Volume 198 - Issue 9

Direct-to-consumer genetic testing — clinical considerations

Author:  Ronald J Trent

Med J Aust 2013; 198 (9): 496-498. || doi: 10.5694/mja12.11019
Published online: 20 May 2013
What doctors should know about do-it-yourself mail-order genetic tests

Do-it-yourself mail-order tests — how should a doctor deal with them?

The DTC genetic testing landscape

A 2003 report by the Australian Law Reform Commission predicted that the number of DTC testing laboratories would grow from the small number operating at the time.1 By 2010, there were over 30 DTC companies, mostly in the United States, whose services were made viable by the robustness of DNA samples sent in the mail, and the growing numbers of available human genetic tests.

The landscape of DTC genetic testing companies is now more complex. Today, there are fewer genetic testing laboratories classified as DTC (about 20) because some companies advertise their DTC tests through the internet, but require a medical practitioner to order them.2 These are not true DTC testing facilities, although they pose some concerns, as will be noted later. It is also important to distinguish DTC genetic tests offered through providers in Australia from those offered by overseas companies, with consequences for regulation and consumer protection.

Rationale and product for sale

Advocates for DTC genetic testing argue that it allows individuals to manage their health more proactively. No one would disagree with this goal. The problem is how DTC tests are advertised and delivered. Opponents of the DTC approach highlight the risks of unproven products being marketed as providing information on clinically significant genetic disorders or traits, where consumers may not be assisted by professionals in assessing the suitability, accuracy or significance of the genetic tests.

Another attraction for consumers of DTC testing, particularly when it is available over the internet, is convenience and greater autonomy in the health system. Traditional genetic testing services provided in Australia are predominantly delivered through public hospitals and can be difficult to access, and they may not be funded through Medicare. In accessing these established services, patients take on a traditional submissive role, which is increasingly at odds with moves toward a doctor–patient collaborative model of care.

DTC genetic testing companies advertise tests for a number of health-related disorders. Sometimes, they provide consumer information about the evidence underpinning the test. For example, one company distinguishes tests based on established research reports, considered to contain reliable findings, or preliminary research reports, considered by the scientific community as needing confirmation.3

Associating a genetic test with medical research gives it some legitimacy but does not indicate whether it can be successfully translated into use in a clinical setting. A term such as “reliable” used by the company to describe the findings of established research reports is imprecise when determining the clinical value of the test. It is even more difficult to see any justification for the use of preliminary research reports in advancing patient care (Box 1).

In response to criticism, companies have relabelled their DTC genetic testing products as “information” rather than as tests for clinical decision making, using various disclaimers. In its sample result for a genetic predisposition DNA testing report, an Australian DTC company (certified to the standard ISO 17025 — see below) states:

Issues for medical practitioners and consumers
Clinical utility

This measure is important for all genetic tests — will the result lead to any meaningful changes in medical management? To consider this it is necessary to review the types of human genetic disorders for which genetic testing is possible.

Mendelian type disorders: Cystic fibrosis (CF) is a single-gene autosomal recessive disorder. A symptomatic newborn child can be confirmed to have CF if he or she is homozygous for the p.Phe508del mutation, the most common one associated with CF. Based on this result, appropriate therapy for CF can be instituted. Therefore this test is clinically useful. Testing for Mendelian disorders is available through conventional genetic testing services in Australia and DTC services.6 Thus, the same genetic test can be provided either in the controlled context of medical advice, or through a DTC mechanism which takes no responsibility for its use in medical decision making. One should also note that not all Mendelian-type genetic tests will have clinical utility. For example, the same CF test sought in a healthy young adult with chronic lung infection will be meaningless because the individual is unlikely to have CF. Even if the individual had, in theory, an extremely mild form of CF, it will not be detected because the range of mutations sought in genetic testing are for severe forms of CF.

Complex genetic disorders: Forty or more genes or genetic loci are implicated in type 2 diabetes. These genes have been identified through population research studies. Tests for complex genetic disorders are not provided by the traditional genetic testing laboratories but can be obtained through DTC laboratories. The problem is whether results from population studies can be translated directly into risks for individuals, without accounting for ethnicity, as many research studies are based on Caucasian subjects. Even if the relative role of the genetic component in disease was precisely understood, we know that environmental contributors to pathogenesis are important. For these reasons, there is very little to no evidence at the moment that genetic testing for complex genetic disorders has any clinical utility.

Some DTC genetic testing scenarios for the medical practitioner
Oversight

Regulatory bodies initially did not consider DTC genetic testing as it was relatively low profile. The DTC industry achieved some notoriety following two enquiries by the US Government Accountability Office, which showed fraudulent practices by some DTC companies.9,10 The United Kingdom’s Human Genetics Commission (HGC) has published recommendations on standards to promote self-regulation, as have the Human Genetics Society of Australasia and the National Health and Medical Research Council of Australia through the production of guidance and information documents. However, the HGC document was criticised because recommendations alone are unlikely to change behaviour without some oversight or incentive to comply.11

A separate regulatory issue is truth-in-advertising. Presumably, DTC companies have received legal advice that selling a product as “information” rather than a medical test is the appropriate way forward. Nevertheless, some websites appear to imply a link between genetic testing, “information” and health outcomes, so the issue of potentially misleading advertising may need to be revisited.

Future

DTC genetic testing will continue to evolve. Linking company services with the requirement for medical practitioners to order the tests is a step in the right direction, but only if medical practitioners have the confidence and eHealth-based tools to determine what tests are clinically relevant and the significance of results. For this it will also be necessary to know about a test’s analytic validity and clinical utility.

1 Selected genetic tests out of the 247 direct-to-consumer tests offered by one United States-based company3*

Test purpose (no. of tests offered)

Examples of available tests

Comments


Carrier status (49)

Cystic fibrosis

Haemochromatosis

Sickle cell anaemia and malaria resistance

Tay–Sachs disease

Relatively straightforward genetic tests dealing with Mendelian genetic disorders and seeking mutations that interfere with gene function. The results should be interpretable, although professional input may be needed. Most of the tests are likely to have clinical utility. Subtle details of the claims made on the website may be misleading, eg, the linking of sickle cell anaemia with malaria resistance on the company’s website is correct from an evolutionary sense but misleading if it suggests that an individual with sickle cell anaemia is resistant to malaria.

Drug response (21)

Abacavir hypersensitivity

Heroin addiction

Naltrexone treatment response

Alcohol consumption, smoking and risk of oesophageal cancer

Testing for the appropriate human leukocyte antigen type before treating with abacavir can reduce the risk of the potentially fatal Stevens–Johnson syndrome. This is an example of how genetic testing can inform treatment options. In contrast, it is not clear how performing the other three tests would lead to changes in drug response or behaviours. Further, these tests are based on preliminary research reports. It is uncertain how the results are interpreted for individual customers.

Disease risk (120)

Asthma

Bipolar disorder

Creutzfeldt–Jakob disease

Gout

Hypertension

Obesity

Diabetes (type 1 and 2)

Testicular cancer

The examples given are only a few of the 120 tests that are identified as assessing disease risk, yet the clinical utility of most genetic tests remains to be demonstrated (even for bipolar disorder, which in this list is considered based on an established research report). They all provide “information”, but it remains to be proven whether it is sufficient or even correct to alter lifestyles based on this information, and whether there will be an impact on outcomes. Many of the tests are for complex genetic disorders, where the relative contributions of genes and the environment are still to be determined.

Traits (57)

Birth weight

Height

Smoking behaviour

Blood glucose

Breastfeeding and IQ

Eating behaviour

The same comments apply here as for the disease risk category, although it would seem even less likely that a genetic test will provide relevant or useful “information” about these traits.


*The disclaimer after the list of genetic tests offered by this company reads: “The tests have not been cleared or approved by the FDA [Food and Drug Administration] but have been analytically validated according to CLIA [Clinical Laboratory Improvement Amendments] standards. The information on this page is intended for research and educational purposes only, and is not for diagnostic use.” Tests considered to be based on established research reports. Funded through Medicare. Tests that are considered to be based on preliminary research reports.

2 Results of a study assessing the performance of two direct-to-consumer (DTC) genetic testing laboratories5

A 2009 study compared the results from two leading United States-based DTC testing facilities that had been sent the same five DNA samples. The results showed an excellent (99.7%) agreement for genetic markers that could be compared. To some extent this would be expected, as most genetic testing laboratories now use sophisticated and automated analytic platforms that reduce the margin for (non-human) error. Thus analytic validity should not be a significant problem in a competent DTC laboratory. In contrast, the study showed disturbing differences in the clinical interpretation of results provided for the same disease and testing the same sample. These inconsistencies included receiving a “high risk” from one laboratory and a “low risk” from the other for prostate cancer, type 2 diabetes, psoriasis and Crohn’s disease.

This example illustrates the importance of distinguishing two aspects of a genetic test: analytic validity and result interpretation. Result interpretation is increasingly becoming the limitation, as the data generated need to be interpreted in terms of biological significance (ie, is this a true DNA mutation leading to a change in gene function?) and clinical significance (ie, what does the genetic test result mean for a patient and his or her family in terms of clinical care?). Hence describing a genetic test with a vague term such as “reliable” is ambiguous as it is not clear whether this refers to the analytic validity or the result interpretation, or ultimately, the impact on patient care.


Author


Competing interests


References


Provenance: Commissioned; externally peer reviewed.