SEARCH

SEARCH BY CITATION

Keywords:

  • pharmacogenomics;
  • depression;
  • antidepressants;
  • service user research;
  • ambivalence

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Data and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

Whilst antidepressant medications are widely used, they are ineffective for nearly 40 per cent of users and cause numerous adverse drug reactions. The pharmacogenomics of depression attempts to better understand the role of genetic variation in antidepressant metabolism in the hope of improving drug efficacy and tolerability. In this paper we present findings from a series of focus groups with the general public and with mental health service users in four European sites. Results indicate broad support for genome-based therapies for depression. Findings, however, also show a wide spread of ambivalence regarding the nature and causes of depression, as well as the use of antidepressant medication. We argue that these uncertainties may negatively impact public and user acceptability of the pharmacogenomics of antidepressants.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Data and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

The prevalence and virulence of depressive disorders are well established. It is estimated, for instance, that at least one in five persons will suffer from depression at some point in their lives (Solomon 2002). By 2020 the World Health Organization approximates that depression will account for nearly six per cent of the global disease burden and will rank second only to ischaemic heart disease as the leading cause of years lost to disability (WHO 2001).

Given the increasing diagnosis of depression, it is not surprising that the use of antidepressant medication has risen sharply in recent years.1 In the United States, antidepressant use has tripled in the last decade. It is estimated that one out of every ten women in the US is taking an antidepressant (Vedantam 2004). In England, 8.2 million prescriptions were dispensed in 1999; four years later, this figure had jumped to over 19 million (MHRA 2004). Yet with more than a dozen different types of antidepressants on licence, there still exists no clear criterion for practitioners to know which drug to give to which patient. Thus, physicians often rely on guesswork and personal preference when prescribing. There is no guesswork, however, as to the profits involved. The drug Prozac alone has earned $2.6 billion annually for its manufacturers Eli Lilly and Company. In total; the European market for psychotropic drugs reached $4,741 million in 2000 (Rose 2005).

Despite their widespread use, there are significant problems with antidepressants. Approximately 40 per cent of people who take them receive no therapeutic benefit. Even when a patient does respond to a particular type of antidepressant, there are often considerable adverse drug reactions, which often lead people to discontinue use, since the therapeutic benefit takes several weeks to begin but the side effects are immediate. Many patients report that the trial and error process of finding an antidepressant that works can often add to their anxiety and suffering.

The promise of pharmacogenomics

Given the pervasiveness of depression and the profitability of antidepressants, it is not hard to see why such drugs are an attractive target for pharmacogenomics – that is, for research which aims to understand better the genetic variations in the metabolic enzymes (in the case of many antidepressants thought to be CYP2D6 and CYP2C19), transporter genes, proteins, and carriers that contribute to antidepressant adverse drug reactions (ADRs). Whilst there is no dominant model as to how pharmacogenomics is, or will, be, adopted in a clinical setting, it is clear at least that progress is slow and uneven2 (Martin et al. 2006). Even when pharmacogenomics is at its best, information derived from tests is probabilistic and relative, not deterministic, providing help for prescribing physicians and their patients, but no patent solutions (Lindpaintner 2002).

Genome-based therapies for depression (GENDEP)

In this paper we aim to explore the range of factors that may impinge upon public and service user acceptability of the pharmacogenomics of antidepressants. We relate these findings both to clinical and sociological literature on depression and antidepressant medication, as well as to broader discussions regarding the development of pharmacogenomics as a promissory technology. We conclude that the uptake of genome-based therapies for depression cannot be separated from wider issues regarding the meanings of mental illness and the significance of taking drugs that have moods as their targets.

The project from which this paper originates formed the ELSI (Ethical, Legal and Social Issues) agenda of a study funded by the European Commission known as Genome-based therapies for depression (GENDEP), whose aim was to identify genetic variations that may affect responses to antidepressants. GENDEP was a three-year grant, begun in 2004, that aimed to recruit 1,000 depressed patients across eight European countries (England, Poland, Slovenia, Italy, Belgium, Denmark, Germany, and Croatia). To be included in the study, participants must have been diagnosed with at least moderate depression according to ICD-10 or DSM-IV criteria; aged 18–65; of either gender and of white ethnicity with European parentage. GENDEP excluded pregnant women, patients with a history of bipolar affective disorder or schizophrenia in first-degree relatives, as well as patients with substance abuse, primary organic disease, or treatment failure with one of the two antidepressants used in the study, Escitalopram or Nortripyline.

GENDEP participants were recruited through GP clinics or psychiatric care units. Researchers took DNA samples via blood, and attempted to correlate a patient's genetic information with their response to whichever study drugs the patient had been placed on. To make correlations between drug response and genetic make-up, GENDEP participants underwent a series of interviews and questionnaires to ascertain how they were feeling and to what degree they were experiencing side effects. Thus, patient volunteers received weekly phone calls and had several face-to-face meetings with GENDEP staff. The questionnaires were extensive and included the HAM-D, the Beck Depression Inventory, the Global Assessment Scale, MADRS, the UKU side-effects check-list, and a sexual functioning questionnaire. After 12 weeks, patients were discharged from the study and returned to the care of their original physician or psychiatrist.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Data and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

In seeking to investigate both general public and mental health service user views regarding the acceptability of the pharmacogenomics of depression, we held a series of focus groups in four European countries associated with the GENDEP study.3

General public focus groups

We organised a series of focus groups with members of the general public in: London, England; Poznan, Poland; Aarhus, Denmark; and Berlin, Germany. Each site held three groups with a random sample of eight participants each. Participants were of mixed gender, ethnicity and age. Participants in Denmark and Germany were recruited via public adverts in local newspapers and magazines. Participants in England and Poland were recruited through a social science marketing agency. Exclusion criteria were anyone with direct experience of depression or antidepressant medication. All group members were paid £25 or its equivalent to compensate for their time and travel costs.

A facilitation guide was developed with the aid of literature reviews. The guide was tested via two pilot focus groups with university students in London. All facilitation materials and consent forms were translated from English into the relevant language. Facilitators from Denmark and Poland attended the groups in London to observe our aims and methods. We travelled to Germany to assist with the organisation of the groups there. All facilitators were native speakers in the language in which the groups were conducted.

In addition to these focus groups, we also organised a half-day workshop (in London only) with recognised experts to discuss the policy-related implications of pharmacogenomics. Attendees included representatives from mental health charities, the pharmaceutical industry, academia, and psychiatry. Discussion was taped and transcribed.

Service user focus groups

The service user focus groups were conducted by researchers from the Service User Research Enterprise (SURE) at the Institute of Psychiatry in London. SURE conducts research from the service user perspective and is staffed mostly by people who are using or have used mental health services. In accordance with this, the service user focus groups in our study were facilitated by service user researchers.

Service user group participants were recruited through user organisations in each country. In Germany and England, participants were recruited consequent upon a survey which asked respondents whether they would take a pharmacogenomic test. In Poland, the participants were recruited direct via the national service user organisation.

The team experienced great difficulty in this recruitment process. One service user organisation in England refused to be part of the study because it dealt with pharmacology and genetics, two ideas that they objected to in principle. In Germany, recruitment was again difficult and laborious. The Board of the German user organisation initially refused to take part in the study but later relented on condition that the SURE researcher attend the group's assembly and present our project aims. At the assembly, our researcher met many objections but it was finally agreed that the research could proceed, an outcome which we partly attribute to the fact that she was well known to the group beforehand. The service user organisation in Denmark has, to date, refused to allow the research to go forward despite more than two years of contact. Again, we attribute this resistance to a rejection of the very principle of pharmacogenomics on behalf of the survivor/user movement.

As responses to the survey in England and Germany indicated high levels of agreement with pharmacogenomics, we became concerned that given the problems cited above with recruitment, our focus group findings might have been biased towards those with positive opinions. For example, in the English groups, responses might have been less supportive if the service user organisation opposed to genetic research had agreed to participate in our groups. We therefore recruited a group of users in Germany who had not completed the survey to ascertain why. We also ran a focus group with Board members of the pan-European user/survivor organisation European Network of (ex) Users and Survivors of Psychiatry (ENUSP). By using these methods of recruitment we hoped to provide a more balanced picture.4

Both public and service user group meetings lasted approximately two hours. Facilitation included an initial introduction to the pharmacogenomics of depression and then a series of prompts designed to get participants talking in general terms about the topic. The second half of public groups used two vignettes to elicit participant views on the main research questions of the study. Tapes were transcribed in their native language, then translated into English.

Analysis

The team worked together during the coding phase to explore key themes and categories in the data. This phase of data analysis helped establish the validity of categories and identify analytical themes. We used action-oriented sequential analysis when appropriate and where quality of the transcription permitted, in order to help provide a wider context for the discussion, group norms, and the processes by which respondents replied. (Kitzinger 1994; Wilkinson 2005). Thematic analysis was supported by NVivo 2. Initial categories were drawn from the topic guide. These were applied to see how far they captured the meanings in the texts. Further themes emerged on re-reading the transcripts and these were added to the coding frame. This process of iteration was repeated until theoretical saturation was reached (Silverman 2000).

Reflexivity in the research process: contributing to expectations

Like other biotechnologies that promise medical breakthroughs, pharmacogenomics is a ‘promissory technology’, built on the expectation that present research will reap future benefits. But since pharmacogenomics testing for depression is not at present a clinical reality, we faced the problem of having to simultaneously introduce the topic to respondents whilst asking their views on it. Inevitably, how we presented pharmacogenomics and our own beliefs regarding the technology will have impacted on the findings. There is some evidence, for example, that service users are more comfortable and more willing to take a critical stance when being interviewed by a service user researcher (Clark et al. 1999, Rose et al. 2003). In our presentation of pharmacogenomics, we tried to maintain a balance by explaining that GENDEP researchers were ‘exploring the genetic factors that are suspected to be involved in the metabolism of antidepressant medications’, and that ‘some people hope that understanding how genetic make-up can affect drug response will result in greater antidepressant efficacy, allowing physicians to prescribe drugs that have fewer adverse reactions’. In the latter half of group discussions, we used case-based vignettes to help us explore some of the ethical issues around pharmacogenomics, such as orphan medicines, racial stratification, and drug access. We amended the cases as appropriate to take into consideration differences between the sites’ various health care delivery systems.

In the course of conducting a large trans-national study such as this, it is possible to forget that we are not merely responding to the consequences of a technology but rather, are playing an active role in contributing to the production of hope surrounding the possible success of that technology (Hedgecoe and Martin 2003). Brown (2003) reminds us to be alert to the ‘situatedness of expectations’ and the role that hype can play in helping to legitimate research agendas to various publics. Indeed, it is the hitherto unfulfilled promise of pharmacogenomics which provided the rationale for the funding of GENDEP. Research such as ours can however plausibly cut both ways – that is, it may serve to help create a vision that mobilises patients to support a promissory technology, and/or it could help define the limits of the technology in terms of its public acceptability.

We note that at times our data are as much about depression and antidepressants as they are about pharmacogenomics. In part this may be a reflection of the fact that people are most comfortable talking about what they know best and, for reasons we have just described, this was not pharmacogenomics. We did attempt to keep discussion focused on pharmacogenomics as much as possible until it became clear that the desire to talk about the wider issues (the significance of an illness and the meanings of medication) was an important finding with implications for the acceptability of genome-based therapies.

Data and discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Data and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

Analysis yielded several clear analytical themes which we describe below. These include general views regarding the acceptability of pharmacogenomics and findings which illustrate how wider perceptions of depression, antidepressants, and genetic research may impact on the reception of genome-based therapies.

Views on the pharmacogenomics of depression

One limit of our study is that we lacked the time and resources to make extensive trans-national comparisons. This would have required in-depth analysis of national and local variations in the four sites of respective health care systems and mental health services. In general terms, however, we make several observations. First, a majority of participants in all the public focus groups felt that pharmacogenomics was a ‘good idea’, felt ‘positive towards it’ and, in one case at least, would ‘have it straightaway’. However, it is clear from our data that support was strongest in Poland where every participant in every group voiced support for the technology, and were on the whole less critical than other sites. Whilst it is possible, of course, that members may have been reluctant to deviate from majority opinion in a group setting, data clearly indicate that the Polish groups saw genetic technology as ‘futuristic’ and largely unproblematic. It is likely that the current state of mental health care in Poland influenced group members. For example, we were told by group members that in Poland antidepressants are relatively expensive compared to other countries. In addition, according to one study, Poland has the longest waiting time between when a patient visits their primary care giver and when they first meet a mental health service provider (Pawlowski and Kiejna 2004). It is not overly surprising then, given these facts, that the notion of a quick and relatively easy treatment through a pill may seem an attractive option.

In comparison with other sites, there were noticeably more reservations about pharmacogenomics in Germany. Although many German participants were supportive of pharmacogenomics in a general sense, this support was not without considerable qualification or concern, as detailed in our findings below. It is worthwhile to note here that Germany has some of the lowest prescription rates for antidepressants in the EU and, until recently, it was possible under German social insurance schemes to spend up to one month in a spa, receiving treatment for nerve and psychiatric-related disorders (Rose 2005, Shorter 2005). In this context, it is not hard to see why there may be some reluctance to embrace a genetic test for antidepressants.

In all the groups, choice was a frequently discussed issue – but it was noticeably more so in the English groups, a finding that we attribute to recent initiatives in the NHS promoting patient choice. In all the sites, service users drew heavily on their personal experience of what taking antidepressants meant for them, their friends, family, and fellow mental health service users. There was an overwhelming desire to eliminate adverse drug reactions, as the following excerpts show:

I don't think I would have minded if I, when I was, quite just before I was pinned down, and injected, if they'd have just done a quick swab and actually got me something that wouldn't have been so horrific in its side-effects because my metabolism contradicted it or whatever it is that they use, I don't think I would have minded that actually in a way (English User Group 2).

We realise that our members often do not die of psychiatric disease but of other diseases which they got in effect of the prolonged use of psychotropic drugs, for example liver diseases, stroke, heart failure and many others. It is not unusual for a psychiatric patient to die between the age of 30 or 40. They die because their body can no longer stand so many drugs and it gets ruined. So what most patients dream of is that a medication is formulated that will not bring about side effects. And it is really not important whether it will be developed based on a patient's genotype or in another way. What we really want is that we stop dying prematurely because of the medications we have to take (Polish User Group 3).

Apart from the devastating side effects of medications, users’ responses also contained a tenor of desperation regarding their desire to escape from the effects of their illness. One participant reported how his bouts with bipolar illness led him to give away £37,000 without knowing where exactly he had given the money. He conveyed a sense of desperation to end the ‘dreadful things’ that his disease brought on and the terrible effect it had on his loved ones:

But I mean, certainly, if you were looking for volunteers for the clinical trials here, I would be rushing and banging on your door, because I've nothing to lose, because I do not want to continue the way I am, and I certainly don't want the, the impact that my condition has on my family. I don't want that to continue any longer (English User Group 1).

The sense of helplessness which pervades family members of sufferers from mental illness is captured by a man who lost his wife to depression only the week before agreeing to participate in the focus group:

I am sure that if my wife was given such a chance, perhaps she would have a medication chosen properly and she would still be alive and we would be together. I really cannot say more. Please, forgive . . . Now it is too late for anything and we cannot change it. . . . It's been a week only and I cannot compose. If she had been administered better medications, if there were better medications, perhaps she would be alive now (Polish User Group 1).

Our data seem to confirm other studies which claim that the public is optimistic about the prospects of genomic-based medications (Gaskell et al. 2006, Nielsen and Moldrup 2007, Rogausch et al. 2006, Rothstein and Hornung 2003). We stress the word ‘claim’ here since, as discussed above, most people had not heard of pharmacogenomics before being asked their views about it. It is not surprising, therefore, that people who are suffering would have favourable views of a technology that offers them a glimmer of relief. One of the more detailed studies into public views was carried out by the British Royal Society who held a series of public workshops with 76 participants (Royal Society 2005a; 2005b). Although only ‘one person per group’ had previously heard of pharmacogenomics before attending the workshop, in general, participants were supportive of pharmacogenomic testing as a means of learning more about treatment options and diseases affecting them.

Whilst our work echoes this general approval of genome-based treatments, we also found a number of important themes which will possibly cast a shadow on the acceptance and clinical adoption of pharmacogenomics. In our view, it is crucial to note that none of these studies asked about pharmacogenomic medicines for psychiatric illness. The Royal Society, for example, limited itself to heart disease, cancer, diabetes and asthma. Our data find that whilst there are significant levels of support for pharmacogenomic drugs, there are considerable and significant caveats which are bound to impact on public and user acceptability of pharmacogenomic tests for antidepressants. One theme from our work is that there is a clear difference between mental and physical illness, a difference which may impact on the reception of pharmacogenomics. Below, we explore these reservations.

Drug reactions and disease risks

A key distinction in pharmacogenomics is whether the test sample derives from an infected tissue, i.e. a tumour genome, or through blood tests, which unlike cancerous tissue, can reveal the inherited genetics of an individual. Tests for Herceptin, for instance, derive from cancerous tissue whilst those for CYP2D6 metabolism (implicated in antidepressant use) require a blood test, which means that secondary information other than drug reaction results could potentially be discovered. This is because the CYP2D6 is also associated in the metabolism of many other different kinds of medicines. In addition, in some cases, the same genes involved in drug metabolism are also associated with increased risk of other disease conditions (Smart, Martin and Parker 2004).

The Royal Society study cited above aimed to distinguish between the likelihood of ‘specific public issues related to pharmacogenomic testing and genetic testing in general’ (2005b: 3). The study concluded that ‘participants were able to understand easily the basic principles of such testing and its distinction from other genetic tests’ (2005b: 6). And yet, crucially, the Royal Society dodged the key issue, since the topic of ‘multiple gene interactions in relation to drug metabolism was not discussed.’ (2005b: 6: fn6) This is important since it is through such interactions that clinically relevant secondary information is likely to emerge (Netzer and Biller-Andorno 2004).

A mainstream view amongst proponents of pharmacogenomics is that there is no real risk of learning clinically relevant information from a pharmacogenomics test (for example, see Roses 2000). A number of sociological studies, however, are increasingly putting this belief in doubt (Pieri and Wynne 2007, Hedgecoe and Martin 2003, Hedgecoe 2004). Our study found that perceptions of the distinction between susceptibility and treatability are tenuous at best. Representatives in our Policy Workshop indicated that ‘there's going to be a blur’ between disease risk and drug reaction and ‘that blur may be quite large’. As one delegate summarised:

I think it's true as to everybody seems to be saying around the table that you know, with some exceptions, it's difficult to separate out questions about treatability from questions about aetiology, from questions about risk. And questions about risk involve questions about inheritance (Policy Workshop).

Whilst focus group participants understood the difference between testing for drug response and testing for disease risk, this separation was nearly impossible to maintain in actual discussions. There was a sense that ‘at the end of the day, a genetic test is a diagnostic method that gives information on a person’ (German Public Group 3). There was a belief that various pressures would conspire to make it hard to keep testing at the level of treatability and that usages of test information could ‘spread like ripples in the water’ (Danish Public Group 3).

A good example of the conflation of disease risk and drug reaction can be found in the exchange below:

  • A:  
    I would want to know that I was getting the best medication. I'd go for a genetic test if he thought I was depressed or is it, yeah, I would want to know. It's in my interest I feel to be treated properly and I wouldn't want to be depressed so you know, that's just the way I feel. I'd want to be treated properly.
  • B:  
    Another concern is well where all this genetic results get held, do they get held in a mainframe computer somewhere (multiple voices)
  • A:  
    I couldn't give tuppence, I couldn't give tuppence where they’re held, stored my (multiple voices)
  • B:  
    Or what they used it for?
  • A:  
    Or what they used it for. If it's going to benefit me and my family in the long run, then I'm all for it. I mean I would hope you know, my granddaughters are going to be tested to show that they've not got the gene that I may well have passed on to them. And if there's any medication they can take when they get to a certain age that would prevent them from probably getting the same disease, I'd be really happy. I'm only sorry that I didn't know about it sooner (English Public Group 2).

This exchange is useful, for it highlights the interactional contextual of how people react to being asked for a pharmacogenomic test. ‘A’ is forced to defend his support of pharmacogenomics by a participant who raises her concerns about the storage and future use of test results. Whereas he first expressed his desire to have a test to be ‘treated properly’ (with the hesitation, ‘or, is it’, which implies he may not have been absolutely certain as to the aim of the test), ‘A’ then seems to respond to the possibility of these concerns by conflating susceptibility and treatability and by saying that he would want his granddaughters tested to make sure they were not carriers of a gene that would pre-dispose them to his same disease5. We highlight this exchange to bring attention to the point that medical treatment is not a two-person exchange system. Pill taking is invariably a social act (Karp 2004) and it is through a network of significant others that a patient will come to question the wisdom of medicating themselves or the effects of a test on their family and friends.

Difficulty, however, in maintaining a clear distinction and concerns over what a test may yield are only one element that may have an impact on the uptake of pharmacogenomic tests for depression. At the heart of the issue for many is the nature of the illness itself and beliefs about its cause and treatability.

The medical model of depression and antidepressants

Kramer (2005) writes tellingly:

The anatomical account of depression has changed the way doctors view their patients. The depressed person sits before us. She speaks, wretchedly, of the trivial disappointment that threw her life into a living hell. Hearing of vulnerability in daily life, we imagine vulnerability at the level of the neuron. No need to look to the face and habit. We can imagine glial cells retreating, and neurons withering, at this very moment (2005: 61–2).

The biomedical explanation of depression and antidepressant medication, which Kramer so colourfully embraces, implies that a chemical deficiency in the brain can be treated with drugs which raise the level of amines and alleviates the depression. Perhaps, as Karp argues, the ‘necessary condition for widespread psychiatric drug use is a cultural induced readiness to view emotional pain as a disease requiring medical intervention’ (Karp 2004: 15). But as an increasing number of commentators are arguing, it does not follow that if enhancing the transmission of serotonin improves depression, then a deficiency in the serotonin system is necessarily responsible for the emergence of the condition in the first place (Horwitz and Wakefield 2007).

For all the support in our focus groups for pharmacogenomics, criticism of the medical model was reflected in a number of responses. There was a feeling that pharmacogenomics represented an extension of the medical model that already pervades the diagnosis and treatment of depression and that ‘there might be a bigger incentive to use medicine instead of conversation therapy’ if genome-based therapies reached the clinic (Danish Public Group 1). As one member put it:

But the question is whether one should not invest all this money that is used to develop these tests and drugs in other things, like . . . I don't know . . . like social structures that would make people less sick in the first place . . . What is important is that more people start thinking more about these themes and that they maybe have the idea that it is maybe not necessarily the solution to find more efficient drugs. Rather [the solution is to] try to live in another more human society so that one gets less sick and less depressive in the first place (German Public Group 3).

Some service users in the ENUSP organisation (European Network of (ex) Users and Survivors of Psychiatry) explicitly used the term ‘the medical model’ thus tying their discussion into user/survivor politics:

If I got this questionnaire, going back a couple of months ago, when I was quite unwell, I would have jumped at the chance of a blood test to make me better. But when I would be getting a bit better and I'd start thinking about it more logically, I'd realise that I'm going back into the medical model again and I'm going back into all this, you know, taking drugs (ENUSP).

This quote highlights a related theme that a number of people expressed – that is, a continual fluctuation and/or contradiction in attitude towards their condition and medication. For some users at least, desperation gets entangled with hesitation, depending where/when in the process they are confronted about the possibility of pharmacogenomics.

Ambivalence

In a recent study, Grime and Pollock (2002) interviewed 32 primary care patients diagnosed with mild to moderate depression in the UK and found that many patients held ‘shifting perspectives’ towards their medication and were full of doubt, ambivalence and uncertainty (2002: 517). On the one hand, people wanted to continue taking antidepressants if it kept their symptoms away and yet simultaneously felt that they were ‘weaker’ for needing to stay on the drugs and desired to ‘sort themselves out’ without recourse to medicine. As a result, patients self-adjusted their dosages and concealed this fact from their physician. Feelings of ambivalence were also heightened by the fact that patients could not fathom staying on antidepressants indefinitely, even when they were effective.

Feelings of doubt, uncertainty and ambivalence can also be seen in both public and users’ views – not only towards antidepressants but towards depression itself, and towards genetic research. Of course, to some extent this is not surprising since ethical issues surrounding genetics and debates over therapeutic enhancement are well rehearsed. But it seems worth reflecting on this issue since shifting views towards medication targeted for pharmacogenomics research and ambivalence regarding the condition those medicines are designed to treat may well have an impact on the reception of pharmacogenomics tests in the clinic. This is a factor that clinicians, psychiatrists, and health care providers will find hard to ignore when consenting for and prescribing genome-based tests for antidepressants and then monitoring patients’ use afterwards.

As one participant put it:

I have friends and colleagues who are taking antidepressants, and I think it has very different effects on them, there are some of them where I can't tell the difference, but some of them have completely lost their personality, and some where it has helped them in a positive way, some of those who have lost their personality, you cannot ‘talk’ to them any more, I think that is really sad (Danish Public Group 1).

Another indicated that:

I am actually quite morally ambivalent, because I can see that the individual needs it when that person gets out there, there is really a need for it, when you are completely desperate, when you look at a person with this illness etc. But generally speaking – when you look at it from above – then I think it is dangerous (Danish Public Group 2).

Part of the ambivalence we are aiming to describe relates to depression itself. There was a sense amongst many that depression was ‘subjective’, that it was not always in need of immediate cure and could in fact be a source of personal growth:

I think behind [genetic research] is such a conception of human being and also of the illness – now you will be fixed, all will be all right. And I have to say my psychosis was also somehow a process of insight. It was also an experience and I know, not everyone thinks that way but I have undergone it and I didn't want to be repaired at all costs (German User Group).

It is ambivalence. Only in the cases of small communities is it a sympathetic ambivalence, a positive one. We don't get in your way, you don't get in ours, but okay. You are a member of our community. Whatever happens, we’ll stick together. In a city, people are concerned with their own matters, they don't want to think about, engage or look after people who are ‘relatively abnormal’ (Polish Public Group 1).

Many of our results confirm research done by Karp (1996, 2004). His work, based on one-to-one interviews with users of antidepressants, raised several key themes. First, the decision to go on an antidepressant was rarely taken lightly and was often connected to others’ views of the mental condition itself. Secondly, people sought to give meaning to and impose order on their illness and medication use, and these meanings shifted over time. Thirdly, Karp also found that there was a reluctance to see oneself as mentally ill or as needing medication. People reported a general uneasiness about controlling one's feelings with a pill, as against wanting to ‘tough it out’ (2004: 104).

One can easily surmise that having to undergo a pharmacogenomic test may add another layer of ambivalence to this equation. If patients associate feelings of self-esteem and integrity with their ability to handle personal problems, then what impact will this have on the uptake of the pharmacogenomics of antidepressants?

Depression is not like cancer

A final related theme which emerged from both the public and user groups concerns the difference between psychiatric and physical illness. Whilst the distinction between the two is debatable in medical terms, in social terms the divide is clear. There is an assumption that the symptoms of mental disorders are in some sense less ‘real’ than those of physical disorders which have a tangible local pathology (Kendell 2001). One group explicitly tied views of depression to those of cancer:

  • A:  
    If someone goes to an oncologist, people feel sorry for them.
  • B:  
    Exactly, there isn't, in Polish society such a . . . with the other afflictions, that you sympathise, you say ‘it's a shame that you are ill’. And yet with psychological problems, then you don't feel sorry for them, you just reject the person (Polish Public Group 3).

Another drew similar conclusions that somehow physical ailments were more ‘real’:

Just think for people it's difficult to come to terms with it because if you've got a lump or a pain, it's very real. You can go to a doctor and you can show them the lump and you can tell them where the pain is. But we’re talking about something that's in somebody's head and I think that has all kinds of kind of taboos in it (English Public Group 1).

And underlying these discussions, of course, is the pervasiveness of stigma that is associated with mental illness:

  • A:  
    I think it is much worse to be mentally ill than being physically ill, because you become another person. It is more difficult to feel compassion and people become tired of you
  • B:  
    That is exactly it
  • A:  
    Your surroundings have a hard time relating to it
  • C:  
    And the mentally ill can be a pain in the ass sometimes, phew
  • A:  
    And it is hard to know what is right to do, should you pad them, or should you say stop it, damn it (Danish Public Group 2).

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Data and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

In his case study of Herceptin and Tacrine, Hedgecoe (2004) argues that:

The role of the clinical context is key in shaping the final form of a pharmacogenetic technology, and since clinical context is in turn formed by social, cultural and, most of all, economic factors, it is these that we should pay attention to when considering the broader aspects of personalised medicine (2004: 175).

Whilst pharmacogenomics has broad support, several strong social and cultural themes have emerged from our study, which lead us to argue that discussions of the clinical acceptability of genome-based therapies for depression cannot ignore some of the wider issues regarding depression and antidepressants. These themes include:

  • 1
    a tendency to conflate the notions of a pharmacogenomic test for antidepressants with a genetic test for depression, a conflation which could cause patients to refuse a pharmacogenomic test if they thought (rightly or not) that the results would have wider implications for themselves or their relatives.
  • 2
    doubts about the medical model of depression, a model which in the minds of some people at least does not adequately explain life events as causal factors. More importantly for the uptake of pharmacogenomics, there is a related view that the treatment of depression has been overly medicalised, leading to a neglect of alterative therapies which could be just as efficacious.
  • It is worth lingering on this point since, in the UK, the Mental Health Foundation has recently advocated that GPs prescribe exercise for mild to moderate depression and not antidepressants. Similarly, the National Institute for Health and Clinical Excellence (NICE) issued a set of guidelines for treating depression that recommended in cases of mild depression, cognitive behavioural therapy ought to be favoured over antidepressants since the benefit-risk ratio for drugs was poor (NICE 2003). However, it remains to be seen how these recommendations are to be implemented given the lack of qualified therapists and that, according to one study, English GPs are more influenced by the promotion efforts of drug companies than independent sources (Dobson 2003).

  • 3
    a deep ambivalence regarding the use of anti-depressants. On the one hand, there is a strong hope, sometimes borne out of desperation as our data show, that new drugs will alleviate patient suffering. Pharmacogenomics may well be seen as part of what Novas (2006) calls a political economy of hope – where science, activism, and capital come together to promote expectations of success in biomedicine. Yet alongside these hopes for a cure, anti-depressant users consistently report that the decision to start taking the drug is not taken lightly, that they feel a sense of ‘giving in’ by having to swallow a pill each day to help manage their mood. Users also report, as our data showed, that anti-depressants sometimes affected one's personality so severely that they began to question if they were really ‘themselves’ anymore. Again, the pharmacogenomics of depression cannot separate itself from these issues or from the political and economic context in which the drugs and genetic tests are developed and promoted. It is likely that the ambivalence captured in our groups is re-enforced by media coverage of claims of withheld evidence from clinical trials, antidepressant-induced suicide and murder, class-action law suits, and parliamentary investigations into their power and influence – all of which have recently tarnished the reputation of Eli Lilly and Glaxo Smith Kline (House of Commons 2005).
  • 4
    a belief that psychiatric illness is different from physical illness and that depression carries with it a certain cultural value. It is well known that many believe there is a connection between melancholy and artistic talent. As we have seen, some patients consider depression to be ‘part of themselves’, which they are not sure they would want to lose – despite its considerable stigma. This seems to be very different from the way people talk about having diabetes, heart disease or cancer – a difference, we argue, that may well carry implications for the uptake of pharmacogenomic tests.

Based on data presented here, there exists an ambivalence regarding the personal and cultural significance accorded to depression and anti-depressants (not to mention genetic technology, of course). It is our view that requiring a genetic test in order to be given an anti-depressant adds another reason to question what is already, for some, a life-changing decision, another reason to doubt the medical model of disease causation, and another reason to ask if they are better off living with their depression than treating it with a pill.

Our findings suggest several implications for emerging medical technologies. It seems likely that how, where, and when a technology is adopted will depend, in part, on more than its clinical utility. Social attitudes towards the condition that a technology aims to treat, as well as beliefs about alternative therapies and their perceived efficacy may also influence their introduction into the clinic. Just as important, it is worth noting that patients can hold inconsistent views on these issues. As we have seen, they may place great hope on the prospect of a successful technology, and yet simultaneously may be wary of the assumptions and scientific models underpinning the development of the same technology. In our view, this means that advocates of a new therapy cannot assume that because it promises to work better, it will necessarily be greeted with widespread enthusiasm.

The sociology of expectations reminds us that the future of a technology is co-constructed (Hedgecoe and Martin 2003). In the case of the pharmacogenomics of depression, then, it will not only be the pharmaceutical companies, scientists, and policy makers that shape the field's development. As we have seen, Prozac is not Herceptin and depression is not cancer. Amongst the public and amongst mental health service users, there exists a widespread uncertainty about the nature, cause and meaning of depression as well as a critical reservation about drugs that have moods as their target. The message seems clear: the development and reception of pharmacogenomics of depression will not be reduced to mere efficacy and adverse drug reactions.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Data and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

Genome-based therapies for depression (GENDEP) was funded by the European Commission, contract number LSHB-CT-2003-503428, and we gratefully acknowledge their support. We would also like to thank the GENDEP ELSI team members and interview participants who contributed to the study and helped make this paper possible.

Notes
  • 1

    We by no means attribute the rise in antidepressant use merely to higher rates of depression. Antidepressants are commonly prescribed for other conditions such as panic attacks, social phobias, and obsessive-compulsive disorders. Again, we highlight this topic at the end of the paper.

  • 2

     Options include the design of new drugs aimed at persons with particular genetic sub-types, the ‘rescue’ of drugs that have been researched but never made it to market because clinical trial failure, the refinement of drugs already on trial to target particular genetic groups, or tests to screen patients prior to prescription.

  • 3

     We are aware that the term ‘public’ is problematic and we use it here only to distinguish non mental health service users.

  • 4

    As explored in the paper, data show that whilst these groups held strong opinions against pharmacogenomics, some of the themes they raised also emerged in our public groups and with other user organisations.

  • 5

     We set aside the fact, of course, that such a test is at present an impossibility for depression and that family history would be a far better indicator anyway.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Data and discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  • Brown, N. (2003) Hope against hype: accountability in biopasts, presents and futures, Science Studies, 16, 321.
  • Clark, C.C., Scott E.A., Boydell, K.M. and Goering, P. (1999) Effects of client interviewers on client reported satisfaction with mental health services, Psychiatric Services, 50, 9613.
  • Dobson, R. (2003) Pharmaceutical industry is main influence in GP prescribing, British Medical Journal, 326, 301.
  • Gaskell, G., Allansdottir, A., Allum, N., et al . (2006) Europeans and biotechnology in 2005: patterns and trends., Eurobarometer, 64, 3, Brussels.
  • Grime, J. and Pollock, K. (2003) Patients’ ambivalence about taking antidepressants: a qualitative study, The Pharmaceutical Journal, 271, 51618.
  • Hedgecoe, A. (2004) The Politics of Personalised Medicine: Pharmacogenetics in the Clinic. Cambridge: Cambridge University Press.
  • Hedgecoe, A. and Martin, P. (2003) The drugs don't work: expectations and the shaping of pharmacogenetics, Social Studies of Science, 33, 32764.
  • Horwitz, A. and Wakefield, J. (2007) The Loss of Sadness: how Psychiatry Transformed Normal Sorrow into Depressive Disorder. Oxford: Oxford University Press.
  • House of Commons (2005) The Influence of the Pharmaceutical Industry. Health Committee Fourth Report of Session 2004–2005 Volume I. London: The Stationery Office Limited.
  • Karp, D. (1997) Speaking of Sadness: Depression, Disconnection, and the Meanings of Illness. Oxford: Oxford University Press.
  • Karp, D. (2006) Is it Me or my Meds? Living with Antidepressants. Cambridge: Harvard University Press.
  • Kramer, P. (2005) Against Depression. New York: Penguin.
  • Kendell, R. (2001) The distinction between mental and physical illness, The British Journal of Psychiatry, 178, 4903.
  • Kitzinger, J. (1994) The methodology of focus groups: the importance of interaction between research participants, Sociology of Health and Illness, 16, 1, 103–21.
  • Lindpaintner, K. (2002) The impact of pharmacogenomics and pharmacogenomics on drug discovery, Nature Reviews, 1, 4639.
  • Martin, P., Lewis, G., Smart, A. and Webster, A. (2006) False Positive? The Commercial and Clinical Development of Pharmacogenetics. IGBis, SATSU, University of Nottingham.
  • Medicines and Healthcare Products Regulatory Agency (2004) Selective serotonin reuptake inhibitor (SSRI) antidepressants – findings of the Committee on Safety of Medicines (CSM) Working Group. Annex A – prescribing trends for SSRIs and related antidepressants. London: MHRA.
  • National Institute of Clinical Excellence (2003) Depression: Management of Depression in Primary and Secondary Care, Clinical Guideline 23. London: National Institute of Clinical Excellence.
  • Netzer, C. and Biller-Andorno, N. (2004) Pharmacogenomic testing: informed consent and the problem of secondary information, Bioethics, 18, 34460.
  • Nielsen, L.F. and Moldrup, C. (2007) The diffusion of innovation: factors influencing the uptake of pharmacogenetics, Community Genetics, 10, 23141.
  • Novas, C. (2006) The political economy of hope: patients’ organisations, science, and biovalue, Biosocieties, 1, 289305.
  • Pawlowski, T. and Kiejna, A. (2004) Pathways to psychiatric care and reform of the public health care system in Poland, European Psychiatry, 19, 16871.
  • Pieri, E. and Wynne, B. (2007) A qualitative study of personalised medicine: public priorities and values. ‘Genetics and Society: Retrospects and Prospects’. CESAGen/CSG 4th International Conference, 26th–28th March 2007, The Royal Society, London.
  • Rogausch, A., Prause, D., Schallenberg, A., Brockmöller, J. and Himmel, W. (2006) Patients’ and physicians’ perspectives on pharmacogenetic testing, Pharmacogenomics, 7, 4959.
  • Rose, D., Wykes, T., Leese, M., Bindman, J. and Fleischmann, P. (2003) Patients’ perspectives on electroconvulsive therapy: systematic review, British Medical Journal, 326, 13636.
  • Rose, N. (2005) Psycho-pharmaceuticals in Europe. In Knapp, M., McDaid, D., Mossialos, E. and Thornicroft, G. (eds) Mental Health Policy and Practice across Europe. Maidenhead: Open University Press.
  • Roses, A.D. (2000) Pharmacogenomics and the practice of medicine, Nature, 405, 85765.
  • Rothstein, M.A. and Hornung, C.A. (2003) Public attitudes about pharmacogenomics. In Rothstein, M.A. (ed.) Pharmacogenomics: Social, Ethical, and Clinical Dimensions. Hoboken, NJ: John Wiley and Sons.
  • Royal Society (2005a) Personalised Medicines: Hopes and Realities. London: Royal Society.
  • Royal Society (2005b) Pharmacogenetics: Dialogue. London: Royal Society.
  • Shorter, E. (2005) The historical development of mental health services in Europe. In Knapp, M., McDaid, D., Mossialos, E. and Thornicroft, G. (eds) Mental Health Policy and Practice across Europe. Maidenhead: Open University Press.
  • Silverman, D. (2000) Doing Qualitative Research. London: SAGE.
  • Smart, A., Martin, P. and Parker, M. (2004) Tailored medicine: whom will it fit? Ethics of patient and disease stratification, Bioethics, 18, 32243.
  • Solomon, A. (2002) The Noonday Demon: an Anatomy of Depression. London: Vintage Books.
  • Vedantam, S. (2004) Antidepressant use by US adults soars, Washington Post 3rd December, A15.
  • Wilkinson, S. (2005) Analysing interaction in focus groups. In Drew, P., Raymond, G. and Weinberg, D. (eds) Talking Research. London: SAGE.
  • World Health Organization (2001) World Health Report 2001. Geneva: World Health Organization.