An Environmental Scan of Consumer-Initiated Germline Genetic Testing for Health Risks

Abstract

As patient access to laboratory testing outside the clinic grows, healthcare providers can expect to confront increasing questions about the utility and interpretation of consumer-initiated genetic testing for health risks. We sought to characterize the current marketplace diversity of consumer-initiated germline genetic testing options.

An environmental scan was conducted to identify germline genetic testing companies that offer testing for at least one diagnosable health condition and are available for purchase by consumers in the U.S. market without a visit to one’s healthcare provider. We limited our scope to tests available between October 1, 2019, and September 30, 2021. We characterized variability in the content and processes used by 21 companies offering 74 distinct test products that met our inclusion and exclusion criteria. A small minority (8 out of 21 companies) offered tests that assessed the presence of at least 1 CDC Tier 1 Condition, for which detection can impact an individual’s clinical care and for which evidence-based guidelines for detection and management exist.

Introduction

The increasingly abundant diversity of consumer-initiated genetic testing products has yet to be adequately characterized. In the setting of increasing public interest in genetic testing,¹ and decreased costs of DNA sequencing,² the stage has been set for an explosion of consumer-initiated germline genetic testing products in the U.S. market, which patients can order and complete from home without a visit to a healthcare provider. This is further facilitated by an evolving regulatory landscape that now allows for the marketing of genetic risk assessments and testing for monogenic disease risks³ as well as strategies by companies to evade FDA regulation by providing MDs who can order tests for consumers as well as supplying genetic counseling services as part of the return of results.⁴

Additionally, test content and delivery models for providing consumer-initiated genetic testing have evolved.⁴ Many tests have shifted their focus from providing data about one’s non-disease traits and ancestry, to focusing on health risks, such as identifying genetic variants implicated in both monogenic diseases and complex (multigenic) diseases, whether a participant is a carrier of a heritable disorder they could pass on to future offspring (carrier screening), and whether they have genetic variants that could cause them to metabolize and respond to medications differently (pharmacogenetics).

These different offerings may use different techniques to assess for genetic variation. These may be generally grouped into genotyping versus sequencing-based methodologies. While genotyping assesses for the presence or absence of prespecified genetic variants of interest, sequencing the DNA includes reading the sequence of nucleotides making up a gene, multiple genes, the protein-coding portion of the DNA (whole exome sequencing), or both the coding and non-coding portions of DNA (whole genome sequencing). While genotyping will assess for the presence of absence of a genetic variant of interest, resulting in a positive or negative test result, the results of sequencing will be compared to a reference and results can include the detection of a genetic variant implicated in disease (pathogenic variant), a genetic variant that is not associated with disease (benign variant), or a variant for which it is currently unknown as to whether the variation is associated with disease or not (variant of uncertain significance). Genotyping-based tests may be subject to a higher rate of false negative results than sequencing-based tests, as only selected genetic variants will be assessed for their presence of absence. However, there are still many cases where sequencing alone will not be adequate to detect all relevant disease-associated changes and additional methodologies may be required to decrease the false negative rate (e.g. methods for detection of copy number variants or intronic changes).

The variability in consumer-initiated tests for health risks, available at widely different price points, advertising testing for risk of various disease predispositions, and with different patient-facing processes for both obtaining testing and returning results (e.g. returning only pathogenic variants versus including variants of uncertain significance), could impact consumer willingness to purchase a test, their experiences as a participant, and their perceived utility of testing.

This variability in product offerings and models also can be confusing not only for consumers, but also for clinicians, who may grapple with evaluating the quality and comprehensiveness of diverse test products. The ACCE Model⁵, which gets its name from each of the four considerations for evaluating test quality, can provide a framework to analyze test quality by considering whether a test accurately and reliably tests for the genetic change(s) it is intended to test for (Analytic validity), the test’s ability to detect the clinical disorder associated with the test (Clinical validity), and the extent to which the results are clinically useful as indicated by the appropriate authorities (Clinical utility). Additionally, the model considers the Ethical, legal, and social implications of genetic testing. This environmental scan will not be providing individual measures of clinical validity and utility across the various tests identified, as this is beyond the scope of this scan.

However, extracting the data needed to inform this framework can be challenging to obtain and organize. For example, documentation of CLIA (Clinical Laboratory Improvement Amendments) accreditation⁶ for the laboratory performing testing could provide reassurance about the analytic validity of the tests. Assessing the content and methods used for testing to interrogate risk for specific clinical phenotype(s) of interest could be helpful in determining the clinical validity of the test. Furthermore, clinical utility could be assessed by assessing whether testing for a specific health risk is supported by clinical guidelines or governmental organizations, such as the US Preventive Services Task Force and the Centers for Disease Control and Prevention (CDC)’s Office of Public Health Genomics. For instance, the CDC has identified three “Tier 1” genetic conditions, for which identifying individuals with a pathogenic variant in the associated genes could have a significant impact on public health,⁷ thus speaking to the clinical utility of identifying individuals with these variants. These conditions and the genes associated with them include hereditary breast and ovarian cancer (HBOC, BRCA1 and BRCA2), Lynch syndrome (MLH1, MSH2, MSH6, PMS2, and EPCAM), and familial hypercholesterolemia (LDLR, APOB, and PCSK9).⁸ A genetics assessment for women at increased risk of HBOC is further supported by the US Preventive Services Task Force.⁹ Finally, considering how companies handle participants’ data and whether they report results that may have uncertain clinical meaning at this time (e.g. variants of uncertain significance) could be evaluated among many other ethical, social, and legal implications of consumer-initiated testing.

We performed an environmental scan describing the breadth of consumer-initiated germline genetic testing for health risks both to understand the products being marketed to consumers and to extract data both about test content, processes, costs, and proxies for quality. Our overarching strategy was to identify commercial companies offering genetic tests of interest and to abstract catalog and organize data from the public domain to provide valuable insights for a breadth of stakeholders in healthcare, including consumers and their clinicians.

Study Data and Methods

Identification of Companies and Genetic Tests

We sought to identify genetic tests meeting the following inclusion criteria: 1) were available for purchase by the consumer without a visit to a clinician outside of the testing company, 2) were available to consumers in the U.S. market, 3) assessed for the presence of germline genetic variants,4) and marketed testing for at least one disease predisposition/health risk. We included tests that marketed being able to predict one or more predispositions to a specific disorder (e.g., cancer, diabetes, lactose intolerance, high triglyceride levels, low vitamin levels) but excluded tests that only reported on traits and wellness (e.g., cilantro taste aversion and caffeine metabolism). Tests solely assessing for gene-drug interactions (pharmacogenetic (PGx) tests) or screening only for carrier status for recessive conditions were excluded; however, if PGx testing, carrier screening, or trait and wellness-related testing was offered in combination with testing for personal disease risks, these tests were included. We excluded tests that were unavailable for purchase between October 1, 2019, and September 30, 2021.

Search Strategy

Our overall search strategy was to use a multi-pronged approach to compile a list of candidate companies offering human germline genetic testing and subsequently review genetic test offerings by a company to exclude genetic tests that did not meet our inclusion and exclusion criteria. We conducted our initial search between August and October 2021. Our initial list of candidate companies was compiled using the National Center for Biotechnology Information’s Genetic Testing Registry (GTR),¹⁰ the U.S. Food and Drug Administration’s (FDA) list of nucleic acid-based tests,¹¹ and the International Society of Genetic Genealogy Wiki.¹² Within the GTR, the search was limited to companies offering germline genetic testing in humans with the purpose categorized as either being for risk assessment or presymptomatic assessment, using the GTR’s preexisting search filter options.¹⁰ The search within the FDA’s nucleic acid-based tests was narrowed down to the companies and companies on the list of human genetic tests.¹¹ All companies listed on the International Society of Genetic Genealogy Wiki were considered for inclusion.¹² Finally, because these products are marketed directly to consumers, we also conducted two different internet searches for company products that potentially meet our criteria. We searched both the top three U.S. search engines (Google, Bing, and Yahoo)^13,14 and the top three online retailers (Amazon, Walmart, and eBay)¹⁵ using the keywords “direct-to-consumer DNA” and “home DNA.” We reviewed the first page of advertised tests. We searched hits for new company offerings on each of the search engines and the first 20 hits that appeared in each search of online retailers and added any new companies and their test offerings to our initial search results.

Data Abstraction

The overarching data collection plan was to systematically abstract publicly accessible data about company test offerings meeting our inclusion and exclusion criteria from websites and sample test reports. The authors met to determine domains of interest for data abstraction. Several fields of interest relating to test characteristics and ordering processes were identified. These domains were refined iteratively during the initial review of a subset of company offerings until the final data abstraction fields were codified with working definitions used to guide abstraction (Codebook, Supplemental Table 1).

Data abstraction was performed independently by LEH and HK, who each individually abstracted test data into separate matrices and then met serially to review and reconcile any differences in abstraction. HR was consulted to resolve any residual uncertainties. Attempts were made after initial data abstraction to contact companies to corroborate abstracted findings. The team would reach out twice electronically and, if unsuccessful, once by phone when company contact information was publicly available. Sample test reports were also requested from companies when unavailable on web pages to aid in data abstraction.

Data abstraction and attempts to contact the companies were originally performed between August 2021 to February 15, 2022.¹ If companies did not reciprocate communication upon contacting them after two emails and a phone call, it was noted in the matrix that contact was not made. The final attempts to confirm test details with the companies were on February 15, 2022. Finally, Genome Web¹⁶ an online organization for news about genomics in business and scientific spheres, was also utilized when we had difficulty determining the status of a company (e.g., whether a company had closed or stopped offering tests).

Data from the initial matrix were reviewed by the study team and presented descriptively. The original matrix findings were divided into tables based on test characteristics, ordering processes, and, for a subset testing for one or more CDC Tier 1 Conditions.

Limitations

Although we used multiple approaches to identify and abstract data in our environmental scan, some limitations persist. First, there is the possibility that some companies offering consumer-initiated germline genetic testing for health predispositions during the study period were not identified. Possible reasons for evading our search strategy could include: 1) if the companies or companies had ceased operations and their websites were taken down, 2) if the companies were not registered in the databases that were searched, 3) if the companies were not actively advertising these products, or 4) if some companies pivoted their advertising to alternate tests during the test identification period (for example, COVID-19 testing as a result of the pandemic), such that their older products would not have been as visible in the online search. However, our structured search of multiple registries, search engines, and online resources was designed to mitigate this risk.

Another limitation was the inability to locate and abstract certain data of interest. Reasons for this include lack of public data posted on company websites, the inability to contact companies, and/or companies declining to share requested information. We both report information that we were able to abstract from available documentation (websites, sample test reports) in the public domain, but also denote which companies could not be contacted to corroborate this abstraction, to provide transparency into this limitation. Furthermore, because the data collection was completed before September 30, 2021, and given the rate at which this industry is growing and changing, company offerings described may have changed by the time that these findings are reported.

Additional limitations include that in certain cases we used crude groupings for which more detail might be desired by individuals considering purchasing a specific test. For example, we described whether tests used genotype vs. sequencing-based methodologies. While some labs may perform additional techniques on top of sequencing to enhance the detection of pathogenic variation for specific conditions of concern, we did not report on these. We rather aimed to provide a larger picture overview of techniques used, but a patient or clinician might investigate the specific test performance characteristics for a particular concern in more detail. Furthermore, we did not report on specific privacy policies that may vary between companies and could be of concern to consumers. Reviewing the fine print of any consumer-initiated test is another area into which consumers might seek to dive deeper than in our overview.

Results

Our search strategy identified 308 candidate companies from 4 distinct sources that we included in our environmental scan (Figure). Of these 308 companies, only 21 offered ≥1 test meeting our search criteria. For our final data abstraction, we included 74 distinct consumer-initiated test offerings from these 21 companies about which to abstract further data (Figure). Of these 21 companies, we were able to successfully contact 14 companies to corroborate the data abstracted about their test offerings.

Figure.

Flow Diagram for Identification of Companies

^a We were able to contact 114/21 companies to corroborate the data abstracted from public sources about their test offerings, as noted in subsequent tables.

The companies and specific test offerings meeting our criteria are characterized in Table 1, which is organized to highlight test characteristics and testing processes. All U.S.-based companies endorsed using either internal or external CLIA-certified laboratories, with the exception of Nebula Genomics. Genotyping (determination of alleles at predetermined sites of the genome) was used by 58/74 of the tests and was the most common methodology for analyzing the customers’ DNA. Of the 58 tests that used genotyping methods, 48 of the tests analyzed known variation in multiple genes, whereas 10 of the tests assessed for 1 or more variants in a single gene. Of the 16 tests that used sequencing, 5 of the tests sequenced multiple genes, 7 of the tests used whole exome sequencing, and 4 of the tests used whole genome sequencing for analysis.

Table 1.

Study Characteristics and Processes^a,b

Company	Uses CLIA-certified laboratory(ies)?	Test Name	Test Characteristics							Test Processes
			Methodology	Scope of test (Single-or multi-gene WES, WGS)	Types of diseases assessed (Monogenic diseases, Complex Diseases, or Both)	Out-of-pocket cost	Does the test include testing for any of the following? (Yes/No)			What is Reported? (Individual Variants, Risk Score, or Both)	Is the report customizable? (Yes/No)	Can genetic data be downloaded? (Yes/No)
							CDC Tier 1 condition(s)	Carrier status	PGx
23andMe	Yes	Health + Ancestry 23andMe+	Genotyping	Multi	Both	$1-$250	Yes	Yes	Yes	Both	Yes	Yes
24 Genetics	No (International Company)	Health DNA Test	Genotyping	Multi	Both	$1-$250	Yes	Yes	Yes	Both	Yes	Yes
		Nutrigenetics Test DNA Sport Test; Skin Care Genetic Test	Genotyping	Multi	Complex Diseases	$1-$250	No	No	No	Both	No	Yes
		Exome Sequencing	Sequencing	WES	a	$501+	Yes	Yes	Yes	a	No	Yes
		Whole Genome Sequencing	Sequencing	WGS	a	$501+	Yes	Yes	Yes	a	No	Yes
Ancestry	Yes	AncestryHealth Core; AncestryHealth Plus	Genotyping	Multi	Both	$1-$250	Yes	Yes	Yes	Both	Yes	Yes
CircleDNA	No (International Lab)	Vital DNA Test	Sequencing	WES	Both	$1-$250	No	No	No	Risk Score	Yes	Yes
		Health DNA Test	Sequencing	WES	Both	$251-$500	Yes	No	No	Risk Score	Yes	Yes
		Family Planning DNA Test	Sequencing	WES	Monogenic diseases	$251-$500	No	Yes	No	Risk Score	Yes	Yes
		Premium DNA Test	Sequencing	WES	Both	$501+	Yes	Yes	Yes	Risk Score	Yes	Yes
Color Genomics, Inc.	Yes	Color Extended	Sequencing	Multi	Monogenic diseases	$1-$250	Yes	No	Yes	Individual Variants	Yes	Yes
CRI Genetics	Yes	Ancestry + Health Allergy + Health Weight Loss + Health Weight Loss	Genotyping	Multi	Complex Diseases	$1-$250	No	No	No	Both	Yes	No
DDC	Yes	HomeDNA Healthy Weight	Genotyping	Multi	Complex Diseases	$1-$250	No	No	No	Risk Score	No	a
		HomeDNA Food and Pet Sensitivity	Genotyping	Multi	Complex Diseases	$1-$250	No	No	No	Both	No	a
Dynamic DNA Labs	Yes	Nutrition DNA Test Fitness DNA Test Skin DNA Test Personality DNA Test	Genotyping	Multi	Both	$1-$250	No	No	No	Both	No	Yes
Genebase	Yes	DNA Nutrition Test DNA Fitness Test	Genotyping	Multi	Complex Diseases	$1-$250	No	No	No	Both	No	No
		DNA Weight Loss Test DNA Celiac Disease Test DNA Alcohol Intolerance Test DNA Osteoporosis Test DNA Skin Health Test
		DNA Lactose Intolerance Test DNA Caffeine Sensitivity Test DNA Alzheimer’s Disease Test Cardiovascular Disease (ApoE) DNA Test DNA Narcolepsy Test	Genotyping	Single	Complex Diseases	$1-$250	No	No	No	Both	No	No
		DNA Type 2 Diabetes Test	Genotyping	Multi	Complex Diseases	$251-$500	No	No	No	Both	No	No
		DNA Hemochromatosis Test	Genotyping	Single	Monogenic diseases	$1-$250	No	No	No	Both	No	No
		DNA Cardiovascular Health Test	Genotyping	Multi	Both	$251-$500	No	No	No	Both	No	No
		DNA Thrombotic Risk Test	Genotyping	Multi	Both	$1-$250	No	No	No	Both	No	No
		Age-Related Macular Degeneration Test	Genotyping	Multi	Complex Diseases	$0b	No	No	No	Both	No	No
		Alpha-1 Antitrypsin Deficiency Test MUTYH-Associated Polyposis Test Hereditary Amyloidosis Test Glucose-6-Phosphate Dehydrogenase Deficiency Test	Genotyping	Single	Monogenic diseases	$0b	No	No	No	Both	No	No
Genomind	Yes	Mental Health Map	Genotyping	Multi	Complex Diseases	$1-$250	No	No	Yes	Individual Variants	No	Yes
Helix OpCo. LLC	Yes	Helix Wellness Test Women’s Wellness test	Sequencing	WES	Complex Diseases	$0b	No	No	No	Both	No	Yes
International Biosciences	No (International Company)	Genetic Predisposition Test	Genotyping	Multi	a	$251-$500	No	No	No	Risk Score	a	a
		Lactose Intolerance DNA Test	Genotyping	Multi	Complex Diseases	$251-$500	No	No	No	Risk Score	a	a
		Diet and Healthy Weight DNA Test	Genotyping	Multi	Complex Diseases	$1-$250	No	No	No	Risk Score	a	a
Invitae	Yes	Invitae Cancer Screen Invitae Cardio Screen Invitae Genetic Health Screen	Sequencing	Multi	Monogenic diseases	$251–500	Yes	No	No	Individual Variants	No	No
Kean Health	Yes	Kean Gene	Sequencing	Multi	Complex Diseases	$1-$250	No	No	No	a	a	a
Living DNA	No (International Lab)	Wellbeing Kit Wellbeing + Ancestry	Genotyping	Multi	Complex Diseases	$1-$250	No	No	No	Both	Yes	Yes
Nebula Genomics	No/Not Documented	Deep Whole Genome Sequencing Ultra Deep Whole Genome Sequencing Basic Whole Genome Sequencing	Sequencing	WGS	Both	$501+	Yes	Yes	Yes	Both	Yes	Yes
Ome Care	Yes	OmeHealth	Genotyping	Multi			No	No	No	Both	No a	No a
		OmeGluten OmeSport OmeNutrition OmeSkin Raw Data Nutrition Report			Complex Diseases	$1-$250a
ORIG3N	Yes	Fitness DNA Test Vitamins DNA Test	Genotyping	Multi	Complex Diseases	$1-$250	No	No	No	Individual Variants	No	No
Selfdecode	Yes	Selfdecode DNA Test	Genotyping	Multi	Both	$1-$250	No	No	No	Risk Score	a	Yes a
TellmeGen	No (International Lab)	Tellmegen Starter DNA Kit	Genotyping	Multi	Both	$1-$250	No	No	No	Both	Yes	Yes
		Tellmegen Advanced DNA Kit	Genotyping	Multi	Both	$1-$250	Yes	Yes	Yes	Both	Yes	Yes
Vitagene	Yes	Health and Ancestry Report Premium Report	Genotyping	Multi	a	$1-$250	No	No	No	a	a	Yes

Companies highlighted in gray were unable to be contacted at all.

Abbreviations: CDC Tier 1 Conditions, Center for Disease Control and Prevention Tier 1 Conditions (Include testing for the Hereditary Breast and Ovarian Cancer Syndrome

Lynch Syndrome, and/or Familial Hypercholesterolemia); PGx, Pharmacogenetic Testing

^aIndicates individual fields for which the data is uncertain (could not be verified with lab) or unknown.

^b$0 tests were add-on options that could be added to another or a previously purchased test without an additional charge.

In terms of the types of disease risks reported for these tests, 39 of the tests reported complex diseases, 10 of the tests reported monogenic disease, and 20 of the tests reported on risk for both complex diseases and monogenic disease; test result reporting data was unable to be obtained for 5 tests. A subset of tests were found to definitely include testing for at least one variant implicated in a CDC Tier 1 condition (18/74 tests). In addition, 13 tests screened for carrier status for at least one condition, and 14 included PGx testing.

Most tests (53/74) cost between $1-$250. None of the 74 tests investigated were eligible for insurance coverage (data not shown). There were 8 tests whose cost fell into the $251-$500 cost category, and 6 distinct tests whose cost fell into the $501+ cost category. All tests priced at greater than $500 used whole exome or whole genome sequencing technologies though it was not always possible to determine the scope of analysis of this data. Of note, 7 of the tests were in the $0 cost category. Most of these tests were “add-ons” at no additional cost if the customer already ordered another test from the same company or if the customer already had DNA sequenced by the same company.

We next characterized variation in the processes reporting test results (Table 1). Most tests offerings (53/74) reported both risk scores and individual variants, 9 reported only risk scores, and 7 reported only individual variants. Customization of test reports was also advertised for 17 of the 74 tests. The options to customize were variable among the tests; for example, allowing the customer to choose which results to include in the final report, enabling the customer to request omission of certain results from the final report before registering the DNA kit, and allowing the customer to request more information (such as VUS’s) after receiving results. For 32/74 tests, raw genetic data could be downloaded after the test by the consumer.

A minority of companies (8/21) offered testing for ≥1 CDC Tier 1 Condition (Table 2). Testing for genetic variants implicated in the hereditary breast and ovarian cancer syndrome (HBOC) was the most common CDC Tier 1 condition offered by a lab, followed by those testing for variant(s) implicated in familial hypercholesterolemia and finally, Lynch Syndrome. Of the 8 companies, 2 required sign off by a company contracted physician of the test order to be able to order the test; however, the consumer could opt-in to this option and therefore still be tested while bypassing the traditional healthcare system. Only one of the eight companies indicated that they return variants of uncertain significance. Post-test genetic counseling services offered by the lab were advertised by 50% of these companies (4/8). Detailed data about the companies offering testing for at least 1 CDC Tier 1 condition is available in the Supplement.

Table 2.

Companies offering testing for At Least One CDC Tier 1 Condition (N=8)

	Frequency	%
CDC Tier 1 conditions offered
HBOC Syndrome	8	100
Lynch Syndrome	6	75
Familial Hypercholesterolemia	7	87.5%
Company Contracted-Physician Review of Test Order Required
Yes	2	25%
No	6	75%
VUS Reported
Yes	1	12.5%
No	2	25%
N/A (Genotype-based testing)	3	37.5%
Post-test genetic counseling is available
Yes	4	50%
No	4	50%

Abbreviations: HBOC, hereditary breast and ovarian cancer syndrome; VUS, Variants of uncertain significance Missing Data: FH=1; VUS=2.

CDC Tier 1 Conditions and the genes analyzed to identify pathogenic variants: HBOC (BRCA1 and BRCA2), Lynch Syndrome (MLH1, MSH2, MSH6, PMS2, and EPCAM), or familial hypercholesterolemia (LDLR, APOB, and PCSK9)

BRCA1 and BRCA2 have a USPSTF Grade B recommendation to refer for genetic counseling and/or testing in high risk populations⁹

Discussion

Our environmental scan identified 21 companies offering 74 different consumer-initiated germline genetic tests to assess health risks. This wealth of offerings has likely been driven by several factors, such as increasing public interest in genetic testing,¹ decreased costs of performing genetic testing,² as well as an evolving regulatory landscape. Increased oversight of consumer-initiated genetic testing for health risks was put in place starting in 2010, when the FDA deemed direct-to-consumer tests for health risks to be medical devices requiring FDA approval,⁴ and in 2013, when the FDA issued the direct-to-consumer personal genetics company 23andMe a warning letter to discontinue marketing and sales of their health-related genetic testing, pending FDA authorization for this device.¹⁸ However, the regulatory landscape for consumer-initiated genetic testing for health risks evolved in 2017 after the FDA approved the first direct-to-consumer tests to provide genetic risks for health conditions in 2017.³ Our findings highlight a resurgent and dynamic marketplace during our more recent search period with some companies obtaining FDA approval and others using strategies to avoid FDA regulation such as onboarding MDs to provide required physician involvement. The direct-to-consumer genetic testing market was estimated to be $1.09 billion in 2019,¹⁹ with substantial growth projected.²⁰

Our findings also highlight the breadth of testing products available to consumers. We identified companies marketing tests for a broad scope of diseases, including monogenic disease, complex disease, and both. The scope of health risk testing promoted by the companies ranged from targeting a single disease risk, such as lactose intolerance or Alzheimer’s risk, to panels analyzing genes implicated in the risk for classes of disease (e.g. cardiovascular disease or cancer risk), to whole genome sequencing.

The processes that companies used to facilitate testing also varied. For companies offering testing for CDC Tier 1 condition(s), we did note that two companies used company-contracted physicians who could order a test on behalf of the consumer, without the consumer having to be seen within the traditional healthcare system, which has previously been characterized as an emerging care model.⁴

Finally, the costs for these different services varied widely, but with a general trend towards higher costs for sequencing-based tests that were either panel tests or whole exome or genome sequencing. Although we do not have insights into the actual sales of these varied products, a recent discrete choice experiment of consumers illustrated increased willingness of consumers to pay for consumer-initiated genetic testing for health risks rather than lifestyle and relationship tests, that allowed for at home company specimen collection, and that included sequencing-based testing over genotyping-only tests.¹⁹

In addition to exploring the consumer-facing variability in testing products and processes, we also sought to understand how the quality of test offerings might be different. To that end, we first sought to identify whether the laboratories performing testing were CLIA-certified, as an indicator of analytic validity. Of the 21 companies, we found evidence that 16 used an internal or external CLIA-certified laboratory, providing some reassurance of the quality of testing (analytic validity).

We also found that the health risks assessed, and the methods used for assessing these risks, varied considerably. Most tests used genotyping (58/74) to assess for the presence or absence of specific genetic variants associated with disease risk, which historically has cost less than DNA sequencing.² One concern related to genotype-based tests is that consumers may not appreciate that they only assess for a limited number of risk alleles and/or pathogenic variants implicated in complex or monogenic disease, respectively, such that a negative test does not rule out a genetic predisposition to the disease. In the cases where testing was being used to understand the risk of complex disease, for which genetic, environment, and other factors all influence the risk of disease, genotyping would be able to report on the presence or absence of risk alleles, which weakly would influence the consumer’s risk of disease.

However, in the case of an individual harboring a monogenic disease risk, a negative genotyping test for complex disease risk could be misinterpreted as absent risk for the untested monogenic disease.²¹ Additionally, the clinical validity of a test is in part dictated by using the appropriate methods to detect the genetic variants implicated in disease. Tests using genotyping to assess for only a subset of risk alleles or pathogenic variants may miss individuals at risk for the phenotype of interest. For example, consumer-initiated testing products that use genotyping to assess for a selected group of genetic variants that are associated with specific ethnic or ancestry groups, such as pathogenic mutations in the BRCA1 and BRCA2 genes that are enriched in individuals with Ashkenazi Jewish heritage, are often not useful for individuals from other backgrounds.²² Additionally, there was not always clarity around how data was analyzed and which information was reported in broader tests, such as WGS-based testing.²³ Fortunately, there has been regulatory movement towards providing greater clarity around the strengths and limitations of these tests. For example, the FDA required 23andMe to show that its test reports could be understood by consumers prior to granting marketing authorization for the three common Ashkenazi Jewish variants in BRCA1 and BRCA2 testing.²⁴

In regards to assessing the clinical utility of the consumer-initiated testing options available, we found that only a small minority of the companies offered testing inclusive of CDC Tier 1 conditions, for which identifying one’s predisposition could be clinically actionable at a population level.⁷ For those companies that did test for at least 1 CDC Tier 1 condition, we found that company-sponsored genetic counseling services were advertised by 50% (4/8 companies). These offerings could be helpful, given data demonstrating that when consumer-initiated genetic test results were returned to the customer without the help of a counselor or physician for interpretation, there was an increase in misinterpretation and unnecessary medical intervention.²⁵ Additionally, physicians may be hesitant to provide an interpretation based on the results of these DTC genetic tests due to their lack of training in genetic medicine and their lack of involvement in ordering the test for their patients.²⁵ Therefore, offering company-sponsored genetic counseling services could be helpful both for initial result interpretation, but also for directing next steps in care, such as seeking a medical consultation, communicating with family members, and encouraging genetics assessments of affected family members.²⁶ However, addressing concerns about inadequate quality of company-sponsored genetic counseling services, ²⁶ potentially including regulation,²⁷ is needed to ensure clinical actionable test results are being interpreted and acted upon appropriately.

Additionally, even if access to quality genetic counseling for consumer-initiated genetic testing improves, primary care and other frontline healthcare providers must be equipped to help patients receiving such test results outside of the healthcare system,²⁸ especially as individuals participating in consumer-initiated testing have not routinely expressed interest in,²⁹ or sought out genetic counseling.³⁰ In the Impact of Personal Genomics Study, consumers were actually more likely to bring external genetic test results to their primary care physician rather than a certified genetic counselor.²⁹ This may be a function of a healthcare system in which primary care physicians are often engaged to facilitate subspecialist referral, and thus consumers would need their physician to help facilitate such referrals, which company-based genetic counselors cannot do. Preparing frontline providers for the diversity of consumer-initiated genetic testing they could encounter through education and systematic initiatives should be invested in given the anticipated further growth in population genetics.³¹

Finally, we noted some areas of variability in testing processes that could raise questions about best practices from an ethical standpoint. For example, many companies allowed consumers to download their genetic data. Although we don’t know how all consumers are using these data, in a prior survey of individuals recruited via social media, the majority of respondents (89%) downloaded their genetic data from consumer-initiated genetic testing, and the vast majority of downloaders used third-party tools for genetic interpretation.³² Several ethical questions about the use of third-party interpretation tools have been raised, including whether adequate informed consent is obtained, how to maintain data privacy and security, and whether the variants and their interpretations are analytically and clinically valid.^33,34

The privacy of personal and genetic test data is not only a concern when third-party tools are used for genomic interpretation, but also in the general use of any consumer-initiated test. In many cases, there is a lack of transparency about different companies’ privacy policies, when it comes to communicating the extent to which personal data and results are shared and sold, which may vary substantially.¹⁷ Although analyzing the variation in privacy and data management policies was beyond the scope of this scan, future comparisons could be considered to analyze the extent of variation.

We also found variation in reporting of variants of uncertain significance (VUS), for which we do not currently know the clinical significance but may in the future. Some companies returned VUS, some opted not to return VUS, and some left the choice to the consumer. While returning VUS could be positive from the perspective of promoting access to all data, it could also present challenges including the risk of misinterpretation of these results as well questions about the need and/or how to recontact consumers if these results are updated. Many individuals and clinicians may not realize that most VUS will later be reclassified to benign.^35,36 These dilemmas and variation in practices are not unique to consumer-initiated testing.³⁷

While variability in test offerings and concerns about the unevenness of quality of consumer-initiated testing products could lead to confusion among both patients and clinicians, there are still potential benefits to consumers of expanded consumer-initiated genetic testing. First, consumer-initiated testing options alleviate some access issues created by a physician-gated healthcare system.³¹ Additionally, democratizing access to genetic data could identify actionable disease risks³⁸ and motivate personal health-oriented behavior-changes based on the results.^39,40 For example, in interviews with 32 carriers of a BRCA1 and BRCA2 variant identified via 23andMe’s Personal Genome Service^®, most sought medical advice to reduce their cancer risk, and their results informed cascade screening that identified 13 additional carriers of this variant.⁴¹ More recently, direct-to-consumer testing was used to facilitate population-based access to genetic testing for breast cancer susceptibility for Canadian adults in The Screen Project.³⁰ These studies suggest that democratizing access to preventive genetic testing can have positive impacts for consumers and their relatives. Further work using tools to assess how participants rate the personal utility⁴² of consumer-initiated genetic testing could provide insights into participant-centered outcomes. Therefore, a continued assessment of the risks and benefits of consumer-driven testing would be prudent as this market continues to expand and evolve.

Future Directions

Our results highlight the breadth of testing available on the consumer market. This market will likely continue to expand, growing to include more test offerings, more and/or new companies offering tests, and more health predispositions being covered. Tools like the National Center for Biotechnology Information’s Genetic Testing Registry (GTR)¹⁰ and the U.S. Food and Drug Administration’s (FDA) list of nucleic acid-based tests¹¹ can be used to help understand these offerings, and models such as the ACCE Model⁵ can provide a framework for clinicians seeking to evaluate novel offerings. Understanding the end-users of these tests and their perceptions of their utility could help to generate more understanding about where demand will shift in the future, informing policy, economics, and preparing healthcare providers. Consumer-perceptions of these varied options could help to generate understanding about where the market is heading in the future, suggesting the value in periodic reassessment of the state of consumer-initiated genetic testing. Accompanying this periodic reassessment is a need for deliberate studies of the risks and benefits of consumer-initiated testing to best understand how to incorporate such testing into the practice of genomic medicine.

Conclusion

There was substantial variability in commercial genetic testing products for health risks observed during this environmental scan spanning October 1, 2019, and September 30, 2021, and it is likely that the diversity of product offerings continues to grow and evolve. Applying frameworks such as the ACCE model for evaluating the analytic validity, clinical validity, clinical utility, and ethical, legal, and social implications of genetic testing can provide a useful tool to characterize the quality and completeness of consumer-initiated testing products. However, this approach must be paired with governmental oversight, education for clinicians, and continually updated resources about these products to provide reliable information for clinicians and consumers.

Article Highlights.

• Decreasing costs for genetic testing, increasing public awareness of genetic predisposition for diseases, and an evolving regulatory environment have created a fertile marketplace for companies to offer a diversity of consumer-initiated testing products that do not require a visit with a clinician.

• An environmental scan of consumer-initiated genetic testing products for health risks available between October 1, 2019, and September 30, 2021 identified 21 companies offering 74 distinct laboratory offerings.

• Test offerings varied in terms of health risks assessed, test methodology used, processes for obtaining testing and returning results, and costs to the consumer.

• A minority of companies (8 of 21) included testing for one of the following CDC Tier 1 conditions for which there is greater evidence for clinical utility of preventive testing: the hereditary breast and ovarian cancer syndrome (BRCA1 and BRCA2 testing), Lynch Syndrome, and familial hypercholesterolemia.

Abbreviations:

CDC

Centers for Disease Control and Prevention

CLIA

Clinical Laboratory Improvement Amendments; DNA, deoxyribonucleic acid

DTC

direct-to-consumer

FDA

Food and Drug Administration; GTR, Genetic Test Registry

HBOC

Hereditary breast and ovarian cancer syndrome

PGen

Impact of Personal Genomics Study

PGx

Pharmacogenomics

VUS

Variants of Uncertain Significance