Biomarkers in Psychiatry: Ethical Issues

The use of biomarkers—including blood or cerebrospinal fluid markers, genetic testing, neuroimaging, or other biological patterns or signatures (such as EEG)—has yet to become common practice in psychiatry for prediction, diagnosis, or monitoring of response to treatment (1). Very few tests are approved for use in neuropsychiatric disorders, and those that exist carry important caveats about their use (2).

Moreover, the complexity of the genetics of most psychiatric disorders, as well as the interaction with environmental factors, makes prediction of risk of disease challenging for psychiatrists as well as for genetic counselors. Moreover, psychiatrists have expressed concerns about the impact of genetic tests on patients, generally do not feel adequately prepared to discuss the results of genetic testing with patients, and may lack ready access to referral sources for genetic counseling for their patients (3).

Another development that has begun to affect psychiatrists and psychiatric practice is the growth of direct-to-consumer (DTC) genetic testing, through online platforms such as 23andme, Counsyl, and Color Genomics. For a relatively small sum (e.g., $200), consumers can, per one site’s promotional messaging, “Find out what your DNA says about your health, traits, and ancestry.” A recent JAMA Viewpoint article points out that the market for DTC genetic testing appears to be growing substantially and that physicians will be increasingly conscripted to help patients make sense of their genetic information (4).

In addition to genetic testing, other biological markers have also become increasingly sought after in the quest for methods to accurately predict disease onset, correctly diagnose specific forms of illness, monitor disease progression, or predict or evaluate treatment response. Several approaches have now entered the clinical realm, including amyloid PET imaging to aid in the diagnosis of Alzheimer’s disease.

Genetic testing and other biomarker-based evaluation and treatment modalities are raising a host of new ethical issues. We focus on genetic testing—in particular, DTC genetic testing—and amyloid imaging as two examples of increasingly common tests, the ethical implications of which are now beginning to be felt. These implications include the classic tensions among key ethical principles; we illustrate some of the emerging ethical tensions that may result from these developments.

Case 1

Kelsey, an otherwise healthy 14-year-old girl, is doing well emotionally, academically, and socially. She is being treated with appropriate pharmacologic and psychological modalities for attention-deficit hyperactivity disorder by a child psychiatrist. With 5 minutes left in the parental portion of a routine follow-up visit, as Kelsey waits in the waiting room, the mother suddenly states to the daughter’s psychiatrist, “Oh, I meant to ask you. We sent in our saliva for genetic testing through one of those online services. I found out that Kelsey has the gene for dementia, but I haven’t told her anything about it yet. I was hoping you could talk to her about what it means. I didn’t really understand what was on the report. I was only trying to find out where our ancestors came from.”

Test Yourself.

Which of the following is an appropriate course of action for the psychiatrist to take at this point?

A.	Ask the mother whether she obtained Kelsey’s written informed consent for the DTC genetic testing beforehand.
B.	Express interest in seeing the genetic report but caution the mother that genetic counseling may be indicated to ensure that the family fully understands the genetic findings.
C.	Tell the mother that the findings do not predict dementia accurately enough and that she wasted her money.
D.	Ask Kelsey to step back into the office and ask her what she knows about the genetic testing that was done.
E.	Schedule a follow-up appointment once a genetic counselor can be identified who can be present to discuss all of Kelsey’s genetic testing.

The answer is B. The psychiatrist can express interest to maintain a therapeutic relationship with the family. However, given that the vignette paints a picture of a mother who may not fully understand the genetic report, it would be most helpful to obtain formal genetic counseling.

This hypothetical case illustrates some of the issues that psychiatrists may begin to face as genetic testing—perhaps particularly DTC genetic testing—becomes more widespread. What started as the mother’s curiosity about family ancestry appears to have transformed into an unanticipated journey into genetic predictive testing.

Obtaining APOE ε4 genetic testing remains controversial. In this case, it appears that the mother has utilized a DTC genetic testing kit, which allows consumers to “opt in” to receive certain genetic test results, in particular for alleles that place individuals at increased risk of Parkinson’s or late-onset Alzheimer’s disease. The ability of consumers to obtain this information directly (and without a requirement for pre- or posttest genetic counseling) was made possible only relatively recently, by a 2017 decision by the U.S. Food and Drug Administration (FDA) allowing the marketing of “genetic health risk” tests (5).

On the basis of the information provided so far, it appears that the mother may have received a genetic report stating that Kelsey has either one or two alleles of the APOE ε4 allele, which increases the risk of developing Alzheimer’s disease in a dose-dependent manner. Genetic prediction based on APOE testing is complex, however, and risk estimates vary depending on sex, ethnicity, and age (6). As the frequency of DTC genetic testing increases, cases such as Kelsey’s may become more common. In addition, adults who have had these tests performed may raise questions and concerns with their provider, given the complexity of the information and the potential for lack of understanding.

One of the most important tasks (and one that will certainly require more than 5 minutes) is for the psychiatrist to ascertain the degree of understanding that the mother possesses regarding the purpose of genetic testing—in particular, the difference between diagnostic and predictive testing. A very useful overview of the range of goals of genetic testing is provided by Hoge and Appelbaum in their comprehensive review of ethical issues in neuropsychiatric genetics (2):

Although diagnostic testing is conducted in some instances in psychiatry (e.g., to confirm diagnoses of certain neurodevelopmental or neurodegenerative disorders), the case described here presents an instance of predictive testing. One of the purposes of genetic counseling prior to genetic testing is to ensure that there is adequate understanding of the limits of prediction. On the basis of the information presented, the psychiatrist is concerned about whether Kelsey’s mother understands the genetic test results, in particular the probabilistic versus deterministic nature of genetic tests. The common misunderstanding that there is “a gene for” a specific disorder (i.e., genetic determinism) persists, so there will be a need to address this misunderstanding—with success not necessarily assured. Indeed, assessing and enhancing patients’ and families’ understanding of genetic information are likely to become increasingly important roles for psychiatrists and physicians more generally (4).

Informed consent as traditionally conceived—fully informed, voluntary, and competent—may need to be reconsidered as patients increasingly access these tests outside of the physician-patient relationship. In addition, helping patients make sense of these claims or the reports that patients receive from these companies also will create numerous ethical and practical challenges for psychiatrists.

In the case presented, it is unclear under what circumstances the mother had the daughter’s genetic testing done. Did Kelsey actually understand what the testing was about, and did she agree to learn her APOE allele status (i.e., did she make an informed, voluntary, and competent decision about the testing)? If the daughter did not make such a choice, it will be important to ensure that the mother understands the gravity of essentially doing genetic testing on her daughter without her daughter’s consent—essentially depriving her child of the right to decide for herself about an important health matter.

Because of controversy about the relative benefits versus potential personal, social, and legal consequences of APOE genotyping, numerous organizations have recommended against APOE genotyping and—if testing is conducted—strongly recommend pre- and posttest genetic counseling.

Arguments in favor of disclosure of APOE genotype in research and clinical settings, as well as the approval of DTC genetic testing, have been based on the notion of autonomy—that is, the idea that people have a right to their genetic information if they desire it. In addition, results of several studies of disclosure of APOE genotype found that—under highly controlled conditions—adults who learned of positive ε4 results did not, overall, suffer significant psychological harm (6, 7), and the most robust predictors of posttest anxiety and depression were baseline anxiety and depression (7). The majority of studies that have examined responses to genetic testing for Alzheimer’s disease have included adults with a family history of Alzheimer’s disease (8). It should be noted, however, that such findings cannot be generalized to the circumstances surrounding the present case, because there was no pre- or posttest counseling, and it is unclear how well informed the mother was prior to accessing the genetic reports.

Psychiatrists should also be aware that genetic counseling may be obtained through the DTC genetic test companies themselves, usually for an additional fee. In this case, the psychiatrist’s best course of action is to spend additional time gathering more information from the mother about the circumstances of the genetic testing and ascertaining how much the mother already knows about Alzheimer’s disease and genetic risk. The motivations for the mother’s seeking genetic information should also be explored, because there may be relevant family dynamics at play. Although this tricky case is fictional, the issue of whether adolescents should have a choice about receiving their genetic information has been previously discussed and debated. In general, experts believe that predictive testing should not be conducted in children or adolescents when there are no available interventions (2).

Case 2

Mr. Jones is a 65-year-old married man who has been experiencing gradually worsening difficulties with his memory for the last three years. A previously high-performing manager at a technology company, his functioning at work declined to the point at which the company executives noticed his poor performance, and he was asked to retire, which he did six months ago. He now presents for evaluation, accompanied by his second wife, to whom he has been married for 10 years. His wife initiated the appointment and describes Mr. Jones as experiencing depressive symptoms, apathy, difficulty sleeping, decreased appetite, and social withdrawal. When the psychiatrist meets with the patient alone, the patient reports that he knows he is having trouble with his memory but that his biggest concern is that his wife seems very irritated with him and is frequently away from home. The psychiatrist performs a Montreal Cognitive Assessment, on which the patient scores a 22 out of 30, missing 4 points on delayed recall, 2 points on visuospatial abilities, 1 point on verbal fluency, and 1 point on orientation. After speaking to the patient alone, the psychiatrist speaks to the wife alone, seeking to gather additional collateral information about the patient’s functioning. During this time, the wife asks the psychiatrist, “Can you order that brain scan to see if he has dementia? I just really need to know for sure, so I can make plans for the rest of my life. I don’t think I can deal with this for much longer.”

Test Yourself.

At this point, which of the following considerations is LEAST relevant to guiding the psychiatrist’s initial response to the wife’s request?

A.	Convincing the patient to undergo an amyloid PET scan in order to establish a definitive diagnosis of Alzheimer’s disease versus another form of dementia
B.	Assessing the patient’s current symptoms, functioning, and quality of life
C.	Determining whether the patient may be at risk of financial exploitation
D.	Assessing the patient’s safety at home and possible need for supervision
E.	Helping the wife decide what she will do if the patient has dementia

The answer is E. Patients’ families do not always behave the way we would like. The specific type of dementia is important to establish, but it is not of immediate importance. The psychiatrist’s most important obligations at this time, aligned with the principle of beneficence, are to ensure that the patient is safe at home; is not being exploited, neglected, or abused in any way; and has adequate supervision and care appropriate to his level of impairment.

In the case provided, Mr. Jones shows signs of cognitive decline and functional impairment, making a diagnosis of probable dementia nearly certain (assuming that potentially reversible causes of dementia, though unlikely, have been ruled out). The specific type of dementia is not clear yet, however. Amyloid imaging, now approved by the FDA to aid in the diagnosis of Alzheimer’s disease, raises ethical challenges for clinicians working with adults who are either at risk for dementia or already exhibiting signs of cognitive impairment. Is an amyloid PET scan necessary or sufficient to establish Mr. Jones’s diagnosis?

Amyloid test results may not be as clear-cut as a patient or family would like; therefore an informed-consent process that reviews the limits of the predictive certainty of this modality is needed before ordering the scan. Moreover, determining whether the patient has the capacity to provide informed consent for the scan is also indicated. Although the FDA has required a “positive” versus “negative” reading of amyloid scans, “intermediate” levels of tracer binding have been found in up to 20% of patients in some studies—levels whose significance was unclear (9). Second, among people with positive scans, the level of amyloid burden does not clearly predict rate of cognitive decline.

In 2013, a task force convened by the Alzheimer's Association and Society of Nuclear Medicine and Molecular Imaging made a number of recommendations regarding the consideration of amyloid imaging (10). First, the task force felt that amyloid imaging should be considered for patients with cognitive impairment when diagnostic uncertainty remains, even after comprehensive clinical evaluation by a dementia expert. Second, in patients with atypical presentations or early age of onset, who otherwise meet criteria for Alzheimer’s disease, amyloid imaging could be considered but still may not be enough to establish the diagnosis. Thus, one argument in favor of obtaining amyloid imaging in this case is Mr. Jones’s relatively young age (65 years). As age increases, the rate of positive amyloid scans increases (i.e., from 5% among adults 50 to 60 years old to over 50% among adults 80 to 90 years old). Therefore, specificity of a “positive” scan decreases with increasing age.

At this time, most experts agree that the cases in which amyloid testing is indicated are quite limited. If the outcome of the imaging would help with diagnostic certainty—and when that diagnosis would affect treatment decision making—it could be a valuable addition to the patient’s work-up. In a case such as that of Mr. Jones, comprehensive neuropsychological testing may be indicated first to establish the pattern of cognitive deficits. In addition, a brain MRI—a less expensive imaging study—may also help point toward a diagnosis of Alzheimer’s disease.

In terms of treatment, a positive scan may help direct the treatment of cognitive symptoms by pointing toward a more definitive diagnosis of Alzheimer’s disease (AD). For instance, Zhong et al. surveyed 315 clinicians using hypothetical vignettes that described patients with varying levels of cognitive impairment and “positive,” “negative,” or no amyloid imaging information. Positive amyloid imaging results were associated with a greater likelihood, and negative amyloid imaging results with a lower likelihood, of recommending acetylcholinesterase inhibitors (11).

It should be noted that amyloid imaging for cognitively normal adults poses ethical questions that are both similar to and different from those raised by APOE genotyping. Similar to APOE genotyping, there has been debate surrounding whether to disclose—in a research setting—the results of amyloid testing to individuals who are cognitively normal but who have a family history of AD. Similar to the REVEAL study described above, recent work has examined the responses of cognitively normal adults who received amyloid imaging as part of research protocols (12). Mozersky and colleagues found that among 50 adults (ages 65–85 years, cognitively normal; 80% of participants had a family history of Alzheimer’s disease) who had positive amyloid scans, a majority (62%) understood that the positive scan meant that they were at increased (but uncertain) risk of developing AD (12). However, 12% of participants thought either that they were at imminent risk of developing AD or that they had a diagnosis of AD. Thus, despite extensive education about the implications of testing that was provided in this protocol, some participants still misunderstood crucial information.

Although amyloid imaging and other developing biomarker advances may help improve diagnostic accuracy in specific cases, communicating effectively with patients and families about the appropriate uses—and relevant limitations—of these modalities will become increasingly relevant to clinicians. Further work is needed to describe and examine the ethical implications of biomarkers in psychiatry.