We have identified four disease characteristics that are relevant to discussions on the ethical issues surrounding genetic susceptibility testing for multifactorial diseases: severity, age of onset, actionability and the somatic or psychiatric nature of the disease. These characteristics are linked to important ethical principles and work together to affect the ethical debate. For example, the potential for adverse psychological and social consequences of genetic testing for late-onset diseases (harm) is greater in the context of diseases that are both severe and have no actionable options. In such cases testing poses greater ethical challenges.
As a general ethical rule of thumb, the likelihood and seriousness of possible harms, including psychological and social harms, should weigh up against the likelihood and magnitude of the potential (medical) benefits of testing. Severity and actionability are therefore relevant disease characteristics. Moreover, severity, actionability and the somatic/psychiatric distinction affect the requirements for good pre- and post test counselling, such that, for example, genetic susceptibility testing for psychiatric diseases will require careful psychological counselling. In children, the right not to know must be protected, which means that late-onset disease should not be tested for, unless there is a clear advantage (a positive benefit-risk ratio) for the child.
We have discussed these disease characteristics and the resulting ethical issues for three exemplary diseases, type 2 diabetes, age-related macular degeneration (AMD) and clinical depression. First, a broader perspective may be appropriate on the age of onset in type 2 diabetes to encompass accumulating risk factors and preclinical stages of disease throughout life. Genetic susceptibility testing for type 2 diabetes may eventually become acceptable even in children and minors, depending foremost on the clinical validity of the test, but also on the actionability of the test result, and on the manner in which 'age of onset' is conceptualised. Potential for medical benefit must be weighed against psychological harms and moral wrongs, such as infringements upon the right not to know. For adults, genetic susceptibility testing for type 2 diabetes may be acceptable under certain conditions. Second, we have described AMD as a less severe somatic disease of very late onset. We have concluded that not many ethical issues are to be expected from susceptibility testing for AMD in consenting adults, whether within a clinical context or on the direct-to-consumer market. Third, clinical depression is understood to be a psychiatric disease with a variable age of onset, a relatively high level of severity and unclear actionability. Genetic information on psychiatric diseases is associated with specific ethical issues, such as stigma and possible adverse psychological consequences, that warrant a very careful consideration of genetic testing for psychiatric diseases.
As a general conclusion we contend that a critical attitude is needed towards personal genome testing services that offer 'packages' of risk estimates for a multitude of multifactorial diseases simultaneously, because, as we have argued, different ethical issues apply to different diseases, depending on their characteristics. As some personal genome testing companies are offering genetic test results for a multitude of diseases that differ from one another with regard to the disease characteristics that we have identified [2
], consumers are confronted with test results that vary in emotional impact and thus pose different requirements for pre- and post-test information and counselling. The ethical evaluation of such broad testing is therefore highly complex. Susceptibility tests for some diseases, such as AMD, can justifiably be offered within directly-to-consumer personal genome testing. For other diseases, on the other hand, it may not even be morally acceptable to include a genetic susceptibility test at all, or only on the condition of professional counselling. Finally, many tests may not be morally justifiable in the case of children, because of a late onset of the disease and a lack of actionability. Further research will be needed in order to establish a sensible subdivision of those broad 'packages' into clusters of diseases with similar characteristics, so as to allow for parallel ethical evaluations of clusters of susceptibility tests within a single personal genome test. Such parallel ethical evaluations should point out what clusters of tests may or may not justifiably be offered, and on what conditions.
When whole-genome sequencing becomes widely accessible to patients and consumers, and yield disease risks not only for multifactorial diseases but also for monogenic diseases, these problems are likely to increase even further. It will not be easy to conduct an overall ethical evaluation of personal genome testing on the basis of whole-genome sequencing, or to determine the appropriate and morally required level of genetic counselling, care and psychosocial support.
Although other aspects of genetic susceptibility testing, such as the more technical properties of the test or specific aspects of the context in which testing is offered, may be equally important to its ethical evaluation, we think that an understanding of ethically relevant disease characteristics will prove helpful for further ethical discussions on genetic susceptibility testing and personal genome testing for multifactorial diseases.