%) are for depressed patients in placebo controlled clinical trials. ADRs without percentages are taken from the entire data pool of 7,678 patients ad- ministered Paxil, including 6,145 depressed patients. Table compiled from the label by Peter R. Breggin. young woman acted silly and more outgoing on fluoxetine, and then developed suicidal feelings when she missed doses [8]. Evidence for this stimulant continuum or syndrome of SSRI-induced adverse reactions will be found in many of the following reports. Sometimes the psychoses lack manic qualities and appear more paranoid in nature. SSRI labels tend to be organized in ways that avoid any implication that the medications can cause stimulation; but detailed analyses of the labels disclose that these drugs produce a continuum of stim- ulation (Breggin [13] for an analysis of the Luvox label; Breggin and Breggin [14] for an analysis of the Prozac label). Table 1 was compiled to illustrate the spectrum of SSRI-induced adverse drug reac- tions and illustrate the frequency of stimulant-like effects. All of the effects listed in the table can occur with stimulants such as amphetamine and cocaine, and many are typical of these stimulants, including hypomania/mania, insomnia, nervousness, anxiety, agitation, central nervous system stimulation, emo- tional lability, tremor, sweating, and palpitation, as well as paranoid reaction, psychosis, hostility, and euphoria. Confirmation of the stimulant syndrome was provided in a previously undisclosed internal document from Eli Lilly and Company, the manufacturer of Prozac (fluoxetine) that was obtained during discovery in product liability suits against the company (Beasley [5]; discussed in Breggin [10, pp. 87–88]; Breggin [9], Fentress Trial Exhibit [28]). Charles Beasley of the company’s Division of Clinical Neurosciences evaluated what he called “activation” in patients taking fluoxetine or placebo in the controlled clinical trials used for Food and Drug Administration (FDA) approval of Prozac for depression. Beasley defined activation as including any of the following: nervousness, anxiety, agitation, insomnia. Beasley found that 38% of fluoxetine-treated patients developed “activation,” but only 19% of placebo patients. The 34 P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) proportion of patients “activated” by fluoxetine would have been higher if other expressions of stimula- tion had been included, such as akathisia, hyperactivity, euphoria, and mania. It would have been further increased if many of the patients had not been prescribed sedative tranquilizers to quiet their symptoms of stimulation [14]. Reports authored by Kapit [47,48], the FDA official in charge of evaluating adverse drug effects dur- ing the approval process of Prozac for depression, repeatedly warned that fluoxetine has a stimulant profile similar to amphetamines. He was concerned that stimulant effects such as insomnia, nervousness, anorexia, and weight loss would produce agitated depression and worsen the condition of some depressed patients (unpublished data discussed and quoted in Breggin [10, pp. 79–81]).3 Clinically, agitated depression is an unstable condition that can lead to violence against self or others more frequently than a non-agitated depression. A number of reports cited in the following sections will mention agitation in patients who behave abnormally as a result of antidepressant effects.

4. Case studies related to SSRI-induced suicidality, violence, and extreme abnormal behavior in adults

4.1. Case reports of mania, violence, and suicide

There are many case reports in the scientific literature documenting the capacity of SSRIs to cause mania in adults, often in association with irritability and aggression. Christensen [18], for example, re- ported on the case of 32-year-old man who developed his first manic episode while taking paroxetine. He became psychotic and “threatened his parents with physical harm” (p. 1400). Other reports cite flu- voxamine as the causative agent (e.g., [17,24,61]). Dorevitch et al. [24] described three cases of fluvoxamine-induced mania. Each case was recognized quickly and the drug was reduced in dose or stopped so that potentially disastrous outcomes were avoided. Had the patients been more secretive or the monitoring less effective, the results could have been more drastic in outcome. In the first case, the patient developed a psychotic manic state with audi- tory hallucinations. In the second case, the patient became euphoric, displayed increased energy, inap- propriate behavior with sexual advances toward other patients, irritability, and fears that people were out to kill him. In the third case, the patient developed multiple signs of mania from excessive sexual activi- ties to excessive talking and argumentativeness. Manic patients who are “argumentative” can sometimes become very aggressive when thwarted. Okada and Okajima [61] described three cases of aggressive and violent behavior induced by fluvox- amine. On 150 mg per day, a 32-year-old woman became both irritable and aggressive, and she expressed impulsive violence during arguments with her family. She improved after her fluvoxamine was reduced (but not stopped). Paroxetine with its high impact and short duration of action is more like fluvoxamine than like fluoxetine. A 29-year old woman on 150 mg of fluvoxamine daily became nervous and irritable and then impulsively violent and was admitted to a psychiatric hospital. She improved with discontin- uation of the drug and further treatment with other medications. A 28-year-old woman receiving 150 mg of fluvoxamine daily exhibited signs of irritability and aggressive behavior and expressed violence

3In a discussion of similarities among SSRIs (in Leonard [55, pp. 9–10]), J. Mendlewicz points out that animal studies indicate that some SSRIs can cause “alertness, agitation or anxiety, or interfere with the quality of sleep” and he asks, “Are there differences between SSRIs in terms of these early ‘activating’ effects?” B.E. Leonard answers that there no known differences among the SSRIs in this regard to their capacity to induce “anxiety or agitation.” P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) 35 toward her mother. She improved when the fluvoxamine was stopped and other medications instituted. The authors warned about the existence of impulsive and aggressive behavior induced by fluvoxamine. In another case report, a woman taking fluvoxamine became suicidal and had to be hospitalized [4]. In the hospital the fluvoxamine dose was increased from 50 mg per day to 150 mg per day whereupon her condition worsened and she began to experience auditory hallucinations. The fluvoxamine was dis- continued and she recovered within twenty-four hours, confirming that the medication had caused the depression and psychosis.

4.2. Case reports of SSRI-induced akathisia and aggression

Akathisia is a painful inner agitation that typically manifests as the inability to sit still or to stop moving. The hyperactivity may manifest itself subtly as a feeling of jitteriness or grossly as frantic pacing or repeatedly sitting up and down. The inner agitation associated with akathisia can become extremely uncomfortable, causing the individual to feel tortured from within (see vivid descriptions in Van Putten [74,75] and Breggin [10], leading to extreme irritability and suicide or violence. Although akathisia by definition usually involves a hyperactive movement component, clinical expe- rience indicates that the same jittery, agitated subjective experience, accompanied by irritable, violence or suicidal feelings, can occur without the specific component of feeling driven to move about. Indeed, on earlier occasions, the individual may have experienced the associated compulsion toward hyperactiv- ity. Healy [41] made similar observations. Furthermore the Diagnostic and Statistical Manual of Mental Disorders [3] specifically allows for an alternative diagnosis of “acute akathisia with only subjective or only objective complaints, but not both” (p. 801). Lipinski et al. [57] reported on five cases of akathisia caused by fluoxetine. They also reviewed the literature and found rates of 9.7% to 25% for fluoxetine-induced akathisia. They concluded, “In summary, fluoxetine, and perhaps other antidepressant drugs as well, may produce the side effect of akathisia fairly frequently” (p. 342). The Public Citizen Health Research Group [67] estimated a rate of 15–25%. While studies of SSRI-induced akathisia vary greatly in the frequency with which this disorder is observed, they confirm that it is common. Lane [54] observed, “SSRI-induced akathisia may represent a form of serotonergic overstimulation or serotonin toxicity” (p. 203). He also cited research linking the phenomenon to the impact of SSRIs on the dopaminergic system. He warned, “The emergence of symptoms of akathisia could be mistaken for a worsening of depression, especially the conversion of a non-agitated depression to an agitated form” (p. 206). This error in judgment could lead to the prescription of increased doses of the offending medication, resulting in a severely worsened condition. Lane cited studies indicating that “fluoxetine is not an appropriate choice of antidepressant for depressed patients with agitation and restlessness” (p. 206) because it can lead to increased rates of agitation, anxiety and manic reactions. Rothschild and Locke [69] reported on three cases of fluoxetine-induced suicidality associated with akathisia. Each case of suicidality developed on fluoxetine (challenge) and then resolved when the drug was stopped (dechallenge). The suicidality then returned when the drug was started a second time (rechal- lenge) and stopped again when the drug was stopped (a second dechallenge). During rechallenge each of the patients developed akathisia and reported that this feeling had caused them to become suicidal each time. Wirshing et al. [77] reported on five cases of a fluoxetine-induced syndrome consisting of akathisia and suicidality. In all five cases, the akathisia and the suicidality remitted when the drug was stopped or reduced in dosage. In one case, a rechallenge with an increased dose of fluoxetine again produced 36 P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) the syndrome. They concluded, “Our cases appear to confirm that certain subjects experience akathisia while taking fluoxetine and that this effect is dose-related in the individual patient. ...Furthermore, like the akathisia in the neuroleptic-treated schizophrenic population, ‘fluoxetine akathisia’ can apparently be associated with suicidal ideation, sometimes of a ruminative intensity” (p. 581). Masand et al. [59] reported on two cases of suicidality in association with fluoxetine. One of the patients suffered from akathisia. In both cases, the suicidal feelings subsided shortly after stopping the medication. Neither patient had prior suicidal ideation. Both developed violent fantasies (hanging and jumping out a window). Akathisia will also be found in combination with SSRI-induced mania and aggression (see below).

4.3. Case reports of SSRI-induced obsessive suicidality and aggression

A number of clinical reports have described a syndrome of obsessive SSRI-induced suicidality and aggression that seems particular to these drugs. The characteristics were first described in a report on fluoxetine-induced obsessive suicidality by Teicher et al. [71]. They summarized, “Six depressed patients free of recent serious suicidal ideation developed intense, violent suicidal preoccupation after 2–7 weeks of fluoxetine treatment” (p. 207). Additional cases and potential mechanisms of action were analyzed by Teicher et al. [72]. Dasgupta [21] described a similar case of “intense suicidal preoccupation” (p. 1570) after four weeks of fluoxetine treatment in a woman who had not been previously suicidal. She, too, rapidly recovered on stopping the fluoxetine. The syndrome was also described by Rothschild and Locke [69] in three patients taking fluoxetine, each of whom again reacted with suicidality when rechallenged by a second administration of fluoxetine. Hoover [43] described another similar case who developed intense, violent suicidality on the two occasions that she was exposed to fluoxetine. Creaney et al. [20] described two patients who became suicidal on SSRIs. One patient developed dysphoria and manic symptoms on fluoxetine and then developed a similar syndrome, this time with suicidal feelings, on fluvoxamine. Another patient became intensely and violently suicidal sixteen days after starting fluoxetine. Gualtieri [38] described the “case of a mentally handicapped gentleman whose rates of self-injurious behavior doubled on fluoxetine, and then fell to baseline after the drug was withdrawn” (p. 393). Gualtieri pointed out that fluoxetine can cause apathy and indifference in some patients and, conversely, mania in others. Based on the literature and my clinical experience, the syndrome of SSRI-induced obsessive suicidality and violence includes many and sometimes all of the following: 1. A relatively sudden onset and rapid escalation of the compulsive aggression against self and/or others. 2. A recent (typically within two months) initial exposure to the medication, or a recent change in the dose of the medication, or a recent addition or removal of another psychoactive substance to the regimen. 3. The presence of other adverse drug reactions, often involving akathisia or stimulation along a con- tinuum from irritability and agitation to agitated depression and mania. 4. Resolution of the syndrome after termination of the causative medication, often with a marked overall improvement in the individual’s mental status. 5. An extremely violent and/or bizarre quality to the thoughts and actions. 6. An obsessive, compelling, unrelenting quality to the thoughts and actions. P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) 37

7. An out-of-character quality for the individual as determined by the individual’s history. 8. An alien or ego-dystonic quality as determined by the individual’s subjective report.

5. Epidemiological studies and clinical trials related to SSRI-induced, depression, suicidality, violence, and extreme abnormal behavior in adults

5.1. Epidemiological studies and clinical trials of SSRI-induced mania and aggression

The following studies focus on SSRI-induced manic-like symptoms and mania. The clinical syndrome of mania is commonly associated with increased irritability, aggressiveness, physical violence, and a va- riety of antisocial and criminal behaviors ([3, pp. 357–362]). However, as the following review indicates, many patients will switch from mania to depression and suicidality. As documented in the FDA-approved labels for SSRIs, clinical studies conducted for the FDA- approval process have shown increased rates of mania. For example, even in the relatively short four- to-six week trials used for the approval of Prozac for depression, slightly more than 1% of patients developed hypomania and mania (see, for example, the 1990 label for Prozac for depression). An un- published FDA report obtained through the Freedom of Information Act indicated that fluoxetine caused mania at a three times greater rate than tricyclic antidepressants given in the same studies (Kapit [48]; reviewed in Breggin [10, p. 86]). Furthermore, in 23 of the 33 cases, fluoxetine caused mania in patients with no past history of mania. In no cases did the older antidepressants cause mania in patients with no prior history. This data contradicts the commonly held clinical notion that SSRI-induced mania is limited to patients with an underlying bipolar disorder. Howland [44] found 11 cases of SSRI-induced mania among approximately 184 (6%) patients treated at a university clinic and hospital with a variety of SSRIs, including fluoxetine, paroxetine, and sertraline. The episodes were “generally quite severe” (p. 426). Eight of the 11 patients became psychotic and 4 were so agitated that they had to be put in seclusion, even though they were probably receiving additional medication to control their iatrogenic mania. Ebert et al. [26] attempted to develop a rate for severe mental aberrations caused by fluvoxamine. They carried out a prospective study of 200 inpatients over a total of 8200 treatment days with the SSRI. Fourteen patients (17%) developed hypomania according to DSM-IV criteria. Three patients (1.5%) developed insomnia, agitation, confusion and incoherent thoughts. These patients became po- tentially violent and suicidal. One, a 35-year old man, developed agitation and restless legs that pro- gressed to insomnia, confusion, paranoid ideas, and hallucinations. He recovered after fluvoxamine was stopped. Another patient, a 38-year old man, developed psychomotor agitation with insomnia that pro- gressed to aggressiveness, incoherent thoughts, confusion, auditory hallucinations and paranoid ideas. He also recovered when fluvoxamine was stopped. A third patient, another 35-year-old man, devel- oped insomnia and then became agitated with restless legs, and severely depressed with suicidal ideas. He was also incoherent, and confused with paranoid ideas. He too recovered within a few days after stopping the medication. Based on the clinical descriptions, all three patients probably suffered from akathisia. Ebert and his colleagues [26] summarized the syndrome as consisting of insomnia, confusion, incoher- ent thoughts, agitation, hallucinations and paranoid ideas. They observed that it was especially frequent in combination with other drugs. They considered it rare but their data indicate that it was common. 38 P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs)

Adding up the 14 hypomanic patients and the 3 psychotic and aggressive patients, there were at least 17 severe central nervous system (psychological) reactions among 200 patients for a rate of 8.5%.4 Peyre et al. (1992) reviewed the histories of 189 patients treated with fluvoxamine and found a rate of 2.5% for manic switches, i.e., the development of mania during treatment for major depression. Troisi et al. [73] used 20 mg per day of fluoxetine to treat nineteen retarded inpatients with epilepsy and a current or recent history of aggressive behavior. All of them were taking other medications as well. Using a standardized rating scale for assessing behavior before, during, and after treatment with fluoxetine, they found an increase in aggressive behavior in nine patients. Unexpectedly, the behavior decreased to below pre-treatment levels after the withdrawal of the fluoxetine. The authors conclude that fluoxetine can worsen aggression in retarded patients with impulsive aggressive behavior. The FDA conducted an epidemiological study comparing fluoxetine to trazodone in regard to spon- taneous reports concerning hostility and intentional injury (Food and Drug Administration [31]). When the FDA factored in the greater number of prescriptions for fluoxetine, fluoxetine still had a higher fre- quency of reports for aggressive and violent behavior. Furthermore, the reports began to accumulate before the controversy surrounding fluoxetine and violence (suppression of the data discussed in Breggin [10, pp. 88–89]).

5.2. Epidemiological studies and clinical trials of SSRI-induced depression and suicidality

An unpublished document obtained during discovery in product liability suits against the drug com- pany disclosed that Eli Lilly and Co., the manufacturer of Prozac (fluoxetine), had evaluated the compar- ative rates of suicide attempts on fluoxetine, amitryptyline and placebo. The data was generated during controlled clinical trials conducted for the FDA-approval process for Prozac for depression. Based on the company’s data for controlled clinical trials, patients taking fluoxetine were three-to-six times more likely to attempt suicide than a similar group of patients taking older antidepressants or placebos (see my testimony in Breggin [9]; reviewed in Breggin [10, pp. 89–91]). An evaluation by a consultant to the company, Avery Winokur, concluded that the increased rate might be due to fluoxetine-induced over- stimulation of the depressed patients (unpublished documents reviewed in Breggin [10, pp. 89 ff]; also see Breggin [9]). Healy [39] reviewed and reanalyzed data comparing the number of suicides and suicide attempts per patient in worldwide placebo controlled clinical trials used for the FDA antidepressant approval process (Khan et al. [49] and Khan et al. [50]). The drugs included four SSRIs (sertraline, paroxetine, citalo- pram, and fluoxetine). As a percentage of patient numbers, there was a statistically significant difference between combined suicides and suicide attempts among all SSRIs patients (1.55%) and among all SSRI trial placebo patients (0.48%). There were also a significantly greater number of completed suicides on SSRIs in the combined suicide and suicide attempt group, as well as in the paroxetine group individually, compared to placebo. Donovan et al. [23] found a significantly increased rate of completed suicide among patients treated with SSRIs compared to those treated with tricyclic and other antidepressants. After correcting the data for the number of prescriptions for each drug, SSRIs were 3.5 times more likely to be associated with suicide. The study was conducted in three regions of England and Ireland, and involved 222 suicides. Donovan et al. [22] also conducted a prospective study of 2776 consecutive cases of deliberate self- harm age seventeen and older who were seen at the accident and emergency department of Derbyshire

4According to the FDA, an adverse drug reaction rate of 1% is frequent or common. P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) 39

Royal Infirmary as a consequence of any act of deliberate self-harm during a two year period (1995– 1996). Of the 2776 cases, 307 had received an antidepressant 30 days or less prior to the incident of deliberate self-harm. With the rate of prescribing in Derbyshire taken into account, the relative incidence of deliberate self-harm was significantly higher (P<0.001) in patients who were prescribed the SSRIs fluoxetine, paroxetine, and sertraline compared to patients who were prescribed the tricyclics amitrypty- line, dothiepin and imipramine. The relative incidence of deliberate self-harm per 10,000 prescriptions was broken down in a table as follows: fluoxetine (19.8), sertraline (14.8), paroxetine (12.1), all SSRIs (16.6), imipramine (3.5), amitryptyline (3.0), and all tricyclics (5.6). Compared to amitryptyline, the rel- ative risk for all SSRIs was many times higher: fluoxetine (6.6), sertraline (4.9), paroxetine (4.0), and all SSRIs (5.5). Of interest in regard to causation, the risk for the tricyclic clomipramine was very high as well with a relative incidence of 13.8 and a relative risk compared to amitryptyline of 4.6. Among the tricyclics, clomipramine has the strongest inhibitory effect on serotonin reuptake (see, for example, Drug Facts and Comparisons [25]). Jick et al. [46] conducted an epidemiological study of reports from general practices (primary care) in the United Kingdom involving 172,598 patients who had at least one prescription for one of ten antide- pressants. Rates of suicides were compared for patients on the various antidepressants. Patients taking fluoxetine were twice as likely to commit suicide compared to patients on other antidepressants. In com- parison to three more sedating antidepressants – doxepin, imipramine, and amitryptyline – fluoxetine was four times more likely to be associated with suicide. Taking into account a past history of suicidal be- havior and/or antidepressant treatment, fluoxetine remained twice as likely to be associated with suicide. Nonetheless, the authors attempted to explain away the dramatic differences. Fisher et al. [29] conducted a phone survey of pharmacy patients taking various antidepressants and found a higher rate of suicidality on SSRIs. In a related study, Fisher et al. [30] compared fluoxetine with a more sedating antidepressant, trazodone. They concluded that fluoxetine caused “a higher incidence of psychologic/psychiatric adverse clinical events, including delusions and hallucinations, aggression, and suicidal ideation” (p. 235). Muijen et al. [60] conducted a six-week double-blind study comparing fluoxetine, mianserin, and placebo with 26, 27, and 28 starters respectively, and 14, 14, and 16 finishers respectively. Two of the fluoxetine patients “took an overdose within two weeks of starting the study, and in both cases this was related to a deteriorating clinical state that necessitated hospitalization” (p. 386). None of the patients in the other drug group or the placebo group suffered from this decline and suicidality. Remarkably, the authors do not include these reactions among the adverse drug effects. Gorman et al. [36] conducted an open trial of fluoxetine involving sixteen patients with panic disorder. They reported, “Two of the nonresponders became depressed and had suicidal ideation while taking fluoxetine. Only one of the two had a history of depression” (p. 331). The authors did not comment on this finding. Healy [40] conducted a randomized double-blind crossover study comparing the effects of sertraline to a non-SSRI antidepressant (reboxetine) in a group of healthy volunteers. Many of the 20 individuals developed adverse mental and neurological effects while taking the sertraline and two became severely disturbed. Case A, a 30-year-old woman, became withdrawn and ruminated over impulsive, disinhibited actions. She was also tearful and did not feel like herself. In addition, her diary recorded impulsiveness, irritability, over-sensitivity, and marked suspicion. She became obsessed with killing herself and almost threw herself beneath a car or train. Case B, an otherwise peaceful 28-year old woman, experienced severe road rage and actually grabbed a teenage boy and threatened to knock him down. On the SSRI, 40 P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) she felt aggressive and fearless. While emotionally disturbed and out-of-control (disinhibited), the two individuals nonetheless felt and appeared emotionally blunted. The mixture of apathy and disinhibited aggressiveness reported by Healy is probably a common find- ing in patient’s who act uncharacteristically violent as a result of taking SSRIs. Hoehn-Saric et al. [42] reported on “Apathy and Indifference in Patients on Fluvoxamine and Fluoxetine.” They described apa- thy, indifference, loss of initiative and disinhibition with and without hypomania in five patients. Levine et al. [56] reported that 7% of 59 non-depressed obese patients became depressed following a rapid increase in fluoxetine to a dose of 80 mg per day.

5.3. Coroner studies

Frankenfield et al. [33] conducted a retrospective case review of all deaths in Maryland where either fluoxetine or tricyclic antidepressants was forensically detected. The study covered a three and one-half year period of time. They found a statistically significant increase in violent suicides in association with fluoxetine (65% versus 23%). Violence was defined to include “gunshot or shotgun wounds, suffoca- tion, stabbing, strangulation, drowning, falls and jumping in front of a moving vehicle” (p. 109). The evaluation of the suicide attempts were blind to which medications were involved. Bost and Kemp [7] reviewed a series of coroner’s reports in Dallas, Texas, involving fifteen suicides associated with fluoxetine treatment. The study covered a nine month period. While they appreciated that their data was impressionistic, they warned that the proportion taking fluoxetine and committing suicide was high enough to be of concern to health care providers.

6. Studies related to SSRI-induced suicidality, violence, and extreme abnormal behavior in children

Many cases of SSRI-induced violent or suicidal behavior involve children or young adults. However, even in regard to cases involving older persons, the literature on children and youth is important. Adverse behavioral effects tend to show up more frequently and severely in children, providing a magnified view of the same or similar effects that the drugs are causing on adults.

6.1. Clinical case studies involving children

A single case study involving paroxetine described a sixteen-year-old who became manic with angry outbursts after three weeks on the drug [62]. In another single case study, a 17-year-old mildly retarded youngster was started on fluvoxamine 50 mg when he became depressed and anxious [70]. After a single dose, he developed increasing agitation and insomnia, followed in the next 24 hours by auditory and visual hallucinations, a fearful mood, and paranoid delusions about the devil. He required hospitalization and was treated with an antipsychotic drug. The authors believe that fluvoxamine caused the acute psy- chosis. As a third example of single case clinical reports, Wilkinson [76] described a character change with increased aggression in a fifteen-year-old boy taking fluoxetine. Uncharacteristically, he struck an- other youngster in the face. Fluoxetine was stopped and within a week he was no longer aggressive. The author identified blunting rather than akathisia as the motivational state. Koizumi [52] described a thirteen and one-half year old boy who developed manic symptoms on 40 mg per day of fluoxetine. These side effects disappeared when the dose was lowered to 15 mg per day. However, after fifteen months of fluoxetine treatment he then developed “explosive, angry outbursts over P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) 41 minor matters, which was totally unlike him” (p. 695). He then experienced a “weird” and ego-alien voice telling him to kill himself. He recovered from these symptoms within ten days of stopping fluoxetine.

6.2. Epidemiological studies and clinical trials involving children

Numerous epidemiological and clinical study reports confirm that SSRIs cause suicidal, violent and manic behavior in children and youth. Three controlled clinical trials conducted for the FDA-approval of paroxetine for children under age eighteen demonstrated a three times increased rate of self-harm and suicidal behavior in paroxetine- treated children compared to placebo. Based on this data in 2003 the British Committee on Medicines prohibited the use of paroxetine in children and the U.S. Food and Drug Administration issued a warning [78]. The manufacturer of venlafaxine recently disclosed unpublished data from its controlled clinical trials for major depressive disorder [79]. Individuals below 18 years of age exposed to venlafaxine had more than twice the relative risk than those exposed to placebo in regard to the development of hostility (2% versus <1%) and suicidal ideation (2% versus 0%). According to the FDA-approved label for fluvoxamine (Luvox in the Physicians’ Desk Reference [64]), the SSRI causes a 4% rate of mania in children under age 18, compared to no cases of mania produced in a similar group of children on placebo. The rate was at least four times greater than in adults (see Breggin [13] for a more complete analysis of the Luvox label). A controlled clinical trial found that fluoxetine caused a 6% rate of mania in depressed children and youngsters age 7–17 ([27, p. 1003]). The reactions were severe enough to cause the children to be dropped out of the trials. By contrast, none of the depressed youngsters on placebo developed mania. Jain et al. [45] made a retrospective examination of the medical charts of children and young men age 8–19 who had taken fluoxetine in a university clinic setting. The researchers found that 23% of fluoxetine- treated young people developed mania or manic-like symptoms. Another 19% developed drug-induced hostility and aggression, including a grinding anger with short temper and increasing oppositionalism. Constantino et al. [19] prospectively studied the course of aggressive behavior in nineteen SSRI-treated psychiatrically hospitalized adolescents who were not pre-selected for potential aggressiveness. They reported symptoms of aggression toward self or others in 12 of 19 patients on SSRIs. Of the 19 patients, 13 were assessed both on and off SSRIs. On the SSRIs there was increased verbal aggression (P = 0.04), increased physical aggression toward objects (P = 0.05), and increased physical aggression toward self (P<0.02). No increase was observed in physical aggression toward others. The authors warned against using SSRIs to treat aggression in children. Another study of children and youth age 8-16 in a university setting found that 50% developed two or more abnormal behavioral reactions to fluoxetine, including aggression, loss of impulse control, agita- tion, and manic-like symptoms [68]. The effects lasted until the fluvoxamine was stopped. A second research study from the same university setting described a number of youngsters (6 of 42 or 14% in their cohort) who became aggressive and even violent while taking fluoxetine [51]. The researchers hypothesized that fluoxetine caused aggressive behavior by means of drug-induced activation (stimulation) or a specific serotonergic-mediated effect. The report [51] provided a clinical window into the development of obsessive violence and a school- shooter mentality. A twelve-year-old boy on fluoxetine developed nightmares about becoming a school shooter and then began to lose track of reality concerning these events. This case occurred in a controlled- clinical trial and the investigators did not know that the child was getting fluoxetine until they broke the 42 P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) double-blind code. The child’s reaction occurred long before any of the well-known school shootings had taken place. Therefore, his reaction was not inspired by the school shootings; it was not a “copycat”: Thirty-eight days after beginning the protocol, F. experienced a violent nightmare about killing his classmates until he himself was shot. He awakened from it only with difficulty, and the dream continued to feel “very real.” He reported having had several days of increasingly vivid “bad dreams” before this episode; these included images of killing himself and his parents dying. When he was seen later that day he was agitated and anxious, refused to go to school, and reported marked suicidal ideation that made him feel unsafe at home as well (p. 180). The child was hospitalized first for three days and then for 17 days. He gradually improved. Then three weeks after his last hospitalization, his local physician – not one of the clinical investigators – put him back on fluoxetine. The child became acutely suicidal until the fluoxetine was stopped a second time. This individual report is important for a variety of reasons: (1) It took place in a double-blind controlled clinical trial. (2) Entirely new symptoms related to violence developed on the drug (This stage is called challenge). (3) The symptoms terminated after stopping the drug (called dechallenge). (4) Some of the symptoms resumed on starting the drug again (called rechallenge). (5) The symptoms cleared for a second time after the drug was again stopped (demonstrating dechal- lenge for a second time).

7. Antidepressant-induced mania described in two standard sources

In a variety of forensic activities including criminal and civil cases, the courts sometimes rely on “au- thoritative” or “standard” texts in order to demonstrate that the opinions rendered are generally accepted by a significant portion of the medical or scientific community.

7.1. The Diagnostic and Statistical Manual of Mental Disorders (1994, 2000)

The American Psychiatric Association [2] Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) and the Fourth Edition Text Revision (DSM-IV-TR, [3]) are written by com- mittees made up of professionals considered expert by many of their colleagues in their respective fields. The conclusions therefore provide a professional consensus or body of conventional wisdom in psychi- atry that can at times be useful in clinical practice and in forensics. Many aspects of the DSM-IV are controversial. However, when such an essentially conservative consensus document provides evidence for SSRI-induced adverse reactions related to mania, suicide and violence, it should alert clinicians to the existence of these clinical phenomena and can provide an avenue for communicating in the courtroom concerning these risks. The DSM-IV was published in 1994, several years after the advent of SSRI antidepressants and makes clear that all antidepressants can cause mania. The first SSRI, fluoxetine, was approved by the FDA in December 1987 and was in widespread use when the following observations about antidepressants were published in the manual. DSM-IV makes multiple references to the fact that antidepressants can cause mania or manic-like be- havior. It states, for example, “Symptoms like those seen in a Manic Episode may be due to the direct effects of antidepressant medication ...” [2, p. 329]. Similarly, it observes, “Symptoms like those seen in a Manic Episode may also be precipitated by antidepressant treatment such as medication ...” [2, p. 331]. P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) 43

References to antidepressant-induced mania and mood disorder can also be found elsewhere in the man- ual as well (e.g., pp. 332 [note at bottom of table], 334, 336, 337, 351, 371 and 372). DSM-IV-TR (2000) emphasizes that a diagnosis of mania or bipolar disorder should not be made when the hypomania or mania first appears while the individual is taking a medication that can cause these symptoms and “usu- ally disappear when the individual is no longer exposed to the substance.” Of great clinical importance, it adds, “but resolution of symptoms can take weeks or months and may require treatment” (p. 191). The association between mania and antisocial behavior, including violence, is underscored in the DSM- IV. Aggression is specifically mentioned as a feature of manic behavior. It is noted that “antisocial be- haviors may accompany the Manic Episode,” “Ethical concerns may be disregarded even by those who are typically very conscientious,” “The person may become hostile and physically threatening to others” and “physically assaultive,” and “The mood may shift rapidly to anger or depression” (p. 330). The very next page in the DSM-IV, repeats the reminder that “Symptoms like those seen in a Manic Episode may also be precipitated by antidepressant treatment such as medication...” (p. 331). Mania is characterized by “increased involvement in goal-directed activities” (American Psychiatric Association [2, p. 328]). Therefore, the individual is able to plan and carry out inappropriate or destruc- tive aggressive actions, or to attempt to cover them up once they have been enacted. Individuals under- going mania often feel uncontrollably driven to carry out elaborate plans, however bizarre, destructive, or doomed they may be. According to the DSM-IV, an “elevated, euphoric or irritable mood” is sufficient to qualify for a di- agnosis of Substance-Induced Mood Disorder with Manic Features ([2, pp. 370 and 375]; DSM-IV-TR, 2000, [3, pp. 405–406]). This descriptor for manic features is sufficiently broad to encompass some or all symptoms associated with stimulation and aggression. Therefore, an SSRI-induced stimulant-like or aggressive reaction can often be diagnosed as an SSRI-Induced Mood Disorder with Manic Features. When drug-induced mood swings occur from mania to depression, sometimes accompanied by switches from violence to suicidality, the diagnosis can include both depressive and manic features. Irritability as used in the DSM-IV has a more ominous meaning than irritability as it is used in ordinary language. During a discussion of depression, the DSM-IV refers to the symptom of “increased irritability (e.g., persistent anger, a tendency to respond to events with angry outbursts or blaming others, or an exaggerated sense of frustration over minor matters)” (p. 321). Many individuals who commit aggression while under the influence of SSRIs will qualify for a Substance-Induced Mood Disorder with Manic Features on the basis of their obvious increase in irritability while taking the drug. The capacity for SSRIs to induce akathisia – and for akathisia to cause suicidality, aggression, and a worsening mental condition – are also recognized in the DSM-IV [2] and the DSM-IV-TR [3] in the section dealing with neuroleptic-induced akathisia. DSM-IV-TR observes, “Akathisia may be associated with dysphoria, irritability, aggression, or suicide attempts.” It also mentions “worsening of psychotic symptoms or behavioral dyscontrol.” It then states, “Serotonin-specific reuptake inhibitor antidepressant medications may produce akathisia that appears identical in phenomenology and treatment response to Neuroleptic-Induced Acute Akathisia” (p. 801).

7.2. Practice guidelines for major depressive disorder in adults (1993)

The American Psychiatric Association [1] practice guideline, like the DSM-IV, attempts to arrive at a consensus among experts. The emphasis, however, is on treatment rather than diagnosis. Like the DSM- IV, the practice guideline was published after the SSRIs were in use. Using several citations from the literature, the practice guideline states: 44 P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs)

All antidepressant treatments, including ECT, may provoke manic or hypomanic episodes. Individuals with a history of mania or hypomania are at particular risk for this untoward effect, although it may occur even in patients with no such history; this complication is estimated to occur in 5–20% of depressed patients treated with antidepressants (p. 22). Recognition of antidepressant-induced manic-like reactions and akathisia in the most commonly used manual of psychiatric diagnosis has important implications for clinical practice and forensics. Practition- ers should be aware that these adverse drug reactions occur and that the patient should be diagnosed with a Substance-Induced Disorder or with akathisia rather than with a primary psychiatric disorder, such as Bipolar I Disorder or an anxiety disorder. It should alert practitioners to the need to stop antidepres- sants at the first sign of initial or recurring hypomanic and manic symptoms, or akathisia. In forensics, recognition of the existence of these adverse drug reactions can help to establish causality in malpractice, product liability, and criminal cases when SSRIs induce abnormal mental and behavioral reactions. The body of literature reviewed in this report and the confirmation found in the DSM-IV and DSM-IV-TR help to establish a standard requiring that physicians be aware of the potential for these drugs to cause mania and akathisia with the associated risks of suicidality, violence, and extreme or bizarre behavior.

8. My clinical and forensic experience with similar cases

I have been a medical expert in a number of suits in which children and adults have developed bizarre, irrational, and violent behavior while taking SSRI antidepressants. In one case in California, a man drowned himself and his two small children in a bathtub a few days after starting on paroxetine (see www.breggin.com for this and other legal cases). Also while taking paroxetine, a young adult in South Carolina committed a violent rape and a man in Pennsylvania drove his car into a policeman in order to obtain the officer’s gun in order to kill himself. In a fourth case involving paroxetine, in Vermont a 17- year old boy who had missed one or two doses of paroxetine bludgeoned a close friend for no apparent reason. In Florida a teenage girl taking fluoxetine fired a pistol pointblank at another younger but the gun fortunately failed to function. None of these individuals had any history of violence prior to taking SSRIs. When all of the SSRI antidepressants are included, I have direct clinical and forensic experience with dozens of cases of aggression in association with these drugs.

9. Discussion: “The Drug Made Me Do It”

There is a natural reluctance to attribute “bad behavior” or loss of ethical restraint (dyscontrol, loss of impulse control) to a psychoactive substance. Western philosophy, religion, and tradition tend to hold human beings responsible for their harmful behaviors and eschew “excusing” such behavior on the basis of “mental illness.” Indeed, the concept of mental illness has been subject to challenge by this author and many others. Nonetheless, the weight of considered evidence indicates that psychoactive substances can play a role in causing suicide, violence, and other forms of disinhibited criminal conduct. First, controlled clinical trials comparing any psychoactive drug to a placebo will typically produce evidence for a pattern of central nervous system adverse drug effects with mental symptoms that are specific for the drug and not for the placebo. For example, SSRI-antidepressants and amphetamine-like agents both tend to produce a continuum of central nervous system stimulation. This physical stimulation will be associated with mental manifestations that range from mild euphoria and irritability to depression and mania, and ultimately to increased rates of both aggression and suicidality. P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) 45

Second, patterns of reports made to the FDA spontaneous reporting system also make apparent that cer- tain drugs are associated with specific patterns of extreme mental and behavioral reactions (for additional examples and an analysis of methodology, see Breggin [10,11]). Even non-psychiatric medications have been implicated in causing depression and suicidality. Isotretinoin (Accutane), a medication used to treat severe acne, has been found to produce depression and suicidality as demonstrated in numerous clinical reports and in individual case studies. In some clinical cases, “depression subsided with discontinuation of the therapy and recurred with reinstitution of therapy” [65, p. 2872]. Third, many physical disorders also affect mental attitudes and behavior. Hyperthyroidism as well as overdoses of thyroid hormone can increase anxiety, irritability, and other emotions that the individual would not ordinarily experience and that can lead to behavioral abnormalities. There are, of course, many similar examples involving hormones such as testosterone and cortisone. More to the point, accidental brain injury to the frontal lobes and surgical lobotomy usually impair judgment, ethical restraint and self-reflection. The character of the individual is often viewed as “changed” and “worsened.” Fourth, as an expert in criminal and civil cases, I have studied the lives of many individuals who – under the influence of psychoactive drugs, such as SSRIs, NSRIs, and benzodiazepines – have committed acts of aggression that were wholly alien to their character and antithetical to their prior behavior. It is, of course, well-known that the illegal use of stimulant drugs, such as methamphetamine and cocaine, can be associated with paranoid reactions and violence. As Preda et al. [66] suggest, the SSRIs and hallucinogens such as lysergic acid diethylamide (LSD) may cause psychosis through similar effects on serotonin receptors. The example of involuntary intoxication under the law helps elucidate the issue of responsibility while under the influence of psychoactive substances. Under the law, an individual is usually held responsible for behavior committed under the influence of alcohol or other non-prescription intoxicants because it is presumed that the individual knew that he was taking a psychoactive substance that can impair judgment and self-restraint. However, in most states an individual can claim involuntary intoxication as a mitigating or exonerating factor in a criminal case. For example, if the individual unknowingly drank alcohol from “spiked” punch, the involuntary nature of the intoxication might become a mitigating or exonerating factor under the law. Similarly, when an individual takes an antidepressant without knowing that it can cause mania, he or she may be exonerated from the consequences of manic-like behavior. If an individual involuntarily intoxicates another person, the perpetrator may be guilty of a crime and the victim may be absolved of any contributory responsibility. For example, a man can be judged guilty of rape if he has impaired the consciousness and self-restraint of his victim by surreptitiously slipping a sedative into her water glass. The victim, even if physically conscious during the sexual act, may be exonerated of seeming acquiescence to the assault on the basis of the involuntary intoxication. The debate over human responsibility will always remain at root ethical and philosophical, as well as a legal. However, empirical data must be taken into account. A mountain of experimental and clinical data, some of it reviewed in this report, supports the concept that psychoactive substances are frequently associated with an increased rate of disturbed mental and behavior reactions, causing some individuals to act as if they have lost their customary ethical restraint and self-control. It may be argued that some individuals will not lose ethical restraint regardless of the nature or intensity of an involuntary intoxication. However, even if some individuals are immune to behaving badly under the influence of drugs, while others seem especially susceptible, this merely reflects human variation, a factor that complicates most research in medicine and behavioral science. The reality of human variation does not undermine the validity of the association between certain drugs and the relatively frequent production of certain kinds of dangerous mental states and behaviors. 46 P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs)

Drug-induced disturbances in mood or in behavior should be viewed as genuine neurological disorders rather than as vague “mental illnesses.” The capacity of speculative “biochemical imbalances” or “genetic factors” to cause or contribute to mania or depression remains unproven. Nor do we know the specific biochemical or neurological mechanisms whereby psychoactive substances cause mental disturbances. But the capacity for psychoactive substances to disrupt brain function and hence mental function is beyond dispute. Furthermore, a great deal of empirical data confirms their capacity to cause disinhibition, mania, depression and other mental phenomena associated with violence toward oneself and others, and other destructive behaviors.

10. Conclusions

There are many reports and studies confirming that SSRI antidepressants can cause violence, suicide, mania and other forms of psychotic and bizarre behavior. Overall, the SSRIs produce violence, suicide and extremes of abnormal behavior by a variety of mechanisms. Teicher et al. [72] suggest nine pos- sible mechanisms: (1) energizing the depressed and suicidal patient, (2) paradoxically worsening the individual’s depression, (3) causing akathisia, (4) causing panic and anxiety, (5) causing manic or mixed manic-depressive states, (6) causing insomnia or disturbances in the sleep architecture, (7) causing ob- sessive suicidal preoccupations, (8) causing borderline states with hostility, and (9) causing alterations in EEG activity. Teicher et al. document each of these phenomena in their review of the literature and, as this paper indicates, the scientific evidence has grown considerably stronger in the intervening decade. With the exception of the alteration in EEG activity, my clinical and forensic work has confirmed that each of above SSRI- and NSRI-induced phenomena can cause violent and suicidal behavior. However, my clinical and forensic experiences and reviews of the literature indicate that four syndromes encompass most of the phenomena and describe most of the individual cases: (1) The production of a stimulant continuum that often begins with lesser degrees of insomnia, ner- vousness, anxiety, hyperactivity and irritability and then progresses toward more severe agitation, aggression, and varying degrees of mania. Mania or manic-like symptoms include disinhibition, grandiosity, sleep disturbances, and out-of-control aggressive behavior, including cycling into de- pression and suicidality. (2) The production of a combined state of stimulation and depression –anagitated depression – with a high risk of suicide and violence. Often the overall depression is markedly worsened. (3) The production of obsessive preoccupations with aggression against self or others, often accom- panied by a worsening of any pre-existing depression. (4) The production of akathisia, an inner agitation or jitteriness that is usually (but not always) ac- companied by an inability to stop moving. It is sometimes described as psychomotor agitation or restless leg syndrome. The state causes heightened irritability and frustration with aggression against self or others, and often a generally worsening of the mental condition. The above syndromes often appear in combination with each other. Often the syndromes will abate within days after stopping the SSRI but sometimes they persist, leading to hospitalization and additional treatment over subsequent weeks or months. Reported rates for these syndromes very widely but each of them appears to be relatively common. They frequently occur in individuals with no prior history of violence, suicidality, psychomotor agitation, or manic-like symptoms. P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) 47

References

[1] American Psychiatric Association, Practice Guidelines for Major Depressive Disorder in Adults, American Psychiatric Association, Washington, DC, 1993. Also published in the American Journal of Psychiatry (April) (1993). [2] American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, Fourth edition (DSM-IV), American Psychiatric Association, Washington, DC, 1994. [3] American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, Fourth edition, Text Revision (DSM-IV-TR), American Psychiatric Association, Washington, DC, 2000. [4] J. Bastani, M. Troester and A. Bastani, Serotonin syndrome and fluvoxamine: A case study, Nebraska Medical Journal 81 (1996), 107–109. [5] C. Beasley, Activation and sedation in fluoxetine clinical studies. Unpublished in-house document generated by Eli Lilly and Company during the FDA-approval process of Prozac for depression and obtained during discovery for Fentress v. Shay Communications et al., Fentress Trial Exhibit 70, 1988. [6] S. Borg and K. Brodin, Antidepressant drugs, in: Meyler’s Side Effects of Drugs, Thirteenth edition, M.N.G. Dukes, ed., Elsevier, New York, 1996, pp. 31–80. [7] R. Bost and P. Kemp, A possible association between fluoxetine use and suicide, Journal of Analytic Toxicology 16 (1992), 142–145. [8] P. Breggin, A case of fluoxetine-induced stimulant side effects with suicidal ideation associated with a possible withdrawal reaction (“crashing”), International Journal of Risk & Safety in Medicine 3 (1992), 325–328. [9] P. Breggin, Testimony in Joyce Fentress et al. vs. Shea Communications et al. [The Wesbecker Case]. Jefferson Circuit Court, Division, 1, Louisville, Kentucky, No. 90-CI-06033, Volume XVI, 1994. [10] P. Breggin, Brain-Disabling Treatments in Psychiatry: Drugs, Electroshock, and the Role of the FDA, Springer, New York, 1997. [11] P. Breggin, Analysis of adverse behavioral effects of benzodiazepines with a discussion of drawing scientific conclusions from the FDA’s Spontaneous Reporting System, Journal of Mind and Behavior 19 (1998), 21–50. [12] P. Breggin, The Antidepressant Fact Book, Perseus Books, Cambridge, MA, 2001. [13] P. Breggin, Fluvoxamine as a cause of stimulation, mania, and aggression with a critical analysis of the FDA-approved label, International Journal of Risk & Safety in Medicine 14 (2002), 71–86. [14] P. Breggin and G. Breggin, Talking Back to Prozac: What Doctors Aren’t Telling You About Today’s Most Controversial Drug, St. Martin’s Press, New York, 1994. [15] P. Breggin and D. Cohen, Your Drug May be Your Problem: How and Why to Stop Taking Psychiatric Medications,Perseus Books, Cambridge, MA, 1999. [16] British Medical Association and Royal Pharmaceutical Society of Great Britain, British National Formulary (BNF),the Pharmaceutical Press, London, England, 1991. [17] C. Burrai, A. Bocchetta and M. Zompo, Mania and fluvoxamine, American Journal of Psychiatry 148 (1991), 1263–1264. [18] R. Christen, Paroxetine-induced mania, American Journal of Psychiatry 152 (1995), 1439–1500. [19] J. Constantino, M. Liberman and M. Kincaid, Effects of serotonin reuptake inhibitors on aggressive behavior in psychiatric hospitalized adolescents: Results of an open trial, Journal of Child and Adolescent Psychopharmacology 7 (1997), 31–44. [20] W. Creaney, I. Murray and D. Healy, Antidepressant induced suicidal ideation, Human Psychopharmacology 6 (1991), 329–332. [21] K. Dasgupta, Additional cases of suicidal ideation associated with fluoxetine, American Journal of Psychiatry 147 (1990), 1570. [22] S. Donovan, A. Clayton, M. Beeharry S. Jones, C. Kirk, K. Waters, D. Gardner, J. Faulding and R. Madeley, Deliberate self-harm and antidepressant drugs: Investigation of a possible link, British Journal of Psychiatry 177 (2000), 551–556. [23] S. Donovan, M. Kelleher, J. Lambourn and T. Foster, The occurrence of suicide following the prescription of antidepres- sant drugs, Archives of Suicide Research 5 (1999), 181–192. [24] A. Dorevitch, Y. Frankel, A. Bar-Halperin, R. Aronzon and Zilberman, Fluvoxamine-associated manic behavior: A case series, Annals of Pharmacotherapy 27 (1993), 1455–1457. [25] Drug Facts and Comparisons, Facts and Comparisons, St. Louis, 2003. [26] D. Ebert, R. Albert, A. May, A. Merz, H. Murata, I. Stosiek and B. Zahner, The serotonin syndrome and psychosis-like side effects of fluvoxamine in clinical use – An estimation of incidence, European Neuro-Pharmacology 7 (1997), 71–74. [27] G.J. Emslie, A.J. Rush, W.A. Weinberg, R.A. Kowatch, C.W. Hughes, T. Carmody and J. Rintelmann, A double-blind, randomized, placebo-controlled trial of fluoxetine in children and adolescents with depression, Archives of General Psy- chiatry 54 (1997), 1031–1037. [28] Fentress et al. vs. Shea Communications et al. [The Wesbecker Case]. Jefferson Circuit Court, Division, 1, Louisville, Kentucky, No. 90-CI-06033, Volume XVI, 1994. [29] S. Fisher, T. Kent and S. Bryant, Postmarketing surveillance by patient self-monitoring: Preliminary data for sertraline versus fluoxetine, Journal of Clinical Psychiatry 56 (1995), 288–296. 48 P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs)

[30] S. Fisher, S. Bryant and T. Kent, Postmarketing surveillance by patient self-monitoring: Trazodone versus fluoxetine, Journal of Clinical Psychopharmacology 13 (1993), 235–242. [31] Food and Drug Administration (FDA), Transcript of Psychopharmacologic Drugs Advisory Committee: 34th Meeting, Department of Health and Human Services, Public Health Service, Food and Drug Administration, Rockville, MD, Ob- tained through Freedom of Information Act, 1991. [32] Food and Drug Administration (FDA), Adverse Event Reporting System (AERS): Freedom of Information (FOI) Report, Center for Drug Evaluation and Research, Food and Drug Administration, Rockville, MD, Obtained through Freedom of Information Act, 1999. [33] D. Frankenfield, S. Baker, W. Lange, Y. Caplan and J. Smialek, Fluoxetine and violent death in Maryland, Forensic Science International 64 (1994), 107–117. [34] O. Fumihiko and K. Okajima, Violent acts associated with fluvoxamine treatment, Journal of Psychiatry and Neuroscience 26 (2001), 339–349. [35] J. Glenmullen, Prozac Backlash, Simon & Schuster, New York, 2000. [36] J. Gorman, M. Liebowitz, A. Fyer, D. Goetz, R. Campeas, M. Fyer, S. Davies and D. Klein, An open trial of fluoxetine in the treatment of panic disorder, Journal of Clinical Psychopharmacology 7 (1987), 329–332. [37] S. Grimsley and M. Jann, Paroxetine, sertraline, and fluvoxamine: New selective serotonin reuptake inhibitors, Clinical Pharmacy 11 (1992), 930–957. [38] C. Gualtieri, Paradoxical effects of fluoxetine, Journal of Clinical Psychopharmacology 11 (1991), 393–394. [39] D. Healy, Lines of evidence on the risks of suicide with selective serotonin reuptake inhibitors, Psychotherapy and Psy- chosomatics 72 (2003), 71–79. [40] D. Healy, Emergency of antidepressant induced suicidality, Primary Care Psychiatry 2000 6 (2000), 23–28. [41] D. Healy, The fluoxetine and suicide controversy: A review of the evidence, CNS Drugs 1 (1994), 223–231. [42] R. Hoehn-Saric, J. Lipsey and D. McLeod, Apathy and indifference in patients on fluvoxamine and fluoxetine, Journal of Clinical Psychopharmacology 10 (1990), 343–345. [43] C. Hoover, Additional cases of suicidal ideation associated with fluoxetine, American Journal of Psychiatry 147 (1990), 1569–1570. [44] R. Howland, Induction of mania with serotonin reuptake inhibitors, Journal of Clinical Psychopharmacology 16 (1996), 425–427. [45] J. Jain, B. Birmaher, M. Garcia, M. Al-Shabbout and N. Ryan, Fluoxetine in children and adolescents with mood disorders: A chart review of efficacy and adverse reactions, Journal of Child and Adolescent Psychopharmacology 2 (1992), 259– 265. [46] S. Jick, A. Dean and H. Jick, Antidepressants and suicide, British Medical Journal 310 (1995), 215–218. [47] R. Kapit, Safety Review of NDA 18-936 [Prozac]. Internal document of the Department of Health and Human Services, Public Health Service, Food and Drug Administration, Center for Drug Evaluation and Research. Obtained through the Freedom of Information Act, 1986. [48] R. Kapit, Response to Dr. Laughren’s Q’s regarding Review of Safety Update, Memo #3.” NDA 18-936. Internal document of the Department of Health and Human Services, Public Health Service, Food and Drug Administration, Center for Drug Evaluation and Research. Obtained through the Freedom of Information Act, 1986. [49] A. Khan, S. Khan, R. Leventhal and W. Brown, Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: a replication analysis of the Food and Drug Administration database, International Journal of Neuropsychopharmacology 4 (2001), 113–118. [50] A. Khan, H. Warner and W. Brown, Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: An analysis of the Food and Drug Administration database, Archives of General Psychiatry 57 (2000), 311–317. [51] R. King, M. Riddle, P. Chappell, M. Hardin, G. Anderson, P. Lombroso and L. Scahill, Emergence of self-destructive phenomena in children and adolescents during fluoxetine treatment, Journal of the American Academy of Child and Adolescent Psychiatry 30 (1991), 179–186. [52] J. Koizumi, Fluoxetine and suicidal ideation, Journal of the American Academy of Child and Adolescent Psychiatry 30 (1991), 695. [53] The Lancet, Editorials: 5-HT blockers and all that, 336 (1990), 345–346. [54] R. Lane, SSRI-induced extrapyramidal side-effects and akathisia: Implications for treatment, Journal of Psychopharma- cology 12 (1998), 192–214. [55] B. Leonard, Pharmacological differences of serotonin reuptake inhibitors and possible clinical relevance, Drugs 43(Supplement 2) (1992), 3–10. [56] L. Levine, S. Rosenblatt and J. Bosomworth, Use of a serotonin re-uptake inhibitor, fluoxetine, in the treatment of obesity, International Journal of Obesity 11(supplement 3) (1987), 185–190. [57] J. Lipinski, Jr., G. Mallaya, P. Zimmerman and H. Pope, Jr., Fluoxetine-induced akathisia: clinical and theoretical impli- cations, Journal of Clinical Psychiatry 50 (1989), 339–352. P.R. Breggin / Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs) 49

[58] L. Marangell, S. Yudofsky and J. Silver, Psychopharmacology and electroconvulsive therapy, in: The American Psychi- atric Press Textbook of Psychiatry, R. Hales, S. Yudofsky and J. Talbott, eds, Chapter 27, American Psychiatric Press, Washington, DC, 1999, pp. 1025–1132. [59] P. Masand, S. Gupta and M. Dewan, Suicidal ideation related to fluoxetine treatment, New England Journal of Medicine 324 (1991), 420. [60] M. Muijen, D. Roy, T. Silverstone, A. Mehmet and M. Thristie, A comparative clinical trial of fluoxetine, mianserin and placebo in depressed outpatients, Acta Psychiatrica Scandinavica 78 (1988), 384–390. [61] F. Okada and K. Okajima, Violent acts associated with fluvoxamine treatment, Journal of Psychiatry & Neuroscience 26 (2001), 339–340. [62] J. Oldroyd, Paroxetine-induced mania, Journal of the American Academy of Child and Adolescent Psychiatry 36 (1997), 721–722. [63] R. Peyre, H. Verdous and M. Bourgeois, Fluvoxamine: Study of treatment effect on a group of 189 hospitalized patients with depression, (French), Encephale 18(1) (1992), 73–74 (in French). [64] Physicians’ Desk Reference, Medical Economics, Montvale, NJ, 2001. [65] Physicians’ Desk Reference, Medical Economics, Montvale, NJ, 2003. [66] A. Preda, R. MacLean, C. Mazure and M. Bowers, Antidepressant-associated mania and psychosis resulting in psychiatric admission, Journal of Clinical Psychiatry 62 (2001), 30–33. [67] Public Citizen Health Research Group, Re: Citizens Petition for revision of fluoxetine (Prozac) labeling. Letter to David Kessler, Commissioner of the Food and Drug Administration, Public Citizen, Washington, DC, 1991. [68] M. Riddle, R. King, M. Hardin, L. Scahill, S. Ort, P. Chappell, A. Rasmusson and J. Leckman, Behavioral side effects of fluoxetine in children and adolescents, Journal of Child and Adolescent Psychopharmacology 1 (1990/1991), 193–198. [69] A. Rothschild and C. Locke, Reexposure to fluoxetine after serious suicide attempts by three patients: The role of akathisia, Journal of Clinical Psychiatry 52 (1991), 491–493. [70] F. Sim, A single dose of fluvoxamine associated with an acute psychotic reaction, Canadian Journal of Psychiatry 45 (2000), 762. [71] M. Teicher, C. Glod and J. Cole, Emergence of intense suicidal preoccupations during fluoxetine treatment, American Journal of Psychiatry 147 (1990), 207–210. [72] M. Teicher, C. Glod and J. Cole, Antidepressant drugs and the emergence of suicidal tendencies, Drug Safety 8 (1993), 182–212. [73] A. Troisi, E. Vicario, F. Nuccetelli, N. Ciani and A. Pasini, Effects of fluoxetine on aggressive behavior in adult inpatients with mental retardation and epilepsy, Pharmacopsychiatry 28 (1995), 73–76. [74] T. Van Putten, Why do schizophrenic patients refuse to take their drugs? Archives of General Psychiatry 31 (1974), 67–72. [75] T. Van Putten, The many faces of akathisia, Comprehensive Psychiatry, International Journal of Neuropsychopharmacol- ogy 2 (1975), 165–172. [76] D. Wilkinson, Loss of anxiety and increased aggression in a 15-year-old boy taking fluoxetine, Journal of Psychopharma- cology 13 (1999), 420. [77] W. Wirshing, T. Van Putten, J. Rosenberg, S. Marder, D. Ames and T. Hicks-Gray, Fluoxetine, akathisia, and suicidality: Is there a connection? Archives of General Psychiatry 49 (1992), 580–581. [78] Food and Drug Administration (FDA), FDA statement regarding the anti-depressant Paxil for pediatric population (June 19, 2003). www.fda.gov/eder/drug/infopage/paxil/default.htm. [79] Wyeth Pharmaceuticals, Dear Health Care Professional, August 22, 2003. CURRICULUM VITAE Dr Yolande Lucire PhD MBBS DPM (FRANZCP-resigned) FORENSIC & MEDICO-LEGAL PSYCHIATRY Website www.lucire.com.au Most papers are available in full on the website.

Senior Research Associate: Department of Government and International Relations, University of Sydney. Conjoint Senior Lecturer in Rural Health, 2003-2005. Qualifications 1996 PhD UNSW in Arts and Social Sciences in History and Philosophy of Science 1970 Member then Fellow, Royal Australian & New Zealand College of Psychiatry 1967 Diploma of Psychological Medicine, London, 1964 MB BS University of Sydney Work: 2009 to 2011 Medico-legal psychiatry related to workers’ compensation claims and research into genetic links to adverse reactions to psychiatric drugs 1972 to 2008 Private practice, originally child and family, then general, forensic and medico-legal psychiatry in Sydney; medico-legal work relating to psychiatric defences in criminal proceedings and workers’ compensation/personal injury claims; research resulting in the papers listed below 2001-2006 Conjoint Senior Lecturer, Psychiatry, Rural Medical School. 1997-2005 Consultant Psychiatrist, Nolan House, Albury, 1994/5 Consultant Forensic Psychiatrist, London, Devonshire Place Locum Consultant Psychiatrist: East Ham Memorial Hospital Fellow: Wellcome Institute for the History of Medicine. (Competing PhD) 1983 -1995 Senior Forensic Psychiatrist, Consultant to Department of Corrective Services NSW, and Long Bay Prison Hospital. 1972 - 1980 Consultant (VMO) South Sydney Hospital, (including Rehabilitation) Psychiatrist, Rozelle Hospital, Sydney 1972- to present Consultant Psychiatrist in private practice, Sydney 1967-1972 General Practice 1968 - 1970 Senior Registrar in Child Psychiatry, Royal Alexandria Hospital for Children 1967 Registrar, Sutton & Belmont Hospital, Surrey, UK. 1965 SRMO, Netherne Hospital, Surrey, UK. 1964. RMO, Prince Henry Hospital, d

Book Constructing RSI; Belief and Desire UNSW Press 2003. Publications in medical and forensic journals 1) Factors Affecting Conception in Women Seeking Termination of Pregnancy. Medical Journal of Australia 1975 (pages 824-27). 2) I Fear the Greeks. Legal Service Bulletin, Feb. 1981. A medico-political expose of Social Security Conspiracy (prepared originally for a conference of psychiatry and law but deemed sub-judice at the time). Submission to the Minister and Commission of Inquiry into Social Security Prosecutions. 3) The Medical Evidence in the First 50 Administrative Appeals Tribunal Decisions. Legal Service Bulletin, Dec. 1982. (Australian) Analysis of the difficulties of evaluation of Invalid Pension applicants. 4) Neurosis in the Workplace. 1986 Medical Journal of Australia 145: 323-7. This paper has been given many citations both in medical and in social science journals. 5) PhD Ideology and Aetiology: RSI: an epidemic of craft palsy 6) Medea - Anatomy of a Multicide. Journal of the Australian Academy of Forensic Sciences. December 1993. 7) Social Iatrogenesis of Epidemic Neurosis. (RSI) Journal of Community Health Studies XX (2) 1988. 8) Analysis of the Function of the Expert, in "The Expert Witness Self- Examined" in book, The Expert Medical Witness, Federation Press 1989. 9) Sex and the practitioner: the Victim Presented at the Plenary Session of the Australian Academy of Forensic Sciences September 11, 2001. Australian Journal of Forensic Sciences. Vol. 34 no.1 18-22. . 10) Comparison Codes Medical Practice Act and Common Law. Whither 200 years of due process? Australian Journal of Forensic Sciences. Vol 34 No.1 22-24 2002 11) The Bearing of Daubert on Sexual Abuse Allegations. Journal of the Australian Academy of Forensic Sciences. Vol 32 no. 245-59. 2001 12) By invitation: Editorial: From personalized medicine to personalized justice: the promises of translational pharmacogenomics in the justice system Wong SH, Happy C, Blinka D, Gock S, Jentzen JM, Donald Hon J, Coleman H, Jortani SA, Lucire Yolande, Morris-Kukoski CL, Neuman MG, Orsulak PJ, Sander T, Wagner MA, Wynn JR, Wu AH, Yeo KT. Pharmacogenomics. 2010 Jun;11(6):731-7

Conference papers, Submissions other Publications

13) The Adversary System or a Better Way? Read at the RANZCP Conference 1983. Prepared as a submission on the Invalid Pension problem for Senator Grimes in 1982. 14) When emotions get converted. On the genesis of RSI as Conversion Disorder Safety Australia, Feb. 1986. Read at Medical Mythology conference, Nov. 1985. 15) Institutionalised & Rewarded Neurosis: RSI, the Australian Disease. Australian Institute of Management Journal, Apr. 1986. 16) Differential Diagnosis of Conversion. Read at RANZCP Annual Conference, May 1986. 17) Resistance to paradigm shift, The Injury Theory versus the Psychosocial Model of Causation in Epidemic RSI. Read at RANZCP Annual Conference, May 1986. Analysis of sources of resistance to the psychosocial model.

2

18) The Use of Proforma for Disability Evaluation. Unpublished but widely read. Presented RANZCP conference, Hobart, May 1985. 19) The First Forensic Interview, "RSI" - the Use of a Pre-Printed Proforma. Presented November 1985 and available in video from the Institute of Psychiatry, Rozelle Hospital. Also available in print. 20) Square Pegs in Round Holes: A Comparison of Medical & Legal Concepts of "Causation" in Epidemic Neurosis, using the epidemic of RSI Proceedings of Conference of the Medico-legal Society of Victoria. Kotakinabalu 1986. 21) Workers' Compensation: A New Approach: Submission to writers of white paper on workers' compensation in NSW (1986). (Unpublished). 22) Theory and Philosophy of Assessment: An analysis of the sources of variance in expert opinion evidence. Forensic Psychiatry Bulletin 1986. 23) RSI, an Epidemic of Craft Palsy. A chapter commissioned by Dr. (now professor) Professor Ivor Jones, then Snr. lecturer in Psychiatry, Melbourne University, for text book, "Essentials of Australian Forensic Psychiatry", 1986. (This book was never published) 24) The Role of the Psychiatric Assessor in Personal Injury Claims. Presented at RANZCP Forensic Psychiatry Conference Leura, November 1990. 25) Repetitive Strain Injury - An Epidemic of Craft Palsy. Proceedings of the Medico-Legal Society of NSW. Vol. 8, pages 134-146. 26) Chronic Fatigue Syndrome: What is a disease? Debate with the Prince of Wales Hospital, Presented in November 1990 at the Institute of Psychiatry in NSW for Continuing Medical Education. 27) The NSW Mental Health Review Tribunal, first seven years of operations. Presented RANZCP Forensic Section Conference Nov 15-20, 1991. 28) Life events and getting sick with "RSI." Presented RANZCP Forensic Section Conference Nov 15-20, 1991. 29) The Narcissist in the Culture of Compensation. Presented 1992 RANZCP Conference Brisbane. 30) The five-colour theorem: A model to elucidate the components of illness, disease and morbidity. Presented at Philosophy and Psychiatry Conference 1996. 31) Can a Linear Medical Model Identify Causation of Illness? (Draft only) 32) How to do a Sex Abuse Evaluation. RANZCP Conference Forensic Section, June 2000, Port Douglas 33) The Politicization of Medicine and the Medicalization of Industrial Relations. Presented at Garran and Baxter conference on Psychological Injury. 34) Comparative Analysis of Paradigmatic Assumptions of the True Believers and The Sceptics contributing to Moral Panic about Child Sexual Abuse. Published on website at www.lucire.com.au 35) Health Status and Predicament in Claimants for RSI 1986-1992. RANZCP Forensic Section Conference, 2001. 36) The Social Construction of the War Neuroses: Are We Being Served? Commissioned paper for 11th Brigade Senior Medical Officers Conference, 14 July, 2001 Townsville Presented again RANZCP Forensic Section Conference, 2001. Published in BMJ online. http://www.bmj.com/cgi/eletters/322/7278/95

3

37) Politicising Medicine and Medicalizing Industrial Relations (repeated) RANZCP Forensic Section Conference, 2001 38) Towards A Taxonomy of Confabulation RANZCP Conference Brisbane June 2001. 39) Submission To Proceedings Before Standing Committee on Law and Justice Inquiry Into Child Sexual Assault Matters. At Sydney on Friday 10 May 2002 40) Confabulation: Forensic Issues. ANZAPPL Conference July 2002. Darwin 41) Submission to Ipp Inquiry: Review of the Law of Negligence: http://www.google.com.au/search?q=cache:fLSVEzzxGHQJ:www.pc.gov.au/inquiry/w orkerscomp/subs/sub102.rtf+negligence+Lucire&hl=en&start=4&ie=UTF-8 (no longer accessible) 42) SSRIs: Forensic Issues. Risk Benefit Analysis and Potential for Litigation In Australia. Duty To Warn?. (PowerPoint) Presented at RANZCP Forensic Section Conference October 2003 Geelong 43) The Use of Textual Analysis in Differentiating True from Fabricated Sex Abuse Allegations. (PowerPoint). Presented at RANZCP Forensic Section Conference October 2003 Geelong 44) Submission to Productivity Commission on adverse responses to antidepressants and the increase in demand for mental health services. http://www.pc.gov.au/__data/assets/pdf_file/0003/17814/subpr047.pdf. 45) Pharma Fraud, Pharmacological Iatrogenesis and the Crisis In Mental Health. Precendent: (the Law magazine) 2005. 46) SSRIs: Do they cause suicide? The Science: Daubert Admissible Evidence. Australian Academy of Forensic Sciences, May 19, 2004. Also presented at International conference of Medical Law, Sydney, 2004. 47) SSRIs and their effects on Mental Health Presentations: A plausible Hypothesis, (PowerPoint). Presented at RANZCP Forensic Section Conference October 2004, Fremantle. 48) New Drugs New Problems: PowerPoint, presented Section of Forensic Psychiatry, April 9 2005. 49) New Drugs New Problems Medico-political Expose of the Suicide Crisis in Mental Health, Australian Journal of Forensic Sciences. 50) The Ethics of the Solitary Empiricist: How Pharmas Changed Common Human Unhappiness into a Deficit Disease. Blackheath Philosophy Forum May 9 2005 51) Do Second Generation Antidepressants Cause Suicide? A Daubert Hearing. Health, Australian Journal of Forensic Sciences. May 19. 2004. 52) Effects of Second Generation Antidepressants and Antipsychotics on Mental Health Services in Australia. Royal Australian and New Zealand College of Psychiatrists 40th Conference, Convention Centre, Sydney 22 to 27 May. 2005. 53) Akathisia and Crime: Product Liability Issues. Royal Australian and New Zealand College of Psychiatrists 40th conference, Convention Centre, Sydney 22 to 27 May. 2005. 54) Submission to Commission of Inquiry: Acute Care Services in NSW Public Hospitals on the prevalence and costs of adverse drug reactions to the health service: focus on psychiatric drugs and Vioxx. 55) Re-focussing Upstream: New Psychiatric Drugs, Genetic Polymorphisms and Public Health. Prepared For Submission Special Commission of Inquiry: Acute Care

4

Services in NSW Public Hospitals. (Garling Inquiry. This was ignored on the basis that "the health department does not agree with her.” 56) The Effects of Pharmaceutical Industry Fraud, and the Texas Medication Algorithm Project on Mental Health Costs And Demand, RANZCP Conference Gold Coast 2006. 57) Submission To Commonwealth Minister of Health concerning the implementation of the Deloitte report commissioned by the Australian Centre for Health Research. Improving the Quality Use of Medicines in Australia; Realising the Potential of Pharmacogenomics. October 2008. http://www.tga.gov.au/alerts/medicines/pdseap- report2009.htm 58) Genetic Polymorphisms, Antidepressants, Akathisia Homicide and the Crisis in Mental Health: Prototype for a Project to Provide Adequate Defences. Poster at The Bureau of Crime Statistics and Research’s 40th Anniversary Symposium, 18 – 19 February 2009. 59) Submission to Complaints Handling Inquiry NSW Department of Health. http://www.scribd.com/doc/2220637/Review-Of-Inquiry-Into-Complaints-Handling-In- NSW-Health 60) CYP450 TESTING MAY BE ESSENTIAL IN PSYCHIATRY. Presented at ASCEPT conference 2009. During 2003-4 193 reports were made to ADRAC from one Health Area of persons who were akathisic and suicidal, homicidal, hallucinated and violent on new antidepressants and atypicals and required hospitalization. A clinically identical sample of 72 was tested for CYP450 genes and 66 had diminished CYP450 capacity (16 homozygous PMs, 16 compound heterozygous PMs, and 34 a huge increase in multiple mutations over non clinical populations). Psychiatric drugs have a narrow therapeutic window, are ineffective below and both toxic and ineffective above that blood level. Diminished Metabolizer status has the same clinical consequences as overmedication and as co-prescribing of inhibitors and competing substrates. The side effects seen are those in Product Information (PI). Testing the other 191 seems to be a matter of urgency, so they do not again develop these conditions, which seem more dangerous when they occur for the second or third time, unheeded. Acknowledgement of thanks to Healthscope Molecular and Diversity Health Institute Research Laboratory for testing. (There will be amendments as numbers have increased as 3 more having 2D6*2 are to be deemed normal metabolizers, resulting in slight alteration of figures and a further 3 homicides have come in, three more with mutations.) 61) Antidepressant-induced akathisia homicides in the absence of metabolizing alleles of the CYP450 family. In Press. Accepted by Journal of Pharmacogenomics and Personalized Medicine. 62) Submission to Senate Suicide inquiry 2010 on the Prevalence of Medication-Induced Suicide and its Relationship to Demand for Services, and Public Health. http://www.aph.gov.au/Senate/committee/clac_ctte/suicide/submissions/sub243.pdf In preparation: Looking for gene-drug response and finding multiple diminishing mutations and polypharmacy. a paper on genetic testing of a population (over 120 persons mainly from a workers compensation, personal injury and criminal offences referral stream) who had become akathisic, suicidal and homicidal on psychiatric drugs metabolized by CYP450: The tested population is clinically identical to that reviewed in the Report of the Psychiatric Drug Safety Expert Advisory Panel 24 December 2009. This investigation was conducted on 90 of

5

Dr. Lucire's 90 reported cases of akathisia in 2007. Associations: of Forensic Psychiatry 1988-1993. Committee Member Australian New Zealand Association of Psychology, Psychiatry and the Law, NSW Branch. 2001-2005. Council member, Australian Academy of Forensic Sciences, 2001- Fellow of the RANZCP, Member Forensic Section (Now resigned). Member: International Centre for the Study of Psychiatry and Psychology (ICSPP). Member: Healthy Skepticism (Countering false and misleading advertising by Pharma) http://www.healthyskepticism.org/global/about/us Member: Australian Society for Clinical and Experimental Pharmacologists and Toxicologists. ASCEPT. Invited Speaker Conference of the Medico-legal Society of Victoria. Kotakinabalu 1986. Square Pegs in Round Holes: A Comparison of Medical & Legal Concepts of "Causation" in Epidemic Neurosis, using the Epidemic of RSI as an example, Published in Proceedings Invited Speaker: Disease Mongering Conference, Newcastle, Australia, April 2006. Constructing RSI Iatrogenesis of an Epidemic. Invited Speaker: Royal College of Psychiatrists Annual Conference, Glasgow, Scotland, 8th to 10th July 2006 Constructing RSI Iatrogenesis of an Epidemic. Invited Speaker Conference, The Proliferation of Diseases which cannot be objectified. Fribourg, Switzerland, 14-15 September 2006. Constructing RSI: Iatrogenesis of an Epidemic,

Internet Communities membership BMJ alerts on Adverse Drug Reactions FDA alerts: warnings and advisories and changes to Product information weekly. Biojest (By invitation: community dedicated to exposing pharmaceutical industry fraud comprising around 200 multidisplinary professionals dedicated to getting drug companies to tell the truth and sharing information) SSRI research, Antipsychotics research Pharmalot http://www.pharmalot.com/news Psych Rights: http://psychrights.org/index.htm I receive about 60 informative emails each night as well as digests of new papers put together by Pharmaceutical Company interests. Google alerts and akathisia and pharmacogenetics. I access Medline and Web of Science and communicate with others with similar interests.

6

BRITISH JOURNAL OF PSYCHIATRY (2006),(2006),188,122^127 188, 122^127

Schizophrenia, neuroleptic medication examined 1–6 weeks later by the Mobile Clinic of the Social Insurance Institution and mortalityy (SII) in two phases: a screening phase and a diagnostic (clinical) phase. The measure- ments of the screening phase, the methods MATTI JOUKAMAA, MARKKU HELIOHELIOVAARA,«VA AR A, PAUL KNEKT, used for studying chronic diseases and the ARPO AROMAA, RAIMO RAITASALO and VILLE LEHTINEN basic results of the Mini-Finland Health Survey have been previously described in detail (Aromaa et al, 1989; MaMakela¨kela¨ et al, 1993). People with a disease history, symp- toms or findings suggestive of chronic dis- eases were asked to participate in the Background There is an excess of Excess mortality among people with schizo- diagnostic (clinical) phase, which was on death from natural causes among people phrenia is partly a result of suicide (Brown, average 3.5 months after the screening 1997; Harris & Barraclough, 1998; Heila¨ examination. Cardiovascular and respira- with schizophrenia. &Lo& Lonnqvist,¨ nnqvist, 2003) but is also attributable tory diseases, diabetes and other somatic Aims Schizophrenia and its treatment to natural deaths (Brown, 1997; Morten- conditions were diagnosed on the basis of sen, 2003), obviously explained by dietary medical history, symptoms, physical with neuroleptics were studied for their and lifestyle factors (Brown, 1997) and examination and findings of the screening prediction of mortalityinmortality in a representative possibly by use of neuroleptics (Montout phase (Aromaa et al, 1989; MaMakela¨kela¨ et al, population sample of 7217 Finns aged 530 et al, 2002; Ray & Meador, 2002). Most 1993). years. studies have included small and non- representative series of patients, short Screening procedure Method A comprehensive health follow-up and no control for confounding The methods used and the basic findings factors. In a representative population examination was carried out at baseline. concerning mental disorders in the Mini- sample over a long follow-up period, we Schizophrenia was determined using the Finland Health Survey have been described found schizophrenia to be associated with in detail elsewhere (Lehtinen et al, Present State Examination and previous excess mortality. In males this was owing 1990a,b). The screening for mental dis- medicalrecords. to cardiovascular diseases (especially orders comprised several parts. We tried coronary disease), suicides and respiratory to make the screen as sensitive as possible Results During a17-year follow-up, 39 diseases and in females to cerebrovascular in order to minimise the number of false- oftheof the 99 people with schizophrenia died. and respiratory diseases (Joukamaa et al, negative cases, as no screen-negative people 2001). In the present study schizophrenia Adjusted for age and gender, the relative were invited to participate in the clinical and its treatment with neuroleptic medi- mortalityriskbetweenthose with schizo- phase of the study. The most important cation were studied for their prediction of phrenia and others was 2.84 (95% CI part of the screen was the 36-item version mortality. We aimed to assess the effects of the General Health Questionnaire (Gold- 2.06^3.90),andwas2.06^3.90),andwas2.25(95%CI1.61^3.15) 2.25(95%CI1.61^3.15) of socio-demographic and lifestyle factors berg, 1972). People were also invited to after further adjusting for somatic and somatic health on the mortality participate in the clinical phase if the differentials. diseases,bloodpressure, bloodpressure,cholesterol,bodycholesterol,body records in the SII indicated that they were mass index, smoking, exercise, alcohol receiving a disability pension because of a mental disorder or were entitled to re- intakeandeducation.Thenumberofneuro- METHOD lepticslepticsused used atatthe the time ofofthe the baseline imbursement for medication for such dis- orders (according to the National Health survey showed a graded relationto The current study was based on the Mini- Insurance Scheme in Finland, all psychoses Finland Health Survey (Aromaa et al, mortality.Adjusted for age, gender, entitle the individual to free medication 1989). This comprehensive survey, carried somatic diseases and other potential risk for their treatment). People were also se- out between 1978 and 1980, was designed lected if they reported having used health factors for premature death, the relative to assess the health of adult Finns. The services (including those provided by a gen- riskwasrisk was 2.50 (95% CI1.46^4.30) per study population was a two-stage cluster eral practitioner) for a mental disorder or sample drawn from the population register increment of one neuroleptic. there was a self-perceived mental disorder. and stratified to represent Finns aged 30 A total of 35% of those who participated Conclusions Thereis anurgentneedto years or over. The first stage comprised in the screening phase were identified by the selection of 40 representative areas. In ascertainwhetherascertain whether thehighmortalityinthe high mortalityin various screening instruments; 95% of the second stage a systematic sample of schizophreniais attributabletothe disorder these agreed to participate in the clinical inhabitants was drawn from each area. phase of the health examination. itself or the antipsychotic medication. The sample size was 8000 persons, of whom 7217 (90%) participated in the Declaration of interest None. health examination (Aromaa et al, 1989). Case-finding procedure The participants were first interviewed The most important method of psychiatric ySee invited commentary,p.128, thisissue. at home by local public health nurses and case identification in the clinical phase

122 SCHIZOPHRENIA, NEUROLEPTICS AND MORTALITY

was use of the short version of the ninth schizophrenia was also recorded in the RESULTS edition of the Present State Examination screening phase and was categorised (PSE; Wing et al, 1974), which was admi- according to the number prescribed. At At baseline a number of lifestyle-related nistered by a psychiatric nurse. Training the time of the baseline assessment the factors and chronic diseases were asso- in the use of the PSE was provided by those following conventional neuroleptics were ciated with the prevalence of schizophrenia. who developed the method. The diagnoses the only neuroleptic drugs in use in Finland: Since some of the associations seemed to were obtained via the computer program phenothiazines with an aliphatic side-chain differ between men and women, all these CATEGO–ID (Wing et al, 1978). Diag- (promazine, chlorpromazine, levome- results were stratified according to gender nostic information was also obtained from: promazine); phenothiazines with a methyl- (Tables 1 and 2). Heavy smoking and case notes of participants in in-patient psy- piperazine side-chain (thioproperazine, obesity were significantly associated with chiatric care; SII records on disability pen- trifluoperazine); phenothiazines with a schizophrenia in both men and women, sions granted because of a mental disorder piperazine-ethanol side-chain (perphen- whereas being underweight proved to be a and on reimbursement for medication to azine, fluphenazine, thiopropazate); phe- significant determinant only in men, and treat psychoses; examination by the physi- nothiazines with a piperidine side-chain diabetes and other somatic diseases in cian in the clinical phase of the survey; (thioridazine, pericyazine, pipotiazine); women. Inverse associations with schizo- and assessment by the psychiatric nurse thioxanthenes (chlorprothixene, clopen- phrenia emerged for alcohol intake and conducting the PSE interview. A panel of thixol, flupenthixol, thiothixene); butyro- serum HDL cholesterol in both men and two psychiatrists reviewed the case notes phenones (haloperidol, melperone, women; in women there was an inverse of in-patients and the psychiatric records moperone); piperazines (pimozide); benza- association for exercise and in men for entitling individuals to SII benefits (disabil- mides (sulpiride); and indoles (oxyperine). systolic pressure. ity pensions and reimbursement for medi- A number of factors may be associated The age- and gender-adjusted relative cines) to validate the diagnoses. For the both with schizophrenia and mortality and risk of total mortality between people with final diagnostic assessments, a special com- could therefore confound the analysis. schizophrenia and others was 2.84 (95% CI puter program was designed that combined Age, gender, level of education, smoking, 2.06–3.90). The risk of mortality was in- information from various sources, taking alcohol intake, exercise, body mass index, creased among people with schizophrenia into account the degree of certainty and systolic and diastolic pressure, serum total even after controlling for potential risk eliminating mutually incompatible and HDL cholesterol, diabetes, cardio- factors for premature death (low level of assessment combinations. There were 99 vascular and respiratory diseases and other education, smoking, alcohol intake, exer- people with a diagnosis of schizophrenia, somatic diseases were therefore considered cise, body mass index, systolic and diastolic giving an age-adjusted prevalence of 1.3% potential confounders or effect modifiers pressure, and total and HDL cholesterol) with no gender difference (Lehtinen et al, in the present study (Brown, 1997; and coexistent somatic diseases (Table 3). 1990a); this was in accordance with an ear- Mortensen, 2003). As there was no difference between men lier population study in Finland (Vaisanen, and women in the association between 1975). schizophrenia and mortality, they were The mortality of those examined has combined in these analyses. been systematically monitored since the There were only four unnatural deaths baseline assessment. This information was Statistical analysis among those with schizophrenia. Adjusted obtained from the Central Statistical Office Logistic regression was used to estimate the for age and gender, the relative risk of of Finland. The principal causes of death associations between the potential con- natural death between people with schizo- were coded according to ICD–8 (World founding and effect-modifying factors and phrenia and others was 2.80 (95% CI Health Organization, 1974). During a the prevalence of schizophrenia. The gener- 2.00–3.93). follow-up from 1978 until 1994 the al linear model was used to compute multi- Of the 99 people with schizophrenia, numbers of deaths (in parentheses deaths of ple partial correlation ratios between those 20 were taking no neuroleptic drug at base- people with schizophrenia) were as follows factors and the number of neuroleptic drugs line, 31 one drug, 34 two drugs and 14 (see Joukamaa et al, 2001): a total of 1597 among participants with schizophrenia. three or more drugs. The most commonly (39) deaths occurred; 876 (17) were caused Cox’s proportional hazards regression used neuroleptic was thioridazine (34%), by any cardiovascular disease; 130 (8) by model (Cox, 1972) was used to estimate followed by perphenazine (20%), chlor- respiratory diseases; 341 (7) by cancers; the strength of the association between promazine (19%), levomepromazine (14%), 72 (2) by injuries; and 20 (2) by suicides. schizophrenia and mortality and between chlorprothixene (13%) and haloperidol The definitions and measurement of neuroleptic drug use for schizophrenia and (12%); use of other neuroleptics was less lifestyle and related factors (level of edu- mortality. The number of neuroleptic drugs than 10%. Among participants with schizo- cation, exercise, smoking, alcohol intake, for schizophrenia was included both as a phrenia, there was a strong inverse re- body mass index, systolic and diastolic categorical variable and as a continuous lationship between serum HDL cholesterol blood pressure, serum total and high- variable in the model. Potential confound- and the number of neuroleptic drugs pre- density lipoprotein (HDL) cholesterol) have ing and effect-modifying factors were scribed (correlation coefficient¼70.41, been described in more detail elsewhere entered into the Cox models. Adjusted rela- P50.001) thatremained statistically (Aromaa et al, 1989; MaMakela¨kela¨ et al, 1993). tive risks and their 95% confidence inter- significant after adjustmentadjustment for age, gender, Participants were asked about drugs pre- vals (CIs) were estimated based on this all lifestyle-related factors and coexistent scribed by their physician. The use of model. The SAS software package version somatic diseases (partial corre- neuroleptic drugs among people with 6.12 (SAS Institute, 1997) was used. lation¼70.31, P¼0.007). Smoking was

123 JOUKAMAA ET AL

Ta b l e 1 Relative risk for schizophrenia according to putative determinants of premature death and coexistent somatic diseases

VariableMen Women

n Schizophrenia, n RR95% CI n Schizophrenia, n RR95% CI

Age, years 30^441343 15 1.00 1 137316 1.00 1 45^54781 16 1.85 0.91^3.77 828 21 2.21 1.15^4.25 55^64603 10 1.49 0.67^3.34 745 80.92 8 0.920.39^2.16 65^74436 30.61 3 0.610.18^2.13 642 70.93 7 0.930.38^2.28 575159 21.13 2 1.130.26^5.0 307 1 0.28 0.04^2.10 Education Lower2245 35 1.00 2660 41 1.00 Average692 80.72 8 0.720.33^1.59 814 10 0.71 0.34^1.46 Higher385 30.48 3 0.480.15^1.60 421 2 0.26 0.06^1.11 Smoking Never980 10 1.00 3029 38 1.00 Quit1136 90.75 9 0.750.30^1.87 369 40.86 4 0.860.30^2.44 Pipe, cigar or 520 cigarettes per day677 14 2.04 0.90^4.62 385 51.03 5 1.030.40^2.67 520 cigarettes per day529 13 2.50 1.08^5.76 112 64.41 6 4.411.77^10.97 Alcohol intake None915 27 1.00 2381 47 1.00 1^99 g/week1666 12 0.22 0.11^0.44 1414 60.17 6 0.170.07^0.40 100^249 g/week237 10.12 1 0.120.02^0.91 48 0 5250g/week250 g/week 504 60.35 6 0.350.14^0.86 52 0 Body mass index (kg/m2) 520.0124 10 9.51 4.11^22.01 235 31.37 3 1.370.39^4.75 20.0^24.91346 15 1.00 1600 17 1.00 25.0^29.91464 14 0.81 0.38^1.68 1351 12 0.85 0.40^1.82 30.0^34.9351 51.19 5 1.190.43^3.32 558 12 2.24 1.03^4.85 535.037 24.84 2 4.841.06^22.19 151 95.98 9 5.982.54^14.04 Exercise None1085 19 1.00 1551 33 1.00 Low level or occasional1642 21 0.72 0.38^1.35 1835 18 0.42 0.23^0.76 High level and regular591 50.47 5 0.470.17^1.28 503 20.16 2 0.160.04^0.69 Chronic diseases Cardiovascular No2196 32 1.00 2628 34 1.00 Yes1126 14 0.82 0.42^1.62 1267 19 1.32 0.69^2.53 Respiratory No2492 30 1.00 3488 49 1.00 Yes830 16 1.64 0.88^3.06 407 40.70 4 0.700.25^1.97 Diabetes No3147 43 1.00 3658 44 1.00 Yes175 3 1.26 0.38^4.21 237 94.15 9 4.151.85^9.32 Other No2570 33 1.00 2728 28 1.00 Yes752 13 1.36 0.70^2.64 1167 25 2.33 1.32^4.11

RR, relative risk. 1. Unadjusted. also associated with the number of neuro- multifactorial analysis (partial correlation was related to the subsequent mortality. Of leptics prescribed (correlation coefficient¼ ¼0.27, P¼0.14). people with schizophrenia taking one, two 0.35, P¼0.007), but the association did The number of neuroleptic drugs and three or more neuroleptic drugs, 11 not reach statistical significance in the prescribed at the time of the baseline survey (35%), 15 (44%) and 8 (57%) respectively

124 SCHIZOPHRENIA, NEUROLEPTICS AND MORTALITY

Ta b l e 2 Odds ratio for schizophrenia according to blood pressure and cholesterol levels

VariableMen Women

Means.d. OR 95% CI1 Means.d. OR 95% CI1

Diastolic pressure, mmHg88.6 11.8 0.82 0.61^1.11 86.1 11.7 1.22 0.92^1.60 Systolic pressure, mmHg144.8 21.2 0.53 0.37^0.77 147.8 26.0 0.82 0.57^1.18 HDL cholesterol, mmol/l1.59 0.4 0.70 0.50^0.98 1.79 0.4 0.47 0.34^0.65 Total cholesterol, mmol/l6.9 1.3 1.04 0.78^1.39 7.0 1.4 1.17 0.88^1.55

OR, odds ratio; HDL, high-density lipoprotein. 1. Per an increment of 1 s.d.

Ta b l e 3 Relative risk of mortality in people with schizophrenia (n¼99) compared with others (n¼7118) after estimates were 1.69 (0.42–6.80), 4.75 adjustment for potential confounding factors (1.95–11.53), 2.53 (0.63–10.21) and 5.35 (1.33–21.55). Factors adjusted forRR 95% CI DISCUSSION None (unadjusted)2.01 1.46^2.76 Age, gender2.84 2.06^3.90 Main findings Age, gender, level of education, systolic pressure, diastolic pressure, 2.211.58^3.08 The present study demonstrated a graded total and HDL cholesterol, body mass index, smoking, alcohol intake relationship between the number of neuro- and exercise leptic drugs prescribed and mortality of Age, gender, cardiovascular disease, respiratory disease, diabetes and 2.892.10^3.98 those with schizophrenia. This relationship other somatic disease and the excess mortality among people with schizophrenia could not be explained All the factors and diseases above2.25 1.61^3.15 by coexistent somatic diseases or other RR, relative risk; HDL, high-density lipoprotein. known risk factors for premature death. To our knowledge this was the first study Ta b l e 4 Relative risk of mortality among people with schizophrenia (n¼99) per an increment of one to analyse such associations in the general neuroleptic drug after adjustment for potential confounding factors population using a prospective design. Although the excess mortality of people Factors adjusted forRR 95% CI with schizophrenia was first demonstrated before the Second World War, the causes None (unadjusted)1.43 1.03^1.99 have varied over the years, especially before Age, gender1.70 1.20^2.41 the neuroleptic era (see Brown, 1997). It Age, gender, level of education, systolic pressure, diastolic pressure, 2.291.38^3.80 has been claimed that the contemporary total and HDL cholesterol, body mass index, smoking, alcohol intake high natural mortality in schizophrenia and exercise results from a variety of lifestyle factors Age, gender, cardiovascular disease, respiratory disease, diabetes 1.691.17^2.45 (Brown, 1997; Mortensen, 2003). Of these factors we were able to consider several and other somatic disease (smoking, exercise, body mass index, blood All the factors and diseases above2.50 1.46^4.30 pressure, serum total and HDL cholesterol), RR, relative risk; HDL, high-density lipoprotein. but not all (e.g. dietary factors). However, after adjustment for these factors the excess died during follow-up, whereas the drugs in the whole of the study population, mortality of people with schizophrenia corresponding rate was 5 (20%) among the 68 people taking neuroleptics for other persisted. Similarly, comorbid somatic those without neuroleptic medication. psychoses were excluded. Adjusted for age diseases did not account for the mortality. Table 4 shows the relative risk of dying and gender, people with schizophrenia tak- Obviously there are other factors associated among those with schizophrenia per incre- ing no neuroleptic, one, two and three or with mortality in schizophrenia and the ment of one neuroleptic drug adjusted for more neuroleptic drugs had relative risks association with neuroleptic drugs was very age, gender, lifestyle-related factors and (95% CI) of 1.29 (0.53–3.11), 2.97 (1.64– clear. This remained after adjustment for chronic somatic diseases. Irrespective of 5.38), 3.21 (1.93–5.35) and 6.83 (3.40– many different potentially confounding the factors modelled, the relationship 13.71) respectively compared with those factors. At the time of the baseline survey between number of neuroleptic drugs and without schizophrenia or any antipsychotic only classic neuroleptic drugs were in use mortality remained strong and statistically treatment with neuroleptics. The associa- in Finland. significant. tion remained stable throughout the obser- It is now well known that most or all To assess the combined effect of schizo- vation period. Even after excluding the first classic neuroleptics can cause prolongation phrenia and the number of neuroleptic 10 years of follow-up, the corresponding of the QT interval of the electrocardiogram

125 JOUKAMAA ET AL

which is associated with the potentially fatal arrhythmia torsade de pointes CLINICAL IMPLICATIONS (Witchel et al, 2003). Recently, in a large study Ray et al (2001) showed that & The somatic health of people with schizophrenia is an important issue. prescription of moderate doses of antipsy- chotics was associated with large relative & The use of classical neuroleptic drugs is associated with mortality in schizophrenia. and absolute increases in the risk of sudden & A combination of neuroleptic drugs seems to increase the risk of mortality. cardiac death. It has been proposed that antipsychotic drugs might also be asso- LIMITATIONS ciated with venous thrombosis, pulmonary embolism (Thomassen et al, 2001) and & The statistical power of the present study was insufficient for subgroup analyses. asthma mortality (Joseph et al, 1996). These & associations could explain at least some of Since the baseline assessment was completed in the1970s, it was not possible to the deaths from respiratory diseases in the use standardised criteria for schizophrenia. present study. & We were not able to analyse the total period of neuroleptic use and the total dosage. Neuroleptic polypharmacy The risk of dying increased in our study if the participants were taking more than one neuroleptic drug. This is in line with previous findings. Waddington et al MATTI JOUKAMAA, MD,PhD,Tampere School of Public Health,University of Tampere and Department of « (1998) found that receiving more than one Psychiatry,TamperePsychiatry,TampereUniversityHospital,Tampere;MARKKUHELIOVAARA,MD,PhD,PAULKNEKT,PhD, University Hospital,Tampere; MARKKUHELIOVAARA,MD,PhD,PAULKNEKT,PhD, ARPO AROMAA, MD, PhD, National Public Health Institute, Helsinki; RAIMO RAITASALO, PhD, Social antipsychotic concurrently was associated Insurance Institution, Helsinki; VILLE LEHTINEN, MD,PhD,NationalMD, PhD, National Research and Development Centre for with reduced survival in a 10-year follow- Welfare and Health, Helsinki, Finland up of 88 people with schizophrenia. In a French study the mortality among patients CorrespCorrespondence:Drondence:Dr Matti Joukamaa,DepartmentofJoukamaa,Department of SoSocialcial Psychiatry,Tampere School of Public Health, with chronic schizophrenia was associated University of Tampere, FIN-33014, Finland. E-mail: [email protected] with the dosage of neuroleptics (Bralet et (First received 28 October 2004, final revision 15 March 2005, accepted 18 May 2005) al, 2000). Why is more than one neuro- leptic drug prescribed? We can presume that one reason might be the lack of re- sponse to a single antipsychotic, possibly Because of the comprehensive nature of Conclusions related to the severity of the illness. Is it also the Mini-Finland Health Survey we were More attention should be paid to the possible that the most severe forms of able to take many potential confounding somatic health of people with schizo- schizophrenia carry a higher risk of dying factors into account. The main screening phrenia. Our results indicate a need to than less severe schizophrenia. However, instrument used, the General Health modify the deleterious lifestyle factors of it must be borne in mind that antipsychotic Questionnaire, is most suitable for screen- patients with schizophrenia previously polypharmacy can also lead to increased ing for non-psychotic psychiatric illnesses recommended by Brown et al (2000). The side-effects because of the potential influ- such as anxiety and mood disorders. recognition and medical management of ence on many receptors in addition to brain However, other aspects of our screening somatic diseases among people with schizo- D dopamine receptors (Waddington et al, 2 procedure, especially review of the SII phrenia is a challenge to psychiatrists. Fu- 1998). records concerning disability pensions and ture research needs to determine whether If classic neuroleptics were the cause of reimbursement for medicines used to treat the high mortality among those with premature deaths among people with severe mental disorders, identified those schizophrenia is mainly attributable to the schizophrenia it could be assumed that the with psychotic disorders. Another limita- disorder per se or to the antipsychotic medi- excess mortality would have decreased with tion was that we were not able to define cation. Study of the association between the the introduction of the newer atypical anti- the total period of neuroleptic use or total newer antipsychotic drugs and mortality psychotics. However, a Swedish study dosage or to assess the impact of prior in patients with schizophrenia will be found an increase in mortality in patients treatment. In addition, the number of important after a sufficient period of treat- with schizophrenia (Osby et al, 2000). people with schizophrenia was relatively ment has elapsed to allow long-term The newer atypical antipsychotic agents low, only 99, of whom 39 died during follow-up. can affect somatic health in different ways follow-up. In spite of these limitations we (Wirshing et al, 2003). consider the results valid. The association of the use of neuroleptic drugs and REFERENCES Study strengths and limitations mortality poses questions which could not be answered in this study. A careful quanti- Aromaa, A., HelioHeliovaara,«vaara, M., Impivaara, O., et al The strengths of the present study included tative analysis of the association of neuro- (19 8 9) Aims, methods and study population. Part 1.In The Execution of the Mini-Finland Health Survey (eds A. a long follow-up and a representative leptic drug dosage and mortality would be Aromaa, M. HelioHeliovaara,O.Impivaara,«vaara,O.Impivaara, et al). Publication sample with a high participation rate. especially interesting. no. ML:88. Helsinki: Social Insurance Institution.

126 SCHIZOPHRENIA, NEUROLEPTICS AND MORTALITY

Bralet, M. C.,Yon,V., Loas, G., et al (2000) Causes de Finland. Basic results from the Mini-Finland Health SAS Institute(19 (1997) 9 7) SAS/STAT Software: Changes and la mortalite¤chez les schizophreschizophrenes:' nes: eetude¤tude prospective Survey. Acta Psychiatrica Scandinavica, 81,,418^425. 418^425. Enhancements through Release 6.12.Cary,NC: SAS sur 8 ans d’une cohorte de 150 schizophreschizophrenes' nes Institute. chroniques. EnceEncephale¤phale, 26,32^41. Lehtinen,V., Joukamaa, M., Jyrkinen, E., et al (19 9 0b) Need for mental health services of the adult Thomassen, R.,Vandenbroucke, J. P. & Rosendaal, Brown, S.(19 (1997) 9 7) Excess mortality of schizophrenia. A population in Finland: results from the Mini-Finland F. R.(2001) (2 0 01) Antipsychotic medication and venous meta-analysis. British Journal of Psychiatry, 171, 502^508. Health Survey. Acta Psychiatrica Scandinavica, 81, thrombosis. British Journal of Psychiatry, 179,63^66., 63^66. Brown, S., Inskip, H. & Barraclough, B.(2000) 426^431. Vaisanen, E.(19 (1975) 75) Psychiatric disorders in Finland. Causes of the excess mortality of schizophrenia. British « « « Acta Psychiatrica Scandinavica Supplementum, 263, Journal of Psychiatry, 177, 212 ^217. MaMakela,kela,M.,Helio M., Heliovaara,vaara, M., Sievers, K., et al (19 93) Musculoskeletal disorders as determinants of disability in 22^33. Cox, D. R.(19 (1972) 72) Regression models and life-tables. Finns aged 30 years or more. Journal of Clinical Waddington, J. L.,Youssef, H. A. & Kinsella, A. Journal of the Royal Statistical Society B, 34,187^220. Epidemiology, 46, 549^559. (19 9 8) Mortality in schizophrenia. Antipsychotic Goldberg,Goldberg,D.P. D. P.(19 (1972) 72) The Detection of Psychiatric Illness Montout, C., Casadebaig, F., Lagnaoui, R., et al polypharmacy and absence of adjunctive anticholinergics by Questionnaire. Maudsley Monograph 21.London: (2002) Neuroleptics and mortality in schizophrenia: over the course of a 10-year prospective study. British Oxford University Press. prospective analysis of deaths in a French cohort of Journal of Psychiatry, 173,,325^329. 325^329. Harris, E. C. & Barraclough, B.(19 (1998) 9 8) Excess schizophrenic patients. Schizophrenia Research, 57, Wing, J. K.,K.,Cooper, Cooper, J. E. & Sartorius, N.(19 (1974) 74) The mortality of mental disorder. British Journal of Psychiatry, 147147^156. ^156. Measurement and Classification of Psychiatric Symptoms. 173,11^53., 11^53. London: Cambridge University Press. Mortensen, P.B.P. B. (2003) Mortality and physical illness in HeilaHeila,«,H.&Lo H. & Lonnqvist,«nnqvist,J.(2003) J. (2003) The clinical schizophrenia. In The Epidemiology of Schizophrenia (eds Wing, J. K., Mann, S. A., Leff, J. P., et al (19 78) The epidemiology of suicide in schizophrenia. InThe R.Murray,P.Jones,E.Susser,R. Murray,P.Jones, E. Susser, et al), pp.275^287. concept of a case in psychiatric population surveys. Epidemiology of Schizophrenia (eds R. Murray,P.Jones, E. Cambridge: Cambridge University Press. Psychological Medicine, 8,,203^217. 203^217. Susser, et al), pp. 288^316.Cambridge: Cambridge University Press. Osby,U., Correia, N., Brandt, L., et al (2000) Wirshing, D. A., Pierre, J. M., Erhart, S. M., et al Mortality and causes of death in schizophrenia in (2003) Joseph,K.S.,Blais,L.,Ernst,P.,Joseph, K. S., Blais, L., Ernst, P., et al (19 9 6) Increased Understanding the new and evolving profile of Stockholm county,Sweden. Schizophrenia Research, 45, morbidity and mortality related to asthma among adverse drug effects in schizophrenia. Psychiatric Clinics 21^28. asthmatic patients who use major tranquillisers. BMJ, 31, of North America, 26,165^190. 79^82. Ray,W. A. & Meador, K. G.(2002) Antipsychotics and Witchel, H. J., Hancox, J. C. & Nutt, D. J.(2003) Joukamaa, M., HelioHeliovaara,«vaara, M., Knekt, P., et al (2001) sudden death: is thioridazine the only bad actor? British Psychotropic drugs, cardiac arrhythmia, and sudden Mental disorders and cause-specific mortality. British Journal of Psychiatry, 180, 483^484. death. Journal of Clinical Psychopharmacology, 23, 58^77. Journal of Psychiatry, 179, 498^502. Ray,W. A., Meredith, S., Thapa, P. B., et al (2001) WorldHealthOrganization(1974)WorldHealthOrganization(19 74) Eighth Revision of Lehtinen,V., Joukamaa, M., Jyrkinen, E., et al Antipsychotics and the risk of sudden cardiac death. the International Classification of Diseases and Related (19 9 0a) Prevalence of mental disorders among adults in Archives of General Psychiatry, 58,,1161^1167. 1161^1167. Health Problems (ICD^8).Geneva: WHO.

127 META-ANALYSIS A Systematic Review of Mortality in Schizophrenia Is the Differential Mortality Gap Worsening Over Time?

Sukanta Saha, MSc, MCN; David Chant, PhD; John McGrath, MD, PhD, FRANZCP

Context: Despite improvements in mental health ser- and pooled selected estimates using random-effects vices in recent decades, it is unclear whether the risk of meta-analysis. We identified 37 articles drawn from 25 mortality in schizophrenia has changed over time. different nations. The median SMR for all persons for all-cause mortality was 2.58 (10%-90% quantile, 1.18- Objective: To explore the distribution of standardized 5.76), with a corresponding random-effects pooled SMR mortality ratios (SMRs) for people with schizophrenia. of 2.50 (95% confidence interval, 2.18-2.43). No sex dif- ference was detected. Suicide was associated with the high- Data Sources: Broad search terms were used in est SMR (12.86); however, most of the major causes-of- MEDLINE, PsychINFO, Web of Science, and Google death categories were found to be elevated in people with Scholar to identify all studies that investigated mortality schizophrenia. The SMRs for all-cause mortality have in- in schizophrenia, published between January 1, 1980, and creased during recent decades ( =.03). January 31, 2006. References were also identified from P review articles, reference lists, and communication with authors. Conclusions: With respect to mortality, a substantial gap exists between the health of people with schizophrenia Study Selection: Population-based studies that re- and the general community. This differential mortality ported primary data on deaths in people with schizo- gap has worsened in recent decades. In light of the po- phrenia. tential for second-generation antipsychotic medications to further adversely influence mortality rates in the de- Data Extraction: Operationalized criteria were used to cades to come, optimizing the general health of people extract key study features and mortality data. with schizophrenia warrants urgent attention.

Data Synthesis: We examined the distribution of SMRs Arch Gen Psychiatry. 2007;64(10):1123-1131

T IS NOW WIDELY ACKNOWL- group of disorders now labeled schizo- edged that schizophrenia con- phrenia, increased mortality rates have tributes substantially to the global been the object of scrutiny since the early burden of disease.1,2 It is also well 20th century.6-8 The quality of research on known that schizophrenia is this topic has improved greatly in recent Iassociated with elevated suicide rates.3 decades, with access to larger, better- Less widely appreciated is the fact that characterized samples and the availabil- people with schizophrenia are at in- ity of high-quality mortality data for the creased risk for premature death associ- general population. Access to these data ated with comorbid somatic conditions.4 allows the calculation of the standard- Apart from adverse effects related to medi- ized mortality ratio (SMR), which com- cation, schizophrenia can trigger a cas- pares mortality in people with schizophre- Author Affiliations: cade of socioeconomic and lifestyle fac- nia vs the general population. The SMRs Queensland Centre for Mental tors that, in turn, can translate into adverse are calculated by dividing the observed Health Research, The Park physical health outcomes. These comor- mortality rates in a given population (eg, Centre for Mental Health, bid physical conditions contribute to in- the number of deaths in a group of indi- Wacol (Mr Saha and Drs Chant and McGrath), and Department creased mortality risks among people with viduals with schizophrenia) by the ex- of Psychiatry, The University of schizophrenia. pected mortality rates in that same group Queensland, St Lucia The association between severe men- as predicted by age- and sex-specific mor- (Drs Chant and McGrath), tal illness and increased mortality rates has tality rates for a standard population. Thus, Australia. long been recognized.5 With respect to the an SMR of 2.0 would indicate that people

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 64 (NO. 10), OCT 2007 WWW.ARCHGENPSYCHIATRY.COM 1123 Downloaded from www.archgenpsychiatry.com at University of Sydney Library, on September 2, 2008 ©2007 American Medical Association. All rights reserved. with schizophrenia are twice as likely to die compared quality studies would be more likely to identify deaths with the general population. The SMRs can be calcu- in schizophrenia, we predicted that SMRs derived from lated for overall mortality (all-cause) or for more spe- such studies would be higher compared with those from cific, widely used categories (eg, cancer, cardiovascular lower-quality studies. On the basis of previous system- disease, endocrine disorders, or suicide). atic reviews and commentaries, we predicted that SMRs In recent years, several scholarly reviews4,9-11 have noted would increase over time. higher mortality in schizophrenia compared with the gen- 4 eral population. A meta-analysis, based on 18 studies pub- METHODS lished between 1969 and 1996, reported an all-cause SMR for people with schizophrenia of 1.51. Another meta- DATA SOURCES analysis,11 based on 20 studies published between 1973 and 1995, reported a similar SMR for people with schizo- Most mortality studies are based on record linkage. People with phrenia (1.57). Although these 2 systematic reviews schizophrenia are identified via psychiatric case registers and agreed on the size of the pooled SMR associated with then subsequently linked to registers of cause of death. Some schizophrenia, there were discrepancies in the sex dif- studies13,22 report mortality ratios based on hospital inpatient ference of overall mortality ratios. Brown4 found a small cohorts. Other studies23,24 have used community-based fol- but significant male excess in the overall mortality ra- low-up data for people with schizophrenia who are first iden- tio, whereas other studies12,13 reported either no sex dif- tified through community surveys and then followed up for ex- ference11 or higher mortality ratios in females compared tended periods. with males. In collating data from different sites, systematic re- IDENTIFICATION OF STUDIES viewers need to appreciate the structure of the underly- Guidelines outlined by the Meta-analysis of Observational Stud- ing data. In light of the differing population age struc- ies in Epidemiology25 were followed to identify and collate mor- 1,14 ture and disease profile among sites, we would expect tality studies. The broad search string of (schizo* or psych*) substantial variation in mortality ratios among sites. For and (mortality or outcome or follow-up) was used in MEDLINE, example, one would predict that SMRs for people with PsychINFO, Web of Science, and Google Scholar to identify schizophrenia would differ between developed and de- all research studies that investigated mortality in schizophre- veloping nations, where the profiles of disease and the nia. Potentially relevant articles (in all languages) were ac- access to services vary markedly. cessed to review the full text. Citations from significant ar- Because of the increased focus on mental health care ticles and review articles were scrutinized to locate additional seen in many countries during the last few decades, one relevant articles, book chapters, and conference papers. The Web of Science Cited Reference Search system was also used to lo- might predict that SMRs associated with disorders such 15,16 cate relevant articles. Finally, letters or e-mails were sent to the as schizophrenia should be decreasing over time. How- senior authors of articles that met the inclusion criteria. These ever, several authors have suggested that SMRs in schizo- authors were provided with an interim list of included studies phrenia have been increasing during recent decades. For and asked to nominate missing studies. example, Osby et al17 found a linear increasing trend of mortality during 5-year periods from 1976 to 1995 among INCLUSION AND EXCLUSION RULES people with schizophrenia. The meta-analysis by Brown4 also reported significantly higher mortality in the 1980s Studies were included if they met all the following criteria: (1) compared with the 1970s. Deinstitutionalization may have published and/or available between January 1, 1980, and Janu- influenced recent secular changes in mortality rates in ary 31, 2006, (2) reported deaths in people with schizophre- schizophrenia. Although deinstitutionalization started in nia as diagnosed by any criteria, (3) studied a population 15 the 1950s, findings on its relationship to mortality have years and older, (4) reported primary data on all-cause mor- been inconsistent.10,11,18 tality and/or cause-specific mortality, and (5) reported SMRs and/or data on observed and expected deaths sufficient to cal- The aims of this study were to undertake a system- culate SMRs. Studies were excluded if they (1) involved people atic review of mortality in schizophrenia and to exam- with a diagnosis other than schizophrenia (ie, studies that re- ine a limited number of planned sensitivity analyses. In ported on broader categories of psychosis were excluded), (2) keeping with our previous systematic review of the in- reported duplicate data, (3) reported SMRs solely attributable cidence19 and prevalence20 of schizophrenia and consid- to suicide (this was the focus of a recent systematic review and ering that variability is to be expected in systematic re- meta-analysis3), and (4) reported mortality in subgroups of the views of SMRs,4,21 we sought to preserve the expected population (eg, homeless people,26 twins,27 and those in- variation in the data rather than to focus only on pooled volved in clinical trials). values derived from meta-analysis. Thus, for the main analyses, we present distributions of mortality esti- DATA ABSTRACTION mates with measures of central tendency (eg, median or means) and quantiles (10% and 90% quantiles). On the Once a study was included, data were extracted and entered basis of all-cause SMR, we predicted that the SMRs of into a 3-level, normalized database that included study-level variables (eg, authors, year of publication, and site), middle- males and females would not differ significantly. We also level variables (eg, age group, recruitment duration, case- predicted that SMRs from the developed world would dif- finding method, and diagnostic criteria), and estimate-level vari- fer from those from the developing world (nondirec- ables (eg, general and specific-cause SMRs for all persons, males, tional hypothesis). We wished to explore the impact of or females). Two or more of the authors checked all data used study quality on SMRs. With the assumption that higher- in the analysis. When disagreements arose, these were re-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 64 (NO. 10), OCT 2007 WWW.ARCHGENPSYCHIATRY.COM 1124 Downloaded from www.archgenpsychiatry.com at University of Sydney Library, on September 2, 2008 ©2007 American Medical Association. All rights reserved. solved by consensus. If required, we contacted the original au- To assess secular trends, we used meta-regression to exam- thors for clarification of issues. The full data set is available from ine the relationship between the midpoint of the follow-up pe- the authors (www.qcmhr.uq.edu.au/epi). riod and all-cause SMR for persons. Study quality scores were To assess the impact of overall quality of the distribution divided into tertiles, and the distribution of all-cause SMR for of SMRs, we devised a quality score. On the basis of operation- persons were compared according to these 3 levels. alized criteria, this score rewarded studies that (1) used supe- We performed statistical analyses for the test of signifi- rior research design features (eg, more thorough case ascer- cance between distributions of different SMRs. These analyses tainment, published diagnostic criteria, methods to confirm take into account (1) the need to control for within-study varia- diagnosis, and longer periods of follow-up) and (2) provided tion (estimates drawn from the same study tend to be more alike comprehensive reporting of the study results (eg, provision of than SMRs drawn from different studies) and (2) the use of a numerator, denominator, SMRs, details of subject attrition, and log transformation to analyze distributions that are often posi- description of the completeness of the data source). Full de- tively skewed. Analyses were performed with SAS statistical soft- tails of the quality score used in this review are available from ware, version 9.2 (SAS Institute Inc, Cary, North Carolina). the authors (www.qcmhr.uq.edu.au/epi). We also undertook a secondary analysis based on conven- In systematic reviews, it is important to avoid double count- tional meta-analytic techniques. Because SMRs are known to ing of the index variable (deaths) by the same or different stud- vary widely among sites because of population and disease fre- ies. Thus, a key feature of this review is the application of se- quency differences, we adopted a random-effects model to es- quential filters to identify discrete mortality estimates. We applied timate a pooled SMR for all-cause mortality for persons.21 When a similar sorting algorithm to that used in our previous re- necessary, 95% confidence intervals (CIs) were generated ac- views of schizophrenia.19,20 Briefly, the mortality estimates were cording to the formula detailed by Rothman and Greenland.21 sorted into different causes of death. Study-level and middle- Heterogeneity among the studies was tested using the Coch- level filters were applied to isolate data from multiple studies ran heterogeneity statistic.32 Apart from the specific analyses that overlapped in both time and place. The third filter was used related to sex differences, we restricted the analyses to per- to select 1 representative mortality estimate for inclusion in the sons to limit the number of planned comparisons. The fund- cumulative distribution using the “most informative” rule. For ing source played no part in the design, analyses, writing, or example, if 1 study presented multiple overlapping ratios, the submission of this study. ratios based on the largest sample were preferred (ie, the wid- est age range was preferred over narrower age strata). The highest-order (and most reliable) category of death, all- RESULTS cause mortality, can be further subdivided according to rules such as those codified by the International Classification of Dis- eases, Ninth Revision (ICD-9).28 Almost all included studies in The electronic search identified 1726 articles, whereas this review were coded with the ICD-9. Although death can re- manual reference checking identified an additional 26 sult from the combination of many different health problems, references. We received responses from 16 authors, who in circumstances in which several codes may be suitable, em- provided an additional 11 references. Four articles from phasis is given to the underlying cause of death. More specific languages other than English were included after trans- causes of death can be allocated to categories according to or- lation. Eleven studies33-43 were excluded because they com- gan systems (eg, cardiovascular or gastrointestinal) or nature pletely overlapped with other included studies. Further of disease (eg, cancers are coded together). Apart from codes details of the results of the search strategy and key fea- for these specific domains, studies occasionally report SMRs tures of the included studies are available from the au- for middle-level categories such as all-unnatural (ICD-9 codes thors (www.qcmhr.uq.edu.au/epi). E800-E999) (which includes codes for suicide, accident, and 9,12,13,18,22-24,44-73 homicide) and all-natural (ICD-9 codes 001-799; the remain- Thesystematicreviewidentified37studies der from all-cause when all-unnatural cause is excluded). that provided data on 561 SMRs for different causes The SMRs were extracted from the publications or calcu- of deaths drawn from 25 different countries: Australia lated by dividing the sum of observed deaths by the sum of ex- (n=2),59,68 Brazil (n=1),61 Bulgaria (n=1),53 Canada pected deaths (when sufficient data were available to calculate (n=3),50,51,65 China (n=1),53 Columbia (n=1),53 Czech Re- these). The distributions of SMRs were assessed in cumulative public (n=1),53 Denmark (n=2),63,64 Finland (n=3),18,22,23 plots, with every SMR contributing to the distribution. The dis- France (n=2),46,48 Germany (n=1),57 Hong Kong (n=1),53 tribution of the data was assessed in rank order for SMRs (low- India (n=2),12,53 Indonesia (n=1),58 Ireland (n=2)53,62 est to highest ranks) with the cumulative percentage of SMRs Israel (n=1),73 Italy (n=2),60,67 Japan (n=3),53,69,71 Norway shown on the vertical axis. Key features of these distributions (n=1),52 Russia (n=1),53 Sweden (n=2),9,66 Taiwan are presented (eg, median, mean, geometric mean, standard de- 49 13 viation, and quantiles at 10%, 25%, 50%, 75%, and 90%). (n=1), the Netherlands (n=1), the United Kingdom 44,47,53,54,56 24,45,53,55,70,72 For all-cause death, we were often able to extract data on (n=5), and the United States (n=6). case fatality rate (CFR). The CFR is calculated by dividing the The SMRs were based on an estimated total of 22 296 number of deaths in people with schizophrenia during a cer- discrete deaths. Thirty-seven studies9,12,13,18,22-24,44-73 pro- tain period by the number of people with that disorder at the vided SMRs for all-cause mortality for either all per- beginning of the period. An annualized CFR was derived to al- sons, males, or females. 14 low comparisons among studies of different durations. Figure 1 shows the distribution for all-cause SMRs for In keeping with definitions from our previous systematic 20,29 all persons, males, and females. The median all-cause SMR reviews of schizophrenia, we divided studies according to for all persons (based on 38 SMRs) was 2.58, with 10% the per capita gross national product of the study site (based on 2004 data)30 and used a standard World Bank definition of and 90% quantiles ranging from 1.18 to 5.76 (Table 1). country status31: (1) least developed countries,mean income In other words, people with schizophrenia had 2.5 times of less than US $2995; (2) emerging economy countries,mean the risk of dying compared with the general population, income between US $2995 and $9266; and (3) developed coun- and the central 80% of all SMRs varied over a 4-fold range. tries,mean income of greater than US $9266. The median annualized all-cause CFR for all persons was

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 64 (NO. 10), OCT 2007 WWW.ARCHGENPSYCHIATRY.COM 1125 Downloaded from www.archgenpsychiatry.com at University of Sydney Library, on September 2, 2008 ©2007 American Medical Association. All rights reserved. 95.4, and 108.3 for the 1970s, 1980s, and 1990s, respec- 100 tively. The CFRs for the 3 decades were not statistically

90 significantly different (F2,23=0.38; P=.38). The 38 studies that report all-cause SMRs for all per- 80 sons are shown in a traditional forest plot with a pooled 70 estimate based on a random-effects model in Figure 2. 60 Using traditional meta-analytic techniques, we found that 50 the pooled random-effects all-cause SMR (based on 37 SMRs

40 with finite standard errors) for all persons was 2.50 (95% CI, 2.18-2.83). The Cochran Q test found a marginally ac- SMR, Cumulative % 30 ceptable level of heterogeneity (Q =50.72; P=.06). We un- All persons 36 20 Males dertook several post hoc analyses to explore potential 10 Females sources of variation (eg, published vs unpublished diag- 0 nostic criteria, cohorts based on first-episode patients vs 1 2 3 4 5 6 7 8 9 all patients, cohorts based on inpatient and/or outpatient SMR samples, sites clustered according to World Health Orga- nization regions, and SMRs attributable to suicide sorted Figure 1. Cumulative plots of standardized mortality ratios (SMRs) for by decade). However, none of the post hoc comparisons all-cause mortality associated with schizophrenia by sex. resulted in significantly different SMR distributions (data not shown). 95.4 per 10 000 population, with 10% and 90% quantiles ranging from 57.2 to 301.7 (5-fold range). The median all-cause SMR for males (3.02) was slightly COMMENT higher than females (2.37); however, these 2 distribu- tions were not statistically significantly different People with schizophrenia have a substantially in- (F1,18=0.0003; P=.99). For all persons, the median SMR creased risk of death compared with the general popu- for natural causes of death was 2.41, and the 10% and lation. Overall, people with schizophrenia have 2.5 times 90% quantiles ranged from 0.99 to 4.10 (Table 1). El- the risk of dying. This review was able to extract data from evated median SMRs were found in all of the specific 37 studies that were conducted in 25 countries. As pre- causes of death apart from cerebrovascular diseases. dicted, the distribution of all-cause SMRs showed promi- Seven studies18,47,49,51,56,65,66 published data for the sum- nent variability. mary category of unnatural causes of death for all per- Confirming the hypothesis that the relative mortality sons, males, or females. Table 1 gives the distributions risk associated with schizophrenia is increasing, we found of SMRs for unnatural causes of death. People with schizo- that SMRs have increased in a linear fashion during the phrenia had 12 times the risk of dying of suicide com- 3 decades examined in this study. This finding is con- pared with the general population (median SMR, 12.86). sistent with earlier studies.4,17 Considering that (1) CFRs Twenty-two studies* were identified that contrib- for schizophrenia did not significantly differ among the uted 28 SMRs for developed countries, 3 studies53,58,61 con- decades and (2) age-standardized mortality rates are gen- tributed 6 SMRs for emerging economy countries, and 1 erally decreasing in most nations,74-76 these findings sug- study53 contributed 4 SMRs for least developed coun- gest that people with schizophrenia have not fully ben- tries. When divided according to this criterion, the all- efited from the improvements in health outcomes available cause SMR distributions were not significantly different to the general population. The SMRs are ratio measures (F2,34=0.30; P=.74); the median all-cause SMRs for least and thus reflect differential mortality. If mortality rates developed, emerging economy, and developed coun- in the general population decrease over time at a faster tries were 2.02, 2.19, and 2.79, respectively (Table 2). rate than those for people with schizophrenia, then SMRs When the all-cause SMRs for all persons were di- for people with schizophrenia will increase over time. The vided into study quality score tertiles, no significant dif- evidence from the current study suggests that this dif- ferences were found between SMR distributions ferential mortality gap has widened over time. (F2,24=0.61; P=.55). On the basis of follow-up periods, Mental health services have advanced in many parts we identified 8 studies24,45,51,54,55,63,71,72 with SMRs from the of the world during the past few decades. Apart from a 1970s, 10 studies47,53,57-60,65-67,73 with SMRs from the 1980s, different mix of community-based care, the introduc- and 7 studies22,46,48,53,61,62,68 with SMRs from the 1990s. Con- tion of the second-generation antipsychotic medica- cerning secular change, meta-regression confirmed a sig- tions in the early 1990s was initially found to be associ- nificant positive association between the follow-up pe- ated with better quality of life and reduced risk of riod midpoint year and all-cause SMR (slope coefficient, relapse.77-79 More recent trials have questioned the clini- 0.06; 95% CI, 0.01-0.11; z=2.21; P=.03). The median cal superiority of second-generation antipsychotic medi- SMRs for the 1970s, 1980s, and 1990s were 1.84, 2.98, cation,80,81 and concern is now widespread about the ad- and 3.20, respectively. Concerning CFRs, the median rates verse effects associated with these medications.82 In per 10 000 population (all-cause mortality) were 162.2, particular, compared with typical antipsychotics, sev- eral of the second-generation antipsychotics are more *References 22, 24, 45-48, 51, 53-55, 57, 59, 60, 62, 63, 65-68, likely to cause weight gain and metabolic syndrome.83 71-73. Because the metabolic syndrome is associated with a 2-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 64 (NO. 10), OCT 2007 WWW.ARCHGENPSYCHIATRY.COM 1126 Downloaded from www.archgenpsychiatry.com at University of Sydney Library, on September 2, 2008 ©2007 American Medical Association. All rights reserved. Table 1. SMRs for Schizophrenia by Cause of Death for All Persons

Quantile No. of Geometric Causes of Death SMRs 10% 25% Median 75% 90% Mean (SD) Mean All-Cause and Middle-Level Categories All-cause (ICD-9 codes 001-799/E800-E999) 38 1.18 1.87 2.58 3.64 5.76 2.98 (1.75) 2.68 All-natural cause (ICD-9 codes 001-799) 6 0.99 1.04 2.41 2.90 4.10 2.31 (1.18) 2.03 All-unnatural cause (ICD-9 codes E800-E999) 3 5.56 5.56 7.50 12.73 12.73 8.60 (3.71) 8.10 Natural Causes, Cause Specific Cardiovascular diseases (ICD-9 codes 390-429) 7 1.11 1.40 1.79 2.49 3.60 2.01 (0.83) 1.88 Cerebrovascular diseases (ICD-9 codes 430-438) 3 0.61 0.61 0.69 1.30 1.30 0.87 (0.38) 0.82 Digestive diseases (ICD-9 codes 520-579) 5 1.79 2.24 2.38 2.50 17.50 5.28 (6.84) 3.34 Endocrine diseases (ICD-9 codes 250-259) 3 2.20 2.20 2.63 11.66 11.66 5.50 (5.34) 4.07 Infectious diseases (ICD-9 codes 001-139) 3 1.60 1.60 4.29 7.80 7.80 4.56 (3.11) 3.77 Genitourinary diseases (ICD-9 codes 580-629) 3 1.54 1.54 3.70 4.29 4.29 3.18 (1.45) 2.90 Neoplastic diseases (ICD-9 codes 140-239) 7 0.71 1.00 1.37 2.01 2.40 1.44 (0.60) 1.33 Nervous diseases (ICD-9 codes 345-349) 4 1.60 1.95 4.22 6.57 7.00 4.26 (2.70) 3.55 Respiratory diseases (ICD-9 codes 460-519) 6 2.20 2.39 3.19 3.80 9.30 4.01 (2.66) 3.51 Other diseases (ICD-9 codes 1-389/630-799) 3 1.45 1.45 2.00 3.40 3.40 2.28 (1.01) 2.14 Unnatural Causes, Cause Specific Accident (ICD-9 codes E800-E949) 6 1.20 1.63 1.73 5.10 8.40 3.30 (2.88) 2.51 Suicide (ICD-9 codes E950-E959) 10 0.66 5.90 12.86 21.43 174.25 43.47 (95.11) 16.13

Abbreviations: ICD-9, International Classification of Diseases, Ninth Revision (ICD-9); SMRs, standardized mortality ratios.

Table 2. SMRs for Schizophrenia of All-Cause Mortality by Economic Development Status for All Personsa

Quantile No. of Geometric Economic Development Status SMRs 10% 25% Median 75% 90% Mean (SD) Mean Least developed countries 4 1.88 1.89 2.02 2.75 3.36 2.32 (0.70) 2.25 Emerging economy countries 6 1.04 1.31 2.19 5.98 8.43 3.52 (3.03) 2.57 Developed countries 28 1.18 1.97 2.79 3.74 5.69 2.96 (1.52) 2.77

Abbreviation: SMRs, standardized mortality ratios. a F2,34 = 0.30; P = .74. to 3-fold increase in cardiovascular mortality and a 2-fold tality associated with schizophrenia affects men and increase in all-cause mortality,84 these adverse effects women equally. would be expected to contribute to even higher SMRs in Of the specific-cause SMRs, suicide was associated with the next few decades.85,86 Unfortunately, we are unable the highest estimate: 12 times greater than expected from to explore the role of atypical medications as a contrib- the general population. In keeping with previous re- uting factor for the increasing SMRs associated with views, the SMRs associated with many different types of schizophrenia (eg, deaths related to clozapine-induced natural causes of death were elevated in people with agranulocytosis or deaths related to atypical antipsychotic- schizophrenia. Curiously, the category neoplastic disor- induced weight gain). Adverse health outcomes associ- der had one of the lowest median SMRs (1.37). Al- ated with weight gain and/or metabolic syndrome (eg, though the median was still greater than 1, several rec- myocardial infarction, cerebrovascular accidents, or can- ord linkage studies89 have suggested that cancers may be cer) may take decades to fully emerge. Thus, it seems likely significantly less prevalent in people with schizophre- that studies undertaken in the 1990s (ie, the most re- nia. The current review examines only mortality, and stud- cent studies included in this review) would capture only ies that examine morbidity would be better able to ex- a small fraction of the eventual burden of mortality as- plore this issue.90 sociated with the adverse effect profile of the second- We found no significant difference in SMRs among generation antipsychotic medications. In light of the ris- sites when sorted by economic status. However, this meta- ing secular trends in SMRs already identified by this analysis identified just 3 studies53,58,61 that provided dis- review, the prospect of further increases in mortality risks crete SMRs from the least developed and emerging for schizophrenia is alarming. economy countries; thus, caution should be exercised in In keeping with the findings of Harris and Barra- the interpretation of this finding. Furthermore, a single clough11 and Simpson,10 we found no significant sex dif- derived variable was used to define economic status, which ference in all-cause SMRs. Thus, although many well- was applied at the ecological level. documented sex differences exist in the epidemiological What factors have contributed to the differential mor- features of schizophrenia,19,87,88 the increased risk of mor- tality risk associated with schizophrenia? Many demo-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 64 (NO. 10), OCT 2007 WWW.ARCHGENPSYCHIATRY.COM 1127 Downloaded from www.archgenpsychiatry.com at University of Sydney Library, on September 2, 2008 ©2007 American Medical Association. All rights reserved. Menezes and Mann61 (Brazil) Hewer and Rossler57 (Germany) Kurihara et al58 (Indonesia) Harrison et al53 (Hong Kong) Harrison et al53 (Japan) Ruschena et al68 (Australia) Räsänen et al22 (Finland) Harrison et al53 (Ireland) Harrison et al53 (Czech Republic) Lesage et al60 (Italy) Harrison et al53 (India) Harrison et al53 (United Kingdom) Casadebaig and Philippe48 (France) Harrison et al53 (United States) Politi et al67 (Italy) Martin et al24 (United States) Brown et al47 (United Kingdom) Harrison et al53 (China) Newman and Bland65 (Canada) Osby et al66 (Sweden) Black45 (United States) Bralet et al46 (France) Tsuzuki and Yuasa71 (Japan) Harrison et al53 (India) Zilber et al73 (Israel) Morgan et al62 (Ireland) Harrison et al53 (India) Harrison et al53 (India) Wood et al72 (United States) Haugland et al55 (United States) Eastwood et al51 (Canada) Harrison et al53 (Russia) Hassell et al54 (United Kingdom) Harrison et al53 (Columbia) Mortensen and Juel63 (Denmark) Lawrence et al59 (Australia) Harrison et al53 (Bulgaria) Harrison et al53 (United States)

Summary

0 1 2 3 4 5 6 7 8 9 10 SMR

Figure 2. Forest plot of standardized mortality ratios (SMRs) for all-cause mortality in people with schizophrenia. Individual SMRs and the pooled estimate are shown with 95% confidence intervals (error bars). For each of the individual studies, the central open box symbol indicates the relative weighting on the pooled estimate (larger symbols indicate greater influence on the summary estimate). Standard errors cannot be calculated when the SMR is0; thus, these values do not contribute to the pooled value. The upper 95% confidence limits for 3 studies extend beyond 10 (indicated with right arrow symbol). See also Supplementary Table S3 at www.qcmhr.uq.edu.au/epi. graphic, clinical, political, and cultural factors mediate associated comorbid somatic conditions may be down- pathways and barriers to health care in general (eg, avail- stream expressions of common genetic or environmen- ability of services, stigma, and disease profiles).91 With tal factors.92,93 For example, it is feasible that polymor- respect to schizophrenia, the onset of the illness can re- phisms in genes may increase the susceptibility to both sult in a cascade of unhealthy lifestyle factors that el- schizophrenia and diabetes94 or that de novo germline evate the risk of various somatic diseases and conse- mutations across many generations could result in an in- quently increase the risk of death. People with creased risk of schizophrenia95 and a wide range of ad- schizophrenia are thought to be less inclined to seek health verse health outcomes. Prenatal nutritional disruptions care, to consume less medical care, to engage in high- may equally affect brain development and general meta- risk behaviors, and to be less compliant with their treat- bolic functioning.96,97 Although the current review can- ments.82,90,92 However, in addition to factors that oper- not address these issues directly, the worsening SMRs as- ate on the pathway to care, schizophrenia and its sociated with schizophrenia noted in recent decades

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 64 (NO. 10), OCT 2007 WWW.ARCHGENPSYCHIATRY.COM 1128 Downloaded from www.archgenpsychiatry.com at University of Sydney Library, on September 2, 2008 ©2007 American Medical Association. All rights reserved. suggest that this already disadvantaged group is not ben- terial is available at www.qcmhr.uq.edu.au/epi: Figure S1: efiting from the improved health of the community in Flow Diagram (Selection Strategy) of Included Studies an equitable fashion. A systematic approach to monitor- in the Mortality of Schizophrenia; Table S2: Quality Re- ing and treating the physical health needs of people with porting Scale; Table S3: Summary Table of All-Cause Mor- schizophrenia is clearly warranted.98 tality and Standardized Mortality Ratio for Schizophre- Several important caveats to this review should be nia (1980-2006); Table S4: Standardized Mortality Ratios noted. Publication bias is always an issue in systematic (SMRs) for Schizophrenia by Different Causes of Death reviews. We endeavored to address this by obtaining data for Males and Females; Table S5: Standardized Mortal- from all available sources, including those from elec- ity Ratios for 3 Quality Score Tertiles of All-Cause Death; tronic databases, citations and authors, and publica- Table S6: Standardized Mortality Ratios for Schizophre- tions in languages other than English. Factors such as nia of All-Cause Mortality for Various Post Hoc Analy- the reliability of psychiatric diagnoses and admission prac- ses (for All Persons); and Microsoft Excel spreadsheet of tices (between sites and across time) could contribute to the primary data for this systematic review, plus associ- the variability identified in this systematic review. The ated labels and formats. reliability of the categorization of cause of death is also Additional Contributions: Dozens of researchers from a cause for concern. With respect to specific-cause mor- around the world assisted in locating the data for this sys- tality, changes in the coding rules for the ICD-9 and be- tematic review, and the staff of the Queensland Centre tween-site variability in the application of these rules also for Mental Health Research assisted in extracting the data need to be taken into account.99,100 However, these is- and preparing the original manuscript. sues do not affect all-cause SMRs (which were used for the main analyses in this review). The current study found a higher all-cause SMR (median SMR, 2.58; pooled meta- REFERENCES analysis SMR, 2.50) compared with the 2 previous re- 4 11 1. Murray CJ, Lopez AD. Global mortality, disability, and the contribution of risk views, which reported all-cause SMRs of 1.51 and 1.57. factors: Global Burden of Disease Study. Lancet. 1997;349(9063):1436- The 2 previous systematic reviews were based on stud- 1442. ies published before 199511 and 19964 compared with the 2. Murray CJ, Lopez AD. Mortality by cause for eight regions of the world: Global current systematic review, which included 18 addi- Burden of Disease Study. Lancet. 1997;349(9061):1269-1276. tional studies published after 1995. 3. Palmer BA, Pankratz VS, Bostwick JM. The lifetime risk of suicide in schizo- phrenia: a reexamination. Arch Gen Psychiatry. 2005;62(3):247-253. In conclusion, compared with the general popula- 4. Brown S. Excess mortality of schizophrenia: a meta-analysis. Br J Psychiatry. tion, people with schizophrenia have a 2- to 3-fold in- 1997;171:502-508. creased risk of dying. Suicide contributes to the in- 5. Graunt J. Natural and political observations mentioned in a following index, and creased mortality associated with schizophrenia; however, made upon the bills of mortality. In: Hunter RA, MacAlpine I, eds. Three Hun- dred Years of Psychiatry 1538-1860. London, England: Oxford University Press; people with schizophrenia have increased mortality risks 1963:1662-1666. attributable to a wide range of somatic conditions. The 6. Kraepelin E. Dementia Praecox and Paraphrenia. Edinburgh, Scotland: Living- increased mortality risk affects both sexes equally. Sub- stone; 1919. stantial variation occurs in all-cause SMRs among sites. 7. Alstrom CH. Mortality in mental hospitals with special regard to tuberculosis. In recent decades, the differential mortality gap associ- Acta Psychiatr Neurol Scand Suppl. 1942:24. 8. Bleuler E. Dementia Paecox or the Group of Schizophrenias. New York, NY: In- ated with schizophrenia has been increasing. It is ternational University Press; 1950. sobering to reflect on this paradox of schizophrenia 9. Allebeck P. Schizophrenia: a life-shortening disease. Schizophr Bull. 1989;15(1): treatment. As we become better at detecting and treat- 81-89. ing the core symptoms of schizophrenia, patients have 10. Simpson JC. Mortality studies in schizophrenia. In: Tsuang MT, Simpson JC, eds. Handbook of Schizophrenia, Volume 3: Nosology, Epidemiology and Ge- worsening SMRs. Given the potential for an even netics of Schizophrenia. Amsterdam, the Netherlands: Elsevier; 1988:245- greater disease burden as a result of the introduction 274. of second-generation antipsychotic medications, 11. Harris EC, Barraclough B. Excess mortality of mental disorder. Br J Psychiatry. research aimed at optimizing the physical health of 1998;173:11-53. people with schizophrenia needs to be undertaken 12. Dube KC, Kumar N, Dube S. Long term course and outcome of the Agra cases in the International Pilot Study of Schizophrenia. Acta Psychiatr Scand. 1984; with a sense of urgency. 70(2):170-179. 13. Brook OH. Mortality in the long-stay population of Dutch mental hospitals. Acta Submitted for Publication: November 4, 2006; final re- Psychiatr Scand. 1985;71(6):626-635. vision received January 16, 2007; accepted March 12, 14. Murray CJ, Lopez AD, Jamison DT. The global burden of disease in 1990: sum- mary results, sensitivity analysis and future directions. Bull World Health Organ. 2007. 1994;72(3):495-509. Correspondence: John McGrath, MD, PhD, FRANZCP, 15. Department of Health and Ageing. National Mental Health Report 2005: Sum- Queensland Centre for Mental Health Research, The Park mary of Ten Years of Reform in Australia’s Mental Health Services Under Na- Centre for Mental Health, Wacol Q4076, Australia (john tional Mental Health Strategy 1993-2003. Canberra: Commonwealth of Austra- [email protected]). lia; 2005. 16. World Health Organization. The World Health Report 2001: Mental Health: Author Contributions: Mr Saha has full access to all of New Understanding, New Hope. Geneva, Switzerland: World Health Organi- the data in the study and takes responsibility for the in- zation; 2001. tegrity of the data. 17. Osby U, Correia N, Brandt L, Ekbom A, Sparen P. Time trends in schizophrenia Financial Disclosure: None reported. mortality in Stockholm County, Sweden: cohort study. BMJ. 2000;321(7259): 483-484. Funding/Support: The Stanley Medical Research Insti- 18. Heila¨ H, Haukka J, Suvisaari J, Lonnqvist J. Mortality among patients with schizo- tute supported this project. phrenia and reduced psychiatric hospital care. Psychol Med. 2005;35(5): Additional Information: The following additional ma- 725-732.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 64 (NO. 10), OCT 2007 WWW.ARCHGENPSYCHIATRY.COM 1129 Downloaded from www.archgenpsychiatry.com at University of Sydney Library, on September 2, 2008 ©2007 American Medical Association. All rights reserved. 19. McGrath J, Saha S, Welham J, El Saadi O, MacCauley C, Chant D. A systematic parison with general population mortality. Psychosomatics. 1988;29(1): review of the incidence of schizophrenia: the distribution of rates and the in- 55-60. fluence of sex, urbanicity, migrant status and methodology. BMC Med. 2004; 46. Bralet MC, Yon V, Loas G, Noisette C. Mortality in schizophrenia: a 8-year fol- 2(1):13. low-up study in 150 chronic schizophrenics. Encephale. 2000;26(6):32-41. 20. Saha S, Chant D, Welham J, McGrath J. A systematic review of the prevalence 47. Brown S, Inskip H, Barraclough B. Causes of the excess mortality of schizophrenia. of schizophrenia. PLoS Med. 2005;2(5):e141. Br J Psychiatry. 2000;177:212-217. 21. Greenland S. Meta-analysis. In: Rothman KJ, Greenland S, eds. Modern Epi- 48. Casadebaig F, Philippe A. Mortality among schizophrenic patients. Encephale. demiology. 2nd ed. Sydney, Australia: Lippincott Williams & Wilkins; 1998: 1999;25(4):329-337. 643-674. 49. Chen WJ, Huang YJ, Yeh LL, Rin H, Hwu HG. Excess mortality of psychiatric 22. Ra¨sa¨nen S, Hakko H, Viilo K, Meyer-Rochow VB, Moring J. Excess mortality inpatients in Taiwan. Psychiatry Res. 1996;62(3):239-250. among long-stay psychiatric patients in Northern Finland. Soc Psychiatry Psy- 50. Curkendall SM, Mo J, Glasser DB, Rose Stang M, Jones JK. Cardiovascular dis- chiatr Epidemiol. 2003;38(6):297-304. ease in patients with schizophrenia in Saskatchewan, Canada. J Clin Psychiatry. 23. Joukamaa M, Heliovaara M, Knekt P, Aromaa A, Raitasalo R, Lehtinen V. Men- 2004;65(5):715-720. tal disorders and cause-specific mortality. Br J Psychiatry. 2001;179:498- 51. Eastwood MR, Stiasny S, Meier HM, Woogh CM. Mental illness and mortality. 502. Compr Psychiatry. 1982;23(4):377-385. 24. Martin RL, Cloninger CR, Guze SB, Clayton PJ. Mortality in a follow-up of 500 52. Hansen V, Arnesen E, Jacobsen BK. Total mortality in people admitted to a psy- psychiatric outpatients 1: total mortality. Arch Gen Psychiatry. 1985;42(1): chiatric hospital. Br J Psychiatry. 1997;170:186-190. 47-54. 53. Harrison G, Hopper K, Craig T, Laska E, Siegel C, Wanderling J, Dube KC, Ganev 25. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, Moher D, K, Giel R, an der Heiden W, Holmberg SK, Janca A, Lee PW, Leo´n CA, Malhotra Becker BJ, Sipe TA, Thacker SB; Meta-analysis Of Observational Studies in Epi- S, Marsella AJ, Nakane Y, Sartorius N, Shen Y, Skoda C, Thara R, Tsirkin SJ, demiology (MOOSE) Group. Meta-analysis of observational studies in epide- Varma VK, Walsh D, Wiersma D. Recovery from psychotic illness: a 15- and miology: a proposal for reporting. JAMA. 2000;283(15):2008-2012. 25-year international follow-up study. Br J Psychiatry. 2001;178:506-517. 26. Babidge NC, Buhrich N, Butler T. Mortality among homeless people with schizo- 54. Hassall C, Prior P, Cross KW. A preliminary study of excess mortality using a phrenia in Sydney, Australia: a 10-year follow-up. Acta Psychiatr Scand. 2001; psychiatric case register. J Epidemiol Community Health. 1988;42(3):286- 103(2):105-110. 289. 27. Kendler KS. A twin study of mortality in schizophrenia and neurosis. Arch Gen 55. Haugland G,CraigTJ,GoodmanAB,SiegelC.Mortalityintheeraofdeinstitutionali- Psychiatry. 1986;43(7):643-649. zation. Am J Psychiatry. 1983;140(7):848-852. 28. National Centre for Health Statistics. Mortality data from the National Vital Sta- 56. Herrman HE, Baldwin JA, Christie D. A record-linkage study of mortality and tistics System. In: International Classification of Diseases, Ninth Revision (ICD- general hospital discharge in patients diagnosed as schizophrenic. Psychol Med. 9). http://www.cdc.gov/nchs/about/major/dvs/icd9des.htm. Accessed Janu- 1983;13(3):581-593. ary 12, 2004. 57. Hewer W, Rossler W. Mortality of patients with functional psychiatric illnesses 29. Saha S, Welham J, Chant D, McGrath J. Incidence of schizophrenia does not during inpatient treatment [in German]. Fortschr Neurol Psychiatr. 1997; vary with economic status of the country: evidence from a systematic review. 65(4):171-181. Soc Psychiatry Psychiatr Epidemiol. 2006;41(5):338-340. 58. Kurihara T, Kato M, Kashima H, Takebayashi T, Reverger R, Tirta IGR. Excess 30. Central Intelligence Agency. CIA World Factbook 2004. Washington, DC: Cen- mortality of schizophrenia in the developing country of Bali. Schizophr Res. 2006; tral Intelligence Agency, Office of Public Affairs; 2004. http://www.cia.gov/cia 83(1):103-105. /publications/factbook/. Accessed January 12, 2004. 59. Lawrence D, Jablensky AV, Holman CDJ, Pinder TJ. Mortality in Western Aus- tralian psychiatric patients. Soc Psychiatry Psychiatr Epidemiol. 2000;35 31. Soubbotina TP. Beyond Economic Growth: An Introduction to Sustainable (8):341-347. Development. 2nd ed. Washington, DC: World Bank; 2004. 60. LesageAD,TrapaniV,TansellaM.ExcessmortalitybynaturalcausesofItalianschizo- 32. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in phrenic patients. Eur Arch Psychiatry Neurol Sci. 1990;239(6):361-365. meta-analyses. BMJ. 2003;327(7414):557-560. 61. Menezes PR, Mann AH. Mortality among patients with non-affective functional 33. Allebeck P, Wistedt B. Mortality in schizophrenia: a ten-year follow-up based psychoses in a metropolitan area of south-eastern Brazil. Rev Saude Publica. on the Stockholm County inpatient register. Arch Gen Psychiatry. 1986; 1996;30(4):304-309. 43(7):650-653. 62. Morgan MG, Scully PJ, Youssef HA, Kinsella A, Owens JM, Waddington JL. 34. Amaddeo F, Bisoffi G, Bonizzato P, Micciolo R, Tansella M. Mortality among Prospective analysis of premature mortality in schizophrenia in relation to health patients with psychiatric illness: a ten-year case register study in an area with service engagement: a 7.5-year study within an epidemiologically complete, ho- a community-based system of care. Br J Psychiatry. 1995;166(6):783-788. mogeneous population in rural Ireland. Psychiatry Res. 2003;117(2):127-135. 35. Baxter DN. The mortality experience of individuals on the Salford Psychiatric 63. Mortensen PB, Juel K. Mortality and causes of death in schizophrenic patients Case Register, I: all-cause mortality. Br J Psychiatry. 1996;168(6):772-779. in Denmark. Acta Psychiatr Scand. 1990;81(4):372-377. 36. Black DW, Fisher R. Mortality in DSM-IIIR schizophrenia. Schizophr Res. 1992; 64. Mortensen PB, Juel K. Mortality and causes of death in first admitted schizo- 7(2):109-116. phrenic patients. Br J Psychiatry. 1993;163:183-189. 37. Buda M, Tsuang MT, Fleming JA. Causes of death in DSM-III schizophrenics 65. Newman SC, Bland RC. Mortality in a cohort of patients with schizophrenia: a and other psychotics (atypical group): a comparison with the general population. record linkage study. Can J Psychiatry. 1991;36(4):239-245. Arch Gen Psychiatry. 1988;45(3):283-285. 66. Osby U, Correia N, Brandt L, Ekbom A, Sparen P. Mortality and causes of death 38. D’Avanzo B, La Vecchia C, Negri E. Mortality in long-stay patients from psy- in schizophrenia in Stockholm County, Sweden. Schizophr Res. 2000;45 chiatric hospitals in Italy: results from the Qualyop Project. Soc Psychiatry Psy- (1-2):21-28. chiatr Epidemiol. 2003;38(7):385-389. 67. Politi P, Piccinelli M, Klersy C, Madini S, Segagni LG, Fratti C, Barale F. Mor- 39. Montout C, Casadebaig F, Lagnaoui R, Verdoux H, Philippe A, Begaud B, Moore tality in psychiatric patients 5 to 21 years after hospital admission in Italy. Psy- N. Neuroleptics and mortality in schizophrenia: prospective analysis of deaths chol Med. 2002;32(2):227-237. in a French cohort of schizophrenic patients. Schizophr Res. 2002;57(2-3): 68. Ruschena D, Mullen PE, Burgess P, Cordner SM, Barry-Walsh J, Drummer OH, 147-156. Palmer S, Browne C, Wallace C. Sudden death in psychiatric patients. Br J 40. Ringba¨ck Weitoft G, Gullberg A, Rosen M. Avoidable mortality among psychi- Psychiatry. 1998;172:331-336. atric patients. Soc Psychiatry Psychiatr Epidemiol. 1998;33(9):430-437. 69. Saku M, Tokudome S, Ikeda M, Kono S, Makimoto K, Uchimura H, Mukai A, 41. Simpson JC, Tsuang MT. Mortality among patients with schizophrenia. Schizophr Yoshimura T. Mortality in psychiatric patients, with a specific focus on cancer Bull. 1996;22(3):485-499. mortality associated with schizophrenia. Int J Epidemiol. 1995;24(2):366-372. 42. Valenti M, Necozione S, Busellu G, Borrelli G, Lepore AR, Madonna R, Altobelli 70. Tsuang MT, Woolson RF, Fleming JA. Causes of death in schizophrenia and E, Mattei A, Torchio P, Corrao G, Di Orio F. Mortality in psychiatric hospital pa- manic-depression. Br J Psychiatry. 1980;136:239-242. tients: a cohort analysis of prognostic factors. Int J Epidemiol. 1997;26(6): 71. Tsuzuki H, Yuasa S. An epidemiological study of deaths in mental hospitals [au- 1227-1235. thor’s transl] [in Japanese]. Seishin Shinkeigaku Zasshi. 1981;83(5):275- 43. Waddington JL, Youssef HA, Kinsella A. Mortality in schizophrenia: antipsy- 304. chotic polypharmacy and absence of adjunctive anticholinergics over the course 72. Wood JB, Evenson RC, Cho DW, Hagan BJ. Mortality variations among public of a 10-year prospective study. Br J Psychiatry. 1998;173:325-329. mental health patients. Acta Psychiatr Scand. 1985;72(3):218-229. 44. Anderson C, Connelly J, Johnstone EC, Owens DGC. Cause of death. Br J Psy- 73. Zilber N, Schufman N, Lerner Y. Mortality among psychiatric patients–the groups chiatry Suppl. 1991;13(13)(suppl 13):30-33. at risk. Acta Psychiatr Scand. 1989;79(3):248-256. 45. Black DW. Mortality in schizophrenia: the Iowa Record-Linkage Study: a com- 74. Wamala S, Blakely T, Atkinson J. Trends in absolute socioeconomic inequali-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 64 (NO. 10), OCT 2007 WWW.ARCHGENPSYCHIATRY.COM 1130 Downloaded from www.archgenpsychiatry.com at University of Sydney Library, on September 2, 2008 ©2007 American Medical Association. All rights reserved. ties in mortality in Sweden and New Zealand: a 20-year gender perspective. BMC 87. Aleman A, Kahn RS, Selten JP. Sex differences in the risk of schizophrenia: evi- Public Health. 2006;6:164. dence from meta-analysis. Arch Gen Psychiatry. 2003;60(6):565-571. 75. Murray CJ, Lopez AD. Alternative projections of mortality and disability by cause 88. Castle DJ. Sex differences in brain development, organisation and degenera- 1990-2020: Global Burden of Disease Study. Lancet. 1997;349(9064):1498- tion: are they relevant to sex differences in schizophrenia? In: Castle DJ, McGrath 1504. J, Kulkarni J, eds. Women and Schizophrenia. Cambridge, MA: Cambridge Uni- 76. Jemal A, Ward E, Hao Y, Thun M. Trends in the leading causes of death in the versity Press; 2000:5-18. United States, 1970-2002. JAMA. 2005;294(10):1255-1259. 89. Cohen M, Dembling B, Schorling J. The association between schizophrenia and 77. Leucht S, Barnes TR, Kissling W, Engel RR, Correll C, Kane JM. Relapse pre- cancer: a population-based mortality study. Schizophr Res. 2002;57(2-3): vention in schizophrenia with new-generation antipsychotics: a systematic re- 139-146. view and exploratory meta-analysis of randomized, controlled trials. Am J 90. Coghlan R, Lawrence D, Jablensky A. Duty to Care: Physical Illness in People Psychiatry. 2003;160(7):1209-1222. with Mental Illness. Perth: The University of Western Australia; 2001. 78. Leucht S, Wahlbeck K, Hamann J, Kissling W. New generation antipsychotics 91. Murray CJ, Lopez AD. Global and regional cause-of-death patterns in 1990. Bull versus low-potency conventional antipsychotics: a systematic review and World Health Organ. 1994;72(3):447-480. meta-analysis. Lancet. 2003;361(9369):1581-1589. 92. Mitchell AJ, Malone D. Physical health and schizophrenia. Curr Opin Psychiatry. 79. Awad AG, Voruganti LN. Impact of atypical antipsychotics on quality of life in 2006;19(4):432-437. patients with schizophrenia. CNS Drugs. 2004;18(13):877-893. 93. Eaton WW, Byrne M, Ewald H, Mors O, Chen CY, Agerbo E, Mortensen PB. 80. Lieberman JA, Stroup TS, McEvoy JP, Swartz MS, Rosenheck RA, Perkins DO, Association of schizophrenia and autoimmune diseases: linkage of Danish na- Keefe RS, Davis SM, Davis CE, Lebowitz BD, Severe J, Hsiao JK. Effectiveness tional registers. Am J Psychiatry. 2006;163(3):521-528. of antipsychotic drugs in patients with chronic schizophrenia. N Engl J Med. 94. Gough SC, O’Donovan MC. Clustering of metabolic comorbidity in schizophre- 2005;353(12):1209-1223. nia: a genetic contribution? J Psychopharmacol. 2005;19(6)(suppl):47-55. 81. Jones PB, Barnes TR, Davies L, Dunn G, Lloyd H, Hayhurst KP, Murray RM, 95. McClellan JM, Susser E, King MC. Maternal famine, de novo mutations, and Markwick A, Lewis SW. Randomized controlled trial of the effect on quality of schizophrenia. JAMA. 2006;296(5):582-584. life of second- vs first-generation antipsychotic drugs in schizophrenia: Cost 96. Barker DJ. Maternal nutrition, fetal nutrition, and disease in later life. Nutrition. Utility of the Latest Antipsychotic Drugs in Schizophrenia Study (CUtLASS 1). 1997;13(9):807-813. Arch Gen Psychiatry. 2006;63(10):1079-1087. 97. Barker DJ, Clark PM. Fetal undernutrition and disease in later life. Rev Reprod. 82. Hennekens CH, Hennekens AR, Hollar D, Casey DE. Schizophrenia and increased 1997;2(2):105-112. risks of cardiovascular disease. Am Heart J. 2005;150(6):1115-1121. 98. Marder SR, Essock SM, Miller AL, Buchanan RW, Casey DE, Davis JM, Kane JM, 83. Remington G. Schizophrenia, antipsychotics, and the metabolic syndrome: is Lieberman JA, Schooler NR, Covell N, Stroup S, Weissman EM, Wirshing DA, there a silver lining? Am J Psychiatry. 2006;163(7):1132-1134. Hall CS, Pogach L, Pi-Sunyer X, Bigger JT Jr, Friedman A, Kleinberg D, Yevich 84. Lakka HM, Laaksonen DE, Lakka TA, Niskanen LK, Kumpusalo E, Tuomilehto J, SJ, Davis B, Shon S. Physical health monitoring of patients with schizophrenia. Salonen JT. The metabolic syndrome and total and cardiovascular disease mor- Am J Psychiatry. 2004;161(8):1334-1349. tality in middle-aged men. JAMA. 2002;288(21):2709-2716. 99. World Health Organization. Manual of the Statistical International Classifica- 85. Allison DB, Casey DE. Antipsychotic-induced weight gain: a review of the literature. tion of Diseases, Injuries, and Causes of Death, 1975 Revision. Vol 1. Geneva, J Clin Psychiatry. 2001;62(Suppl 7):22-31. Switzerland: World Heath Organization; 1977. 86. Fontaine KR, Heo M, Harrigan EP, Shear CL, Lakshminarayanan M, Casey DE, 100. Goldacre MJ, Duncan ME, Griffith M, Cook-Mozaffari P. Psychiatric disorders Allison DB. Estimating the consequences of anti-psychotic induced weight gain certified on death certificates in an English population. Soc Psychiatry Psychi- on health and mortality rate. Psychiatry Res. 2001;101(3):277-288. atr Epidemiol. 2006;41(5):409-414.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 64 (NO. 10), OCT 2007 WWW.ARCHGENPSYCHIATRY.COM 1131 Downloaded from www.archgenpsychiatry.com at University of Sydney Library, on September 2, 2008 ©2007 American Medical Association. All rights reserved. Editorial For reprint orders, please contact: [email protected] From personalized medicine to personalized justice: the promises of translational pharmacogenomics in the justice system

“Personalized justice complements personalized medicine and the overlapping practice of translational medicine, which holds that individual differences are caused primarily by genetic and environmental factors.”

Enabled by pharmacogenomics (PGx), mole­ syndrome, was treated with methyl­phenidate, cular imaging and other molecular biomarkers, clonidine and fluoxetine [10]. Over a 10‑month personalized medicine (PM) promises to opti‑ period, he developed gastrointestinal toxicity, Steven HY Wong†, mize therapy while minimizing side effects. It incoordination and disorientation, and sei‑ Christopher Happy, may also dramatically impact the justice system zures. He died from a cardiac arrest. Post‑ Dan Blinka, Susan Gock, in ways we are only beginning to understand. mortem toxicology showed high fluoxetine Jeffrey M Jentzen, Personalized medicine has already entered the and norfluoxetine concentrations, and PGx Hon. Joseph Donald, curricula of well-regarded medical schools such as revealed a poor CYP2D6 metabolizer geno‑ Howard Coleman, type, resulting in fluoxetine accumulation that of Johns Hopkins University (MD, USA) [1], Saeed A Jortani, but law schools offer no analog. Although clinical and toxicity. Subsequently, the boy’s parent acceptance of PM has proved slow even with US were absolved from involvement in fluoxetine Yolande Lucire, Cynthia FDA support [2,101], PM’s legal ramifications are intoxication. Another example is genotyping L Morris-Kukoski, evident. Recently, for example, the FDA relabeled uridine 5´-diphosphate-glucuronyltransferase Manuela G Neuman, some drugs with companion PGx [2], such as war‑ 1A1 for patients medicated with irinotecan Paul J Orsulak, farin withCYP2C9 and vitamin K epoxide reduct‑ to avoid hematopoietic toxicity [11]; Tara Sander, ase complex 1 to reduce bleeding [3,4,101]. If PGx ƒƒDrug sensitivity: in addition to warfarin, one Michael A Wagner, retrospectively reveals that the warfarin patient should genotype HLA-B*5701 [12] and Jennifer R Wynn, was at high risk and testing was not initially per‑ HLA-B*1502 [13] for patients medicated with Alan HB Wu & formed, litigation may follow. Indeed, some law‑ abacavir [12] and some antiepileptics [13], Kiang-Teck J Yeo yers advertise on the internet for cases involving respectively, to avoid Stevens–Johnson syn‑ †Author for correspondence: warfarin-related errors [102]. Consequently, PGx drome. Lawyers use internet advertising to Medical College of Wisconsin, may become part of defensive medicine. Milwaukee, WI 53226-0509, USA reach persons who may be affected [103]. Tel.: +1 414 805 6971 Personalized justice (PJ) complements PM Fax: +1 414 805 6980 and the overlapping practice of translational In establishing PJ, a firm foundation should [email protected] For a full list of affiliations, please see [5–8] medicine , which holds that individual dif‑ be based on sound legal principles as well as the back of this article. ferences are caused primarily by genetic and reliable and valid evidence-based studies, not environmental factors. The acronym ‘TSPB’ on ‘junk’ science and unsubstantiated case captures its elements in relation to adverse drug reports. This lesson resonates in the defi‑ reactions (ADRs): toxicity, sensitivity, impaired ciencies that beset various forensic sciences performance (e.g., driving under the influence recently reported by the National Academy of drugs) and behavioral changes. Future legal of Science [14,104]. The American Academy applications may include molecular imaging and of Forensic Sciences supports the National analyses – genomic, proteomic, metabolomics Academy of Science’s 13 recommendations and and epigenetics/imprintomics. By comparison, the following principles: the need for strong molecular DNA fingerprinting for identity testing scientific foundations; laboratory accredita‑ is well accepted [9]. Conceptually, Figure 1 proposes tion; certification of technicians; the stand‑ a social-balance relationship for PM and PJ [8]. ardization of terminology; ethical protocols; In assessing PJ, consider two index scenarios: governmental oversight; and the education of legal professionals, including judges, in foren‑ ƒƒDrug toxicity: a 9‑year old boy, diagnosed sic scientific methods and principles [15,16,105]. with attention-deficit hyperactivity disorder, It is imperative that PJ heeds these recommen‑ obsessive–compulsive disorder, and Tourette’s dations, including the study of the relationship

10.2217/PGS.10.63 © 2010 Future Medicine Ltd Pharmacogenomics (2010) 11(6), 731–737 ISSN 1462-2416 731 Editorial Wong, Happy, Blinka et al.

has become the ‘gold standard’ for forensic sci‑ Inevitable social balance? ences. Thus, the DNA channel may provide a Check and balance helpful port of entry for PJ.

“It will be important then to educate Personalized Personalized judges, lawyers and legal academics about medicine justice the explanatory power of personalized justice and personalized medicine.” Toxicity PGx Efficacy Sensitivity Behavior The prime questions though will relate Performance to the role PJ will play in the legal system. DNA evidence thus far is narrowly confined to trace evidence: was this biological evi‑ dence left by the defendant or someone else? A thornier problem occurs when we attempt to apply biological evidence to moral culpa‑ bility, which pertains to an accused’s personal blameworthiness. The Supreme Court recog‑ Figure 1. Complementary relationship of personalized medicine and nized in Penry v. Lynaugh that punishment for personalized justice. a criminal offense should be directly related to Reproduced and modified with permission from [8]. the defendant’s personal culpability [21]. The of PGx biomarkers to TSPB and the educa‑ concept of personal culpability acknowledges tion of interested parties, including forensic that human choices are shaped by many fac‑ pathologists and toxicologists, those engaged tors: genetic, neurological, intellectual, edu‑ in molecular diagnostics and, of course, the cational, social and environmental. It follows legal community. Based on the aforementioned then that an individual’s blameworthiness for assessment, this article ushers in the practice criminal conduct may vary depending on the of PJ by differentiating between science and factors that shaped his moral development or myth, proposing a legal framework, updat‑ compromised his choices. ing the reader on rapidly developing techno­ logical advances, and illustrating scenarios and “...should courts consider identifiable published cases. biological conditions that predispose a person to criminal behavior in weighing Legal framework moral culpability? Legal precedent suggests While personalized medicine is rapidly tak‑ that it should.” ing root among the medical sciences, one may reasonably expect a slower, more begrudging Thus, from a PJ perspective, the question acceptance by the legal profession. Law is becomes something like this: should courts innately conservative and reluctant to accom‑ consider identifiable biological conditions modate dramatic change. ‘Cutting-edge’ that predispose a person to criminal behavior developments of all sorts often take decades in weighing moral culpability? Legal prec‑ to gain a foot­hold [17]. It will be important edent suggests that it should. Consider Roper then to educate judges, lawyers and legal aca‑ v. Simmons, where the Supreme Court held demics about the explanatory power of PJ and that persons under the age of 18 years could PM. The law’s incredibly rich experience with not be subjected to the death penalty because DNA developments may, however, facilitate their brains were not yet fully developed [22]. this task [18]. MRIs and neuroimaging showed that neuro‑ One set of barriers consists of evidence rules, nal changes in the brain continued to develop particularly those involving expert opinion into the early twenties. Since the brain affects testimony. The vaunted ‘Daubert standard’, behavior, the justices ruled that punishing a pioneered by the federal courts and adopted person for behavior caused by an underdevel‑ by many states, demands that judges serve as oped brain (of which the defendant had no gatekeepers who will ensure that only ‘reliable’ choice) violated the prohibition against cruel science is admitted [19]. Although courts have and un­usual punishment. Similar logic was been distressingly inconsistent in how they applied in Atkins v. Virginia, which prohibited scrutinize most sciences [20], DNA evidence subjecting the mentally retarded defendants to

732 Pharmacogenomics (2010) 11(6) future science group Translational pharmacogenomics in the justice system Editorial

the death penalty [23]. Roper and Atkins illus‑ chip (Affymetrix, CA, USA), Autogenomics trate the principle that criminal defendants INFINITI™ Analyzer (Autogenomics, CA, with brain-based deficits are not as morally cul‑ USA), Osmetech eSensors® (WA, USA), pable as those without. As such, they deserve a ParagonDx (NC, USA), and ABI PRISM® lesser penalty. This is a legal springboard for PJ. SNaPshot™ (Applied Biosystems, CA, USA) and TaqMan® assays (Applied Biosystems). Forensic pathology perspectives Thus, the laboratory can rapidly develop, vali‑ For several medical examiner/coroner offices date and perform PGx testing in-house within in the USA and Europe, PGx has served as an months, further enhanced by readily available adjunct for drug death certification – an emerg‑ quality-control products and survey programs. ing practice of molecular autopsy [7,8]. Previous The limitations include existing evidence to studies showed a higher prevalence of CYP2D6 demonstrate significant and medically rel‑ genetic variations, corresponding to inter­ evant correlations for many disease-causing mediate and slow metabolizers with decreased genes and variants, limited detection of genetic or without enzymatic activity, in the decedents variants within the context of each testing intoxicated with methadone, oxycodone and platform, clinical interpretation of genotype antidepressants [7,8,24,25]. Thus, PGx might aid results including environmental factors, and in the interpretation of the effect of impaired transplanted organs interfering with testing. drug metabolism due to genetic variations. If potentially lethal medications are identified at “...in the USA and Europe, the scene of a crime with correspondingly toxic pharmacogenomics has served as an drug concentrations of the decedent, and subse‑ adjunct for drug death certification – an quent PGx testing confirms an ‘extensive’ (nor‑ emerging practice of molecular autopsy.” mal) metabolizer, death is certified as a suicide. If the deceased’s genotype is a variant – result‑ ing in decreased drug metabolism – death is cer‑ „„Drug hypersensitivity tified as accident. Recent indications of PJ for in vitro diagnostics forensic pathology include a PGx section in the In vitro lymphocyte toxicity assays (LTAs) forensic toxicology texts for medical examiners compares the peripheral blood lymphocytes of by Molina [26] and Karch [27]. Future molecular patients with a history of an ADR with control diagnostic biomarkers of interest might include individuals who take the same drug in the same epigenetics/imprintomics and gene expression dose and do not present any ADR [31]. LTA is in understanding suicide [28,29], metabolomics based upon the dysfunction of mitochondria and proteomics [8]. in people hypersensitive to certain drugs, such as sulfonamides, NSAIDs, protease inhibitors “...criminal defendants with brain-based and antiepileptics. This test can also detect pos‑ deficits are not as morally culpable as those sible drug–drug interactions. Dysfunction of without. As such, they deserve a lesser mitochondria has severe cellular consequences penalty. This is a legal springboard for and is linked to a lack of detoxification of drugs personalized justice.” in humans. Several surveillance strategies have evolved that limit mitochondrial damage and ensure cellular integrity. Intra-organellar pro‑ Molecular diagnostics teases conduct protein quality control and exert The detection of individual genetic variants is regulatory functions, allowing the mitochon‑ at the heart of PM and PJ. SNPs, the most dria to protect against apoptosis. LTA can be common type of genetic variation, might affect used in PJ when several drugs are implicated drug metabolism [30]. Several SNP genotyp‑ in an ADR in order to enable a distinction ing technologies facilitate rapid PGx testing between the drug that produced the reaction in clinical laboratories. The three main steps, and the other drugs, which were taken in the DNA extraction, amplification and detection, same period of time but did not contribute to may be performed by automated platforms. the ADR. Biotechnology companies offering PGx testing platforms, some with FDA approval, include: Illustrative cases & scenarios Luminex xTag ® (Luminex Corporation, „„Alcohol TK, USA), Roche AmpliChip® (Roche, Alcoholism, with up to 30–40% inherit‑ Basel, Switzerland), Affymetrix DMET® ability, is a complex and controversial disease

future science group www.futuremedicine.com 733 Editorial Wong, Happy, Blinka et al.

with both environmental and genetic compo‑ „„Warfarin nents. Genetic variations influence alcohol’s Oral warfarin anticoagulation is widely used pharmaco­kinetics/metabolism and pharmaco­ to prevent thromboembolic events. Dosing dynamics. Alcohol dehydrogenase and acetal‑ selection is due to a narrow therapeutic range dehyde dehydrogenase are two main polymor‑ with a large interindividual variation (20-fold) phic enzymes involved in alcohol metabolism, affected by genetic and nongenetic factors [35,36]. with a minor contribution from CYP2E1. Approximately 10–17% of patients experience Pharmacodynamic systems influenced by PGx bleeding [35]. Genotyping of CYP2C9, CYP4F2, are: GABA A/B receptors, glutamate (NMDA VKORC1 and relevant clinical factors account and a-amino-3-hydroxyl-5-methyl-isoxazole- for up to approximately 56% of dosing vari‑ propionate), serotonin, voltage-activated cal‑ ability [37,38]. In 2007, the FDA relabeled war‑ cium channels, dopamine/norepinepherine/ farin to suggest genotyping, followed by the acetylcholine and opioid systems. For exam‑ 22 January 2010 relabeling [4]: ple, naltrexone, used for detoxification, binds “The patient’s CYP2C9 and VKORC1 geno- to opioid receptor µ1, and the variants of the type information, when available, can assist in candidate gene of this receptor may affect selection of the starting dose.” addiction treatment [32,33]. Previously, in May 2009, the Centers for „„Antidepressants & antipsychotics Medicare and Medicaid Services recommended Personalized justice might address the effect against reimbursement [106]. Potential legal of antidepressant and/or antipsychotics on culpability was addressed in the introduction. behavioral changes. A recent review examined the relationship of violent behavior to the anti‑ „„Pain management depressants: paroxetine, sertraline and fluox­ In addressing pain management with safety, etine. Different verdicts in a series of medi‑ Dr Woodcock of the FDA discussed the balance colegal cases reflected the different judicial of providing patients with efficacious analge‑ processes, without considering drug-induced sics and the associated risks [39]. For example, violence [34]. Incidentally, co-author Lucire in ultrarapid metabolizers, greater CYP2D6 studied patients medicated with antidepres‑ activity can lead to poisoning after opioid sants and antipsychotics metabolized by poly‑ administration. A 2007 case report detailed a morphic CYPs, and assessed the development breast-feeding infant who suffered respiratory of akathisic, suicidal and/or homicidal idea‑ depression and died as a result of toxic amounts tions, and their relationship to CYP gene vari‑ of morphine being present in the milk [40]. The ations [Lucire Y, Pers. Comm.]. The validity of these mother, later identified with multiple copies preliminary observations are pending publica‑ of CYP2D6 genes corresponding to an ultra­ tion in peer-reviewed journals and ­validation rapid metabolizer, ‘overconverted’ codeine to a by other investigators. ‘high’ amount of morphine. This was excreted into breast milk, resulting in the baby’s high morphine concentrations, which was identified Box 1. Advantages and disadvantages of using pharmacogenomics in postmortem ana­lysis. Consequently, guide‑ as an adjunct biomarker in personalized justice. lines were developed for breast-feeding mothers Advantages: medicated with codeine. ƒƒ DNA is stable in postmortem settings ƒƒ It provides a personalized approach for assessing drug response Conclusion ƒƒ It can assist in the interpretation of drug concentrations in postmortem toxicology In recognizing the complementary, check and and drug death certification balance relationship of PM and PJ, translational ƒƒ It can assess patient compliance ƒƒ The turnaround time is suitable for medicolegal and/or forensic applications PGx may serve the promising role of an adjunct ƒƒ The cost of pharmacogenomics is ‘low‘ in comparison with the legal settlement biomarker for interpreting drug-related toxicity ƒƒ It might differentiate between chronic versus acute toxicity and sensitivity. Currently, robust, scientific and Disadvantages: clinical studies substantiating the relationship ƒƒ Data are available in clinical cases but are limited in postmortem cases between PGx and behavioral and/or perform‑ ƒƒ Legal interpretation is challenging owing to its complexity ance changes are lacking [34]. These desired PJ ƒƒ Drug inhibitors, inducers of enzymes and environmental factors studies are challenging to perform because of complicate interpretation ethical and legal considerations and a lack of ƒƒ It does not account for post-translational modifications funding. Consequently, interpretations may ƒƒ Multiple enzyme systems are involved in metabolism be extrapolated from case reports, or clinical

734 Pharmacogenomics (2010) 11(6) future science group Translational pharmacogenomics in the justice system Editorial behavioral and performance studies; for exam‑ foundation could support the emerging con‑ ple, studies related to ‘driving under the influ‑ cept of PJ becoming a reality to enhance patient ence of drug’. Other advances include automated safety and maintain social justice. platforms and the potential use of oral fluid for toxicology and PGx. Oral fluid, currently Disclosure being evaluated for forensic drug testing [41–44] This article reflects the consensus of the co-authors and not and therapeutic drug monitoring, is easily col‑ the official position of opinion of their respective employers. lected for PM and PJ pending outcome studies to demonstrate efficacy comparable to blood Financial & competing samples. In considering PGx for PJ, advantages interests disclosure and disadvantages are listed in Box 1. The authors have no relevant affiliations or financial In ushering PJ practice in with PGx, a work‑ involvement with any organization or entity with a finan- ing group should consist of colleagues from cial interest in or financial conflict with the subject matter inter-related disciplines in order to probe and or materials discussed in the manuscript. This includes keep abreast of recent developments, to grade employment, consultancies, honoraria, stock ownership or evidence of case reports and outcome studies, options, expert testimony, grants or patents received or and to develop inclusion and exclusion crite‑ pending, or royalties. ria. With sound scientific and legal principles No writing assistance was utilized in the production of and correct interpretation, a firm and lasting this manuscript.

(FPTMPGxSG): Personalized medicine 13 Chung WH, Hung SI, Hong HS et al.: Bibliography enabling personalized justice: methadone Medical genetics: a marker for Papers of special note have been highlighted as: pharmacogenomics as an adjunct – for Stevens–Johnson syndrome. Nature 428, 486 n of interest molecular autopsy, and for addiction and (2004). nn of considerable interest driving under the influence of drugs (DUID). nn Reports the association of Stevens–Johnson 1 Miller ED: A bold leap into the future – Clin. Chem. Lab. Med. 46, A118 (2008). personalized medicine is key to the new genes syndrome with the HLA-*B1502 allele. 8 Wong SHY: Pharmacogenomics as molecular to society curriculum. In: Hopkins Medicine. 14 Committee on Identifying the Needs of the autopsy – an adjunct to forensic pathology/ Pasquale SE (Ed.). Johns Hopkins Medicine, Forensic Science Community – National toxicology: from Gregor Mendel to Baltimore, MD, USA, 48 (2009). Research Council of the National Academies: personalized medicine and personalized justice. Strengthening Forensic Science in the United 2 Wu AHB, Babic N, Yeo KTJ: Implementation In: Clarke’s Analysis of Drugs and Poisons (4th States: A Path Forward. National Academy of of pharmacogenomics into the clinical Edition). Moffat AC, Osselton DM, Widdop B Sciences. The National Academies Press, practice of therapeutics: issues for the (Eds). Royal Pharmaceutical Society Washington, DC, USA, 1–352 (2009). clinician and the laboratorian. Per. Med. 6, Publishing, London, UK (2010) (In Press). 315–327 (2009). nn National Academy of Science identified n First publication of the concept of 3 Wadelius M, Pirmohamed M: deficiencies in several areas of forensic personalized justice. Pharmacogenetics of warfarin: current status sciences and offered recommendations and future challenges. Pharmacogenomics 9 Jeffreys AJ, Wilson V, Thein SL: for the future. Individual‑specific ‘fingerprints’ of human J. 8(7), 99–111 (2007). 15 American Academy of Forensic Sciences: DNA. Nature 316(6023), 76–79 (1985). 4 Ray T: FDA updates warfarin labeling with The American Academy of Forensic Sciences PGx-guided dosing ranges. Pharmacogenomics nn Beginning of the practice of approved position statement in response to Reporter 3 February 2010. DNA fingerprinting. National Academy of Sciences’ “Forensic Needs” Report. American Academy Of Forensic 5 Wong SH: Pharmacogenomics and 10 Sallee FR, DeVane CL, Ferrell RE: Sciences, CO, USA (2009). personalized medicine for drug addiction and Fluoxetine-related death in a child with toxicology: towards personalized justice? cytochrome P-450 2D6 genetic deficiency. 16 Bohan TL: President’s message. Academy Presented at: 11th Asian Pacific Congress of J. Child. Adolesc. Psychopharmacol. 10, 327–334 News 39(1), 34–35 (2009). (2000). Clinical Biochemistry. Beijing, China, 14–19 17 Gianelli P: Understanding Evidence October 2007. n Published index case that included (3rd Edition). 323 (2003). nn Introduces for the first time the concept of pharmacogenetics for interpreting 18 Federal Judicial Center: Reference Manual on personalized justice at an international fluoxetine toxicity. Scientific Evidence (2nd Edition). Washington, scientific congress. 11 Cote JF, Kirzin S, Kramar A et al.: UGT1A1 DC, USA, 487 (2000). 6 Wong SHY, Happy C: Personalized justice, polymorphism can predict hematologic 19 Federal Rule of Evidence, 702, Testimony translational pharmacogenomics and toxicigy inpatients treated with irinotecan. by Experts. Clin. Cancer Res. 13, 3269–3275 (2007). personalized medicine – relevant to the 20 Kassirer JP, Cecil JS: Inconsistency in forensic sciences? Tox. Talk 33, 22–23 (2009). 12 Mallal S, Phillips E, Carosi G et al.: evidentiary standards for medical testimony: 7 Wong SH, JM Jentzen, RZ Shi; the Forensic HLA-B*5701 screening for hypersensitivity to disorder in the courts. JAMA 288, 1382–1387 Pathology/Toxicology Methadone abacavir. N. Engl. J. Med. 358, 568–579 (2002). Pharmacogenomics Study Group (2008).

future science group www.futuremedicine.com 735 Editorial Wong, Happy, Blinka et al.

21 Penry v. Lynaugh, 429 US 302 (1989). A systematic review of PGX dosing of 106 Centers for Medicare and Medicaid Services 22 Roper v. Simmons, 543 US 551 (2005). warfarin. J. Gen. Intern. Med. 24, 656–664 www.cms.hhs.gov/mcd/ (2009). viewdraftdecisionmemo. 23 Atkins v. Virginia, 536 US 304 (2002). 37 Caldwell MD, Awad T, Johnson JA et al.: asp?from2=viewdraftdecisionmemo. 24 Wong SHY, Wagner MA, Jentzen JM et al.: CYP4F2 genetic variant alters required asp&id=224& Pharmacogenomics as an adjunct of warfarin dose. Blood 111, 4106–4112 molecular autopsy for forensic (2008). pathology/toxicology: does genotyping Affiliations CYP 2D6 serve as an adjunct for certifying 38 Gage BF, Eby C, Johnson JA et al.: Use of ƒƒ Steven HY Wong, PhD methadone toxicity? J. Forensic Sci. 48, PGX and clinical factors to predict the Professsor of Pathology, Psychiatry and 1406–1415 (2003). therapeutic dose of warfarin. Clin. Pharmacol. Ther. 84, 326–331 (2008). Behavioral Medicine, and Population 25 Wong SHY, Gock SB, Shi RZ et al.: Health‑Epidemiology and Director of 39 Woodcock J: A difficult balance – pain Pharmacogenomics as an aspect of molecular Toxicology and TDM, Medical College of management, drug safety and the FDA. autopsy for forensic pathology/toxicology. Wisconsin, Milwaukee, N. Engl. J. Med. 361, 2105–2107 (2009). In: Pharmacogenomics and Proteomics: WI 53226–0509, USA Enabling the Practice of Personalized Medicine. n US FDA perspective centered around pain Tel.: +1 414 805 6971 Wong SHY, Linder M, Valdes R Jr (Eds). management and drug safety. Fax: +1 414 805 6980 AACC Press, Washington, DC, USA, 40 Madadi P, Koren G, Cairns J et al.: [email protected] 311–320 (2006). Safety of codeine during breastfeeding: fatal and 26 Molina DK: Handbook of Forensic Toxicology morphine poisoning in the breastfed Scientific Director, Toxicology Department for Medical Examiners. CRC Press, Boca neonate of a mother prescribed codeine. and Pharmacogenomics, Milwaukee County Raton, FL, USA, 1–370 (2010) (Appendix Can. Fam. Phys. 53, 33–35 (2007). Medical Examiner’s Office, WI, USA C – PGXs, 343–347). ƒƒ Christopher Happy, MD 41 Bosker WM, Huestis MA: Oral fluid testing Medical Examiner of Milwaukee County and 27 Karch SB: Karch’s Pathology of Drug Abuse for drugs of abuse. Clin. Chem. 55, Assistant Professor of Pathology, Medical (4th Edition). CRC Press, Boca Raton, FL, 1910–1931 (2009). USA, 1–709 (2009). College of Wisconsin, Milwaukee, 42 Cone EJ, Huestis MA: Interpretation of oral WI 53233, USA 28 Ernst C, Mechawar N, Turecki G: Suicide fluid tests for drugs of abuse. Ann. NY Acad. ƒƒ Dan Blinka, JD, PhD neurobiology. Prog. Neurobiol. 89, 315–333 Sci. 1098, 51–103 (2007). Professor, Marquette Law School, (2009). 43 Pil K, Verstraete AZ: Current developments Milwaukee, WI 53233, USA 29 Golgin E: Epigenetic suicide note. in drug testing in oral fluids. Ther. Drug ƒƒ Susan Gock, MSc The Scientist, 18–19 August (2009). Monit. 30, 196–202 (2008). Technical Director, Toxicology Department, 30 Kim S, Misra A: SNP genotyping 44 Bush DM: The U.S. Mandatory Guidelines Milwaukee County Medical Examiner’s technologies and biomedical applications. for Federal Workplace Drug Testing Office, WI, USA Annu. Rev. Biomed. Eng. 9, 289–320 (2007). Programs: current status and future and Instructor of Pathology, Medical College of 31 Neuman MG, Malkiewicz IM, Shear NH: considerations. Forensic Sci. Int. 174, Wisconsin, Milwaukee, WI 53233, USA A novel lymphocyte toxicity assay to assess 111–119 (2008). ƒƒ Jeffrey M Jentzen, MD, PhD drug hypersensitivity syndromes. Clin. Professor of Pathology and Director of Biochem. 33, 517–524 (2000). „„Websites Autopsy Services, University of Michigan, 32 King AC, Volpicelli JR, Frazer A, 101 Coumadin® Tablets (Warfarin Sodium Ann Arbor, MI, USA O’Brien CP: Effect of naltrexone on Tablets, USP) Crystalline Coumadin® For ƒƒ Honorable Judge M Joseph Donald, JD subjective alcohol response in subjects at high Injection (Warfarin Sodium for Milwaukee County Circuit Court – and low risk for future alcohol dependence. Injection, USP) Branch 2, Milwaukee County Safety Psychopharmacology 129, 15–22 (1997). www.accessdata.fda.gov/drugsatfda_docs/ Building, Room 423, Milwaukee, WI 53233, 33 Wang JB, Imai Y, Eppler CM, Gregor P, label/2010/009218s108lbl.pdf USA Spivak CE, Uhl GR: µ-opiate receptor: ƒƒ Howard Coleman 102 Coumadin Overdose, Warfarin Overdose, cDNA cloning and expression. Proc. Natl Genelex Corporation, 3000 First Avenue, Law Examples Acad. Sci. USA 90(21), 10230–10234 (1993). Suite One, Seattle, WA 98121, USA www.spanglaw.com/medical-malpractice/ ƒƒ Saeed A Jortani, PhD 34 Healy D, Herxheimer A, Menkes DB: medication-errors/coumadin Antidepressants and violence: problems at the Associate Professor, Department of Pathology 103 SSJ Law interface of medicine and law. PloS 3, and Laboratory Medicine, University of http://ssjlaw.mobi 1478–1487 (2006). Louisville, 511 S. Floyd Street (Room 217A), 104 Strengthening Forensic Science in the Louisville, KY 40202, USA n Comprehensive review of candidate United States: A Path Forward ƒƒ Yolande Lucire, PhD, MBBS antidepressant cases that showed potential www.nap.edu/catalog/12589.html Forensic and Medico-Legal Psychiatry, legal implications. Level 5, 203–233 New South Head Road, 105 The American Academy of Forensic Sciences NSW 2027, Australia 35 Stehle S, Kirchheiner U, Lazar A, Fuhr U: Approves Position Statement in Response to ƒƒ Cynthia Morris-Kukoski, PharmD PGXs of oral anticoagulants. Clin. the National Academy of Sciences’ “Forensic Forensic Examiner Toxicology, Pharmacokin. 47, 565–594 (2008). Needs” Report FBI Laboratory, Quantico, VA, USA 36 Kangelaris KN, Bent S, www.aafs.org/pdf/AAFS_Position_ and Nussbaum RL, Garcia DA, Tice JA: Statement_for_Press_Distribution_ Clinical Pharmacologist/Toxicologist, United Genetic testing before anticoagulation? 090409.pdf States Navy Reserves, USA

736 Pharmacogenomics (2010) 11(6) future science group Translational pharmacogenomics in the justice system Editorial

ƒƒ Manuela Neuman, PhD ƒƒ Michael A Wagner, PhD ƒƒ Alan HB Wu, PhD Associate Professor of Pharmacology, Associate Professor, Department of Professor of Pathology and Laboratory Biophysics and International Health, Pharmacology and Toxicology and Medicine, University of California San Department of Pharmacology & Institute of Department of Pathology, Indiana University Francisco, San Francisco General Hospital, Drug Research, USA School of Medicine, USA 1001 Potrero Avenue, San Francisco, and and CA 94110, USA Director of In vitro Drug Safety & Director, Indiana State Department of ƒƒ Kiang-Teck J Yeo, PhD BioTechnology, MaRS Discovery District, Toxicology, 550 W. 16th Street, Indianapolis, Professor of Pathology, Director, Clinical 101 College Street, Suite 300, Laboratory IN 46202, USA Chemistry, UC Med Laboratory, and 351, Toronto, ON, M5G 1L7, Canada ƒƒ Jennifer R Wynn, PhD Pharmacogenomics Program, University of ƒƒ Paul J Orsulak, PhD, MBA Assistant Professor of Criminal Justice, Chicago, Chicago, IL 60637, USA Consultant, Clincal and Forensic Toxicology, Coordinator, Joint Criminal Justice Program, Minneapolis, MN, USA Department of Law, Police Science and ƒƒ Tara Sander, PhD Criminal Justice Administration, LaGuardia Associate Professor of Pathology, Medical Community College and John Jay College of College of Wisconsin, USA Criminal Justice, 899 10th Avenue, and New York, NY 10019, USA Director, Molecular Diagnostics of Children’s Hospital of Wisconsin, Milwaukee, WI 53226, USA

future science group www.futuremedicine.com 737

Lucire and Crotty Akathisia-related homicide CASE SERIES and REVIEW

Antidepressant-induced akathisia-related homicides associated with diminishing mutations in metabolizing genes of the CYP450 family Yolande Lucire Christopher Crotty Edgecliff Centre, Edgecliff, NSW, Australia

Correspondence: Forensic psychiatry: Yolande Lucire Level 5, 203–233 New South Head Road, Edgecliff Centre, Edgecliff, NSW 2027, Australia Tel +61 2 9327 1499 Fax +61 2 9327 4555 Email [email protected]

Abstract Purpose: To examine the relation between variant alleles in 3 CYP450 genes (CYP2D6, CYP2C9 and CYP2C19), interacting drugs and akathisia in subjects referred to a forensic psychiatry practice in Sydney, Australia. Patients and methods: This paper concerns 10/129 subjects who had been referred to the first author's practice for expert opinion or treatment. More than

1 120 subjects were diagnosed with akathisia/serotonin toxicity after taking psychiatric medication that had been prescribed for psychosocial distress. They were tested for variant alleles in CYP450 genes, which play a major role in Phase I metabolism of all antidepressant and many other medications. Eight had committed homicide and many more became extremely violent while on antidepressants. Ten representative case histories involving serious violence are presented in detail. Results: Variant CYP450 allele frequencies were higher in akathisia subjects compared to random primary care patients tested at the same facility. 10 subjects described in detail had variant alleles for one or more of their tested CYP450 genes. All but 2 were also on interacting drugs, herbals or illicit substances, impairing metabolism further. All those described were able to stop taking antidepressants and return to their previously normal personalities. Conclusion: The personal, medical, and legal problems arising from overuse of antidepressant medications and resulting toxicity raise the question: how can such toxicity events be understood and prevented? The authors suggest that the key lies in understanding the interplay between the subject’s CYP450 genotype, substrate drugs and doses, co-prescribed inhibitors and inducers and the age of the subject. The results presented here concerning a sample of persons given antidepressants for psychosocial distress demonstrate the extent to which the psychopharmacology industry has expanded its influence beyond its ability to cure. The roles of both regulatory agencies and drug safety pharmacovigilantes in ensuring quality and transparency of industry information is highlighted.

Keywords: adverse drug reaction, drug therapy, safety pharmacogenetics, CYP1A2, CYP3A4 CYP2D6, CYP2C9, CYP2C19, drug metabolism, public health, suicide, violence, human rights.

2 Introduction Many, if not most, medicines used to alter brain chemistry interact with the Cytochrome P450 (CYP450) system of enzymes, as substrates, inducers or inhibitors. The activity of this enzyme system is genetically determined and is also affected by other intrinsic factors such as age, gender, and comorbidity as well as extrinsic factors such as nutrition and drug–drug interactions. The CYP450 superfamily of genes is the major pathway of the phase 1 oxidative metabolism of many drugs currently marketed. All antidepressants are metabolized by the enzymes CYP2D6, CYP2C19, CYP3A4, and CYP1A2 to varying degrees, with the first three contributing most to metabolism.1–3 Variability in metabolism and treatment response to antidepressants can be affected by genetic polymorphisms. The general population can be broadly divided by CYP450 genetic testing into poor, intermediate, or extensive (normal or “wild- type”) metabolizers for CYPs 2C9, 2C19 and 2D6, with an additional category of ultra-rapid metabolizers for CYP2C19 (due to *17 allele) and CYP2D6 due to duplications of extensive metabolizer alleles.1–3 The Karolinska group, led by Ingelman-Sundberg, maintains a website summarizing the relevant literature on the metabolizing functions of different alleles.4 Table 1 shows the nomenclature conventions for naming the non-variant and variant alleles, with examples of some of the clinical consequences that may result from having one or more variant alleles for a given CYP450 gene.5 Drug metabolism can be affected by the medications taken, for example, genetically extensive metabolizers will metabolize drugs like poor metabolizers after taking CYP inhibitors (notably fluoxetine and paroxetine) that reduce or eliminate enzyme activity.6,7 With some drugs and prodrugs such as codeine, ultra-rapid metabolism producing an active metabolite was found to be problematic.8 Ultra-rapid metabolism at CYP2D6 (due to allele duplications), seems to be strongly associated with a large number of deaths from intoxication and suicide.9-11 The ultra-rapid metabolizer allele CYP2C19*17 is likely to cause therapeutic failures in drug treatment with antidepressants, but no reports of serious adversity associated with it have been published to date.12 High or fast-

3 changing levels of psychotropic substances can cause unpredictable toxicity leading to violent behavioral effects, including akathisia.13,14 These effects are listed in product information (PI) or “label”; product information for drugs used in the USA can be found on the website of the US Food and Drug Administration.15 Extrapyramidal side effects (EPSEs), including akathisia, may develop very fast in persons with slower CYP450 metabolism.16 These effects were first identified in children whose cytochrome systems and other metabolic pathways were not yet fully developed or may have not yet been induced. Further information on CYP450 attribution for the drugs and substances described in this paper is presented in Table 2, adapted from Wynn et al, whose Clinical Manual of Drug Interaction Principles for Medical Practice has been the authors’ invaluable tool for understanding drug–gene and drug–drug interpretations.1 It accords with Flockhart’s Cytochrome P450 Drug Interaction Table.17 Information about the therapeutic window comes for Baumann et al.18 We present here a summary of findings from a group of subjects who were referred to the forensic psychiatry practice of Yolande Lucire for psychiatric assessment. They were found to be experiencing adverse drug reactions, manifestations of neurotoxicity/akathisia, and were tested for variant alleles in CYP450 genes. We present some essential features, including genotype, drug regime, and outcomes, along with rationales based on the research literature in CYP450 drug metabolism, for a selected group of subjects who represent the extreme violent end of the spectrum for akathisic subjects. Patients and methods Study setting and population description In an ongoing, naturalistic study, 129 people were referred or self-referred to Yolande Lucire’s forensic psychiatry practice for expert opinion and were diagnosed as having medication-induced disorders, serious adverse drug reactions to drugs metabolized by CYP450 enzymes. They were tested for variant alleles for three CYP450 genes (2D6, 2C9, 2C19); nine more were tested because they were first-degree relatives. The total number of persons tested was 138. Four swabs produced incomplete results.

4 Sources of referral Of the 129, 92 were referred by lawyers for various issues; 29/92 were being treated for compensable work stress; 9/92 for compensable physical injuries; and a further 30/92 faced criminal charges because they had committed a variety of offences, ranging from trespass, running amok, and malicious damage to violence resulting in homicide; 6/92 had been brought before professional tribunals for behavioral changes; 7/92 had issues with child custody; and the remainder had a variety of problems. Following akathisic, suicidal, self-mutilating, and/or homicidal behavior on psychiatric drugs, 29/129 persons had self-referred, requesting to be research subjects and they conformed to the study's criteria for testing. Another 11/129 were friends and neighbors who requested testing, asking if CYP450 genotyping could shed light on their adverse drug reactions. Reasons for testing Five of the 129 had experienced lack of response or adversity with nonpsychiatric drugs: metoclopramide (an akathisia inducer) (1), clopidogrel (1), opioids (3). The other 124 were found to be experiencing adverse drug reactions for drugs used in psychiatry and/or zolpidem. They had reported symptoms or shown signs of akathisia-related suicidal and/or homicidal impulses, dysphoria, which overlapped with manic shift with behavioral disturbance and symptoms of serotonin toxicity, in the range described by Breggin.19 Three were investigated after bizarre sleep activities while on zolpidem. Clinical characteristics of the tested population Of the tested population, 86 seemed to have been fully functioning workers or students experiencing various degrees of psychosocial stress, who were being treated for compensable injuries and/or stressors. Four had been diagnosed with mental illness before they used substances, medicinal or other: two with schizophrenia and two with bipolar disorder. At least 27 were given psychiatric medications following or during the use of illicit substances such as cannabis, amphetamines, and ecstasy (MDMA), three had taken hypericum before taking antidepressants and had developed similar adversity on that. Three

5 had received attention for anorexia nervosa, three for the effects of sexual abuse, two for brain injuries, one for organophosphate poisoning, and one for epilepsy. The expected array of human problems was disclosed, with overlapping alcohol overuse (2), personality disorders (6), and bereavements (4). The premedication conditions of six could not be reliably assessed. Outcomes of psychiatric medications in this population Out of the 129, eight had committed homicide, three had committed suicide and one had sleepwalked to her death. The first author restored 16 to their previously normal personalities by supervising a very slow and safe withdrawal of medications in the correct order. Fifty-six others had stopped taking the drugs, with or without medical support, after intervals varying from one day to several years. Some had stopped taking the drugs after consulting pharmacists or protocols available on the Internet. These reported recovery. Thirty reported back that their treating doctors were unresponsive to genetic information. Some were forced by court orders at the instigation of the prescribing physician of tribunals) to continue taking the same drugs by injection. Those who were not allowed to stop medications included 4 of the 16 who had been imprisoned, while 12 of those who had been imprisoned for akathisia-related events reported recovery. Twelve had manifested neuroleptic brain damage, in one case demonstrable on magnetic resonance imaging, and, in two, deterioration involving loss of cognitive function confirmed by intelligence tests. Two suffered strokes and one had become diabetic when olanzapine was added (without success) in an attempt to quell her antidepressant-induced hallucinosis. Those taking zolpidem experienced various unmotivated sleep activities, including a death that occurred while sleepwalking and a homicide. Twenty-six were known to be taking illicit or herbal psychoactive substances metabolized by CYP450 and had been treated with antidepressants, antipsychotics, atypicals, and enforced zuclopenthixol depot injections (all also metabolized by CYP450) and remained under psychiatric care until they stopped medication or committed suicide. Their treating doctors are among the 30 unresponsive to genetic information. This attitude may be related to a popular but

6 contested belief that cannabis accelerates the onset of mental illness.20 It has not been not possible to follow up the fates of 26 of them. Genetic testing The CYP450 genetic testing for all but one of the described subjects was carried out at Healthscope Molecular in Melbourne, Australia. Healthscope Molecular tested for CYP2D6: *2 (EM), *9, *10, *17, *41 (partial function alleles) and *3, *4, *5, *6, *7, and *8 (non-functional alleles) as well as multiple copies (potential ultra-rapid metabolizers); CYP2C9: *2 (non-functional allele) and *3 (partial function alleles); CYP2C19: *2 (non-functional) and *3 (partial function alleles) and after 2007, 2C19*17 (ultra-rapid metabolizer allele), using the Applied Biosystems (Foster City, CA) ABI3130. In the other case (Subject 5 in Table 3), gene testing was carried out by Genelex (Seattle, WA). Results The association among variant alleles, drug metabolism, and severe akathisia Of 138 persons tested for CYP450 genes, 129 had experienced adversity, mainly akathisia, due to psychiatric drugs and 9 were their first-degree relatives, also with a history of adversity on other drugs. 49 akathisia subjects were tested for six alleles in three CYP450 genes at Diversity Health Institute of Westmead Hospital and the significantly increased frequency of multiples of diminished activity alleles in akathisia subjects compared to a sample of Western Sydney clinical patients has been previously reported.21 85 akathisia subjects fully tested (or retested) at Healthscope Molecular (see Patients and Methods for allele coverage) were compared to a sample of 150 randomly selected primary care patients referred for other tests and tested for CYP450 genes at the same laboratory in 2007 (Keith Byron, personal communication). This comparison revealed significantly higher frequencies of variant alleles in akathisia subjects in comparison with the primary care sample, and significantly lower frequencies of subjects with extensive metabolizer alleles (Figure 1). All but 1/85 (i.e. 98.8%) of akathisia subjects possessed at least one variant allele for the three CYP450 genes tested (CYPs 2C9, 2C19 and 2D6) while 81% of random patients possessed at least one various allele; 19% of

7 random primary care patients had no variant alleles compared with 1.2% of akathisia patients (Figure 1). The odds ratio for this comparison was 19.3 (95% CI: 2.57-144.5; P=0.00014; Fisher 2-tailed exact probability test) therefore the null hypothesis that this difference in variant allele frequency between the two samples occurred by chance alone can be safely dismissed.The 2007 random patient data from Healthscope Molecular identified 2D6 *2 as a “variant” and 2C19 *17, since it was not tested for at that time, was subsumed into the non- variant 2C19*1 category; the comparison in Figure 1 was made on this basis. An analysis of CYP450 alleles after 2007 (the case for all akathisia subjects tested at Healthscope and which included 2C19*17) that considered their functionality as recognised presently, revealed the following: 10/85 had no diminishing (as opposed to variant) mutations, 37/85 had 1, 29/85 had 2, 8/85 had 3 and 1/85 had 4 diminishing mutations. Among the 10/85 with no diminishing mutations, 2D6*2 which is nominally, but not invariably, extensive22 appeared in 4/10 and was the sole variant allele. 4/10 carried the ultra-rapid 2C19*17 in conjunction with 2D6*2. 2C19*17 appeared in 28/85, 2 of these were homozygous (*17/*17) and *17 was combined with diminishing mutations in 21 of these 28 subjects. The possible significance of the *17 allele is elaborated further in the Discussion. 1 subject carried the gene multiplication, 2D6 *1XN (ultra-rapid) as a sole mutation and another carried 2D6*2XN, but not on its own. Only one akathisia subject carried the *1 allele exclusively (homozygous) across all three genes. The 10 subjects described in detail in Table 3 suffered from severe akathisia, induced by antidepressants prescribed for various forms of psychosocial distress. In 9/10 cases shown in Table 3, the individual was on drugs for which genotyping would predict impaired metabolism, as they were intermediate or poor metabolizers for those drugs. Eight of the ten were taking other interacting drugs, substances, or herbal preparations. The subjects None of the 10 subjects described had any history of mental illness; none had been violent before. All recovered from akathisia after stopping the

8 medication without assistance or supervision and, frequently, against medical advice. Subject 1 (35-year-old female; genotypes: CYP2D6 *4/*41, CYP2C9 *1/*2, and CYP2C19 *1/*17) was prescribed a low dose of the 2D6-metabolized tricyclic antidepressant nortriptyline (12.5 mg/day) 3 days prior to committing homicide. She was severely impaired in metabolizing this drug owing to the 2D6 combination of *4 (null allele) and *41 (reduced activity allele). This subject also had reduced metabolism through 2C9 (*1/*2) and increased (ultra-rapid) metabolism through 2C19 (*1/*17), although the significance of these genotypes for nortriptyline metabolism is unclear. Subject 1, in her own words: My husband was drinking. I took small doses of valerian for a month and had weird dreams and premonitions. When I took nortriptyline, I immediately wanted to kill myself, talked myself out of it. I’d never had thoughts like that before. My husband was angry, shouting. I walked outside a lot, with palpitations, trouble breathing, and became more depressed. My smoking went up to 25 a day, no alcohol. I didn’t sleep for two nights, dreamt, then slept maybe three hours, felt awful. I dreamt that my daughter had dark teeth and I saw a black halo around her head, a spear hanging over it. I felt like a zombie. I believed I had to help my daughter, that a bad spirit possessed her. I picked up a knife and stabbed her and woke up. I was not myself. I was looking on from the outside, controlled by dark forces. She said: “Mum, what are you doing here?” I realized what I’d done. I asked my husband to kill me. He called the police. I felt better in the police cells without the pills, but the pills started again and thoughts of killing myself returned. Nortriptyline is extensively metabolized in the liver, primarily by hydroxylation.23,24 There is a strong correlation between total plasma clearance of nortriptyline and 2D6 enzyme activity. 24–27 Reduced clearance resulting from this subject’s 2D6 genotype alone would lead to a build up of serum nortriptyline, which has been shown to be associated with loss of efficacy, toxicity, and

9 increased adversity in several studies.28-30 Nortriptyline is also significantly metabolized through 3A4 as an additional pathway, according to in vitro 31 and in vivo data. 32-35 In the case of severely reduced metabolism through 2D6, as in this subject, 3A4 would have been the major pathway for metabolism. However, the nortriptyline was superimposed on valerian, which has been shown to be a 3A4 inhibitor in vitro and to significantly increase the maximum serum levels of a 3A4 substrate in vivo.36-38 The addition of a substrate to an inhibitor is one drug– drug interaction scenario (Pattern 2), as described in Armstrong et al.39 The addition of nortriptyline to valerian immediately provoked a toxic delirium in this subject; she became akathisic and suicidal and 3 days later she committed the homicide. Subjects 2, 3, and 4 were prescribed the SSRIs fluoxetine or paroxetine as monotherapy. Both these drugs are metabolized extensively in the liver by 2D6 and these subjects all had variant alleles for 2D6, being either intermediate or poor metabolizers. Fluoxetine and paroxetine are also both potent inhibitors of 2D6, converting both extensive and intermediate metabolizers into poor metabolizers over time.40 Subject 2 (18-year-old male; genotypes: CYP2D6 *4/*5, CYP2C9 *1/*1, and CYP2C19 *1*17), with no CYP2D6 activity due to his carrying a null activity allele (*4) and a gene deletion (*5), was prescribed half the standard dose of fluoxetine, 10 mg/day, for 14 days as his sister was comatose after a car accident. He had used cannabis, whose active ingredients and metabolites have a long half-life. He had normal metabolizing genes for 2C9, which also metabolizes the psychoactive (R)-enantiomer of fluoxetine.41-44 He was an ultra- rapid metabolizer for 2C19 (*1/*17). He killed his father 4 days after he ran out of fluoxetine. Subject 2, in his own words: I felt better, energized, then restless, no appetite, no sexual desire; gouged my lip then became aggressive, paranoid, violent, could not sleep, walked around thinking constantly about suicide. Tried to kill myself twice, tying a sheet around my neck, cutting my wrist, became depressed, and

10 cried. I ran out of fluoxetine and felt like a ticking time bomb. I had no thoughts for consequences. Fearless, I challenged strangers to fight, punched out windows, threw a toolbox through windscreen and sped, running over street signs, crashed my truck; stood at the top of a bridge wanting to jump. I wrote messages on the shattered windscreen of my vehicle. Wanting to die, I walked a long way to a friend’s house where I found a pistol, walked eight miles to my father’s house, talked to him for a few minutes. Felt very small, as if watching myself from above. I remember my dad, then the sound of a gun. No argument, no provocation. Don’t remember pulling it out of my backpack or pointing it. Couldn’t understand what had happened. What had I done? I wanted to shoot myself. I confessed immediately. While (S)-fluoxetine is primarily metabolized by 2D6, (R)-fluoxetine is metabolized preferentially by 2C9 and, to a lesser extent, by 2D6.43 In the absence of 2D6 activity, which was the case in this subject, CYPs 2C9, 2C19, and 3A4 would become important metabolic pathways.41-44 The CYP450 metabolic pathways for tetrahydrocannabinol and cannabinol (the major psychoactive ingredients) include 3A4, 2C9, and cannabis inhibits 3A4.45–48 In this subject, the total lack of 2D6 metabolizing capacity, along with interference with other metabolic pathways through the use of cannabis, would have led to rapid increases in fluoxetine serum levels, triggering an akathisic state. He was 18, in an age group now subject to a black box warning in all antidepressant labels. We consider that his age alone placed this subject into the category of persons at increased risk of the adverse events noted by the US Food and Drug Administration (FDA) and this was additional to his impaired metabolism of fluoxetine and the use of cannabis. For poor metabolizers, Eli Lilly, in its product information, recommends Prozac® Weekly (Eli Lilly and Co, Indianapolis, IN) capsules, a delayed-release formulation; containing enteric-coated pellets of fluoxetine hydrochloride equivalent to 90 mg (291 μmoles) of fluoxetine.49 Subject 3 (35-year-old male; genotypes CYP2D6 *2/*4, CYP2C9 *1/*2, and CYP2C19 *1/*2) and Subject 4 (46-year-old male; genotypes CYP2D6

11 *1/*41, CYP2C9 *1/*1, and CYP2C19 *17*17) were on paroxetine. Both subjects were intermediate metabolizers for 2D6, the major metabolic pathway for paroxetine.50,51 Subject 3 had three diminishing mutations across the three CYP genes tested (intermediate metabolizer each of 2D6, 2C9 and 2C19) and was on single dose (20 mg/day), whereas Subject 4 had one diminishing and two ultra- rapid mutations and was taking treble dose (60 mg/day). Paroxetine is also known to be an extremely potent inhibitor of 2D6 and has long been known to convert genotypic extensive and intermediate metabolizers to phenotypic poor metabolizers, at steady state.52-54 Over time, these subjects were thereby converted to phenotypic poor metabolizers through the inhibition of 2D6 activity by paroxetine. Subject 3, a 35-year-old tradesman with one child (genotypes: CYP2D6 *2/*4, CYP2C9 *1/*2, CYP2C19 *1/*2), had diminished metabolizing capacity in all three genes, and was prescribed 20 mg/day paroxetine for relationship difficulties. In his words: I couldn’t leave my house, had trouble going to work. Anxiety was like nothing I’ve ever had before. Feelings were strange, peculiar, anger unjustified. I felt bad about it. I tried to punch her boyfriend through the car window. Police said I tried to run him down. I thought death was imminent, had an impulse to commit a violent suicide, hoping someone would kill me. I was irritable with family, clients. I had racing thoughts, like a video on fast-forward. Weird, violent dreams, I drove around a lot, sometimes in her street, visiting three or four people in a day, unable to settle. Thought I was bullet-proof, god-like, a special person. My sex drive increased; I had an extraordinary erection, lasting for hours, then I lost interest in everything, stayed in bed, plummeted, depressed, suicidal. Drove to the park. I did not expect to find them there. My son ran up to me. His mother said, “He’s not going with a crazy person like you.” Last thing I remember was getting the knife from my toolbox, walking in slow motion. A guy cracked a wet towel at me. I reconstructed what happened

12 from what I was told. I got it wrong, for example, where I stabbed her. I knew I had done it because I had blood all over me. This person committed homicide in a state of automatism following a psychological blow after 11 weeks of akathisia caused by paroxetine toxicity. Subject 4, an independent and successful businessman, was prescribed paroxetine because he became anxious about the possibility of not making enough money to support his family; he had lost weight and sleep over this. On paroxetine he experienced suicidal urges and stopped it. Later, he resumed taking paroxetine and soon entered a violent manic-suicidal delirium that was accompanied with akathisia. Thinking the drug was not working, he gradually increased the dose to 60 mg/day. His mental condition deteriorated and, 40 days into a homicidal akathisia delirium-like state and affected by a manic shift, he believed that he had to kill his son, (to whom he attributed extraordinary powers) and did so, believing he had committed an altruistic, even heroic, act. Like others on psychoactive substances, more so when mood is elevated and cognitive impairment present, he suffered from medication induced-anosognosia, which is defined as "unawareness or denial of a neurological deficit."55 Subject 5 (16-year-old male; genotypes: CYP2D6 *1/*6, CYP2C9 *1/*1, and CYP2C19 *1/*1) was living in a group home, struggling at school, and his girlfriend had left him. He was prescribed sertraline (unknown dose) and became agitated, described himself as "restless energetic and anxious." He felt playful at first, and then became "obsessive, impulsive and self-destructive." He was hospitalized and became uncharacteristically violent, irrational, and compulsive. He began self-mutilating and writing pages of disorientated escape plans, had unwelcome thoughts of destroying property and designing weapons. He felt like he was "caught in whirlwind of feverish, angry and fearful thoughts, losing track of time and having no will power," and that he was losing his memory. He reported palpitations and tremors on sertraline and experienced violent impulses to hurt himself and others, eventually making a suicide attempt. Sertraline is metabolized mainly by 2C19.56 This subject’s 2C19 alleles were normal, but sertraline immediately provoked in him a range of akathisic behaviors, and this

13 bears further inquiry, given his age and circumstances. Following the attempted suicide, he was switched to fluoxetine, 20 mg/day, and attempted suicide again. After this, fluoxetine was increased first to 40 then 60 mg/day with no improvement in his akathisia. He became emotional and labile and pitted himself against his family and the therapist. Valproate was added and increased to 1000 mg/day. He became agitated then described himself as restless, energetic, and anxious. After 11 weeks in hospital, he killed his therapist and lit a fire. Valproate inhibits 2C9 in vivo.57 As previously noted, fluoxetine is metabolized by both CYPs 2D6 and 2C9. In this case fluoxetine would be expected to increasingly inhibit his already genetically impaired 2D6 activity, more so as the dose was increased over time. His 2C9 activity would have been inhibited by valproate prescribed at 1000 mg/day. The combined effect of diminished metabolism, to begin with at 2D6, with the inhibition of 2C9 by valproate, would have been to drastically reduce the metabolism of fluoxetine in this subject. The addition of an inhibitor, valproate, to a substrate, fluoxetine, is also a classic drug-interaction scenario (Pattern 1) as described in Armstrong et al.39 In addition to impaired 2D6 metabolism and the addition of valproate, the authors note here again that this subject’s young age would have placed him in the higher risk category for serious adversity on antidepressants, as cautioned by the FDA advisories. This subject recovered fully upon withdrawal of psychiatric drugs. In only one case, shown in Table 2, was the subject an extensive (or normal) metabolizer for the antidepressant drug in use before committing or attempting homicide, but he had been treated with polypharmacy throughout. Subject 6 (50-year-old male; genotypes: CYP2D6 *1/*2, CYP2C9 *1/*1, and CYP2C19 *1/*2) was employed as an administrator and prescribed variously fluoxetine, amitriptyline, diazepam, occasional metoclopramide, and continuously zolpidem or zopiclone for a year to treat distress from his divorce. Fluoxetine was at no time prescribed on its own. He was then prescribed venlafaxine at 300 mg/day, a dose higher than the maximum recommended by the manufacturer Wyeth58, (225 mg/day) for 2 years and became more akathisic over this period.

14 He was also put on valproate 1000 mg/day and continued with zolpidem alternating with zopiclone. He was genetically an extensive metabolizer at 2D6 (genotype *1/*2), which metabolizes venlafaxine.59,60 He carried the "null activity" allele 2C19 *2, which may have slowed the metabolism of venlafaxine.61 3A4 is also considered to be important in venlafaxine metabolism.62,63 Taking a friend's advice, he stopped taking venlafaxine, exchanging it for a vitamin preparation containing serotonin. At the same time, his doctor prescribed the antibiotic ciprofloxacin, which inhibits CYPs 1A2 and 3A4.64,65 At the time of the homicide, he was taking ciprofloxacin, valproate, and the vitamin mixture containing serotonin, as well as dosing repeatedly with zolpidem. While sleep driving and cleaning a friend's gun, he shot two strangers. As in other cases of aberrant sleep behaviors associated with zolpidem, he remembered none of this time period. While the subject’s venlafaxine, serotonin, or zolpidem levels during the venlafaxine withdrawal and subsequent homicide cannot be known, it has been shown that zolpidem and venlafaxine (along with other serotonin-boosting drugs) interact leading to serotonin syndrome, a myriad of mental state changes and hallucinatory states.66 Ciprofloxacin, which inhibits the major metabolic pathway of zolpidem, 3A4, would further have increased the level of zolpidem as these also interact pharmacokinetically, with an estimated 46% increase in bioavailability of zolpidem.67 The subject suffered the zolpidem side effect of sleep activity, while serotonin toxicity is associated with violence. His treating doctors kept him taking antidepressants for some years. He stopped taking medication in prison, whereupon he returned to his normal personality. While all the subjects had variant alleles and diminished CYP450 metabolism, subjects 7, 8, 9, and 10 all had variant alleles in the major pathways for the antidepressants they were taking while they were experiencing violent episodes, which, in these cases, involved attempted homicide. Furthermore, they were also using, alone or in combination, either other medications, illicit substances, or herbal remedies that had the combined effect of exacerbating the impaired metabolism caused by their variant alleles.

15 Subject 7 (25-year-old male; genotypes: CYP2D6 *1/*41, CYP2C9 *1/*2, and CYP2C19 *1/*2) was prescribed 10 mg escitalopram (double the standard dose) for anxiety following the birth of his child and to help him stop illicit substance use (which may have included cannabis). He was also on hypericum 900 mg/day. Escitalopram, the (S)-enantiomer of citalopram, has been shown to be the most selective of the SSRIs and at least 30 fold more potent than its (R)- isomer, based on in vitro studies.68 Escitalopram is demethylated by CYPs 2C19, 3A4, and 2D6 to (S)-desmethylcitalopram and then demethylated further, at a much slower rate, by 2D6 to (S)-didesmethylcitalopram, which are both psychoactive.1,69,70 The relative contributions to net intrinsic clearance of escitalopram from in vitro data, after correction for relative hepatic abundance, have been estimated as: 2C19 (37%), 3A4 (35%), and 2D6 (28%).71 In keeping with this, escitalopram interacts clinically with both 3A4 and 2D6 substrates. This subject had an inactive 2C19 allele (*2) and a reduced-function allele for 2D6 (*41) and 3A4 is induced by hypericum.72 He reported that his personality changed in that he immediately became depressed, suicidal, agitated, paranoid, and very aggressive, he also felt “manic,” impulsive, and despairing and complained of ejaculation failure. Starting on the sixth day after commencing escitalopram, he made several suicide attempts and several assaults on his partner over 40 days, the last causing her near-lethal injuries. The effects of the cumulative dose of a serotonin reuptake blocker, escitalopram, in combination with hypericum at high dose were synergistic on serotonin levels and immediately caused violent serotonin toxicity, which overlaps and coexists with akathisia.73 Subject 8, a mother (26-year-old-female; genotypes: CYP2D6 *2/*4, CYP2C9 *1/*1, CYP2C19 *2/*17), was initially prescribed paroxetine for stress. On paroxetine, she experienced an uncharacteristic episode of rage and attempted suicide by inhalation of carbon monoxide from a car exhaust, so she stopped taking it. Two years later, she was prescribed paroxetine again and reassured about its safety. The second time she experienced intense restlessness, surges of rage and anger, marital troubles, panic attacks, impulsive

16 spending sprees, and constant suicidal ideation. She reasoned that her low self- esteem, insomnia, and suicidal behavior were due to difficulties with her in-laws. She overdosed and was admitted to hospital where paroxetine was increased. She tried to kill herself again and was diagnosed with an “adjustment disorder," a stress-induced state. She was switched to venlafaxine, 37.5 mg/day and it was increased to 300 mg/day over 3 months. Each upward dose adjustment occasioned a week spent in bed with exhaustion, as she was unable to get up (akinesia). Her mental state deteriorated and violent outbursts and suicidal ideation became frequent and severe. Unable to stay in one place, she drove several hundred miles with her children and tried to kill them and herself by car exhaust. A few days later she tried to kill her children and herself again. There were no interacting drugs in her regimen, however, her genotype for 2D6 (*2/*4) places her in the intermediate metabolizer category, for which there is evidence of at least twofold increases in serum-venlafaxine levels in both single dose and steady-state studies.74-76 Moreover, case reports have documented serious venlafaxine-related adverse events, which are either suspected or confirmed to have resulted from above-therapeutic-range serum-venlafaxine levels; the events ranged from psychosis77,78 to hallucination, psychomotor agitation, and delirium.79–81 The subject pleaded diminished responsibility to four charges of attempted murder and was not required to complete time in prison. She recovered fully on withdrawal. Wyeth provides the following information in product information for venlafaxine, Effexor® XR, USA, including homicidal ideation, since 2005. (Note: frequent is 1/ 10; infrequent, 1/100; rare, 1/1000; suicide attempts and homicidal thinking are catastrophic effects, irrespective of frequency.58 Body as a whole – Frequent: chest pain substernal, chills, fever, neck pain; Infrequent: face edema, intentional injury, malaise, moniliasis, neck rigidity, pelvic pain, photosensitivity reaction, suicide attempt, withdrawal syndrome; Rare: appendicitis, bacteremia, carcinoma, cellulitis.

17 Nervous system - Frequent: amnesia, confusion, depersonalization, hypesthesia, thinking abnormal, trismus, vertigo; Infrequent: akathisia, apathy, ataxia, circumoral paresthesia, CNS stimulation, emotional lability, euphoria, hallucinations, hostility, hyperesthesia, hyperkinesia, hypotonia, incoordination, manic reaction, myoclonus, neuralgia, neuropathy, psychosis, seizure, abnormal speech, stupor, suicidal ideation; Rare: abnormal/changed behavior, adjustment disorder, akinesia, alcohol abuse, aphasia, bradykinesia, buccoglossal syndrome, cerebrovascular accident, feeling drunk, loss of consciousness, delusions, dementia, dystonia, energy increased,facial paralysis, abnormal gait, Guillain-Barre Syndrome, homicidal ideation, hyperchlorhydria,hypokinesia, hysteria, impulse control difficulties, libido increased, motion sickness, neuritis,nystagmus, paranoid reaction, paresis, psychotic depression, reflexes decreased, reflexes increased, torticollis. Very little of the above-quoted information appears in Australian Product Information, which contains no black box warning and denies differences in metabolism between poor and extensive metabolizers, and lists only: Common: Abnormal dreams, decreased libido, dizziness, dry mouth, increased muscle tonus, insomnia, nervousness, paraesthesia, sedation, tremor. Uncommon: Apathy, hallucinations, myoclonus. Rare: Convulsion, manic reaction, neuroleptic malignant syndrome (NMS), serotonergic syndrome.82 Subject 9 (52-year-old female; genotypes: CYP2D6 *4/*4, CYP2C9 *1/*1, and CYP2C19 *1/*1), a teacher, was prescribed a series of antidepressants following harassment at work. After 3 years on a series of antidepressants, always superimposed on polypharmacy containing CYP450 substrates and inhibitors and drugs with similar side effects and causative of pharmacodynamic interactions, she tried to kill her two children while taking venlafaxine XR with other medicines. She suffered from mild chronic bilirubinemia, which may have been Gilbert's syndrome.

18 She was suffering from stress at work and paroxetine was prescribed and increased to 60 mg/day. Paroxetine is metabolized by 2D6 and she had no 2D6 activity, with two null alleles (*4/*4). After paroxetine had occasioned a hospital admission for suicidality, she was switched to citalopram and this was titrated up to 60 mg/day, which was three times the standard dose, while her restlessness, aggression, and suicidality continued. Citalopram is metabolized by CYPs 2C19, 2D6, and 3A4.1,83,84 Its major metabolic pathway, 2C19, was inhibited by her treatment for menopause, conjugated estrogens, which also inhibit CYPs 3A4 and 2C19, as do naturally occurring estrogens.1 Citalopram was replaced by venlafaxine, which is metabolized by CYPs 2D6, 2C9, and 2C19, and inhibits 2D6 mildly, but it was prescribed at 300 mg/day, 8 times a standard dose. Her body pain, diagnosed as "atypical somatoform disorder" occasioned the use of tramadol, metabolized by 2D6, and celecoxib, which is metabolized by 2C9 and inhibits 2D6.85-87 Her hypertension was treated with irbesartan, metabolized by 2C9.88 Her elevated cholesterol was treated with atorvastatin, metabolized by 3A4 and inhibits 2C9 and 3A4, and this drug is associated with painful muscle degeneration.1,89 Tramadol and celecoxib interact adversely. Irregular doses of the akathisia-inducing and 2D6-metabolized drugs metoclopramide and tramadol, which can cause serotonin syndrome, made it impossible for her to recall exactly which of these two co-prescribed drugs she had taken prior to attempting to kill her children and herself, first by gas and then by almost jumping with them from a great height, an impulse she resisted at the last minute. She had been hospitalized eleven times over 3 years for suicidal behaviors and for the homicide attempts, but she no longer needed psychiatric attention after she stopped taking medicines. She came to the first author's attention a year after the homicide attempts as she had been charged with "entering enclosed lands," while delirious after missing doses of venlafaxine, which were left at home over a holiday weekend. She returned to work briefly after following withdrawal protocols from the Internet, but further prescriptions of lovastatin, tramadol, and estrogens all induced further

19 bouts of akathisia and she lost her job. She was never "herself" again or able to work, but hypertension did resolve. She reported this: I’m more depressed than I’ve ever been in my life. I throw things when someone comes into my room. I had surges of rage with everyone, with the medical people who could not fix me. It was strange to feel like that, with violent thoughts all the time. I drove all day to find a silo I was going to jump. I took the children on a holiday and gassed us all to kill them and myself. I was not charged. The thoughts of suicide are there all the time. I’ve now started a suicide book. The psychiatrist doesn’t understand me. I fidget, cannot keep my legs still, I am apprehensive, irritable, uncomfortable with unstable mood. I’ve got a hit list in my head. It’s listed, one, two, three. I think how to get access to these people. I feel violent. Before all this, I was a big reader. I can’t focus. Subject 10 (25-year-old female; genotypes: CYP2D6 *2/*4, CYP2C9 *1/*3, and CYP2C19 *1/*1) was in an abusive relationship, using cannabis, and was prescribed citalopram, unknown dose. She immediately became very disturbed, aggressive, and made multiple suicide attempts. She tried to jump in front of a train with her child and lost custody as a result. Her medication was changed to venlafaxine to treat her worsening depression. More of the same disturbed behavior followed and went on for several years. She was prescribed diazepam and other benzodiazepines to counteract her akathisia and it was then alleged that she was addicted to them. Cannabis use continued throughout. She recalled having constant thoughts of self-harm, and was being told she was “attention seeking” when she was arrested for running in front of a car in one of her many suicide attempts. She subsequently made a good recovery after slowly stopping the venlafaxine. In this case, venlafaxine metabolism would have been severely diminished due to her 2D6 genotype (*2/*4) and the alternative pathways, 2C19 and 3A4, would also have been compromised through the combination of 2C19 genotype (*1/*3) and the inhibition of 3A4 by components of cannabis.

20 Discussion In all of the cases presented here, the subjects were prescribed antidepressants that failed to mitigate distress emerging from their predicaments, which encompassed psychosocial stressors such as bereavement, marital and relationship difficulties, and work-related stress. Every subject's emotional reaction worsened while their prescribing physicians continued the “trial and error” approach, increasing from standard or higher dose and/or switching to other antidepressants, with disastrous consequences. In some cases the violence ensued from changes occasioned by withdrawal and polypharmacy. In all of these cases, the subjects were put into a state of drug-induced toxicity manifesting as akathisia, which resolved only upon discontinuation of the antidepressant drugs. This paper has detailed and substantiated in specific terms how the metabolism of each of the antidepressant drugs used by the subjects would have been seriously impaired both before and at the time they committed or attempted homicide. They were experiencing severe reported side effects, adverse drug reactions due to impaired metabolism complicated by drug–drug interactions against a background of variant CYP450 alleles. It is important to consider that while the literature often reports on “major” pathways of metabolism based on studies including subjects with known polymorphisms, their “minor” pathways can become more important in certain circumstances, such as when major pathways are unavailable due to inhibition or genetic impairment. In addition, in some cases, psychoactive metabolites have different metabolic pathways that warrant investigation. It also bears mentioning here that CYP2C19 *17 occurred in 28/85 akathisia subjects and in 4 of the more severe cases, including one homozygous *17/*17, that are detailed in Table 3. This particular allele was previously concluded from a meta-analysis as likely having a limited clinical impact, however the results reported here point to a possible role in some of the severe adversity occasioned by the prescribing regimes as detailed in Table 3.90 Notably, CYP2C19 *17 has been previously cited as potentially being problematic for antidepressant therapy.12 In all the cases in Table 3 where a

21 2C19 *17 was present, 2D6 had genetically impaired metabolic capacity as well and this combination of ultra-rapid for 2C19 and diminished for 2D6 bears further investigation as to its impact on metabolism of psychiatric medications. Healy's evidence in a series of Daubert hearings concerning antidepressant-induced akathisia homicide cases was substantiated by documents obtained through court-ordered access to drug companies' archives.91 Akathisia suicides and homicides, particularly when they involved children, gave rise to the first antidepressant suicide advisories in 2003, amid a public outcry. Healy published the watershed review in 2003 “Lines of evidence on the risks of suicide with selective serotonin reuptake inhibitors.92 This challenged but did not arrest the ongoing denial by their manufacturers in the USA. Healy concluded “The data reviewed here, which indicate a possible doubling of the relative risk of both suicides and suicide attempts on SSRIs compared with older antidepressants or non-treatment, make it difficult to sustain a null hypothesis, i.e. that SSRIs do not cause problems in some individuals to whom they are given. Further studies or further access to data are indicated to establish the magnitude of any risk and the characteristics of patients who may be most at risk.” 92 The US Food and Drug Administration (FDA) was aware of successful litigation based on suicide epidemiology and, having held some public hearings, it further ordered that a black-box warning about their effects in children, part of a “class suicidality labelling language for antidepressants” be incorporated in each drug’s product information.93 The first advisory concerning adults was posted on March 22, 2004. It warned of, among other conditions and side effects, “worsening depression or the emergence of suicidality” and “hostility, aggressiveness, impulsivity, akathisia” in both “adult and pediatric patients being treated with antidepressants for major depressive disorder as well as for other indications, both psychiatric and nonpsychiatric”.94 In June 2005, the FDA conceded "causation" of suicide by antidepressants in that relative risk of suicide vs. untreated was doubled or more.

22 The FDA later extended suicide warnings to other medicines. In 2006, it reviewed 273 clinical trials but ignored suicides committed in the withdrawal period, which it still refers to and counts as "placebo suicides."93 By the criteria used by the FDA, the two homicides committed 2 and 4 days into withdrawal by Subjects 2 and 6 would not be counted as selective serotonin reuptake inhibitors (SSRI) events, as FDA calculations only count suicides on active substance or one day after stopping. These advisories and their history are published on the website of the FDA.94 Deteriorating outcomes in mental illness, deaths, violence, and suicide rates have been documented by epidemiologists and have increased up to 20- fold since 1924.95–99 Some people taking psychiatric drugs develop akathisia and some people who develop akathisia kill themselves or others. Yet the drugs can be effective in persons suffering serious depression, provided their doses are adjusted according to their ability to metabolize them normally and there is informed monitoring. Australian prescribers and patients have still not been advised of antidepressant suicide or homicide by relevant authorities, including the Therapeutic Goods Administration (TGA). The TGA did not follow or adopt any of the public health advisories produced by the FDA, nor did it demand that drug companies notify users or prescribers of these major changes. Further, drug companies were not forced to include the "class suicidality labelling language for antidepressants" in full or accurately, in Australian prescribing information. Australian institutions; health departments, both state and federal; medical boards; medical students; courts and coroners; and colleges are still being advised that the notion that an antidepressant induces suicide or homicide is a strange idea (Yolande Lucire, personal experience). The result of this lack of knowledge has been a trebling of demand for mental heath services, paralleled by rising costs since 1990, rising numbers of non-recovering patients and of suicides and homicides by patients under mental health care.100 Failing to make clear that these effects are not related to the psychiatric diagnosis, are not evidence of bipolar mania, schizophrenia, or major depression the criteria for

23 which stipulates an exclusion that the diagnosis to be made is not caused by a substance, a medication or other treatment101 causes some Australian prescribers to believe they are seeing an exacerbation of a psychiatric condition or a new disorder. Akathisia, a toxidrome, is a neurotoxic side effect, sometimes accompanied by a delirium, reported as a disjointed narrative resembling a natural dream. It is entirely different from a functional psychosis that manifests in a clear sensorium, often associated with delusions that rationalize the hallucinations. Census data from 2001 indicated that 4.7% of Australians were on or had recently used antidepressants and a total of 9.6% were similarly using psychotropic medications.102 Olfson and Marcus (2009) reported that in 2005, 1/10 Americans over the age of 6 was on antidepressant medication and these were more likely be given antipsychotic medications as well.103 The authors believe that they have identified some of the major problems that underpin adversity and akathisia induced by improper dosing of antidepressants. Attending to these through careful consideration of the published evidence on diagnosis and treatment using pharmacogenetic information and consideration of drug interactions can help reverse this public health catastrophe. The authors consider that public health advisories warning of the effects of interactions between illicit substances and medications may also be warranted. Medication-induced suicide and homicide Drugs as diverse in structure and function as hypericum (St John's wort),104 varenicline (Chantix®; Pfizer, New York, NY), oseltamivir (Tamiflu®; Genentech USA, Inc, San Francisco, CA), isotretinoin (Roaccutane®; Hoffman-La Roche, Basel, Switzerland), mefloquine (Lariam®; Hoffman-La Roache), metoclopramide (Maxolon®; Shire plc, St Helier, Jersey, UK; Reglan®; UCB, Brussels, Belgium), zolpidem (Stilnox®; Sanofi-Aventis, Paris, France;) calcium channel blockers, antiepileptic drugs mooted as “mood stabilizers,” reserpine, benzodiazepines, statins, and interferon105, all induce suicidal and homicidal thinking as an occasional side effect. Moore et al (2010) identified 1527 cases of violence, including homicides, disproportionally reported to the FDA for 31 drugs, including

24 varenicline, eleven antidepressants, six sedative/hypnotics and three drugs for attention deficit hyperactivity disorder.106 Akathisia has been known to be associated with suicide since the 1950s and with homicide since 1985.107,108 Shear et al (1983) reported homicides associated with akathisia after treatment with depot fluphenazine.109 Schulte (1985) described cases of suicide an homicide associated with heloperidol.110 Healy et al (2006) described nine cases of murder, suicide, and severe violence in patients who had not been mentally ill before being medicated.111 Bostock and Gardner, pharmacovigilantes, have collated information obtained from over 4600 media stories of massacres, homicides, suicides, and school and college shootings dating back to 1966, involving antidepressants old and new, and drugs prescribed for attention deficit hyperactivity disorder, detailing the drugs and legal defences in some cases.112 In the cases presented in this paper, concerning subjects with abnormal CYP450 metabolism (i.e. ultra-rapid and/or diminished), the antidepressant or its metabolites may have reached a toxic level in hours or days correlating with onset of intense dysphoria and akathisia. The symptoms of toxicity were not recognized, or were ignored by patient and/or treating doctor and, in many cases, the dose of the antidepressant was increased while various “antidotes” to side effects, like sleeping pills, nausea, and pain medications, were added. They were prescribed by clinicians educated by drug company representatives, available information, and key opinion leaders who receive substantial benefits from the makers of these drugs, an issue that is coming to light with whistle-blower (qui tam) cases taken by attorneys, state and federal, against the makers for fraudulent promotion.113 Healy (2006) has documented the details of these fraudulent promotions, but they remain outside the purview of regulatory agencies who approve, even subsidise, drugs and sanction ghost-written product information concerning their use.114 Conclusion A detailed history and mental-state examination in these subjects presented here can exclude functional mental illness (which still has no biological markers) and

25 confirm neurotoxicity. Pharmacogenetic evidence can assist with the re-diagnosis of the population that causes the increased demand for mental health services. Restoring them to normality will reduce that demand and associated costs by taking pressure off ambulances, hospitals, prisons, and forensic wards. Akathisia homicides have been defended as instances of involuntary intoxication both with and without genetic evidence.115 Some perpetrators (and victims) succeed in receiving damages from the manufacturers for failure to warn.116 The Innocence Project examined evidence from crimes that took place before DNA testing was available.117 Personalized medicine in health care can bring about corresponding personalized justice in clinics, tribunals, courts and morgues.118 The most basic textbook of psychiatry, the Diagnostic and Statistical Manual of Mental Disorders, appended with each new edition’s number, provides this Delphic, obscure, and ambiguous guidance about treating akathisia: “Worsening of psychotic symptoms or behavioral dyscontrol may lead to an increase in neuroleptic medication dose, which may exacerbate the problem.” Differentiating akathisia (a toxic psychosis) from schizophrenia would be a crucial differential diagnosis in psychiatry in that the appropriate treatment for one condition can be catastrophic for another. Akathisia can develop very rapidly after initiating or increasing neuroleptic medication.101 However, sudden (as opposed to slow) withdrawal of serotonin-boosting antidepressants (or other substances) in an akathisia sufferer may make the problem worse and staying on them, once akathisia has developed, is equally dangerous. It is the authors’ contention that prescribing antidepressants without knowing about CYP450 genotypes is like giving blood transfusions without matching for ABO groups. Similarly, it is essential that other drugs in a patient’s drug regimen be understood in terms of interacting with CYP450 enzymes as substrates, inducers, or inhibitors. In 2010, Piggott et al reviewed meta-analyses of efficacy trials and the Sequenced Treatment Alternatives to Relieve Depression (Star*D) study. They found that antidepressants were only marginally (2.7%) more efficacious

26 compared with placebos. The same meta-analyses documented profound publication bias, inflating their apparent efficacy as well as bias in failing to report the negative results. The STAR*D analysis found that the effectiveness of antidepressant therapies was probably even lower than the modest one reported by the study authors with a progressively increasing dropout rate across each study phase. The authors argue for a reappraisal of the current recommended standard of care of depression.119 Acknowledgments Thanks are extended to Healthscope Molecular in Clayton, Victoria, and Diversity Health Institute, Westmead Hospital, Westmead, New South Wales, for generosity in providing genetic testing.

Disclosure The authors report no conflicts of interest in this work.

27

New References 1. Wynn GH, Oesterheld JR, Cozza K, Armstrong SC. Clinical Manual of Drug Interaction Principles for Medical Practice. Arlington, VA: American Psychiatric Publishing; 2009. 2. Licinio J, Wong ML. The pharmacogenomics of depression. Pharmacogenomics J. 2001;1(3):175-177. Review. 3. Bertilsson I, Dahl M-L, Tybring G. Pharmacogenetics of antidepressants: clinical aspects. Acta Psychiatr Scand Suppl. 1997;391:14–21. Review. 4. The Human Cytochrome P450 (CYP) Allele Nomenclature Committee [homepage on the Internet]. Karolinska group; [updated 2008 Sep]. Available from: http://www.cypalleles.ki.se. Accessed May 23, 2011. Ingelman-Sundberg M, Daly AK, Nebert DW eds. 5. Ingelman-Sundberg M, Sim SC. Pharmacogenetic biomarkers as tools for improved drug therapy; emphasis on the cytochrome P450 system. Biochem Biophys Res Commun. 2010 May 21;396(1):90–94. Review. 6. de Leon J, Armstrong SC, Cozza KL. Clinical guidelines for using pharmacogenetic testing of CYP450 2D6 and CYP450 2C19 in psychiatry. Psychosomatics. 2006;47:75–85. 7. Zourková A, Hadasová E. Paroxetine-induced conversion of cytochrome P450 2D6 phenotype and occurence of adverse effects. Gen Physiol Biophys. 2003;22:103–113. 8. Gasche Y, Daali Y, Fathi M et al. Codeine intoxication associated with ultrarapid CYP2D6 metabolism. N Engl J Med. 2004;351:2827–2831. 9. Zackrisson AL, Lindblom B, Ahlner J. High frequency of occurrence of CYP2D6 gene duplication/multiduplication indicating ultrarapid metabolism among suicide cases. Clin Pharmacol Ther. 2010;88(3):354– 359. 10. Zackrisson AL, Holmgren P, Gladh AB, Ahlner J, Lindblom B. Fatal intoxication cases: cytochrome P450 2D6 and 2C19 genotype distributions. Eur J Clin Pharmacol. 2004 Oct;60(8):547-52. 11. Zackrisson AL, Holmgren P, Gladh AB, Ahlner J, Lindblom B. Cytochrome P450 2D6 genotyping of fatal intoxications using Pyrosequencing. Progress in Forensic Genetics 10. 2004;1261:36–38. 12. Sim SC, Risinger C, Dahl ML et al. A common novel CYP2C19 gene variant causes ultrarapid drug metabolism relevant for the drug response to proton pump inhibitors and antidepressants. Clin Pharmacol Ther. 2006;79(1):103–113 13. Sachdev P. The epidemiology of drug-induced akathisia: Part I. Acute akathisia. Schizophrenia Bulletin. 1995;21(3):431-49. Review. 14. Sachdev P. The epidemiology of drug-induced akathisia: Part II. Chronic, tardive, and withdrawal akathisias. Schizophrenia Bulletin. 1995;21(3):451-61. Review. 15. Drugs@FDA: FDA approved drug products [webpage]. Silver Spring, MD: US Food and Drug Administration [no date]. Available at:

28 http://www.accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm. Accessed April 28, 2011. 16. Vandel P, Haffen E, Vandel S, Bonin B, Nezelof S, Sechter D, Broly F, Bizouard P, Dalery J. Drug extrapyramidal side effects. CYP2D6 genotypes and phenotypes. Eur J Clin Pharmacol. 1999 Nov;55(9):659- 65. 17. Indiana University School of Medicine Department of Medicine, Division of Clinical Pharmacology. Drug Interactions: cytochrome P450 drug interaction table [website]. Indianapolis, Indiana University School of Medicine; 2011 [updated April 12, 2011]. Available at: http://medicine.iupui.edu/clinpharm/ddis/table.asp. Accessed April 12, 2011. 18. Baumann P, Hiemke C, Ulrich S et al. The AGNP-TDM expert group consensus guidelines: therapeutic drug monitoring in psychiatry. Pharmacopsychiatry. 2004;37(6):243–265. 19. Breggin PR. Suicidality, violence and mania caused by selective serotonin reuptake inhibitors (SSRIs): a review and analysis. Int J Risk Saf Med. 2003;16(1):31–49. 20. Large M, Sharma S, Compton MT, Slade T, Nielssen O. Cannabis use and earlier onset of psychosis: a systematic meta-analysis. Arch Gen Psychiatry. Epub 2011 Feb 7. 21. Piatkov I, Jones T, Rochester C. Cytochrome P450 loss-of-function polymorphism genotyping on the Agilent Bioanalyzer and clinical application. Pharmacogenomics. 2009 Dec;10(12):1987-94. 22. Gaedigk A, Simon SD, Pearce RE, Bradford LD, Kennedy, MJ, Leeder JS. The CYP2D6 activity score: translating genotype information into a qualitative measure of phenotype. Clin Pharmacol Ther 2008. 83: 234-42. 23. Sjöqvist F, Alexanderson B, Asberg M et al. Pharmacokinetics and biological effects of nortriptyline in man. Acta Pharmacol Toxicol (Copenh). 1971;29 Suppl 3:255–280. 24. Bertilsson L, Eichelbaum M, Mellström B, Säwe J, Schulz HU, Sjöqvist F. Nortriptyline and antipyrine clearance in relation to debrisoquine hydroxylation in man. Life Sci. 1980;27(18):1673–1677. 25. Mellström B, Bertilsson L, Säwe J, Schulz HU, Sjöqvist F. E- and Z-10- hydroxylation of nortriptyline: relationship to polymorphic debrisoquine hydroxylation. Clin Pharmacol Ther. 1981;30(2):189–193. 26. von Bahr C, Birgersson C, Blanck A, Göransson M, Mellström B, Nilsell K. Correlation between nortriptyline and debrisoquine hydroxylation in the human liver. Life Sci. 1983;33(7):631–636. 27. Dahl ML, Nordin C, Bertilsson L. Enantioselective hydroxylation of nortriptyline in human liver microsomes, intestinal homogenate, and patients treated with nortriptyline. Ther Drug Monit. 1991;13(3):189–194. 28. Kin NM, Klitgaard N, Nair NP et al. Clinical relevance of serum nortriptyline and 10-hydroxy-nortriptyline measurements in the depressed elderly: a multicenter pharmacokinetic and pharmacodynamic study. Neuropsychopharmacology. 1996;15(1):1–6.

29 29. Perry PJ, Zeilmann C, Arndt S. Tricyclic antidepressant concentrations in plasma: an estimate of their sensitivity and specificity as a predictor of response. J Clin Psychopharmacol. 1994;14(4):230–240. Review. 30. Steimer W, Zöpf K, von Amelunxen S, Pfeiffer H, Bachofer J, Popp J, Messner B, Kissling W, Leucht S. Amitriptyline or not, that is the question: pharmacogenetic testing of CYP2D6 and CYP2C19 identifies patients with low or high risk for side effects in amitriptyline therapy. Clin Chem. 2005 Feb;51(2):376-85 31. Venkatakrishnan K, von Moltke LL, Greenblatt DJ. Nortriptyline E-10- hydroxylation in vitro is mediated by human CYP2D6 (high affinity) and CYP3A4 (low affinity): implications for interactions with enzyme-inducing drugs. J Clin Pharmacol. 1999;39(6):567–577. Erratum in: J Clin Pharmacol. 1999;39(8):866. 32. Brøsen K, Kragh-Sørensen P. Concomitant intake of nortriptyline and carbamazepine. Ther Drug Monit. 1993;15(3):258–260. 33. Krähenbühl S, Smith-Gamble V, Hoppel CL. Pharmacokinetic interaction between diltiazem and nortriptyline. Eur J Clin Pharmacol. 1996;49(5):417–419. 34. Gannon RH, Anderson ML. Fluconazole-nortriptyline drug interaction. Ann Pharmacother. 1992;26(11):1456–1457. 35. Bebchuk JM, Stewart DE. Drug interaction between rifampin and nortriptyline: a case report. Int J Psychiatry Med. 1991;21(2):183–187. 36. Zhou S, Gao Y, Jiang W, Huang M, Xu A, Paxton JW. Interactions of herbs with cytochrome P450. Drug Metab Rev. 2003;35(1):35–98. 37. Lefebvre T, Foster BC, Drouin CE, Krantis A, Livesey JF, Jordan SA. In vitro activity of commercial valerian root extracts against human cytochrome P450 3A4. J Pharm Pharm Sci. 2004;12;7(2):265–273. 38. Donovan JL, DeVane CL, Chavin KD, Wang JS, Gibson BB, Gefroh HA, Markowitz JS. Multiple night-time doses of valerian (Valeriana officinalis) had minimal effects on CYP3A4 activity and no effect on CYP2D6 activity in healthy volunteers. Drug Metab Dispos. 2004 Dec;32(12):1333-6. 39. Armstrong SC, Cozza KL, Sandson NB, Six Patterns of drug-drug interactions. Psychosomatics. 2003;44(3):255–258. 40. Spina E, Santoro V, D'Arrigo C. Clinically relevant pharmacokinetic drug interactions with second-generation antidepressants: an update. Clin Ther. 2008;30(7):1206–1227. Review. 41. Margolis JM, O'Donnell JP, Mankowski DC, Ekins S, Obach RS. (R)-, (S)- , and racemic fluoxetine N-demethylation by human cytochrome P450 enzymes. Drug Metab Dispos. 2000;28(10):1187–1191. 42. Scordo MG, Spina E, Dahl ML, Gatti G, Perucca E. Influence of CYP2C9, 2C19 and 2D6 genetic polymorphisms on the steady-state plasma concentrations of the enantiomers of fluoxetine and norfluoxetine. Bas Clin Pharmacol Toxicol. 2005;97(5):296–301. 43. Ring BJ, Eckstein JA, Gillespie JS, Binkley SN, VandenBranden M, Wrighton SA. Identification of the human cytochromes p450 responsible

30 for in vitro formation of R- and S-norfluoxetine. J Pharmacol Exp Ther. 2001;297(3):1044–1050. 44. Fjordside L, Jeppesen U, Eap CB, Powell K, Baumann P, Brøsen K. The stereoselective metabolism of fluoxetine in poor and extensive metabolizers of sparteine. Pharmacogenetics. 1999;9(1):55–60. 45. Yamaori S, Ebisawa J, Okushima Y, Yamamoto I, Watanabe K. Potent inhibition of human cytochrome P450 3A isoforms by cannabidiol: Role of phenolic hydroxyl groups in the resorcinol moiety. Life Sci. 2011 11;88(15–16):730–736. 46. Sachse-Seeboth C, Pfeil J, Sehrt D et al. Interindividual variation in the pharmacokinetics of Delta9-tetrahydrocannabinol as related to genetic polymorphisms in CYP2C9. Clin Pharmacol Ther. 2009;85(3):273–276. 47. Watanabe K, Yamaori S, Funahashi T, Kimura T, Yamamoto I. Cytochrome P450 enzymes involved in the metabolism of tetrahydrocannabinols and cannabinol by human hepatic microsomes. Life Sci. 2007 20;80(15):1415–1419. 48. Bland TM, Haining RL, Tracy TS, Callery PS. CYP2C-catalyzed delta9- tetrahydrocannabinol metabolism: kinetics, pharmacogenetics and interaction with phenytoin. Biochem Pharmacol. 2005 1;70(7):1096–1103. 49. Prozac Weekly® [Product Information]. Eli Lilly And Company Indianapolis, IN; 2001 for NDA 18-936/S-060/062). http://www.accessdata.fda.gov/drugsatfda_docs/nda/2001/21- 235_Prozac_Prntlbl.pdf. Accessed June 13, 2011. 50. Sawamura K, Suzuki Y, Someya T. Effects of dosage and CYP2D6- mutated allele on plasma concentration of paroxetine. Eur J Clin Pharmacol. 2004;60(8):553–557. 51. Charlier C, Broly F, Lhermitte M, Pinto E, Ansseau M, Plomteux G. Polymorphisms in the CYP 2D6 gene: association with plasma concentrations of fluoxetine and paroxetine. Ther Drug Monit. 2003;25(6):738–742. 52. Sindrup SH, Brøsen K, Gram LF et al. The relationship between paroxetine and the sparteine oxidation polymorphism. Clin Pharmacol Ther. 1992;51(3):278–287. 53. Jeppesen U, Gram LF, Vistisen K, Loft S, Poulsen HE, Brøsen K. Dose- dependent inhibition of CYP1A2, CYP2C19 and CYP2D6 by citalopram, fluoxetine, fluvoxamine and paroxetine. Eur J Clin Pharmacol. 1996;51(1):73–78. 54. Sindrup SH, Brøsen K, Gram LF. Pharmacokinetics of the selective serotonin reuptake inhibitor paroxetine: nonlinearity and relation to the sparteine oxidation polymorphism. Clin Pharmacol Ther. 1992;51(3):288– 295. 55. Breggin PR. Intoxication anosognosia: the spellbinding effect of psychiatric drugs. Ethical Hum Psychol Psychiatry. 2006;8(3): 201–205. 56. Rudberg I, Hermann M, Refsum H, Molden E. Serum concentrations of sertraline and N-desmethyl sertraline in relation to CYP2C19 genotype in psychiatric patients. Eur J Clin Pharmacol. 2008;64(12):1181–1188.

31 57. Gunes A, Bilir E, Zengil H, Babaoglu MO, Bozkurt A, Yasar U. Inhibitory effect of valproic acid on cytochrome P450 2C9 activity in epilepsy patients. Bas Clin Pharmacol Toxicol. 2007;100(6):383–386. 58. Effexor XR® [Product Information]. Wyeth Pharmaceuticals Inc., Philadelphia, PA; November 2005. http://www.accessdata.fda.gov/drugsatfda_docs/label/2005/020699s054, 057lbl.pdf Accessed June 13, 2011. 59. Hermann M, Hendset M, Fosaas K, Hjerpset M, Refsum H. Serum concentrations of venlafaxine and its metabolites O-desmethylvenlafaxine and N-desmethylvenlafaxine in heterozygous carriers of the CYP2D6*3, *4 or *5 allele. Eur J Clin Pharmacol. 2008;64(5):483–487. 60. Shams ME, Arneth B, Hiemke C et al. CYP2D6 polymorphism and clinical effect of the antidepressant venlafaxine. J Clin Pharmacol Ther. 2006;31(5):493–502. 61. Fukuda T, Nishida Y, Zhou Q, Yamamoto I, Kondo S, Azuma J. The impact of the CYP2D6 and CYP2C19 genotypes on venlafaxine pharmacokinetics in a Japanese population. Eur J of Clin Pharmacol. 2000;56(2):175–180 62. Lindh JD, Annas A, Meurling L, Dahl ML, AL-Shurbaji A. Effect of ketoconazole on venlafaxine plasma concentrations in extensive and poor metabolisers of debrisoquine. Eur J Clin Pharmacol. 2003;59(5–6):401– 406. Epub 2003 Jul 25. 63. Fogelman SM, Schmider J, Venkatakrishnan K et al. O- and N- demethylation of venlafaxine in vitro by human liver microsomes and by microsomes from cDNA-transfected cells: effect of metabolic inhibitors and SSRI antidepressants. Neuropsychopharmacology. 1999;20(5):480– 490. 64. Sriwiriyajan S, Samaeng M, Ridtitid W, Mahatthanatrakul W, Wongnawa M. Pharmacokinetic interactions between ciprofloxacin and itraconazole in healthy male volunteers. Biopharm Drug Dis. 2011;32(3):168–174. Epub 2011 Mar 1. 65. Shahzadi A, Javed I, Aslam B et al. Therapeutic effects of ciprofloxacin on the pharmacokinetics of carbamazepine in healthy adult male volunteers. Pak J Pharm Sci. 2011;24(1):63–68. 66. Elko CJ, Burgess JL, Robertson WO. Zolpidem-associated hallucinations and serotonin reuptake inhibition: a possible interaction. J Toxicol Clin Toxicol. 1998;36(3):195–203. Review. 67. Vlase L, Popa A, Neag M, Muntean D, Leucuţa SE. Pharmacokinetic interaction between zolpidem and ciprofloxacin in healthy volunteers. Eur J Drug Metab Pharmacokinet. 2011;35(3–4):83–87. Epub 2010 Oct 5. 68. Owens MJ, Knight DL, Nemeroff CB. Second-generation SSRIs: human monoamine transporter binding profile of escitalopram and R-fluoxetine. Eur J Drug Metab Pharmacokinet. 2001;50(5):345–350. 69. Olesen OV, Linnet K. Studies on the stereoselective metabolism of citalopram by human liver microsomes and cDNA-expressed cytochrome P450 enzymes. Pharmacology. 1999;59(6):298–309.

32 70. Herrlin K, Yasui-Furukori N, Tybring G, Widén J, Gustafsson LL, Bertilsson L. Metabolism of citalopram enantiomers in CYP2C19/CYP2D6 phenotyped panels of healthy Swedes. British J Clin Pharmacol. 2003;56(4):415–421. 71. von Moltke LL, Greenblatt DJ, Giancarlo GM, Granda BW, Harmatz JS, Shader RI. Escitalopram (S-citalopram) and its metabolites in vitro: cytochromes mediating biotransformation, inhibitory effects, and comparison to R-citalopram. Drug Metab Dispos. 2001;29(8):1102–1109. 72. Mannel M. Drug interactions with St John's wort : mechanisms and clinical implications. Drug Saf. 2004;27(11):773-97. Review. 73. Robinson, DS. Serotonin Syndrome. Primary Psychiatry. 2006;13(8):36– 38. 74. Fukuda T, Yamamoto I, Nishida Y et al. Effect of the CYP2D6*10 genotype on venlafaxine pharmacokinetics in healthy adult volunteers. Br J Clin Pharmacol. 1999;47(4):450–453. 75. Fukuda T, Nishida Y, Zhou Q, Yamamoto I, Kondo S, Azuma J. The impact of the CYP2D6 and CYP2C19 genotypes on venlafaxine pharmacokinetics in a Japanese population. Eur J Clin Pharmacol. 2000 May;56(2):175-80. 76. Veefkind AH, Haffmans PM, Hoencamp E. Venlafaxine serum levels and CYP2D6 genotype. Ther Drug Monit. 2000;22 (2):202–208. 77. Safeekh AT, Pinto D. Venlafaxine-induced psychotic symptoms. Indian J Psychiatry. 2009;51(4):308–309. 78. Pfeffer F, Grube M. An organic psychosis due to a venlafaxine- propafenone interaction. Int J Psychiatry Med. 2001;31(4):427–432. 79. Gareri P, De Fazio P, Gallelli L et al. Venlafaxine-propafenone interaction resulting in hallucinations and psychomotor agitation. Ann Pharmacother. 2008 Mar;42(3):434–438. 80. Howe C, Ravasia S. Venlafaxine-induced delirium. Can J Psychiatry. 2003;48(2):129. 81. Alexander J, Nillsen A. Venlafaxine-induced delirium. Aus N Z J Psychiatry. Epub 2011 Mar 24. 82. Efexor®-XR [product information]. Baulkham Hills, NSW: Wyeth Australia Pty Ltd; February, 2010. 83. Rochat B, Amey M, Gillet M, Meyer UA, Baumann P. Identification of three cytochrome P450 isozymes involved in N-demethylation of citalopram enantiomers in human liver microsomes. Pharmacogenetics. 1997;7(1):1–10. 84. Yu BN, Chen GL, He N et al. Pharmacokinetics of citalopram in relation to genetic polymorphism of CYP2C19. Drug Metab Dispos. 2003;31(10):1255–1259. 85. Pedersen RS, Damkier P, Brosen K. Tramadol as a new probe for cytochrome P450 2D6 phenotyping: a population study. Clin Pharmacol Ther. 2005 Jun;77(6):458-67. 86. Tang C, Shou M, Rushmore TH, Mei Q, Sandhu P, Woolf EJ, Rose MJ, Gelmann A, Greenberg HE, De Lepeleire I, Van Hecken A, De Schepper

33 PJ, Ebel DL, Schwartz JI, Rodrigues AD. In-vitro metabolism of celecoxib, a cyclooxygenase-2 inhibitor, by allelic variant forms of human liver microsomal cytochrome P450 2C9: correlation with CYP2C9 genotype and in-vivo pharmacokinetics. Pharmacogenetics. 2001 Apr;11(3):223-35. 87. Werner U, Werner D, Rau T, Fromm MF, Hinz B, Brune K. Celecoxib inhibits metabolism of cytochrome P450 2D6 substrate metoprolol in humans. Clin Pharmacol Ther. 2003 Aug;74(2):130-7. 88. Halberg P, Karlsson J, Kurland L et al. The CYP2C9 genotype predicts the blood pressure response to irbesartan: results from the Swedish Irbesartan Left Ventricular Hypertrophy Investigation vs Atenolol (SILVHIA) trial. J Hypertens. 2002;20(10):2089–2093. 89. Cham S, Evans MA, Denenberg JO, Golomb BA. Statin-associated muscle-related adverse effects: a case series of 354 patients. Pharmacotherapy. 2010 Jun;30(6):541-53. 90. Li-Wan-Po A, Girard T, Farndon P, Cooley C, Lithgow J. Pharmacogenetics of CYP2C19: functional and clinical implications of a new variant CYP2C19*17. Br J Clin Pharmacol. 2010 Mar;69(3):222-30. Review. 91. Healy D. Let Them Eat Prozac: The Unhealthy Relationship between the Pharmaceutical Industry and Depression. New York: New York University Press; 2004. The evidence given in the Daubert Hearings is available at: http://www.healyprozac.com/Trials/default.htm. Accessed May 26, 2011. 92. Healy D. Lines of evidence on the risks of suicide with selective serotonin reuptake inhibitors. Psychother Psychosom. 2003 Mar-Apr;72(2):71-9. Review. 93. Laughren TP. Memorandum: overview for December 13 meeting of Psychopharmacologic Drugs Advisory Committee (PDAC). Silver Spring, MD: US Department of Health and Human Services Public Health Service, Food and Drug Administration Center for Drug Evaluation and Research; November 16, 2006. Available at: http://www.fda.gov/ohrms/DOCKETS/ac/06/briefing/2006-4272b1-01- FDA.pdf. Accessed April 12, 2011. 94. US Food and Drug Administration (FDA). Antidepressant use in children, adolescents, and adults. US Food and Drug Administration [website]. Silver Spring, MD: FDA; [updated Aug 12, 2010]. Available at: http://www.fda.gov/drugs/drugsafety/informationbydrugclass/ucm096273. htm. Accessed April 12, 2011. 95. Healy D, Harris, R, Tranter R et al. Lifetime suicide rates in treated schizophrenia: 1875–1924 and 1994–1998 cohorts compared. Br J Psychiatry. 2006;188:223–228. 96. Lawrence D. Jablensky AV. Holman CD. Pinder TJ. Mortality in Western Australian psychiatric patients. Soc Psychiatry Psychiatr Epidemiol. 2000;35(8):341–347. 97. Saha S, Chant D, McGrath J. A systematic review of mortality in schizophrenia: is the differential mortality gap worsening over time? Arch Gen Psychiatry. 2007;64(10):1123–1113.

34 98. Colton CW, Manderscheid RW. Congruencies in increased mortality rates, years of potential life lost, and causes of death among public mental health clients in eight states. Pre Chronic Dis. 2006;3(2)A42. 99. Burgess P, Pirkis J, Jolley D, Whiteford H, Saxena S. Do nations’ mental health policies, programs and legislation influence their suicide rates? An ecological study of 100 countries. Aus N Z J Psychiatry. 2004;38(11– 12):933–939. 100. New South Wales (NSW) Mental Health Sentinel Events Review Committee. Tracking Tragedy: a systemic look at homicide and non-fatal serious injury by mental health patients, and suicide death of mental health inpatients. Fourth Report of the Committee. NSW Mental Health Sentinel Events Review Committee; March 2008. (Updated and confirmed by question 10218 in NSW parliament.) Available at: http://www.health.nsw.gov.au/pubs/2009/pdf/tracking_tragedy_2008_fourt h_report.PDF. Accessed April 12, 2011. 101. American Psychiatric Association. Schizophrenia and other psychotic disorders and Mood Disorders. Diagnostic and Statistical Manual of Mental Disorders (DSM IV), 4th ed. Washington DC: American Psychiatric Association; 1994: 273-391. 102. Australian Bureau of Statistics, National Health Survey: Mental Health, Australia, 2001. http://www.abs.gov.au/ausstats/[email protected]/mf/4811.0 Accessed June 9, 2011. 103. Olfson M, Marcus SC. National patterns in antidepressant medication treatment. Arch Gen Psychiatry. 2009;66(8):848-856. 104. Nanayakkara PW, Meijboom M, Schouten JA. [Suicidal and aggressive thoughts as a result of taking a Hypericum preparation (St. John's wort)]. Ned Tijdschr Geneeskd. 2005 Jun 11;149(24):1347–1349. 105. Marks DH, Breggin PR, Braslow D. Homicidal Ideation Causally Related to Therapeutic Medications Int J Risk Saf Med. 2008;20(4):231–240 106. Moore TJ, Glenmullen J, Furberg CD, Prescription drugs associated with reports of violence towards others. PLoS One. 2010;5(12):e15337. 107. Healy D, Aldred G. Antidepressant drug use & the risk of suicide. Int Rev Psychiatry. 2005;17(3):163–172. 108. Cem Atbaşoglu E, Schultz, SK, Andreasen NC. The relationship of akathisia with suicidality and depersonalization among patients with schizophrenia. J Neuropsychiatry Clin Neurosci. 2001;13(3):336–341. 109. Shear MK, Frances A, Weiden P. Suicide associated with akathisia and depot fluphenazine treatment. J Clin Psychopharmacol. 1983;3(4):235–236. 110. Schulte JL. Homicide and suicide associated with akathisia and haloperidol. Am J Forensic Psychiatry. 1985;6(2):3–7.

35 111. Healy D, Herxheimer A, Menkes DB. Antidepressants and violence: problems at the interface of medicine and law. PLoS Medicine / Public Library of Science. 2006 Sep;3(9):e372. 112. SSRI Stories: Antidepressant nightmares [homepage on the Internet]. www.ssristories.com Accessed June 9, 2011 113. Newman M. Bitter pills for drug companies. BMJ 2010; 341:c5095 114. Healy. D. Manufacturing Consensus. Culture, Medicine and Psychiatry 30: 135–156, 2006. 115. R. v Hawkins [2001] NSWSC 420, 2001, other similar cases with defences are detailed at http://www.ssristories.com also http://www.healyprozac.com 116. Cassidy vs. Eli Lilly Company, 2002, Civil Action No. 821, United States District Court For The Western District of Pennsylvania http://www.prozactruth.com/genetictesting.htm (accessed 29/04/2011) 117. Innocence Project [homepage on the Internet]. 40 Worth St., Suite 701 New York, NY 10013. http://www.innocenceproject.org accessed June 13, 2011 118. Wong SH, Happy C. Blinka D, Gock, S. Jentzen JM, Donald J. Coleman, et al. Editorial: From personalized medicine to personalized justice: The promises of translational pharmacogenomics in the justice system. Pharmacogenomics 11: (6) p. 731- 2010. 119. Pigott HE, Leventhal AM, Alter GS, Boren JJ. Efficacy and Effectiveness of Antidepressants: Current Status of Research. Psychother. Psychosom.. 2010;79:267–279.

36 Figure 1: Prevalence of individuals with different numbers of variant CYP450 alleles in two groups: Akathisia subjects and randomly selected primary care patients.

40 35 30 25 % of 20 sample 15 10 5 0 01234 Number of variant alleles

Akathisia Group (n=85) Random Sample Group (n=150)

37 Table 1: Drug metabolizing CYP450 phenotypes and polymorphic CYP450 alleles of importance for drug response

Phenotype Genetic basis Clinical Examples of alleles Examples of clinical consequences consequences causing the phenotype UM More than two Lack of response CYP2C19*17 CYP2C19: increased bleeding risk with active gene copies of the parent drug clopidogrel on the same allele, CYP2D6*1*2 x N or increased Increased ADRs (gene duplication/ CYP2D6: nonresponse of antiemetics, expression of a due to increased multiduplication) antidepressants and increased suicide risk single gene metabolite or active drug ADRs after codeine and tramadol treatment production EM Two functional Ordinary CYP2C9*1, Normal metabolism alleles response CYP2C19*1, CYP2D6*1 IM One defective Increased CYP2C9*2, Reduced response to tamoxifen and to allele or two concentration of CYP2D6*10, analgesic treatment partially defective parent drug CYP2D6*41 alleles Decreased metabolite formation PM Two defective Higher levels of CYP2C9*3, Increased ADRs alleles the parent drug CYP2C19*2, CYP2C19*3, Reduced response to analgesics and Increased risk for CYP2D6*4, tamoxifen ADRs CYP2D6*5

Note: Reproduced with permission from Ingelman Sundberg and Sim.5 Abbreviations: ADR, adverse drug reaction; UM, ultra-rapid metabolizer; EM, extensive metabolizer; IM, intermediate metabolizer; PM, poor metabolizer.

38 Table 2: Drugs involved in these cases, with therapeutic windows and CYP450 attribution as metabolic substrates, inhibitors and inducers

Drug Therapeutic CYP450 Metabolism Enzymes inhibited Enzymes induced window Alcohol 2E1d 2E1 (acute) 2E1 (chronic) Cannabis and its 2C9, 3A4 2C9, 3A4, 3A5 2C9 metabolites45-48 Ciprofloxacin 3A4 1A2a, 3A4a Citalopram 30–130 ng/mL 2C19, 2D6, 3A4 2D6c Escitalopram 15–80 ng/mL 2C19, 2D6, 3A4 2D6c Estrogen 2C9 (minor), 3A4 2C19b, 3A4c Fluoxetine 120–300 ng/mL 2C9, 2C19, 2D6, 3A4 1A2c, 2C9b, 2C19b 2D6a, 3A4b, (including 2B6b norfluoxetine) Fluvoxamine 150–300 ng/mL 1A2, 2D6 1A2a, 2B6b, 2C9b, 2C19a, 2D6b, 3A4b, Hypericum 2C9 (long term), 2C19, 3A4 Nortriptyline24-27, 31- 70–170 ng/mL 2D6, 3A4 2C19, 2D6 35 Paroxetine 70–120 ng/mL 2D6, 3A4 1A2c, 2B6a, 2C9c, 2C19c 2D6a, 3A4b, Sertraline 10-50 ng/mL 2B6, 2C9, 2C19, 2D6, 3A4 1A2c, 2B6b, 2C9b, 2C19b, 2D6e, 3A4b Serotonin (5HT) 2C9 Valerian36-38 3A4 Valproate 50–100 μg/mL Complex: 2A6, 2C9, 2C19 2C9, 2D6 Uncertain, possibly 3A4 Venlafaxine 195–400 ng/mL 2C19, 2D6, 3A4 2D6c (including O- desmethylvenlafa xine Zolpidem 90–325 ng/mL 1A2, 2C9, 3A4 (major) Notes: Where not referenced, data is from Wynn et al.1 aPotent inhibition; bmoderate inhibition; cmild inhibition; dother hepatic and extrahepatic pathways important eSertraline is a potent inhibitor at high concentrations (typically at doses>200mg/day) Information about therapeutic window is from Baumann et al.18

39 Table 3: Information on Subjects’ drug regimen, CYP450 genotypes, criminal acts, and outcomes for ten of those suffering violent akathisia

Subject Drug, dose and Genotypes Circumstances of crime Outcome Number treatment duration # of null, Gender diminished or ultra-rapid Age Reason for metabolizer treatment alleles 1 Nortriptyline 2D6 *4/*41 Killed teenage daughter in toxic Pleaded mental illness, (12.5 mg/day) 2C9 *1/*2 delirium after 3 days misdiagnosed with schizophrenia F superimposed 2C19 *1/*17 and treated with more CYP450 on valerian One null allele substrates 35 Two diminished Recovered on withdrawal but more Treated for alleles drugs are still being prescribed on distress due to One ultra-rapid Tribunal orders with near fatal husband’s allele consequences. drinking 2 Fluoxetine (10 2D6 *5/*4 Violent akathisia for 14 days; 60-year sentence, wants to appeal mg/day) 2C9 *1/*1 killed father 4 days after subject Recovered on withrawal. M 2C19 *1/*17 ran out of pills 14 days then 4 Two null alleles 18 days in One ultra-rapid withdrawal allele superimposed on cannabis use.

Treated because twin sister was comatose after

40 a car crash

3 Paroxetine (20 2D6 *2/*4 Stabbed former partner 30+ Suspected of bipolar disorder before mg/day) minimal 2C9 *1/*2 times after 11 weeks of stopping paroxetine M alcohol 2C19 *1/*2 akathisia on paroxetine Convicted of murder Two null alleles Psychological blow automatism, Recovered on withdrawal 35 Distressed by One diminished no memory “on and off” allele relationship with mother of his child 4 Paroxetine (40 2D6 *1/*41 Killed son in a manic-shift Medical supervision mg/day) 40 2C9 *1/*1 akathisia/delirium after 42 days Recovered on withdrawal M weeks. 2C19 *17/*17 on paroxetine and 20 days after Suicidal, but One diminished its increase to 60 mg/day 46 resumed and allele increased to 60 Two ultra-rapid mg/day alleles

Anxiety about not making enough money to support family; insomnia and weight loss

41 5 Fluoxetine (60 2D6 *1/*6 Killed therapist in hospital after 14 years in jail for murder; wants to mg/day, 2C9 *1/*1 11 weeks, in hospital for suicide appeal M valproate 1000 2C19 *1/*1 attempts on sertraline, then on Recovered on withdrawal mg/day) One null allele fluoxetine 16 Treated because depressed, in a group home, struggling at school, and girlfriend left 6 Initially 2D6 *1/*2 3 years of violent akathisia on Appeal by state against short fluoxetine then 2C9 *1/*1 polypharmacy. In withdrawal sentence was unsuccessful M amitriptriptyline. 2C19 *1/*2 from venlafaxine (4 days), Recovered on withdrawal Valproate One null allele subject shot a stranger with 50 (800/mg day) friend's gun; subject recalled throughout nothing (polypharmacy) Homicide occurred in a state of Then somnambulism venlafaxine (300 mg/day) + valproate + zolpidem and/or zopiclone for 2 years Finally, withdrawal from venlafaxine while on a vitamin mixture containing serotonin,

42 ciprofloxacin, and zolpidem (10 mg) taken repeatedly Treated because of distress over divorce. 7 Sertraline, then 2D6 *1/*41 Nearly killed partner after 2 12 years in jail for attempted murder: escitalopram (10 2C9 *1/*2 months of uncharacteristic wished to appeal M mg/day) + 2C19 *1/*2 violence Recovered on withdrawal hypericum (900 One null allele 24 mg/day) caused Two diminished a series of alleles suicide attempts and assaults culminating in near lethal violence

Treated for anxiety and illicit substance use 8 Paroxetine (20 2D6 *2/*4 Several suicide attempts and Pleaded diminished responsibility to mg/day) 2C19 *2/*17 two attempts to kill two children four charges of attempted murder, no F followed by 2C9 *1/*1 jail time. fluoxetine, Two null alleles Recovered on withdrawal 26 sertraline, One ultra-rapid citalopram then allele venlafaxine (300 mg/day) for 2 years

43

Treated for difficulties with in-laws 9 Paroxetine (to 60 2D6 *4/*4 Tried to kill two children after 3 Hospitalized: venlafaxine for another mg/day) then 2C9 *1/*1 years of akathisia on all SSRIs, year until unassisted withdrawal F citalopram, 2C19 *1/*1 against a background of Further episodes of akathisia with fluoxetine then Two null alleles polypharmacy involving up to tramadol, simvastatin, conjugated 52 venlafaxine (300 ten drugs at a time estrogens on their own mg/day) and Recovered from akathisia on intermittent withdrawal but never worked again metoclopramide as needed, tramadol as needed, diazepam 5 mg as needed, conjugated estrogens (Premarin) 0.625 mg, pravastatin 40 mg, sometimes altovastatin, irbesartan (300 mg/day), tenormin (12.5 mg/day), valacyclovir for shingles, seretide, various antibiotics,

44 penicillin, cephalexin, erythromycin, celecoxib (200 mg ad lib), oroxine (100 mg/day), various antibiotics, flu vaccination.

Treated for harassment at work 10 Citalopram (dose 2D6 *2/*4 Personality changed to Subject was switched to venlafaxine unknown), then 2C9 *1/*2 aggressive and suicidal; subject on which she continued to make F venlafaxine 2C19 *1/*1 jumped in front of train with child suicide attempts for 4 years until superimposed on One null allele and lost custody for years withdrawal and full recovery 25 heavy cannabis One diminished use. allele

Treated for marital distress

Note: All subjects were tested at Healthscope Molecular except Subject 5, who was tested for the same alleles at Genelex (Seattle, WA, USA). Abbreviation: SSRI, selective serotonin reuptake inhibitor.

45