Three reasons not to buy the nicotest(tm) genetic test

Three reasons not to buy the NicoTestTM genetic test
GeneWatch UK
December 2004
What is the NicoTestTM?
NicoTest is a new genetic test kit being marketed directly to the public via the internet(www.nicotest.com). It is marketed by a company called g-Nostics Ltd which hopes tosell it more widely in the future, both “over the counter” and via doctors in the NationalHealth Service. G-Nostics is a “spin out” company from Oxford University. The universityis a shareholder in the company and the test is based on research by Dr Robert Waltonin the university’s Department of Clinical Pharmacology1. Dr Walton is Chief ScientificOfficer, Lead Inventor and co-founder of the company2.
NicoTest is a smoking cessation programme, which includes three things: 1. A questionnaire and DNA test which the company claims wil identify which smokersare more likely to respond to nicotine replacement therapy (nicotine patches, gum andinhalers) and less likely to respond to drugs such as bupropion (Zyban) when they try toquit smoking;2. A second DNA test which the company claims gives the customer’s metabolic profile(how their body responds to nicotine) and therefore how much nicotine replacement totake;3. A programme of support to quit, including email, chat rooms and a computerprogramme (based on cognitive behavioural therapy).
The test costs £94.99 and involves sending a drop of blood to the company by post.
Which genes are tested?
The first genetic test involves a gene which the company claims predisposes people tonicotine addiction. Although this gene is not named on the website, it is clear from thereferences given that it is a gene known as the ‘dopamine D2 receptor gene’, DRD2. Acommon variation in this gene is claimed to relate to how useful people find nicotinepatches when they try to stop smoking. People with the ‘addiction gene’ supposedly findnicotine patches more useful than people without it, but people without the ‘addictiongene’ supposedly find the drug bupropion (better known by its trade name, Zyban) moreuseful.
The second genetic test gives a ‘metabolic profile’ and is used to advise how muchnicotine replacement to take. The main gene thought to be involved in nicotinemetabolism is called CYP2A6.
In GeneWatch’s view this genetic test is, at best, a waste of money and
could also be misleading and potentially harmful to health.
We recommend that you do not take this test.
The claims made for this genetic test are based on partial and misleading evidence. Forexample, the claimed “addiction gene” included in the test does not have a statisticallysignificant association with nicotine addiction. GeneWatch believes that an independent regulator should be set up to assess allgenetic tests before they are sold. We also think that genetic tests should not bemarketed directly to the public, but only via medical professionals who can ensure thatthey are properly interpreted. The following is a summary of our concerns about the NicoTestTM.
1. Poor and selective scientific evidence.
1.1 The claimed nicotine “addiction gene” does not have a statistically significant
association with smoking.
The NicoTest website claims that its first genetic test identifies people born with “a genepredisposing them to nicotine addiction” (these are the people advised to use nicotinereplacement therapy when they try to stop smoking). In fact the statistical link betweenthis gene (DRD2) and nicotine addiction is disputed – some studies have found a linkbut others have not. A recent study has attempted to combine the results of all theresearch on the DRD2 gene. It found that the association between the gene andsmoking behaviour (the likelihood of starting, continuing and quitting smoking) was notstatistical y significant when the most sophisticated analysis was used3. This means thelink remains at best uncertain and may not exist at all.
The most recent study in the UK concludes that inherited genetic variations in DRD2“have little or no effect on an individual’s smoking behaviour”4 and states this finding isconsistent with the only earlier study done in the UK. G-Nostics Ltd must know aboutthese studies because its Chief Scientific Officer, Lead Inventor and Co-Founder (DrRobert Walton) is one of the authors on these scientific papers.
It is wrong to tell people that they have a “nicotine addiction gene” when the
evidence for this claim does not meet accepted scientific standards.

1.2 Published scientific evidence suggests that the first DNA test may be no use
at all for deciding which treatment best helps men to quit.
The “FAQs” section of NicoTestTM website cites one journal paper which it says showsthat its DNA test can identify people who are more likely to respond to nicotinereplacement therapy (NRT)5 – these are the people with the so-cal ed addiction gene(the rarer form of the DRD2 gene, which about 35% of people have). The website doesnot cite a more recent paper in the British Medical Journal, which repeats the samestudy looking at differences between men and women with the DRD2 gene6. This studyconcludes: “In women the effectiveness of nicotine patches seems to be related togenotype [genetic makeup]…No significant relationship between genotype and patcheffectiveness was seen for men”. Again, Dr Walton is an author on both these scientificpapers.
The same section of the website cites a different scientific paper which it says shows itsDNA test will identify people less likely to respond to bupropion (Zyban)7. This paperlooks for links between different variations in the DRD2 gene and success in quittingsmoking for people taking this drug. It also concludes that: “Significant associations ortrends were not observed in men”. When the results for men and women werecombined together, the effect of the gene on response to Zyban was not statistical y significant. Another study (not cited on the website) has also found that the effect of theDRD2 gene on quitting smoking using Zyban was not significant when men and womenwere combined together8.
The website does not mention evidence that its genetic test is not likely to be
useful for men. This is a particularly serious omission because more men than
women smoke.
1.3 The evidence that the first DNA test is useful to women is limited and partial.
The evidence that the first DNA test is useful to identify who responds best to nicotinepatches is based on a single study of 1625 people, 752 of whom then had the genetictest5. This scientific paper does not conclude that people should have a genetic testbefore using nicotine patches – it concludes that more research is needed. Although itfound a link between genetic makeup and the effectiveness of nicotine patches (for agroup of 445 women and 307 men combined) this was only observed in the short term.
In fact, the link between the claimed “nicotine addiction gene” and the effectiveness ofnicotine patches was only statistically significant in the first week. A slightly longer effect(up to 12 weeks) was observed when another gene (called DBH) was added to the testand the researchers looked at the effect of the two genes together. The study says:“There were no associations between genotype and patch effectiveness beyond 12weeks”. Because the effect of the genetic difference was small, a further scientific paper basedon the same study looked at the differences between men and women6. In this study,nicotine patches helped women with the ‘good responder’ gene (the so-called ‘addictiongene’) to quit, even in the longer term (up to 8 years of not smoking) but they did nothelp men with this gene to quit. If this conclusion is correct, it means this single genealone does not determine a person’s response to the patches – there must be otherfactors involved, or the effect in women may have occurred by chance or for some otherunknown reason. The study says: “In women the effectiveness of nicotine patchesseems to be related to genotype” and the scientists suggest a possible explanation -that nicotine replacement therapy may be subject to different genetic influences in menand women. This theory has not been tested by further research and there might beother explanations for this finding.
The evidence that the first DNA test is useful to decide who should use bupropion(Zyban) is even weaker7. In this case a single study found no differences in men andonly small differences in women who had a different genetic makeup. This studyconcluded that womens’ response to Zyban “may be partially due” to these geneticdifferences. The authors also admit that the interactions between several differentgenes are likely to be important and it is “quite probable” that the study is not largeenough to identify these effects. A different study (not cited on the website) found thatthe DRD2 gene might be important, but only when combined with the effect of anothergene8. Again, much more research is needed to draw definite conclusions.
Scientists are now finding that most studies linking genes with diseases9 or withbehaviour (including addiction)10 are not confirmed when the studies are repeated.
Usually the gene turns out to be much less important than first thought. Predictingresponse to medicines from a genetic test may be easier than predicting disease orbehaviour, but is stil very complex11. These problems need to be resolved beforegenetic tests are sold. When there are many different possible ways that genes could work (for example: lots of different genes; lots of other factors; the possibility thatdifferent factors are important in men and women) very large and careful studies areneeded to tell which explanation is correct.
Selling a genetic test based on a single study is irresponsible because the
results of this kind of study usually turn out to be misleading when more
research is done.
1.4 There is no evidence provided for the second DNA test.
The “FAQs” section of the website says that a second DNA test will give the user’s“metabolic profile” for nicotine and therefore how much nicotine replacement to take. Itsays: "This is much better than trial and error!”. But the website gives no scientificreference for its claim and later admits: “Until we collect the figures it is very difficult toestimate the benefit in terms of quitting”. Taking this second DNA test thereforeamounts to paying to take part in a research project. The metabolism of nicotine iscomplex12 and it is unlikely that a single genetic test can replace the need for trial anderror. This is because many different biological and other factors (such as social andpsychological factors) are likely to be involved13. Scientists do not yet know if it will bepossible to predict the best dose of nicotine replacement therapy for an individual.
The NicoTest website does not say which genes wil be tested for the ‘metabolic profile’but a gene called CYP2A6 is the most likely one to be included. This gene makes anenzyme that metabolises nicotine (it breaks it down into a different chemical). There issome evidence that different variations in this gene are linked with different likelihoodsof successfully quitting smoking, but this evidence is not conclusive because there aresome errors in one of the studies that has been done3. The genetic variations inCYP2A6 that might be linked with success in quitting are also rather rare in white peoplein Britain.
Taking the second genetic test amounts to paying to take part in a research
project. No evidence yet exists that using this test to choose the dose of
nicotine replacement therapy is better than trial and error.
1.5 NicoTest’s claims about success rates are selective and misleading.
NicoTest quotes success rates for quitting smoking on its website. Some relate topeople who have the claimed ‘addiction gene’ and others to people who do not. Someare for people who have used Zyban or nicotine patches and others are for people whohave not. Some are short-term success rates (people who have abstained from smokingfor only one week) and others are long-term (people who have not smoked for 8 years).
In many cases the website makes false comparisons and fails to make clear thelimitations of the evidence.
The NicoTestTM website says: “Analysis of clinical trials eight years afterwards haveshown that up to 40 percent of smokers with the ‘addiction gene’ can successfully giveup if they use specific and appropriate nicotine replacement therapy (NRT), whilearound 20 percent who don’t have it can still succeed with relevant and appropriatetreatment. This compares to around 4 percent for those who don’t use NRT or followappropriate treatment”14. Elsewhere on the website it is claimed that if this second group of people (those without the ‘addiction gene’) try bupropion (Zyban) their success ratecan increase from 20% to “about the same as for nicotine replacement therapy” (40%)15.
These claims are seriously misleading because the clinical trial that looked at smokers“eight years afterwards” did not show success rates of up to 40% after eight years6. Itshowed success rates this high only for the shortest time period (after one week)5. TheFAQs section of the website refers to success rates after one week, but the front pageimplies this is the eight year figure. The 4% success rate quoted for people who do notuse any treatment is an eight year figure – so it is wrong to compare it with the oneweek figures for people who use treatment.
This means, firstly, that the effects of treatment (with nicotine patches or with Zyban)have been greatly exaggerated. For the study referred to on the NicoTest website,success rates after eight years were 5% (5.9% with nicotine patches and 4.3% without)and after one year were 9.4% (11.2% with patches and 7.7% without)16. Otherresearchers have also looked at success rates after one year. They have found thatsmokers who use nicotine patches to quit typically have success rates after one year ofabout 14% compared to about 8% without (combining the results of 33 studies), andthose who use bupropion (Zyban) have success rates of about 17% compared to about10% without (combining the results of 7 studies)17. Adding counselling can increasethese success rates further18. This means that treatment with nicotine patches or withZyban does help people to quit smoking, but not to the extent that is implied on theNicoTest website. The statement (made in the FAQs section of the website) that of thepeople who stop smoking at one week “about 50%” wil be smoke free after 8 years isalso incorrect – this applies to smokers who have stopped smoking for one year (notjust for one week)16. Secondly, most of these studies do not include any assessment of whether taking agenetic test will help or not – it could improve or worsen these success rates, or simplymake no difference. The conclusion that the genetic test wil help smokers to give up isbased only on the two studies cited by the company (one study for nicotine patches andone study for Zyban): this conclusion is likely to be untrue for men and may beunreliable for women based on this published evidence. The figures quoted in the FAQssection of the website are based on the effect of using nicotine patches after one weekonly: significant differences between those with the “addiction gene” and those withoutwere not seen after longer time periods except when men were excluded from the study.
The one week effect is also smaller than implied on the website (which wronglycompares a 40% success rate for those with the “addiction gene” to a 20% success ratefor those without – this figure should be 34%). These one week success rates are alsobiased because the people who agreed to take part in this part of the study were morelikely to quit than those who refused16,6. This means that both the 40% and 34% figuresmay be misleading (they are probably too high). The difference in the effectiveness of nicotine patches after only one week (betweenthose with the “addiction” gene and those without) is statistically significant, so it may bea real effect5. But it was only a short-term effect observed in a single study. For apopulation that is half men and half women, taking the genetic test does not help toimprove success rates for either nicotine patches or Zyban over the longer term.
The website gives misleading figures for success with treatment – it compares
short-term success in stopping smoking for people given treatment to long-term
success rates without treatment. This is a false comparison which exaggerates
the benefits of treatment. For the genetic test, the website uses figures from
only the first week of a single eight year study and also exaggerates the
difference between smokers with and without the claimed “addiction gene”.
There is no evidence that the gene test makes any difference to success rates in
the longer term, except perhaps in women.
1.6 Some people could be misled by this advice.
Most people who take the NicoTestTM wil not succeed in stopping smoking permanently.
But some people might be persuaded by their genetic test results that it is not worthtrying either nicotine patches or Zyban in the future because these treatments wil notwork for them. This conclusion would be wrong in many cases and might lead to somepeople continuing to smoke who might otherwise be helped to quit.
Telling people they have a “nicotine addiction gene” when the link to smoking behaviouris not statistically significant is wrong. But more research is also needed to find outwhether this kind of information real y helps people to quit. A possible danger is thatpeople could become fatalistic and believe that their addiction to tobacco is genetic andtherefore too hard to beat. One study has found that learning of a genetic predispositionto nicotine dependence may increase desire for pharmacological cessation methods(such as nicotine patches or Zyban), but may undermine the perceived importance ofwillpower in stopping smoking19. Again, more research is needed.
The effect of misleading genetic information on smokers’ future attempts to quit iscurrently unknown. However, it is possible that misleading information could be harmfulto their health by dissuading them from trying other ways to quit.
Taking the NicoTest could mislead smokers because they may be given wrong
advice about which smoking cessation method will work best for them.
Misleading information about genes and nicotine addiction could potentially
harm health if it affects future attempts to quit.
2. The scientific evidence for the test has not been independently
assessed.
The company claims that it operates “within a regulated environment”. In fact, regulationof genetic tests is extremely limited. There is no independent assessment of whether ornot the gene tested actually predisposes people to nicotine addiction, or whether thetest is useful to help choose the best treatment. The “CE mark” and laboratoryaccreditation referred to on the website involves, at most, an independent check ofwhether the company has tested the gene that it says it has and found the right geneticvariation (DNA sequence). It does not involve any independent check of whether thisgene is linked to nicotine addiction or success in using nicotine patches or Zyban. G-Nostics is a “spin out” company from Oxford University and the university and OxfordCapital Partners are its shareholders20. £4 million has been invested in the company byOxford University, Oxford Capital Partners and the management team themselves15. G-Nostics Ltd’s scientific officer is also its lead inventor and co-founder and is a co-authoron many of the relevant scientific papers. There is a clear conflict of interest between his role in promoting a commercial product and at the same time providing and interpretingthe evidence given to G-Nostics’ customers.
3. The information given to potential customers is incomplete and raises
significant ethical concerns.
3.1 Possible links with other conditions
The claimed ‘addiction gene’ (DRD2) has also been linked with alcoholism, although thislink is also far from certain and could be incorrect. Because dopamine is an importantchemical in the brain, scientists are also studying the links between DRD2 and otherdopamine receptor genes and various psychiatric il nesses and response to anti-psychotic medication. For example, one study has found an (unconfirmed) link betweenone (different) form of the DRD2 gene and schizophrenia21. This study is not yetpublished and could be wrong, but NicoTest customers should be aware that furtherresearch could reveal information that they do not want to know. This has happened inthe past with a different genetic test: a gene that had been linked to risk of heart diseasewas later found to have a link with increased risk of Alzheimer’s Disease. Some studies have linked genetic variations in the ‘nicotine metabolism gene’ CYP2A6with risk of cancer, although this link also remains uncertain.
3.2 Lack of medical involvement
Because the test is being sold over the internet (and sales in high street pharmacies areplanned), doctors will have no information about the test. They wil not be able to checkthe evidence for the test; ensure it is properly interpreted; or give further information ifnew studies change the evidence or link one of the genes to a disease. G-Nostics limitall communication to email and could not provide face-to-face information or counsellingif necessary.
GeneWatch believes that genetic tests should only be marketed through doctors andshould be regulated so that they can be properly interpreted.
3.3 Privacy and genetic discrimination
Anyone taking a genetic test should be aware that there is currently no legislation toprevent insurers or employers asking for genetic test results and using them to decidewho gets insurance or a job. There is currently a voluntary agreement between theinsurance industry and the Government not to use most genetic test results for mostpolicies, but this agreement ends in 2006. If requests are made for genetic test resultsafter this date they will be made to the person applying for the insurance policy or job(not to g-Nostics Ltd) and failure to reveal them could render an insurance policy invalid.
Conclusions
Better treatment of nicotine addiction is an area where genetic research might proveuseful in the future22. This is because a better understanding of nicotine addiction mightlead to a better understanding of what helps smokers to quit. It is also possible thatgenetic tests might help to match smoking cessation treatments to genetic make-up atsome point in the future. However, the role of genes in nicotine addiction and treatment is still poorly understood and the role of other factors (including chance) means thatscientists do not yet know if this approach to medicine wil work23. Therefore thesegenetic tests are currently only useful for research - selling them to members of thepublic is at best premature. In its rush to market NicoTestTM, g-Nostics Ltd is misleadingits potential customers.
The NicoTestTM is at best a waste of money. Men in particular have nothing to gain bybuying a test which the evidence already shows is not likely to be useful for them. TheNicoTest could also mislead smokers and potentially harm health because people arelikely to be given wrong advice about how to best increase their chance of quittingsmoking.
More information about the need for independent regulation of genetic tests is availableon GeneWatch’s website24,25. GeneWatch UK, The Mill House, Manchester Road,
Tideswell, Buxton, Derbyshire, SK17 8LN.
Tel: 01298 871898; Fax: 01298 872531;
Email: [email protected]
Website: www.genewatch.org
References
1 http://www.isis-innovation.com/about/news/gnostics.html .
2 http://www.g-nostics.com/?Dr%20Robert%20Walton .
3 Munafo MR, Clark TG, Johnstone EC, Murphy MFG, Walton R (2004) The genetic basis for smoking
behavior: A systematic review and meta-analysis. Nicotine & Tobacco Research, 6(4), 583-597.
4 Johnstone EC, Yudkin P, Griffiths SE, Fuller A, Murphy M, Walton R (2004) The dopamine D2 receptor
C32806T polymorphism (DRD2 Taq1A RFLP) exhibits no association with smoking behaviour in a healthy
UK population. Addiction Biology, 9, 221-226.
5 Johnstone EC, Yudkin, PL, Hey K, Roberts SJ, Welch SJ, Murphy MF, Griffiths SE, Walton RT (2004)
Genetic variation in dopamergenic pathways and short-term effectiveness of the nicotine patch.
Pharmacogenetics, 14(2), 83-90.
6 Yudkin P, Munafo M, Hey K, Roberts S, Welch S, Johnstone E, Murphy M, Friffiths S, Walton R (2004)
Effectiveness of nicotine patches in relation to genotype in women versus men: randomised controlled trial.
British Medical Journal, 328, 989-990.
7 Swan GE, Valdes AM, Ring HZ, Khroyan TV, Jack LM, Ton CC, Curry SJ, McAfee T(2004) Dopamine
receptor DRD2 genotype and smoking cessation outcome following treatment with bupropion SR, The
Pharmacogenomics Journal, advance online publication, 19 Oct 04; doi: 10.1038/sj.tpj.6500281 .
8 Lerman C, Shields PG, Wileyto EP, Audrain J, Hawk Jr LH, Pinto A, Kucharski S, Krishnan S, Niaura R,
Epstein LH (2003) Effects of dopamine transporter and receptor polymorphisms on smoking cessation in a
bupropion clinical trial. Health Psychology, 22(5), 541-548.
9 Editorial. In search of genetic precision. The Lancet, 361, 357.
10 Munafo MR, Clark TG, Moore LR, Payne E, Walton R, Flint J (2003) Genetic polymorphisms and
personality in healthy adults: a systematic review and meta-analysis, Molecular Psychiatry, 8(5), 471-484.
11 Tucker G (2004) Pharmacogenetics – expectations and reality, British Medical Journal, 329, 4-6.
12 Batra V, Patkar AA, Berrettini WH, Weinstein SP, Leone FT (2003) The genetic determinants of smoking.
Chest, 123(5), 1730-1739.
13 Jarvis MJ (2004) ABC of smoking cessation: why people smoke. British Medical Journal, 328, 277-279.
14 www.nicotest.com/auxiliary/nicotest.aspx .
15 www.nicotest.com/auxiliary/faq.aspx .
16 Yudkin P, Hey K, Roberts S, Welch S, Murphy M, Walton R (2003) Abstinence from smoking eight years
after participation in randomised control ed trial of nicotine patch, British Medical Journal, 327, 28-29.
17 Fagerström KO(2003) Clinical treatment of tobacco dependence: the endurance of pharmacological
efficacy, Journal of Clinical Psychiatry Monograph, 18(1), 35-40.
18 Sutherland, G (2003) Evidence for counselling effectiveness for smoking cessation, Journal of Clinical
Psychiatry Monograph, 18(1), 22-34.
19 Wright AJ, Weinman J, Marteau TM (2003) The impact of learning of a genetic predisposition to nicotine
dependence: an analogue study. Tobacco Control, 12, 227-230.
20 http://www.g-nostics.com/?Shareholders .
21 Lawford BR, Young RM, Swagell CD, Barnes M, Burton SC, Ward WK, Heslop KR, Shadforth S, van Daal
A, Morris CP (2005) The C/C genotype of the C957T polymorphism of the Dopamine D2 receptor is
associated with schizophrenia. Schizophrenia Research, 73(1), 31-37.[abstract only, available on PubMed].
22 Hal W, Madden P, Lynskey M (2002) The genetics of tobacco use: methods, findings and policy
implications, Tobacco Control, 11, 119-124.
23 Senn S (2004) Individual response to treatment: is it a valid assumption? British Medical Journal, 329, 966-
968.
24 GeneWatch UK (2004) Genetic tests and health: the case for regulation. Briefing No. 28, Sept 2004.
Available on: http://www.genewatch.org/HumanGen/Publications/Briefings.htm#Brief28 .
25 GeneWatch UK (2003) Pharmacogenetics: better, safer medicines? Briefing No. 23, July 2003. Availableon: http://www.genewatch.org/HumanGen/publications/briefings.htm#Brief23 .

Source: http://www.genewatch.org/uploads/f03c6d66a9b354535738483c1c3d49e4/Nicotest_brief_final.pdf