Saturday, 13 October 2007

Money for an old helix?

Where there's money to be made, you can bet your bottom dollar that in an unregulated market people will exploit ignorance, peddle half-truths and spin-doctor evidence in order to sell their "miracle pills" or some such.

And there's definitely gold in them thar genes. According to this article in the New Scientist (the full article is behind a paywall, sorry...) there is great concern about over-the-counter genetic tests which claim to find out whether you are at risk of killer diseases such as heart disease and cancer. Concern based on the fact that by reducing complex conditions which have several associated risk factors down to a black-and-white test (which costs hundreds of pounds), the public is being duped into a false sense of heightened expectation from medicine and science.

There are thousands of scientific studies dedicated to unearthing genetic variants which increase the risk of contracting serious illnesses (just put the words "gene variant disease" into the PubMed search engine and see for yourself). The majority of these controlled, well executed, peer-reviewed studies are of course invaluable, but it's how companies and individuals exploit the findings which is of concern.

The recent explosion in genetic data has lead to myriad gene variants - often single base differences between your gene that encodes protein X and my sequence - being linked with increased susceptibility to disease Y. Note my use of language here - we're talking about how having a particular variant of gene X makes it n% more likely that you will develop disease Y. This is not, repeat, not, the same as discovering the gene defect that causes an inherited condition. In the popular press, however, this process gets boiled down to "bad version of gene X causes heart disease/diabetes/hair loss/etc." In other words, people are often lead to believe that if there was a way of finding out which version of gene X you had, you could put yourself of high alert of developing Y, perhaps even alter your diet or take a prophylactic drug to avoid the fate seemingly hard-wired into you DNA.

This of course is the ultimate dream in "personalised medicine" - screen people for particular gene variants, tell them to eat less fat, smoke less, or even better, take a pill, and they'll avoid ill health. Not as simple as that, I'm afraid.

New tests recently available over the counter are developed based on peer-reviewed published studies which link a particular gene variant - allele - to a given disease. But let's say such a test looks for whether you have the "bad version" of ApoE4, a gene linked with overproduction of cholesterol. Everyone knows cholesterol leads to heart disease, so if I have myself tested for the bad ApoE4 variant I can prepare for the worst, right? Nope, because the jury's still out on whether this biochemical variation in cholesterol production causes an overall increase in the clinical incidence of heart disease. Think about it - just because a version of the gene X causes me to make lots of cholesterol, doesn't mean I'll necessarily develop heart disease, as there are dozens of steps in between. Mere details don't stop companies such as Genetic Health UK from selling an £825 kit that tests for "bad versions" of ApoE4. What do you do with the information once you've tested positive for the bad variant? eat more sensibly? exercise more? drink less? guess what, save yourself eight hundred quid and do those things anyway, 'cos according to most studies knowing which variant of one gene you carry isn't nearly as important as these lifestyle factors are in preventing what are essentially diseases of affluence.

Much worse still may be the cynical "double play" some companies offer, whereby they test your DNA for particular variants associated with elevated risk of disease Y, and then offer ludicrously over-priced combinations of nutrients, vitamins, minerals and the like to counter the effect of said "bad gene." According to Bijal Trivedi's account of such practices in the New Scientist (again, article behind a paywall I'm afraid...),
An investigation conducted by the US Government Accountability Office in July
suggested that the type of nutrigenetic testing offered by four companies -
Sciona, Genelex, Market America and Suracell - "misled consumers by making
predictions that are medically unproven and so ambiguous that they do not
provide meaningful information". The GAO report also criticised some companies
for selling supplements supposedly tailored to a customer's genetic needs. These
"nutraceuticals" cost anywhere from $1200 to $1800 per year, yet according to
the report they differed little from multivitamins available at the local

Cynical, exploitative, extortionate.

As a scientist, you are trained to interpret data in a particular context. In the case of genetic studies looking at which gene variants are associated with disease - so called "linkage studies" - we interpret the occurrence of variants of gene X in terms of odds ratios. This means that if you think about such studies, you're able to judge that whilst having a given genetic variant may increase your odds of developing (e.g.) lung cancer by two-fold, you can't carry on smoking 40-a-day and blame your genes, as the smoking increases your chances of lung cancer by (e.g.) ten-fold. The public at large, however, are taught by a reductionist press and clever marketing that "variant of X causes disease Y," and moreover that taking $2000-worth of pills can counter that effect.

Is this symptomatic of a black-and-white view of science and medicine, where the public expect easy answers to complex, multi-faceted conditions in the form of pills off the shelf? Maybe. One thing's for sure - until over-the-counter tests for genetic variants are subjected to regulation by an independent authority like the MHRA, people will continue to believe that a drop of blood and a simple test can reveal the deadly future awaiting them.

Truly a 21st century crystal ball.


gimpy said...

seems rather pricey. I reckon designing primers and ordering enough for a few hundred PCR reactions would cost ~£5, Taq is dirt cheap. Restriction enzymes for looking at SNPs would probably cost ~£50 for a few hundred reactions. Sequencing - well that's pretty cheap these days for short sequences ~£20 say. So that £850 test probably costs considerably less than £50 for reagents, sequencing (if they do that), post and packaging, glossy leaflet, etc. Good way to make a lot of money. Of course any post-graduate with access to a university ordering system could perform such tests with ease. Doing so would be unethical of course. But cheap.

Which makes me wonder, is there legislation preventing this? Genetic Counsellors are accredited, regulated and bound by an ethical code aren't they? They take genetic testing very seriously and take a great deal of care when explaining the significance of say a CF allele in a parent and so on. I know you argue that such tests are mainly meaningless but testing someone for CF, a BRCA allele, various Poly-Q polymorphisms, etc can be very very troubling for the individual diagnosed as they have indispuitable links with disease.

teekblog said...

I know you argue that such tests are mainly meaningless but testing someone for CF, a BRCA allele, various Poly-Q polymorphisms, etc can be very very troubling for the individual diagnosed as they have indispuitable links with disease.

i think this is the point. for single-gene disorders, where one point mutation causes an inherited defect (e.g. CF, or the retinal dystrophies i work on), screening has its place. even then, only alongside the appropriate counselling.

for diseases in which genetic polymorphisms or even mutations play a smaller part, in the context of social, dietary and other contributing factors, offering black-and-white tests (which are, as you say, a pricey rip-off) is dubious at best.

coracle said...

Interesting stuff, I've got really high hopes for pharmacogenetics so seeing tricky ideas like these being pedaled is worrying. Interestingly, Dr JB was promoting research by Nestle into metabolic typing. It seems a remarkably pointless bit of research:

NestlĂ© annouce research supporting the concept of ‘metabolic typing’

though perhaps I'm doing it a disservice?

Chemist said...

I would have thought in the UK this would be a medical device. As such the MHRA would be the first point of call

teekblog said...

coracle, pharmacogenomics does indeed have its place, or should i say probably will do. IMHO, pre-screening patients for SNPs which have a definite association with e.g. adverse reaction to certain drugs is a likely application. also there's the very sensible approach to screening at-risk individuals for e.g BRACA1/2 mutations. but screening for a polymorphism that at best has a trenuous link to a multi-factorial (genes+lifestyle+environment+blind luck) disease, followed by prescription of over-priced vitamins, is decidedly not pharmacogenomics.

will read about the nestle stuff and comment later - thanks!

teekblog said...

chemist, not sure that's the case. will dig around to make sure, but i think this sort of testing slips thru the net as of now. which is why according to new scientist new guidlines/legislation is being considered.