It’s Really Easy to Get Your Genetic Results. But Good Luck Making Sense of Them.

Direct-to-consumer genetic tests make DNA information more accessible and affordable. But no one can really be sure what to make of the results.

National Journal
Nov. 8, 2013, midnight

A ge­net­ic test a few years ago told Cath­er­ine Afar­i­an and her hus­band that their fu­ture child would have a 50 per­cent chance of in­her­it­ing brown eyes, 30 per­cent of hav­ing green eyes, and 20 per­cent blue. That same test in­dic­ated that she has a de­creased risk for Alzheimer’s, and an in­creased one for colorectal can­cer.

Afar­i­an’s “ge­net­ic ex­per­i­ment,” as she calls her son, un­sur­pris­ingly ended up with brown eyes. Her risk res­ults though were a bit un­ex­pec­ted: the Alzheimer’s res­ults were a re­lief be­cause she has a fam­ily his­tory; the can­cer res­ults were a good thing to know be­cause she doesn’t.

Afar­i­an is the spokes­per­son for 23andMe, a privately held dir­ect-to-con­sumer, or DTC, ge­net­ic-test­ing com­pany based in Sil­ic­on Val­ley. The com­pany is at the fore­front of a grow­ing move­ment to bring the rap­idly evolving world of ge­net­ics in­to the hands of con­sumers, mak­ing their in­form­a­tion ac­cess­ible and af­ford­able for the first time.

“As a com­pany, we fun­da­ment­ally be­lieve your DNA is yours, you own it, and you should have ac­cess to it,” she says. “It’s our mis­sion to tell you everything sci­ence can.”

However, when it comes to ge­net­ics, sci­ence doesn’t know everything. And since what sci­ence can tell us about ge­net­ics is con­stantly chan­ging, ac­cur­ate in­ter­pret­a­tion of the res­ults is po­ten­tially out of step with the rap­id growth of test­ing tech­no­logy.

23andMe and the World of Ge­net­ic Test­ing

23andMe and oth­er DTC ge­net­ic-test­ing com­pan­ies have boomed in re­cent years, with hun­dreds of thou­sands us­ing the ser­vices to learn about their DNA. In 2007, 23andMe’s first gen­ome ser­vice cost $999 and provided 14 ge­net­ic re­ports; today it is $99 and of­fers 254 dif­fer­ent re­ports. 

These re­ports are lim­ited to dis­eases or con­di­tions that re­search­ers have some in­form­a­tion about. 23andMe of­fers re­ports in four cat­egor­ies: car­ri­er status, dis­ease risk, drug re­sponse, and traits. These range from cyst­ic fibrosis to Par­kin­son’s to an­ti­de­press­ant re­sponse to sex-hor­mone reg­u­la­tion. “As­paragus meta­bol­ite de­tec­tion,” un­der traits, will rank the odds of smelling as­paragus in your ur­ine.

“As­paragus meta­bol­ite de­tec­tion,” un­der traits, will rank the odds of smelling as­paragus in your ur­ine.

A ge­net­ic test through a doc­tor would typ­ic­ally be either a spe­cif­ic test for a spe­cif­ic muta­tion to con­firm a dia­gnos­is, or a broad­er dia­gnost­ic test look­ing for mark­ers across the gen­ome as­so­ci­ated with dis­ease states. But 23andMe ana­lyzes about 1 mil­lion of the more than 10 mil­lion single nuc­le­otide poly­morph­isms in the hu­man gen­ome and re­ports find­ings for a range of char­ac­ter­ist­ics. 

Se­lect­ing cer­tain SN­Ps from the gen­ome helps 23andMe keep costs low. Con­sumers simply mail in their saliva and 23andMe sends it to a cer­ti­fied lab. Two to four weeks later, the re­ports are avail­able se­curely on­line.

“The as­sump­tion is that at some point [23andMe] will move to whole se­quen­cing — but at a price point for con­sumers,” Afar­i­an says. 

The growth of these con­sumer tests holds great sci­entif­ic prom­ise, but also raises the ques­tion of un­cer­tainty on the in­ter­pret­a­tion side.

For 23andMe, in­ac­cur­acy in the test­ing is not a con­cern. “The in­form­a­tion 23andMe provides is in­cred­ibly ac­cur­ate,” Afar­i­an says. “If we tell you you’re a C-C there, you are. From there, the next step is in­ter­pret­a­tion — what does sci­ence know about it?”

It turns out the an­swer isn’t so clear.

(In)ac­cur­acy of In­ter­pret­a­tion

“There is a huge gap in in­form­a­tion about ge­n­om­ic ser­vices,” says James O’Leary, chief in­nov­a­tion of­ficer at Ge­net­ic Al­li­ance, a non­profit ad­vocacy or­gan­iz­a­tion ded­ic­ated to trans­form­ing health through ge­net­ics. “New tech­no­lo­gies de­vel­op quickly and it’s un­clear to people how to ac­cess them, what the pur­poses are, and what is reas­on­able in terms of res­ults.”

Even if the tests them­selves are ac­cur­ate, it’s im­port­ant for con­sumers to take much of the in­ter­pret­a­tion with a grain of salt.

“There is a huge gap in in­form­a­tion about ge­n­om­ic ser­vices.”

“Every­one will have res­ults. Every­one has muta­tions. Wheth­er they are med­ic­ally ac­tion­able or not is a dif­fer­ent story,” O’Leary ex­plains.

Not only are muta­tions not ne­ces­sar­ily in­dic­at­ive of a con­di­tion, they’re just part of the story. “A lot of the in­form­a­tion that is giv­en back to pa­tients on those types of life­style im­pacts of ge­net­ics are not very in­flu­en­tial risk factors,” O’Leary con­tin­ued. “The ge­net­ic por­tion doesn’t ac­count for a large por­tion of risk; ge­net­ic in­form­a­tion doesn’t make a lot of sense without oth­er life­style and en­vir­on­ment­al factors.” 

Afar­i­an re­cog­nizes the lim­its as well. “It’s the ba­sic nature of ge­net­ics that there are very few cases where ge­net­ics are de­term­in­ist­ic, mean­ing if you have a gene, you will have the con­di­tion,” she said. Her son could have had blue eyes, and Afar­i­an’s risk factor res­ults don’t mean that she will or won’t get those dis­eases. “At the end of the day, age is the greatest risk factor for Alzheimer’s. Fun­da­ment­ally, if you don’t want to get Alzheimer’s, don’t get old.”

In rare cases, however, muta­tions are in­dic­at­ive of far high­er risk, and are med­ic­ally ac­tion­able. An oft-cited ex­ample is a harm­ful muta­tion in the BRCA1 or BRCA2 genes, which greatly in­creases the risk of de­vel­op­ing breast or ovari­an can­cer. This is the muta­tion that fam­ously caused An­gelina Jolie to get a pre­vent­at­ive mastec­tomy earli­er this year. Where­as in many cases muta­tions will only in­crease risk by a frac­tion of a per­cent, a muta­tion on the BRCA1 or BRCA2 genes in­creases the risk sev­er­al-fold. If there is a fam­ily his­tory of can­cer that sug­gests someone may have a muta­tion in one of these genes, he or she can take a ge­net­ic test to check, and then de­term­ine how to man­age their risk.

BRCA1 and BRCA2 re­ports are offered through 23andMe, but it is a more lim­ited screen­ing. Since a fam­ily his­tory can de­term­ine wheth­er an in­di­vidu­al ought to take the ge­net­ic test at all, and be­cause it is a com­plex and ser­i­ous ill­ness, BRCA test­ing is prob­ably bet­ter suited for a doc­tor-ad­min­istered ge­net­ic test.

Re­cre­ation­al or med­ic­al?

While much of DTC ge­net­ic test­ing began largely for en­ter­tain­ment — traits like eye col­or, hair type, ear­wax con­sist­ency — the grow­ing em­phas­is on com­plex health and dis­ease factors could make the know­ledge gap a great­er con­cern. 

“It’s im­port­ant to re­cog­nize that for many people this truly is ‘re­cre­ation­al ge­net­ics,’ ” says Howard Levy, an as­sist­ant pro­fess­or in the Di­vi­sion of Gen­er­al In­tern­al Medi­cine and McK­usick-Nath­ans In­sti­tute of Ge­net­ic Medi­cine at Johns Hop­kins Uni­versity. “If they’re get­ting the data and mak­ing med­ic­al de­cisions without med­ic­al coun­sel­ing, that could be a prob­lem. If it’s just cool stuff to know — track­ing an­ces­try, cock­tail-party con­ver­sa­tion, for fun — and they are not mak­ing med­ic­al de­cisions, maybe that’s not a bad thing.”

However, the fact that 23andMe res­ults in­clude all types of con­di­tions means con­sumers are get­ting both “en­ter­tain­ing” and health care-re­lated re­ports.

“Be­cause so many res­ults are of un­cer­tain sig­ni­fic­ance, it’s an in­cred­ibly chal­len­ging thing to deal with and fig­ure out what it means as a con­sumer,” O’Leary says. “We’re find­ing out new in­form­a­tion about the gen­ome all time; man­aging that out­side the health care con­text is really dif­fi­cult for people.”

23andMe aims to al­le­vi­ate some of this dif­fi­culty by provid­ing in­ter­pret­a­tion for the res­ults. Their re­ports cite sci­entif­ic stud­ies and use a star sys­tem to in­dic­ate the level of cred­ib­il­ity. As new stud­ies come out, 23andMe sends an email in­cor­por­at­ing new in­form­a­tion to keep people en­gaged in their ge­net­ics.

But this is where the pace of ge­net­ics re­search gets tricky.

This isn’t a prob­lem with the sci­ent­ists and clini­cians at 23andMe. “What it comes down to,” Levy says, “is if we’re go­ing to live on the cut­ting edge of know­ledge, new dis­cov­er­ies are ac­ted on right away. The down­side is that new dis­cov­er­ies some­times make a mis­take.” 

Levy says he is more open to DTC ge­net­ic test­ing than most phys­i­cians; that it comes down to a philo­sophy of autonomy versus pa­ter­nal­ism, and he’s a fan of autonomy. Yet en­gage­ment re­lies on con­tin­ued par­ti­cip­a­tion and ac­count­ab­il­ity on the part of the con­sumer as well. Keep­ing up with the in­form­a­tion, Levy says, is equally the re­spons­ib­il­ity of the con­sumer and the ge­net­ics-in­ter­pret­a­tion com­pany. 

Levy does re­com­mend that con­sumers bring their res­ults to their doc­tor as well, something that 23andMe and many ge­net­ics pro­fes­sion­als en­cour­age, par­tic­u­larly if con­sumers are con­sid­er­ing a life­style or med­ic­al change based on their test res­ults. “[An in­di­vidu­al] could cause ser­i­ous per­man­ent harm to them­selves by mis­in­ter­pret­ing the in­form­a­tion they get,” Levy says.

Doc­tors and the Edu­ca­tion De­fi­cit

As of now though, stud­ies show that the ma­jor­ity of test con­sumers don’t share res­ults with their doc­tors. In fact, shar­ing the res­ults with a doc­tor has two sig­ni­fic­ant road­b­locks: cost and ex­per­i­ence. 

While the ac­tu­al DNA test­ing is quickly be­com­ing in­ex­pens­ive and ubi­quit­ous, pro­fes­sion­al in­ter­pret­a­tion still re­quires time and ex­pert­ise. This iron­ic­ally makes the phys­i­cian fol­low-up more of a lim­it­ing factor in un­der­stand­ing res­ults than the tests them­selves. 

The second prob­lem is that even doc­tors and spe­cial­ists in the field some­times don’t know or can’t keep up with ge­net­ics re­search. “There’s a sig­ni­fic­ant edu­ca­tion­al de­fi­cit — on the con­sumer and also the phys­i­cian side,” says Mi­chael Dougherty, dir­ect­or of edu­ca­tion at the Amer­ic­an So­ci­ety of Hu­man Ge­net­ics. “If you take [DTC res­ults] to the phys­i­cian, gen­er­ally they don’t know what to do with it.”

“There’s a sig­ni­fic­ant edu­ca­tion­al de­fi­cit.”

As a res­ult, med­ic­al groups are work­ing to in­crease edu­ca­tion of the pub­lic and of phys­i­cians re­gard­ing ge­net­ics and its grow­ing role in medi­cine. Dougherty ex­plains that while doc­tors should refer pa­tients to spe­cial­ists if they sus­pect the in­di­vidu­al has a ge­net­ic con­di­tion, ge­net­ic spe­cial­ists are un­for­tu­nately in short sup­ply. This makes it im­port­ant for phys­i­cians to be­come more fa­mil­i­ar with the field. “Where the greatest [edu­ca­tion­al] need ex­ists is with front­line health pro­viders — nurses phys­i­cians, primary-care phys­i­cians, nu­tri­tion­ists, et cet­era,” he says. “It will be­come in­cum­bent in­creas­ingly on primary-care phys­i­cians to re­cog­nize when the ge­net­ic test res­ults are mov­ing bey­ond area of their ex­pert­ise.” 

In ad­di­tion to edu­ca­tion pro­grams, data through DTC tests them­selves can also con­trib­ute to on­go­ing ge­net­ic re­search to help bet­ter un­der­stand res­ults. Ninety per­cent of 23andMe test con­sumers op­ted in to par­ti­cip­ate in the re­search, al­low­ing 23andMe sci­ent­ists to view their an­onym­ous data and identi­fy hun­dreds of new ge­net­ic as­so­ci­ations and identi­fy tar­gets for po­ten­tial new drugs and treat­ments.

The op­por­tun­ity to con­trib­ute to that re­search is one of the things 23andMe val­ues most. “I’d like to think that some treat­ment 20, 30, 40 years from now — I’d like to think my fam­ily par­ti­cip­ated in mak­ing some of those things avail­able,” Afar­i­an said.

As our con­ver­sa­tion was wrap­ping up, I asked if there was any­thing she’d like to add.

“Well it sounds like you’re not a cus­tom­er,” she said. “I would love to send you a kit!” 

I’m not sure I could handle the un­cer­tainty.

What We're Following See More »
Trump Inauguration Spending Now Under Investigation
59 minutes ago

"Federal prosecutors in Manhattan are investigating whether President Trump’s 2017 inaugural committee misspent some of the record $107 million it raised from donations, people familiar with the matter said. The criminal probe by the Manhattan U.S. attorney’s office, which is in its early stages, also is examining whether some of the committee’s top donors gave money in exchange for access to the incoming Trump administration, policy concessions or to influence official administration positions."

Federal Judges Nix Proposed Atlantic Pipeline
2 hours ago

In a rare rebuke to energy companies in the Trump era, "a panel of federal judges has rejected permits for the Atlantic Coast natural gas pipeline to cross two national forests and the Appalachian trail in Virginia, finding that the national Forest Service 'abdicated its responsibility' and kowtowed to private industry in approving the project. The harshly worded, 60-page decision issued Thursday by three judges from the U.S. Court of Appeals for the Fourth Circuit is part of a string of legal setbacks for the 600-mile pipeline. The $7 billion project, being built by a consortium of companies led by Dominion Energy, is planned to carry natural gas from West Virginia, through Virginia and into North Carolina."

Senate Moves to End Support for Saudi War
3 hours ago
Federal Judge Upholds Ranked-Choice Voting in Maine
5 hours ago

"A federal judge on Thursday rejected Republican U.S. Rep. Bruce Poliquin’s constitutional claims against ranked-choice voting and denied the incumbent’s request for a new election against Democratic Congressman-elect Jared Golden. U.S. District Court Judge Lance Walker ruled that, contrary to the arguments of Poliquin’s legal team, the U.S. Constitution does not require that whichever congressional candidates receives the most votes—or 'a plurality'—be declared the winner. Instead, Walker ruled the Constitution grants states broad discretion to run elections."

Mueller Probing Middle East Countries' Influence Campaigns
5 hours ago

Officials working under Special Counsel Robert Mueller are investigating Middle Eastern countries' attempts to influence American politics, and are set to release the findings in early 2019. "Various witnesses affiliated with the Trump campaign have been questioned about their conversations with deeply connected individuals from the United Arab Emirates, Saudi Arabia, and Israel ... Topics in those meetings ranged from the use of social-media manipulation to help install Trump in the White House to the overthrow of the regime in Iran." Investigators are also probing meetings organized by Lebanese-American businessman George Nader, and Joel Zamel, "a self-styled Mark Zuckerberg of the national-security world with deep ties to Israeli intelligence."


Welcome to National Journal!

You are currently accessing National Journal from IP access. Please login to access this feature. If you have any questions, please contact your Dedicated Advisor.