The role of disease characteristics in the ethical debate on personal genome testing.

0.27 MB·English

by  Bunnik, Eline M

Bunniketal.BMCMedicalGenomics2012,5:4
http://www.biomedcentral.com/1755-8794/5/4
DEBATE Open Access
The role of disease characteristics in the ethical
debate on personal genome testing
Eline M Bunnik1*, Maartje HN Schermer1 and A Cecile JW Janssens2
Abstract
Background: Companies are currently marketing personal genome tests directly-to-consumer that provide genetic
susceptibility testing for a range of multifactorial diseases simultaneously. As these tests comprise multiple risk
analyses for multiple diseases, they may be difficult to evaluate. Insight into morally relevant differences between
diseases will assist researchers, healthcare professionals, policy-makers and other stakeholders in the ethical
evaluation of personal genome tests.
Discussion: In this paper, we identify and discuss four disease characteristics - severity, actionability, age of onset,
and the somatic/psychiatric nature of disease - and show how these lead to specific ethical issues. By way of
illustration, we apply this framework to genetic susceptibility testing for three diseases: type 2 diabetes, age-related
macular degeneration and clinical depression. For these three diseases, we point out the ethical issues that are
relevant to the question whether it is morally justifiable to offer genetic susceptibility testing to adults or to
children or minors, and on what conditions.
Summary: We conclude that the ethical evaluation of personal genome tests is challenging, for the ethical issues
differ with the diseases tested for. An understanding of the ethical significance of disease characteristics will
improve the ethical, legal and societal debate on personal genome testing.
Background part connected with differences between the diseases
A growing number of personal genome testing services tested for.
are presently available that estimate genetic susceptibil- Differences between diseases may have important
ity to multifactorial diseases [1]. Unlike genetic tests for implications for the ethical, legal and societal debate on
monogenic diseases, susceptibility tests for multifactorial genetic susceptibility testing for multifactorial diseases
diseases can be obtained almost exclusively on the (for the remainder of this paper we will use the term
direct-to-consumer market. Personal genome tests are ‘susceptibility testing’), whether it is offered as part of a
non-targeted: they determine a person’s risk for a multi- personal genome test or on its own, for a single disease,
tude of multifactorial diseases simultaneously [2-5], whether within the clinic or on the direct-to-consumer
including cardiovascular disease, Alzheimer’s disease, market. The offering, for instance, of susceptibility test-
osteoporosis, type 2 diabetes, schizophrenia, and many ing for less severe diseases for which there are preven-
types of cancer. One of the leading personal genome tive options available may be morally acceptable,
testing companies is currently estimating personal risks whereas the same test for severe diseases in the absence
for over two hundred diseases and other phenotypic of treatment options may not. Personal genome tests
traits in one single test [2]. The large quantity of test may comprise both severe and less severe diseases and
results will have limited clinical significance for and may as a whole thus be difficult to evaluate. In recent
varying emotional impact on consumers, which are in years, a range of ethical, legal and societal issues has
been touched upon in professional and academic discus-
sions on genetic testing, such as privacy issues, psycho-
logical impact, regulatory issues, and informed consent
*Correspondence:[email protected]
1Dept.ofMedicalEthicsandPhilosophyofMedicine,ErasmusUniversity [6-8]. These issues however may not arise in all suscept-
MedicalCenter,dr.Molewaterplein50,Rotterdam,3015GE,theNetherlands ibility tests that are offered within one personal genome
Fulllistofauthorinformationisavailableattheendofthearticle
©2012Bunniketal;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommons
AttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,andreproductionin
anymedium,providedtheoriginalworkisproperlycited.
Bunniketal.BMCMedicalGenomics2012,5:4 Page2of9
http://www.biomedcentral.com/1755-8794/5/4
test. Privacy issues, for example, may be more pressing normative frameworks, one for clinical genetics [10] and
for genetic tests for psychiatric diseases than for somatic two for population genetic screening programmes,
diseases, and the psychological impact of testing may be namely the Wilson and Jungner criteria [11] and the
more serious in severe diseases than in less severe dis- ACCE model [12], which take into account differences
eases. With this paper, we propose a systematic in disease characteristics and their impact on the ethical
approach to the ethical evaluation of susceptibility test- evaluation of testing. Not all disease characteristics or
ing, which considers differences between diseases and consequences of testing that are mentioned in the three
which explains what ethical issues are to be addressed, frameworks, however, are applicable to genetic suscept-
on the basis of disease characteristics. ibility testing for multifactorial diseases. For example,
Asinthispaperwefocusexplicitlyondiseasecharacter- while characteristics such as the prevalence of the dis-
istics and their impact on the ethical debate, we do not ease [13] or the population health impact [11] are rele-
elaborateupon characteristicsofthe test,such as clinical vant from a public health perspective, they are arguably
validityandclinicalutility.Geneticsusceptibilitytestsdif- less relevant for the ethical evaluation of susceptibility
fer from presymptomatic tests for monogenic diseases: testing in individuals. Further, implications for relatives
they indicate riskratherthandiagnosisandaregenerally or availability of a follow-up plan [10] have been formu-
oflimitedtomoderateclinicalvalidityandutility.Clinical lated in the context of clinical genetic testing for mono-
validity and utility are of paramount importance to the genic diseases and do not readily apply to testing for
ethical, legal and societal debate, but have already been multifactorial diseases, where test results so far have
addressedelsewhere[8,9].Testcharacteristicsanddisease been - and are likely to continue to be - of limited clini-
characteristics, however, are not entirely separable and cal validity and utility.
may interact with one another as well as among them- On the basis of an ethical analysis, we have identified
selves,aswillbepointedoutinthediscussion. four disease characteristics that are specifically relevant
The ethical debate may benefit from a better under- to the ethical evaluation of susceptibility testing: sever-
standingofmorallyrelevantdifferencesbetweendiseases. ity, actionability, age of onset, and the distinction
For national health care systems, policy-makers, health- between psychiatric and somatic diseases. These charac-
careprofessionals,physicians, companiesorother stake- teristics are morally relevant because they are directly
holders who are considering the offering or the connected with three of the four major bioethical princi-
regulation of susceptibility testing, it is important to be ples: beneficence, non-maleficence and respect for
awareofdifferencesbetweendiseasesortypesofdiseases, autonomy [14]. Beneficence is the medical professional’s
because susceptibility tests for different diseases may be duty to care, to do well, and to act in patients’ interests.
connected withdifferent ethical issuesand may warrant The principle of non-maleficence specifies that a medi-
differentethicalevaluations.Thisposeschallengestoan cal professional should refrain from doing needless
industry that is increasingly offering personal genome harm. Respect for a person’s autonomy implies roughly
tests that consist of more and more susceptibility tests that patients should be allowed to make their own
for more and more diseases rather than single suscept- voluntary, informed choices with regard to their health
ibilitytestsforsinglediseases.Theinsightsofferedinthis management. For the purposes of this paper, we leave
paper are also anticipatory of future developments, and aside the fourth principle, that of justice, because we
applyequallytosusceptibilitytestingbasedonarraytech- focus on the justifiability of the provision of susceptibil-
nologiesand tosusceptibility testingbasedonexome or ity testing to an individual rather than on societal, dis-
whole-genomesequencingtechnologies. tributive issues that surround healthcare on a collective
First, we will present and discuss a list of key disease level. Below, we will discuss the four characteristics and
characteristics that are crucial to the ethical appraisal of describe how they give rise to ethical issues.
genetic susceptibility testing for multifactorial diseases. It is important to note at the outset of our discussion
Then we will discuss susceptibility testing for three dis- that internationally recognised guidelines on the provi-
eases, type 2 diabetes mellitus, age-related macular sion of genetic testing [15,16] demand that medical
degeneration and clinical depression, and ethical issues supervision and genetic counselling are made available
that must be taken into account in the evaluation of in genetic testing. The guidelines apply to genetic test-
such tests. ing that is ‘offered in a clinical context’ [15] and genetic
testing ‘for health purposes’ [16], respectively. It is there-
Discussion fore unclear whether these guidelines apply also to
1. Disease characteristics and their associated ethical direct-to-consumer genetic testing or to genetic testing
issues for purposes other than health. Both guidelines do spe-
We have conducted a review of the biomedical and cify that the form and the extent of genetic counselling
bioethical literature and identified three existing “shall be defined according to the implications of the
Bunniketal.BMCMedicalGenomics2012,5:4 Page3of9
http://www.biomedcentral.com/1755-8794/5/4
results of the test” [16] and “should be proportionate of screening: it requires that there be “an effective
and appropriate to the characteristics of the test, the remedy, acceptable action, or other measurable benefit”
test limitations, the potential for harm, and the rele- [12]. In the context of personal genome testing, with its
vance of test results to individuals and their relatives” wide variety of diseases tested for, actionability has
[15]. In our discussion, we will follow the idea that become a more appropriate notion than treatability.
counselling requirments should be proportional and For susceptibility testing, prevention is argued to be
show how disease characteristics play a role in deter- one of the main purposes and potential benefit [2]
mining the appropriate level of counselling and medical [[3,17]. Prevention may indeed constitute a benefit to
care. both the individual and society and an argument for the
1.1 Severity moral acceptability of susceptibility testing. It should be
Severity of a disease refers to the morbidity and mortal- noted that in order for preventive measures to be effec-
ity brought about by the disease. In clinical genetics, the tive, compliance is essential. Compliance is generally
severity of the disease affects the emotional, psychologi- promoted when preventive options are simple and
cal and social impact of genetic test results. Disease morally or psychologically acceptable [19]. The daily
severity therefore has consequences for the ethical taking of supplementary vitamins, for example, is likely
requirement to offer good care (beneficence) [14] in the to be accepted and carried out much more easily than
form of genetic counselling and psychological support. more intrusive measures, such as medication with side-
It is recommended that extensive pre- and post-test effects or a radical change of diet.
genetic counselling is mandated in all cases of severe However, it is not at all clear that susceptibility testing
hereditary disease [10]. One could argue that for severe contributes or will contribute to the prevention of mul-
diseases, where the anticipated psychological impact and tifactorial diseases [20]. If it does not or will not, its
the potential for harm are greater than for less severe major rationale will lack ground, and therewith, it may
diseases, as a general rule, more information, guidance not be able to reach the morally required balance of
and care are morally required. benefits and risks. An alternative rationale for suscept-
The severity of a disease, however, is not always easy ibility testing and potential benefit could be the personal
to determine: diseases tend to express differently from utility [19] that consumers find in knowing their genetic
one individual to the next. High levels of phenotypic information. It has been found that disease risk informa-
variability may cause genetic susceptibility test results to tion obtained from a screening programme may serve
leave patients or consumers with substantial uncertainty psychological or personal purposes, such as relief from
with regard to the severity as well as the onset of the uncertainty [21], solace, the value of knowing. Whether
disease. These uncertainties may pose challenges for the personal utility on its own could provide a sufficient
provision of pre- and post-test information on personal argument for the ethical acceptability of genetic suscept-
genome testing and for its informed consent processes. ibility testing or personal genome testing, is still a topic
Thus, disease severity affects the ethical issues of infor- for discussion [22].
mation, informed consent, genetic counselling, guidance 1.3 Age of onset
and care. Through predictive genetic testing, it has become possi-
1.2 Actionability ble for individuals to know their genetic risk before the
There are potential harms involved in genetic testing, onset of symptoms of disease. Knowledge of genetic risk
such as health risks (e.g. unnecessary follow-up), psy- for late-onset disease may cause anxiety and distress.
chological risk (e.g. anxiety) and social risks (e.g. finan- Not everyone wishes to obtain such knowledge [23].
cial costs, discrimination). Therefore, in the ethical One of the classic ethical principles within clinical
evaluation of genetic testing and screening services, a genetics is therefore the right not to know, which is
favorable balance between risks and benefits, i.e. derived from respect for autonomy [24]. Especially in
between the principles of beneficence and non-malefi- children the right not to know is a recurrent issue,
cence, has traditionally been a central criterion [11]. because adverse psychological and social effects of pre-
The availability of therapeutic interventions constitutes dictive genetic testing for late-onset diseases may be
a major benefit and a rationale [9,11] for genetic testing severe in children [25], and also because children are
or screening. With part of medicine’s focus shifting considered incapable or less capable of making autono-
from cure to prevention over the last few decades [17], mous choices. It is generally agreed upon that in the
however, the notion of treatment has come to include absence of medical necessity, predictive genetic testing
other ‘meaningful actionable options’ [18], which were for late-onset diseases must be postponed until adult-
added primarily to include reproductive decision-mak- hood, when young adults have attained the ability to
ing. The more contemporary ACCE framework intro- make autonomous decisions [25]. Exceptions are some-
duces preventive options and actions as possible benefits times made when the clinical geneticist finds that the
Bunniketal.BMCMedicalGenomics2012,5:4 Page4of9
http://www.biomedcentral.com/1755-8794/5/4
child or adolescent has sufficient cognitive ability to par- Thirdly, it has been suggested that genetic testing
ticipate in decision-making and when the child’s medical itself or a positive test result may become a self-fulfilling
or other interests may weigh up against potential harms prophecy, a ‘trigger event’ for the psychiatric disease
[21]. In clinical genetic practice, however, children and tested for [33]. Moreover, the manner in which the dis-
minors are more often refused than allowed predictive ease is understood by the patient (e.g. “it is in my genes,
genetic testing for late-onset diseases [26]. These ethical therefore it is an inevitable aspect of myself”) is likely to
principles may be equally applicable to genetic suscept- reflect on and modulate the development of the disease
ibility testing for late-onset multifactorial diseases in itself [34]. Thus, genetic testing could have unforeseen
children or minors. negative effects on the disease itself.
The distinction between early-onset and late-onset of It is important to note here that the psychiatric/
disease,however,cannotbesoreadilymadeinbiological somatic distinction has a somewhat different status than
reality. In the pathogenesis of most multifactorial dis- the other disease characteristics. Modulated by action-
eases,geneticand non-geneticriskfactorsact and inter- ability and severity, genetic testing for psychiatric dis-
actover timetocauseevolvingstagesofdisease.Insuch eases may be accompanied by a varying likelihood of
a‘cascademodel’ofdisease[27],theageofonsetcannot adverse psychological consequences and thus bring the
always be pinpointed straight-forwardly. Whereas the principle of non-maleficence into play to a varying
disease itself (e.g. symptomsoftype 2diabetes) may not extent. This way, disease characteristics may sometimes
become manifest before adulthood or even old age, cer- combine to lead to ethical issues. They should not be
tainriskfactors(e.g.overweight)orearlystagesofdisease considered in isolation, but rather with an eye for their
(e.g.pre-diabetes)maybepresentalreadyinyouth.Natu- dynamic relations.
rally,geneticriskfactorsarepresentatconception.Since In conclusion, there is not yet much experience with
the first steps towards disease may already be made (clinical)genetictestingforpsychiatricdiseases.Through
before birthorinearly childhood, the notionof onsetof personalgenometesting,thepracticeofriskpredictionfor
disease could become less clear. In children or minors psychiatricdiseasesiscurrentlytakingshape,butthefield
who already demonstrate certain risk factors or early and its ethical issues are largely unexplored. Due to the
stages of disease, genetic susceptibility testing for these complexitiesinetiology,andthesocietalandpsychological
diseases,eventhoughtheyhavetraditionallybeenconsid- sensitivities that surround psychiatric diseases, genetic
eredlate-onset,couldarguablybemorallyacceptable. testingforsuchdiseasesrequiresacarefulapproach,with
1.4 Psychiatric/somatic distinction specialattentiontoethicalissues.
There are at least three morally relevant differences
betweensomaticandpsychiatricdiseasesthatmayleadto 2. Three diseases: how disease characteristics affect the
ethical issues. Firstly, knowledge ofgenetic risksforpsy- ethical debate
chiatric diseases could “undermine or weakena person’s We have identified four key disease characteristics that
senseofintegrityandwell-being”evenbeforetheappear- are relevant to the ethical evaluation of susceptibility
ance of symptoms [28]. Psychiatric diseases can change testing: severity, actionability, age of onset, and the psy-
patients’perceptionsoftheworld,theirbehaviors,desires, chiatric/somatic distinction. In this section, we analyse
opinionsandgoals,theirrelationships,andwhothey are. how these disease characteristics influence the ethical
The potentially greater psychological impact of genetic issues surrounding susceptibility testing for three multi-
testingforpsychiatricdiseasesmayrequireahigherpoten- factorial diseases, namely type 2 diabetes, age-related
tial for medical benefit and higher standards for genetic macular degeneration (AMD) and clinical depression.
counsellingthanthoseusedforsomaticdiseases[28]. These diseases are illustrative of both the disease char-
Secondly, psychiatric diseases are associated with acteristics and the main ethical issues surrounding sus-
stigma to a greater extent than somatic diseases [29,30]. ceptibility testing, whether it be offered in a clinical
Stigma may lead to social and societal problems, such as setting or directly-to-consumer. Our main question is
disturbed family or personal relations, or discrimination whether and on what conditions it would be morally
at the workplace or in the context of (health) insurance. justifiable to offer individual susceptibility testing to
The level of stigmatisation varies with the disease and children and adults. Key elements of the discussion are
partly determines the level of psychological and social summarised in Table 1.
risk involved [31]. Genetic susceptibility testing is 2.1 Type 2 diabetes: variable severity and possibilities for
thought capable of increasing rather than decreasing early preventive intervention
stigma associated with psychiatric diseases [32]. A high Genetic susceptibility testing services for type 2 diabetes
risk of stigma and subsequent psychological harm will have already been made available directly-to-consumer
increase the required potential for (medical) benefit as both as targeted tests and as part of non-targeted perso-
well as the need for good care and counselling. nal genome tests [2,35]. While both physicians and
Bunniketal.BMCMedicalGenomics2012,5:4 Page5of9
http://www.biomedcentral.com/1755-8794/5/4
Table 1Diseasecharacteristics andtheir relationto ethical principles: Ethicalissues to consider
Disease Ethicalprinciples Relationcharacteristic/principle:
characteristic Ethicalissuestoconsider
Severity Beneficence Highseverity®highpotentialformedicalbenefit,highriskofpsychologicalharm
Non-maleficence
Actionability Beneficence Highactionability®highpotentialfor(medical)benefit
Non-maleficence Moderateactionability®riskofpsychologicalharm(e.g.victim-blamingorfeelingsofguilt)
Lowactionability®riskofpsychologicalharm(emotionalimpactoftestresult)
Ageofonset Respectfor Lateageofonset®rightnottoknowinadultsandchildren
autonomy
Psychiatric/somatic Non-maleficence Psychiatricdiseases:moreriskofharmbecauseofstigma;moreriskofharmbecauseofpsychological
impact
consumers have been found to respond favorably to the found to be at decreased risk may wrongly feel assured
idea of direct-to-consumer genetic testing for type 2 dia- that they will remain free from disease, regardless of
betes susceptibility [36], experts are not convinced of its their lifestyles [42]. They may fail to understand that
current clinical validity and utility [37,38]. general health recommendations are relevant to the
Type 2 diabetes is a somatic disease of varying sever- whole of the population, including low-risk subgroups.
ity. In its initial stages, it generally causes relatively mild Low-risk individuals may ignore these recommendations
symptoms, but on the long term, diabetes may cause and consequently put their health conditions at risk.
kidney failure, blindness and neuropathy of the extremi- The risk of false reassurance should be taken into
ties, and it may cause premature death. As known non- account in the consideration of an offer of susceptibility
genetic risk factors, such as overweight and pre-diabetes, testing for type 2 diabetes to adults or children.
are increasingly affecting the young [39], type 2 diabetes The actionability of a test is not only a characteristic
can no longer simply be considered a late-onset disease. of the disease tested for, but is influenced by other fac-
There are both therapeutic options and well-established tors, as well, notably the clinical validity of the test. As
preventive strategies available for type 2 diabetes, for preventive measures in type 2 diabetes are identical to
children as well as for adults, at the level of lifestyle general health recommendations that apply to the entire
improvements. population, it follows that it is unclear whether there is
The variability of the severity of type 2 diabetes poses any use for susceptibility testing [37]. The actionability
difficulties for the ethical evaluation of susceptibility of susceptibility testing may nonetheless be twofold: it
testing for the disease. From a precautionary perspective, may motivate high-risk individuals to adhere to general
it could be argued that type 2 diabetes should be viewed health recommendations [42] and it may be used to tar-
as a severe disease and require high levels of genetic get preventive strategies, guidance, care and monitoring
counselling and psychological support. However, empiri- by healthcare professionals, at those who are likely to be
cal research has shown that on the short term genetic most in need. Any clinical utility - any potential for
susceptibility testing for type 2 diabetes hardly causes medical benefit - will only be realised on the condition
any psychological harm or emotional impact at all of established clinical validity, which today has not been
[40,41], thus suggesting that consumers may not experi- fulfilled.
ence increased personal risk for type 2 diabetes as Moreover, in the context of a limited or moderate
severe, and that requirements for counselling in order to clinical validity, preventive measures ought not to be
prevent psychological harm may be less stringent. There too burdensome or too strongly associated with health
may be discrepancies between the severity of a disease risks, psychological or societal risks. When the benefits
as perceived by medical professionals and the severity of of preventive action are uncertain, the risks should be
the same disease as perceived by other publics. Ques- minimal in order to assure proportionality. In the case
tions regarding standards of severity may be interesting of type 2 diabetes, preventive measures are likely to be
topics for further (ethical) research. acceptable to most consumers: they may not always be
Further, the existence of preventive options for type 2 easy to adopt, but they do not entail health risks or side
diabetes implies a potential for medical benefits to be effects. Adequate post-test counselling could help to
obtained from susceptibility testing. Preventive options improve the implementation of behavior change and
consist in general health recommendations, such as a thus actionability.
healthy diet, regular physical exercise, weight loss, and At the same time, the window of opportunity for pre-
smoking cessation. As a consequence, if false reassur- ventive action is expanding at present. Given that child-
ance occurs, it may lead to harm. Individuals who are hood obesity may not only cause severe morbidity in
Bunniketal.BMCMedicalGenomics2012,5:4 Page6of9
http://www.biomedcentral.com/1755-8794/5/4
the young [43], but is also an independent risk factor for be of substantial clinical validity, and are expected to
adult obesity [44] and subsequent diseases (e.g. type 2 eventually outperform and improve or replace existing
diabetes), it may be sensible to start the prevention of prediction models [50].
adult obesity or type 2 diabetes already in childhood. It AMD is a leading cause of vision loss worldwide [51].
is still unclear whether and what childhood interven- It is a somatic disease of (very) late onset that generally
tions will prove effective [43,45], and whether genetic progresses slowly. Although end-stage AMD may entail
susceptibility testing for type 2 diabetes will gain clinical serious vision loss and brings along a fair amount of
validity and will prove to be of added value over and morbidity, the disease is not life-threatening. Further-
above clinical risk factors for the identification of at-risk more, there are treatment options (e.g. laser therapy)
individuals [38]. If it can be established that early inter- and preventive options (specific vitamin supplements,
ventions yield levels of medical benefit that could not be smoking cessation, the wearing of sunglasses) available
otherwise attained, and that genetic testing can be useful for AMD [52]. Although the preventive measures are
for the targeting of interventions at subgroups of high- safe, they do not seem to be very effective [53].
risk children, a favourable risk-benefit ration for testing Given the late age of onset and the absence of oppor-
in childhood might ensue. In that case it could become tunities for primary preventive action in childhood or
rational and morally justifiable to test children for their adolescence, there are no medical reasons for childhood
genetic susceptibility to type 2 diabetes. This will also genetic testing. Therefore, it is preferable to postpone
depend on other ethical issues, such as psychological genetic testing for AMD until maturity, when young
harms: at-risk children who do not adhere to lifestyle adults can make a more informed and autonomous
recommendations and develop the disease later in life decision whether or not to undergo testing and can pro-
may blame themselves or be blamed by others. Such vide informed consent. Thus, children’s and minors’
‘victim-blaming’ or feelings of guilt will not always be rights not to know their genetic risk can be protected.
justified in the context of a multifactorial disease for As AMD tends not to strike until old age, psychologi-
which susceptibility testing is of moderate predictive cal and social risks of genetic testing are likely to remain
ability: some at-risk individuals may develop the disease limited. Although there are no effective preventive
even if they take appropriate measures, whereas other options, given the availability of treatment options, the
at-risk individuals may not fall ill despite their failing to moderate severity and the (very) late onset of the dis-
take preventive action. These risks must be weighed ease, susceptibility testing for AMD will not give rise to
against the potential medical benefits. very many ethical issues in consenting adults. Testing
In conclusion, while the age of onset may be difficult for AMD may become a morally acceptable practice in
to determine for type 2 diabetes and the disease is the clinic or on the direct-to-consumer personal genome
rather severe, it is actionable. There is currently insuffi- testing market, depending largely on the ways in which
cient evidence to support an offer of genetic susceptibil- further ethical issues are dealt with, such as quality
ity testing for type 2 diabetes to children or minors. assurance, information provision and informed consent,
Susceptibility tests are presently of insufficient clinical which are beyond the scope of this paper.
validity and utility [38] in order to justify such an offer. 2.3 Clinical depression: psychiatric disease, stigma, and
For adults, however, it could be argued that in the psychological risk
absence of clear harms, the potential for benefit need Potential consumers have expressed high interest in
not be great in order to justify genetic testing, on the genetic susceptibility testing for clinical depression [54].
condition, naturally, of adequate ethical safeguards such They have also indicated to be interested in such testing
as adequate information provision, informed consent, for their children [55]. Several companies are offering
quality assurance and privacy protection, and provided susceptibility testing for clinical depression directly-to-
the test has a reasonable predictive ability. Stringent consumer [56,57]. The clinical validity of susceptibility
requirements for counselling may not be necessary to testing for clinical depression has not been established
protect against psychological harm, but post-test coun- [58].
selling may be helpful to improve actionability and to Clinical depression can present itself at any age [52].
protect against the risk of false reassurance. While non-genetic risk factors for depression, such as
2.2 Age-related macular degeneration: very late onset emotional neglect or negative emotional experiences in
There have been optimistic reports on the feasibility of early youth, high stress levels, major life events, and
testing for genetic susceptibility to age-related macular drug abuse, have been identified, they are notoriously
degeneration (AMD) [46,47], and several personal gen- difficult to avoid. As of yet, no feasible, effective primary
ome testing services including AMD have already been preventive strategies have been established for depres-
put onto the direct-to-consumer market [2,48,49]. sion, neither in children nor in adults. The disease varies
Genetic susceptibility tests for AMD are considered to in severity, but is generally considered to be relatively
Bunniketal.BMCMedicalGenomics2012,5:4 Page7of9
http://www.biomedcentral.com/1755-8794/5/4
severe [52]. For severe diseases, it could be argued, sus- depression may bring more harm than benefit onto per-
ceptibility tests are morally required to be of sufficient sons. As long as uncertainty prevails regarding the psy-
clinical validity, because low or moderately predictive chological implications of genetic susceptibility testing
testing for severe diseases would leave the tested indivi- for psychiatric diseases, a cautious approach may be
dual with high degrees of uncertainty in the light of a warranted, even in consenting adults. More research
fearful medical scenario. This could cause psychological into the psychological and social consequences of perso-
harm to the patient and so violate the principle of non- nal genome testing for depression and other psychiatric
maleficence. Since, as of yet, the level of clinical validity diseases will be needed.
in susceptibility testing for depression has not proven
sufficient, the ethical acceptability of a testing offer is Summary
not clear. We have identified four disease characteristics that are
Companies are increasingly marketing pharmacoge- relevant to discussions on the ethical issues surrounding
nomic testing for drug response directly to consumers, genetic susceptibility testing for multifactorial diseases:
also in the context of psychiatric diseases [2,59]. One severity, age of onset, actionability and the somatic or
company is offering a combination of susceptibility test- psychiatric nature of the disease. These characteristics
ing and pharmacogenomics testing for anti-depressant are linked to important ethical principles and work
response [56]. Such combinations of susceptibility test- together to affect the ethical debate. For example, the
ing and pharmacogenomic testing are a new develop- potential for adverse psychological and social conse-
ment on the direct-to-consumer genetic testing market. quences of genetic testing for late-onset diseases (harm)
Theoretically, they may increase the actionability of the is greater in the context of diseases that are both severe
test and decrease the perceived severity of the disease and have no actionable options. In such cases testing
tested for, and thus bring more benefit and less harm to poses greater ethical challenges.
individuals than would single susceptibility tests. On the As a general ethical rule of thumb, the likelihood and
other hand, however, this development may not be with- seriousness of possible harms, including psychological
out risk. In the absence of evidence of clinical validity - and social harms, should weigh up against the likelihood
and there is no such evidence for pharmacogenomic and magnitude of the potential (medical) benefits of
testing for antidepressant response [60] - potential for testing. Severity and actionability are therefore relevant
medical benefit is not likely to materialise. Sensitivities disease characteristics. Moreover, severity, actionability
and risks surrounding genetic testing for psychiatric dis- and the somatic/psychiatric distinction affect the
eases should be weighed carefully. requirements for good pre- and post test counselling,
It has been hypothesised that susceptibility testing for such that, for example, genetic susceptibility testing for
psychiatric diseases be made available to children [33]. psychiatric diseases will require careful psychological
Children who are tested to be at increased risk could be counselling. In children, the right not to know must be
provided with ‘pre-symptomatic interventions’ [33], it is protected, which means that late-onset disease should
claimed, consisting of advice about avoiding environ- not be tested for, unless there is a clear advantage (a
mental stress, or prophylactic medication. However, in positive benefit-risk ratio) for the child.
the absence of feasible primary preventive options, pre- We have discussed these disease characteristics and
dictive testing is not likely to yield any medical benefit the resulting ethical issues for three exemplary diseases,
for children. Genetic information about psychiatric dis- type 2 diabetes, age-related macular degeneration
eases may have an unknown effect on the child’s “devel- (AMD) and clinical depression. First, a broader perspec-
oping sense of self and future prospects” [31], and may tive may be appropriate on the age of onset in type 2
even become a self-fulfilling prophecy [33]. As long as diabetes to encompass accumulating risk factors and
clinical validity and utility are lacking, and taking into preclinical stages of disease throughout life. Genetic sus-
account moral considerations such as non-maleficence ceptibility testing for type 2 diabetes may eventually
and the right not to know, there are no convincing rea- become acceptable even in children and minors,
sons to justify the offering of genetic testing for clinical depending foremost on the clinical validity of the test,
depression risk to children or minors. but also on the actionability of the test result, and on
In conclusion, in addition to its variable age of onset, the manner in which ‘age of onset’ is conceptualised.
relatively high level of severity, and unclear preventive Potential for medical benefit must be weighed against
options, clinical depression is characterised by its psy- psychological harms and moral wrongs, such as infringe-
chiatric rather than somatic nature. The disease is ments upon the right not to know. For adults, genetic
endowed with stigma, and knowledge about one’s sus- susceptibility testing for type 2 diabetes may be accepta-
ceptibility may cause psychological risks (see section ble under certain conditions. Second, we have described
1,4). Given the lack of ‘actionability’, genetic testing for AMD as a less severe somatic disease of very late onset.
Bunniketal.BMCMedicalGenomics2012,5:4 Page8of9
http://www.biomedcentral.com/1755-8794/5/4
We have concluded that not many ethical issues are to evaluation, we think that an understanding of ethically
be expected from susceptibility testing for AMD in con- relevant disease characteristics will prove helpful for
senting adults, whether within a clinical context or on further ethical discussions on genetic susceptibility test-
the direct-to-consumer market. Third, clinical depres- ing and personal genome testing for multifactorial
sion is understood to be a psychiatric disease with a diseases.
variable age of onset, a relatively high level of severity
and unclear actionability. Genetic information on psy-
Acknowledgements
chiatric diseases is associated with specific ethical issues,
ThistextisaresultofaresearchprojectoftheCentreforSocietyand
such as stigma and possible adverse psychological conse- Genomics(CSG)inTheNetherlands,fundedbytheNetherlandsGenomics
quences, that warrant a very careful consideration of Initiative.
genetic testing for psychiatric diseases.
Authordetails
As a general conclusion we contend that a critical atti- 1Dept.ofMedicalEthicsandPhilosophyofMedicine,ErasmusUniversity
tude is needed towards personal genome testing services MedicalCenter,dr.Molewaterplein50,Rotterdam,3015GE,theNetherlands.
that offer ‘packages’ of risk estimates for a multitude of 2Dept.ofEpidemiology,ErasmusUniversityMedicalCenter,dr.
Molewaterplein50,Rotterdam,3015GE,theNetherlands.
multifactorial diseases simultaneously, because, as we
have argued, different ethical issues apply to different Authors’contributions
Allauthorshavecontributedsubstantiallytotheconceptionanddesignof
diseases, depending on their characteristics. As some
themanuscript.EBhasdraftedthemanuscript.BothMSandAJhave
personal genome testing companies are offering genetic criticallyrevisedit.Allauthorshavereadandapprovedthefinalmanuscript.
test results for a multitude of diseases that differ from
Competinginterests
one another with regard to the disease characteristics
Theauthorsdeclarethattheyhavenocompetinginterests.
that we have identified [2], consumers are confronted
with test results that vary in emotional impact and thus Received:8September2011 Accepted:19January2012
Published:19January2012
pose different requirements for pre- and post-test infor-
mation and counselling. The ethical evaluation of such
References
broad testing is therefore highly complex. Susceptibility 1. BorryP,CornelMC,HowardHC:Whereareyougoing,wherehaveyou
tests for some diseases, such as AMD, can justifiably be been:arecenthistoryofthedirect-to-consumergenetictestingmarket.
JCommunityGenet2010,1:101-106.
offered within directly-to-consumer personal genome
2. 23andme.[http://www.23andme.com].
testing. For other diseases, on the other hand, it may 3. Navigenics.[http://www.navigenics.com].
not even be morally acceptable to include a genetic sus- 4. GeneticHealth.[http://www.genetichealth.com].
5. Knome.[http://www.knome.com].
ceptibility test at all, or only on the condition of profes-
6. AmericanCollegeofMedicalGenetics(ACMG),BoardofDirectors:ACMG
sional counselling. Finally, many tests may not be Statementondirect-to-consumergenetictesting2008.
morally justifiable in the case of children, because of a 7. EuropeanSocietyofHumanGenetics(ESHG):StatementoftheESHGon
direct-to-consumergenetictestingforhealth-relatedpurposes.EurJ
late onset of the disease and a lack of actionability.
HumGenet2010,18:1271-1273.
Further research will be needed in order to establish a 8. BunnikEM,SchermerMHN,JanssensACJW:Personalgenometesting:Test
sensible subdivision of those broad ‘packages’ into clus- characteristicstoclarifythediscourseonethical,legalandsocietal
issues.BMCMedEthics2011,12:11.
ters of diseases with similar characteristics, so as to
9. BurkeW,PinskyLE,PressNA:Categorizinggeneticteststoidentifytheir
allow for parallel ethical evaluations of clusters of sus- ethical,legal,andsocialimplications.AmJMedGenet2001,106:233-240.
ceptibility tests within a single personal genome test. 10. KääriäinenH,HietalaM,KristofferssonU,etal:Recommendationsfor
geneticcounsellingrelatedtogenetictesting.EurogentestUnit32008.
Such parallel ethical evaluations should point out what
11. WilsonJMG,JungnerG:InPrinciplesandpracticeofscreeningfordisease.
clusters of tests may or may not justifiably be offered, Volume34.Geneva:WHO;1968.
and on what conditions. 12. HaddowJE,PalomakiGE:ACCE:Amodelprocessforevaluatingdataon
emerginggenetictests.InHumangenomeepidemiology:Ascientific
When whole-genome sequencing becomes widely
foundationforusinggeneticinformationtoimprovehealthandprevent
accessible to patients and consumers, and yield disease disease.Editedby:KhouryMJ,LittleJ,BurkeW.OxfordUniversityPress;
risks not only for multifactorial diseases but also for 2003:217-233.
13. CentersforDiseaseControlandPrevention(CDC):PublicHealth
monogenic diseases, these problems are likely to
Genomics.[http://www.cdc.gov/genomics/gtesting/ACCE/acce_proj.htm].
increase even further. It will not be easy to conduct an 14. BeauchampTL,ChildressJF:PrinciplesofBiomedicalEthics.6edition.Oxford
overall ethical evaluation of personal genome testing on UniversityPress,USA;2008.
15. OECD:Guidelinesforqualityassuranceinmoleculargenetictesting2007.
the basis of whole-genome sequencing, or to determine
16. CouncilofEurope:AdditionalProtocoltotheConventiononHumanRights
the appropriate and morally required level of genetic andBiomedicine,concerningGeneticTestingforHealthPurposes2008.
counselling, care and psychosocial support. 17. LarsenLT:Ispreventionbetterthancure?Publichealthpolicyandthe
circularstructureoflearning.SSRNeLibrary2010.
Although other aspects of genetic susceptibility test-
18. HealthCounciloftheNetherlands:GenetischescreeningTheHague:Health
ing, such as the more technical properties of the test or CounciloftheNetherlands;1994,[inDutch].
specific aspects of the context in which testing is 19. CouncilofEurope:RecommendationNo.R(94)11onScreeningasaToolof
PreventiveMedicine1994.
offered, may be equally important to its ethical
Bunniketal.BMCMedicalGenomics2012,5:4 Page9of9
http://www.biomedcentral.com/1755-8794/5/4
20. JanssensACJW,vanDuijnCM:Anepidemiologicalperspectiveonthe 48. MaculaRisk.[http://www.macularisk.com/].
futureofdirect-to-consumerpersonalgenometesting.InvestigGenet 49. GeneticDiagnosticLaboratory.[http://www.med.upenn.edu/genetics/core-
2010,1:10. facs/gdl/].
21. RobertsonS,SavelescuJ:Isthereacaseinfavourofpredictivegenetic 50. SeddonJM,ReynoldsR,MallerJ,FagernessJA,DalyMJ,RosnerB:
testinginyoungchildren?Bioethics2001,15:26-49. Predictionmodelforprevalenceandincidenceofadvancedage-related
22. GrosseSD,McBrideCM,EvansJP,KhouryMJ:Personalutilityandgenomic maculardegenerationbasedongenetic,demographic,and
information:Lookbeforeyouleap.GenetMed2009,11:575-576. environmentalvariables.InvestOphthalmolVisSci2008,50:2044-2053.
23. MorrisonP,Harding-LesterS,BradleyA:UptakeofHuntingtondisease 51. WorldHealthOrganization(WHO):TheGlobalBurdenofDisease:2004
predictivetestinginacompletepopulation.ClinGenet2011,8(3):281-286. Update2008.
24. CouncilofEurope:Conventionfortheprotectionofhumanrightsanddignity 52. EvansJ:Antioxidantsupplementstopreventorslowdownthe
ofthehumanbeingwithregardtotheapplicationofbiologyandmedicine progressionofAMD:asystematicreviewandmeta-analysis.Eye2008,
(ConventiononHumanRightsandBiomedicine)1996. 22:751-760.
25. BorryP,StultiensL,NysH,CassimanJ,DierickxK:Presymptomaticand 53. WongIYH,KooSCY,ChanCWN:Preventionofage-relatedmacular
predictivegenetictestinginminors:asystematicreviewofguidelines degeneration.IntOphthalmol2011,31:73-82.
andpositionpapers.ClinGenet2006,70:374-381. 54. WildeA,MeiserB,MitchellPB,SchofieldPR:Publicinterestinpredictive
26. DuncanRE,SavulescuJ,GillamL,WilliamsonR,DelatyckiMB:An genetictesting,includingdirect-to-consumertesting,forsusceptibility
internationalsurveyofpredictivegenetictestinginchildrenforadult tomajordepression:preliminaryfindings.EurJHumGenet2009,
onsetconditions.GenetMed2005,7:390-396. 18:47-51.
27. BoeninkM:Molecularmedicineandconceptsofdisease:theethical 55. LaegsgaardMM,KristensenAS,MorsO:Potentialconsumers’attitudes
valueofaconceptualanalysisofemergingbiomedicaltechnologies. towardpsychiatricgeneticresearchandtestingandfactorsinfluencing
MedHealthCareandPhilos2009,13:11-23. theirintentionstotest.GenetTestMolBiomarkers2009,13:57-65.
28. NuffieldCouncilonBioethics:MentaldisordersandgeneticsTheethical 56. GenePlanet.[http://www.geneplanet.com/].
context1998. 57. MapmyGene.[http://www.mapmygene.com/].
29. SpriggsM,OlssonCA,HallW:Howwillinformationaboutthegeneticrisk 58. DemirkanA,PenninxBWJH,HekK,etal:Geneticriskprofilesfor
ofmentaldisordersimpactonstigma?AustNZJPsychiatry2008, depressionandanxietyinadultandelderlycohorts.MolPsychiatry2011,
42:214-220. 16:773-783.
30. PhelanJ:Geneticbasesofmentalillness-acureforstigma?Trends 59. TheMayoClinic.[http://www.mayoclinic.org/depression/].
Neurosci2002,25:430-431. 60. KeersR,AitchisonKJ:Pharmacogeneticsofantidepressantresponse.
31. HoopJG:EthicalConsiderationsinPsychiatricGenetics.HarvRev ExpertRevNeurother2011,11:101-125.
Psychiatry2008,16:322-338.
32. MitchellPB,MeiserB,WildeA,FullertonJ,DonaldJ,WilhelmK, Pre-publicationhistory
SchofieldPR:PredictiveandDiagnosticGeneticTestinginPsychiatry.Clin Thepre-publicationhistoryforthispapercanbeaccessedhere:
LabMed2010,30:829-846. http://www.biomedcentral.com/1755-8794/5/4/prepub
33. NewsonAJ:Depressionunderstress:ethicalissuesingenetictesting.BrJ
Psychiatry2009,195:189-190. doi:10.1186/1755-8794-5-4
34. BerghmansR,deJongJ,TibbenA,deWertG:Onthebiomedicalization Citethisarticleas:Bunniketal.:Theroleofdiseasecharacteristicsin
ofalcoholism.TheorMedBioeth2009,30:311-321. theethicaldebateonpersonalgenometesting.BMCMedicalGenomics
35. deCODEme.[http://www.decodeme.com/]. 20125:4.
36. GrantRW,HivertM,PandiscioJC,FlorezJC,NathanDM,MeigsJB:The
clinicalapplicationofgenetictestingintype2diabetes:apatientand
physiciansurvey.Diabetologia2009,52:2299-2305.
37. JanssensACJW,GwinnM,ValdezR,NarayanKMV,KhouryMJ:Predictive
genetictestingfortype2diabetes.BMJ2006,333:509-510.
38. MihaescuR,MeigsJ,SijbrandsE,JanssensAC:Geneticriskprofilingfor
predictionoftype2diabetes.PLoSCurr2011,3:RRN1208.
39. WilliamsDE,etal:Prevalenceofimpairedfastingglucoseandits
relationshipwithcardiovasculardiseaseriskfactorsinUSadolescents,
1999-2000.Pediatrics2005,116:1122-1126.
40. BlossCS,SchorkNJ,TopolEJ:Effectofdirect-to-consumergenomewide
profilingtoassessdiseaserisk.NEnglJMed2011,364:524-534.
41. PijlM,TimmermansDRM,ClaassenL,JanssensACJW,NijpelsG,DekkerJM,
MarteauTM,HennemanL:Impactofcommunicatingfamilialriskof
diabetesonillnessperceptionsandself-reportedbehavioraloutcomes:a
randomizedcontrolledtrial.DiabetesCare2009,32:597-599.
42. FroschDL,MelloP,LermanC:Behavioralconsequencesoftestingfor
obesityrisk.CancerEpidemiolBiomarkersPrev2005,14:1485-1489.
43. EbbelingC,PawlakD,LudwigD:Childhoodobesity:public-healthcrisis,
commonsensecure.Lancet2002,360:473-482.
44. MustA,etal:Morbidityandmortalityassociatedwithelevatedbody Submit your next manuscript to BioMed Central
weightinchildrenandadolescents.AmJClinNutr1996,63:445S-447S. and take full advantage of:
45. MonastaL,BattyGD,MacalusoA,RonfaniL,LutjeV,BavcarA,Van
LentheFJ,BrugJ,CattaneoA:Interventionsforthepreventionof
• Convenient online submission
overweightandobesityinpreschoolchildren:asystematicreviewof
randomizedcontrolledtrials.ObesRev2010,5:e107-118. • Thorough peer review
46. DesprietDDG,KlaverCCW,vanDuijnCC,JanssensACJW:Predictivevalue
• No space constraints or color figure charges
ofmultiplegenetictestingforage-relatedmaculardegeneration.Arch
Ophthalmol2007,125:1270-1271. • Immediate publication on acceptance
47. MallerJ,GeorgeS,PurcellS,FagernessJ,AltshulerD,DalyMJ,SeddonJM: • Inclusion in PubMed, CAS, Scopus and Google Scholar
Commonvariationinthreegenes,includinganoncodingvariantin
• Research which is freely available for redistribution
CFH,stronglyinfluencesriskofage-relatedmaculardegeneration.Nat
Genet2006,38:1055-1059.
Submit your manuscript at
www.biomedcentral.com/submit