20091125-6

Fluorescent FT protein in the phloem of an Arabidopsis plant.
Courtesy of Laurent Corbesier and George Coupland

 

Researchers unfold a key step in the process that tells plants to flower, findings that could one day benefit agriculture.

The-Scientist.com, November 23, 2009, by Bob Grant   —  Few acts of nature seem simpler than flowers blooming on the outstretched tips of a plant’s shoots. But the induction of that seemingly simple process baffled plant biologists for almost 60 years.

In the 1930s, Cornell University plant scientist James Knott coined the term “florigen” for a mysterious signal that instructs flowers to begin growing at the tips of stems, called apical meristems.1 Researchers knew and had demonstrated that changes in day length and temperature caused plants to flower, a process essential to plant reproduction. Knott tracked the unidentified florigen traveling through the vascular system of a spinach plant, and other scientists worked out parts of the molecular pathway that allowed plants to sense environmental changes and respond by producing flowers. But the chemical identity of florigen eluded discovery. “It was a technical challenge to put that last nail in the coffin,” says Richard Amasino, a plant scientist at the University of Wisconsin in Madison.

Then in 2007, researchers at the Max Planck Institute for Plant Breeding Research in Cologne, Germany, cracked the case. Plant geneticist George Coupland and colleagues showed that florigen, at least in the flowering Arabidopsis plants they were studying, was a protein encoded by the gene FLOWERING LOCUS T (FT), which behaves like certain types of kinase inhibitors in plant cells. “People were not expecting [florigen] to be a protein or a nucleic acid,” Coupland recalls. “They were expecting it to be a hormone or small molecule,” which typically act as chemical signals in plants.

“It was very clear that FT was the signal,” says Jorge Dubcovsky, a University of California, Davis, plant geneticist who was not involved with the study. Nailing down the identity of florigen meant that science finally identified all the key molecular puzzle pieces involved in flowering-a process of great interest to agriculturalists, for whom an understanding of the molecular machinery of flowering could lead to greater control of their crops by speeding up or slowing down blossom production. “We’ve got the major mysteries solved,” says Amasino. Since Coupland’s long-awaited discovery, his lab and others have continued to answer other lingering questions in flowering pathways, while other plant biologists have corroborated his results and pinpointed the identity of florigen in other, more agriculturally relevant, plants.

False positive

As Coupland and his group sought to provide support for their hypothesis that florigen was a protein, Swedish researchers in 2005 claimed that they had discovered that florigen was mRNA transcribed from the FT gene.2 But the Swedish paper showed an extremely low level of FT mRNA in the meristem, which Coupland says that his group saw as a potential artifact of the real-time PCR method used to detect it. Other researchers shared Coupland’s misgivings. Meanwhile, Science, which published the 2005 paper, was heralding it as a “breakthrough of the year,” and plant biology textbooks began reporting that florigen was FT mRNA.

Coupland and his colleagues continued their experiments, but the 2005 Science paper did cause Coupland to be extra careful about assembling his evidence that the FT protein was florigen. “We couldn’t do exactly the same experiments” as the authors of that paper, such as using PCR to show which molecule was moving from leaves to stem tips to induce flowering, Coupland says.

The Coupland group tagged the FT protein with a fluorescent protein, and using microscopy, showed FT moving from leaves through the phloem-a central vascular tissue that transports water in plants-to the apical meristem, where it induced the growth of Arabidopsis flowers. The researchers also grafted one plant to another and showed that the FT protein was moving from the leaf of one plant to the meristem of the other-further proof that the protein was the long-distance carrier of the flowering signal. They also tracked FT mRNA in these experiments, but failed to find those molecules crossing the junction between the grafted plants.

They submitted their paper to Science; later it came to light that the FT mRNA paper contained serious flaws and was retracted from the journal. The first author on the paper was accused of fudging some crucial data. “There was some fearsome competition there,” says Dubcovsky, “and some people put a priority on speed over certainty.”

Crop confirmation

In the same issue of Science where Coupland published his results, a Japanese team identified an ortholog of the FT protein as the florigen in rice (another Hot Paper).3 William Lucas, a UC Davis plant biologist, confirmed that the FT protein was florigen in pumpkins,4 while Dubcovsky identified a florigen protein homologous to FT protein in wheat.5 The FT protein has also been shown to induce flowering in poplar trees.6

But more mysteries about flowering physiology remain. “[Identifying florigen is] a very important piece of the puzzle, but there’s still a lot to be done,” says Dubcovsky. For example, there is no direct evidence to show how the protein moves through the phloem, according to Coupland. His group is working to characterize some of the mechanistic links between the FT protein and other players in the flowering pathway, such as CONSTANS, a transcriptional regulator that triggers the transcription of FT in leaves and the bZIP transcription factor FD, which interacts with the FT protein to deliver the flower induction signal to its target in the meristem. “There may well be other things to find out in the details of this model,” he says.

 

L. Corbesier et al., “FT protein movement contributes to long-distance signaling in floral induction of Arabidopsis,” Science, 316:1030-33, 2007. (Cited in 144 papers)

 

References

1. J. Knott, “Effect of a localized photoperiod on spinach,” Proceedings of the Society for Horticultural Science, 31:152-54, 1934.

2. T. Huang et al., “The mRNA of the Arabidopsis gene FT moves from leaf to shoot apex and induces flowering,” Science, 309:1694-96, 2005.

3. S. Tamaki et al., “Hd3a protein is a mobile flowering signal in rice,” Science, 316:1033-36, 2007.

4. M. Lin et al., “FLOWERING LOCUS T protein may act as the long-distance florigenic signal in the cucurbits,” Plant Cell, 19:1488-506, 2007.

5. C. Li et al., “Wheat FT protein regulates VRN1 transcription through interactions with FDL2,” Plant Journal, 55:543-54, 2008.

6. T. Igasaki et al., “The FLOWERING LOCUS T/TERMINAL FLOWER 1 family in Lombardy poplar,” Plant Cell Physiology, 49:291-300, 2008.

Editor’s Note:

Developments in pharmacogenomics hold promise for improvements in drug safety and efficacy. David Danar, MD, Scientific Director for MedscapeCME, interviewed pharmacogenomics expert Kathleen M. Giacomini, PhD, Professor and Co-Chair, Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, on breakthroughs in pharmacogenomics and future directions.

MedscapeCME: Dr. Giacomini, how would you define pharmacogenomics?

Dr. Giacomini: There are many definitions of pharmacogenomics, but I would like to stick to the one in which we define pharmacogenomics as the genetic basis for variation in drug response. There we’re talking about how genetic variance in people’s DNA or germline DNA may influence their response to drugs. Some people consider pharmacogenomics in a broader context, and they include tumor expression level differences, etc, that go beyond germline DNA. For this discussion, I’ll focus on germline DNA or the DNA that people inherit from their parents.

MedscapeCME: What are the biggest breakthroughs in pharmacogenomics?

Dr. Giacomini: I would like to talk specifically about the breakthroughs that have to do with important clinical implications for drug therapy. I believe that number one on that list might be abacavir. There is a genetic variant in an HLA locus, and that genetic variant can help guide therapy for abacavir. Abacavir is used in the treatment of HIV, and about 8% of people, depending on their race and ethnicity, who receive abacavir have a hypersensitivity reaction.[1] Hypersensitivity reactions may include skin rash, fevers, gastrointestinal symptoms, and lethargy, and they can be life-threatening. At 8%, it’s pretty high. That alone could cause physicians to hesitate to prescribe the drug. They’ve found an HLA locus where certain people carry HLA-B*5701.[1] If you carry that HLA-B*5701, you are susceptible to a hypersensitivity reaction, whereas if you don’t, you’re almost not susceptible.

It’s a very definitive genotype, and you can now genotype for it; the US Food and Drug Administration (FDA) has put some warnings on the official label of abacavir and recommended genotyping. From my understanding of the physician community now, the prescribers — perhaps mostly physicians who are treating HIV or viral infections — are using this genetic test, and if people have the HLA-B*5701 genotype, then they’re prescribing something else. Effectively, that has caused great improvement in the safety of this drug. I would consider that one of the most important breakthroughs in pharmacogenomics. It’s a discovery that is being used to improve the safety of a drug that previously was very risky. You couldn’t predict who was going to have this reaction.

MedscapeCME: Are there any other breakthroughs that come to mind?

Dr. Giacomini: Another big one that is increasingly being used, and certainly being discussed — it’s a little slow on the uptake in terms of being used — is warfarin. Warfarin is an anticoagulant. There’s a wide variation in the dose of warfarin that people receive to elicit a certain level of anticoagulant effect.

[In regard to] warfarin, genetic variants in CYP2C9 and VKORC1 were discovered.[2] Genetic variants in those 2 genes can account for some of the variation in warfarin dosing.

I would say that that’s another breakthrough, and I think that that will be increasingly used in warfarin dosing in anticoagulant clinics and by physicians who are treating patients with warfarin. Those genetic tests can help get the dosing right. That’s the second breakthrough.

I don’t know that I’d call it a breakthrough, but I think there have been significant advances in what you would call the relationship between genetic variants in drug-metabolizing enzymes and response to a number of different drugs. I’ll address CYP2D6: We’ve known about genetic variants in CYP2D6 for a long time. That’s an enzyme that metabolizes a quarter or a bit less than a quarter of prescription drugs. It’s important, and a number of people carry reduced-function CYP2D6 genetic variants and reduced-to-nonfunctional CYP2D6 genetic variants. A number of people also carry multiple copies of CYP2D6 in their genomes; therefore, the ones who carry reduced-function variants cannot metabolize very well, whereas the ones who carry the multiple copies overmetabolize or metabolize at a very fast rate.

This has led to a number of different problems, for example, tamoxifen, which is used to treat breast cancer. CYP2D6 activates tamoxifen to the active compound in tamoxifen, endoxifen and if you have a low-activity CYP2D6 genetic variant or allele, you will not convert the tamoxifen or will convert tamoxifen to endoxifen at very slow rates.[3,4] When patients with low-activity CYP2D6 genetic variants are treated with tamoxifen, they’re not getting the proper dose of endoxifen, the active ingredient. These women may be on tamoxifen for many years but not responding to the drug, unknown to them, so it’s not really working. Some studies have shown that not only does it not work, but the individuals may have no side effects because it’s not working, so they tolerate it very well and stick with it.[5,6]

However, in the individuals in whom it’s working but who have the side effects, the hot flashes, etc, they may not take the drug. [In cases when the drug is] working, patients may not take the drug, whereas when it’s not working, they may continue to take it. This is one example of which there are more clinical studies; I would not say that physicians genotype for CYP2D6 before they prescribe tamoxifen, but I think that is coming because there are a lot of fairly convincing studies.

Similarly, in that category other drugs are subject to the CYP2D6 ultrarapid metabolism. Codeine is a drug that is activated by CYP2D6 to morphine. When you take codeine, you actually get some morphine from conversion by CYP2D6. If you’re an ultrarapid metabolizer, for example, you have multiple copies of CYP2D6, you will have more morphine.[7,8] In a case report of a woman who was an ultrarapid metabolizer and had multiple copies of CYP2D6 combined with the UGT2B7*2/*2 genotype, her breast-fed baby had a clinical picture consistent with opioid toxicity that led to neonatal death because it was being breast-fed and ingesting higher levels of morphine and possibly its metabolite — because the mother had been taking codeine postpartum.[9,10]

CYP2D6 is one example of a drug-metabolizing enzyme. There are also thiopurine methyltransferase and 6-mercaptopurine as well as irinotecan and UGT1A1.[11] A number of examples are coming down the pike, which will be increasingly translated to patient care and enhancing drug therapy.

MedscapeCME: What have you been working on?

Dr. Giacomini: Our area is a little different. We’re more in the earlier-discovery area. We have been looking at genetic variation in membrane transporters and how membrane transporter genetic variants might relate to clinical drug response. We’ve learned a lot about genetic variation in membrane transporters. I believe that in the area of pharmacogenomics, what I’ve given you as the most exciting breakthroughs are the things that are closest to patient care. An emerging number of breakthroughs will be seen in the coming years, and I consider our laboratory as contributing part of this new wave of new discoveries, which is occurring right now.

We’re working on genetic variants in membrane transporters. The one that has received a lot of attention are genetic variants in OATP1B1 and statin-induced myopathy.[12] In regard to the rhabdomyolysis that people may get with statins, some studies have shown that genetic variants in a transporter lead to increased risk for this devastating adverse effect, and that’s something that we’re very interested in as well.

MedscapeCME: What are the techniques that were used to make these discoveries?

Dr. Giacomini: There is a new wave of technologies. When I look at the discoveries that I have cited as being the most important breakthroughs because they’re closest to clinic, those discoveries are probably using more traditional techniques — traditional genotyping methodologies. They were not necessarily using large genome-wide association studies, although some genome-wide association studies were performed on, for example, warfarin, later. The first discoveries were just candidate genes and traditional genotyping methods. Emerging discoveries and emerging studies are using genome-wide technologies, in which there are large single nucleotide polymorphism (SNP) platforms where you can determine the genotype of, even now, 1 million SNPs in patients, for example, who are on a drug and patients who are on the drug and don’t respond. You can start to discover what’s different about the people who aren’t responding compared with the people who are responding, or the people who are experiencing an adverse drug reaction and those who aren’t. You can do that with these genome-wide platforms. I would say that that technology has been very important and is increasingly important. We’ll see a number of these genome-wide association studies in pharmacogenomics over the next 5 years. Those are just coming.

There have been a lot of genome-wide association studies for disease risk. That is, you take a bunch of people who have diabetes and people who don’t, and you say, “What’s the difference in their genotypes?” You can look at a million different SNPs and say, “Well, what do the patients with diabetes have that patients without diabetes don’t?” Genome-wide association technologies have been applied, and the statistical methodologies that are needed to analyze those data have been applied in a number of studies on disease risk. In pharmacogenomics, you’re looking at a group of people who experience an adverse drug response vs those who don’t, or a group of people who respond to the drug and a group who don’t, and genome-wide methods are just beginning to be applied. That’s why I’m saying that there is a whole new wave of discovery that will appear in the next 5 years.

I believe that genome-wide platforms are number one among methodologies that are going to advance the field. The second will be next-generation sequencing, in which you’ll have the person’s entire genome, or at least the genes in the genome — the sequences. You can look for rare variants. The genome-wide methods look at a million more common SNPs, but when you start to sequence, rare SNPs can be found. I think that that’s going to be another very important advance in pharmacogenomics, but that’s coming down the pike; it’s not yet ready. We’re not seeing studies yet on a lot of next-generation sequencing methods, but we will.

MedscapeCME: Can you outline some of the basic differences between these techniques?

Dr. Giacomini: Classic genotyping involves picking your candidate genes and knowing the SNPs that you want to assay. You run a simple genotyping assay. It can be a polymerase chain reaction. You just sequence that fragment; you pick a number of candidate genes; and you know what you’re looking for.

Genome-wide association methods usually require you to have a more agnostic view. You say, “Well, I’m not sure. It could be anything in the genome. I don’t know why these people are experiencing an adverse drug reaction.” Then you can imagine it as a SNP chip. Millions of SNPs have been discovered over many years. One person has a C and one person has a T, so you put 1 million SNPs on a platform and genotype those 1 million SNPs, and they’re throughout your genome. On every chromosome you’ve got markers on that platform, and you’re measuring. You’re determining what particular polymorphism an individual carries at those 1 million sites. You’re asking whether any of those polymorphisms associate with the adverse effect or the nonresponse when you talk about pharmacogenomics.

With next-generation sequencing, you’re saying, “Well, I’m not even sure whether it’s one of those polymorphisms that we know about. It may be something the person has, but is not very common. Maybe only their family has it or not too many people have it.” Then you just resequence. You sequence their whole genome, and then you’re not looking at the 1 million polymorphisms that you know about, but you’re actually sequencing their whole genome or the genes in their genome, and you’re looking for anything. There’s increasing explorative power as you go from a candidate gene, in which you think you know something and you’re testing that hypothesis, to an agnostic approach in which you think it’s some polymorphism that you know about, or it’s some variant that might be linked to that polymorphism. The final step is when you say, “Let’s just sequence it all and figure out what they’ve got.” It’s an advancing technology, but along with it comes more complicated statistical methodologies.

I’d like to add something that is clearly on the horizon — 2 things that I didn’t discuss. First, there are a number of so-called biobanks. They’re called biobanks because in some way they’re collecting DNA on all patients, for example, that walk in the door at Vanderbilt University Clinic [Nashville, Tennessee], and their DNA are collected. Vanderbilt has a biobank. They also have electronic medical records that go along with the DNA. They’re doing similar studies at Kaiser, etc. In Japan, RIKEN has a huge biobank where they have 300,000 DNA samples.

You can imagine the power of those biobanks. You now have DNA on a lot of patients tied to their electronic medical records. If you get appropriate consent from the patients and you appropriately conduct the studies, you could learn a lot about variation in drug response and the genetic determinants of that. That’s one big emerging [practice]. We’re going to see biobanks and studies of drug response from them, and then we’ll have a larger sample size.

I also wanted to talk about consortia, an international consortium vs a regional consortium or US consortium, but consortia of investigators who are getting together and working on pharmacogenomics problems. The advantage is pooling large numbers of samples so that you can have a highly powerful study to look for genetic determinants of variation in drug response, and you can span multiple ethnic groups. We have one called the Global Alliance in Pharmacogenomics Japan (GAP-J or PGRN-Riken), and that’s a consortium between the Center for Genomic Medicine in RIKEN at Yokohama, Japan, and the Pharmacogenetics Research Network (PGRN) funded by the National Institutes of Health (NIH). That consortium’s gotten together, and there are 14 genome-wide association studies that are ongoing. That’s why I say that there’s a lot of new stuff coming down the pike.

I believe that those international consortia are going to be leveraged to learn a lot about genetic variation and drug response, and then the biobanks, such as at Vanderbilt and Kaiser, one in Northern California, etc. Your audience should know that those are coming because I think they’re going to be exciting.

MedscapeCME: What sort of timeline do you envision?

Dr. Giacomini: Well, I would say that the international consortia are increasing. We have 14 studies, of which they’ve already run the genome-wide association for about 6 studies. They’ve already run the SNP chips. They’re in the data-analysis phase on 6 of those studies in the GAPJ consortia. We’ll be seeing results from that. I would say that they’ll start to be published within the next couple of years. Analysis takes some time. In genome-wide association studies, you have to replicate. If you don’t replicate, it could just be chance that there was a genetic difference between your cases and your controls. It might not be truly the causative alleles that you’ve identified, but just some chance alleles, so you have to replicate your studies. I’d say that in the next couple of years we’ll see a number of genome-wide association studies and then 3 years after that; over the next 5 years we’ll see many, many genome-wide association studies on variation in drug response. We’re all looking forward to those results.

This activity is supported by an independent educational grant from Navigenics.

20091125-3

Rom Houben was involved in a road accident in 1983 

GoogleNews.com, BRUSSELS, November 24, 2009, by Philippe Siuberski  –  The story of a Belgian patient wrongly diagnosed as comatose for 23 years revives the debate on care for those considered in a vegetative state, with the astonishing case far from unique according to a recent study.

Rom Houben, the victim of a road accident in 1983, was believed to be in a vegetative state but was in fact paralyzed and unable to communicate.

20091125-4

Houben could hear what was being said around him, but was unable to respond

 

Houben’s true condition was discovered three years ago when new tests at the University of Liege, led by Professor Steven Laureys, found that his brain was still functioning.

The 46-year-old Houben, whose case has made global headlines in recent days, had been conscious for years but unable to communicate or even make known to his caretakers and family that he was conscious.

The state, which has various levels, is known as “locked-in syndrome” and a recent study carried out in Belgium discovered that doctors get their coma diagnosis wrong “in numerous cases.”

The research, by Laureys and others, found that in too many cases poor coma diagnoses were given — more than 40 percent in certain categories of sufferers.

Former engineering student and martial arts enthusiast Houben told the German weekly Der Spiegel that he had meditated to pass the long years trapped in his own body.

Using a specially-adapted computer to type messages, Houben has been able to describe the ordeal he endured.

“I would scream, but no sound would come out,” he said, “I will never forget the day they finally discovered what was wrong — it was my second birth.”

He could hear what was being said around him throughout, but was unable to respond.

“I became the witness to my own suffering as doctors and nurses tried to speak to me and eventually gave up,” he said.

The notion of the “minimally conscious state,” different from coma, was not known to medicine until 2002, according to the Liege University researchers.

The study revealed that of 44 patients diagnosed by normal methods as being in a vegetative state, 18 were in some way conscious and four of them eventually emerged from their ‘coma’.

Correct diagnosis, often achieved through monitoring brain movement, would find patients responding to pain or speech. “You can’t talk about ‘vegetables’ — they understand,” said neuropsychologist Audrey Vanhaudenhuyse, part of the Liege research team.

The creation of new tests benefited Houben in 2006 and in general is helping to minimize diagnostic errors.

However, the relevant monitoring techniques such as magnetic resonance imaging (MRI) remain expensive and are not available in all hospitals.

Patients from Belgium and elsewhere in Europe have for years been treated at Liege University, which has state-of-the-art equipment.

“The aim is to have an overall view of the functioning of the brain, to determine which areas have been preserved and to decide whether the patient has a good chance of recuperation or not,” Vanhaudenhuyse explained.

“We measure the degree of auditive perception, making them listen to a neutral sound, like a ‘beep’ then saying their name. If the brain reacts differently then there is a level of consciousness there,” she added.

The Liege study also concludes that, despite medical progress, misdiagnosis has not significantly diminished in recent years, due to the lack of common testing procedures.

“Every patient should be tested at least 10 times before they are categorically defined as ‘vegetative’,” according to Laureys, so as to avoid any more cases like Houben’s.

20091125-5

Belgian doctor Steven Laureys discovered in 2006 that Rom Houben’s brain was still functioning

20091125-2

Credit: Technology Review

  

A startup aims to calculate the value in the onslaught of genetic tests. 

 

MIT Technology Review, November 23, 2009, by Emily Singer  —  Per Lofberg wants to bring genomic medicine to the masses by overcoming one of the field’s biggest barriers–getting insurers and other payers to cover the growing numbers of genetic tests reaching the market. To achieve that, he founded Generation Health, a health benefit management company that aims to sift through the data on these tests, which range from those that predict an individual’s risk of heart disease or cancer to those that determine how well a patient metabolizes a certain drug. Lofberg’s goal is to find the ones that provide the greatest medical utility and economic value.

Earlier this month, the startup, based in Upper Saddle River, NJ, announced a partnership with CVS Caremark, which manages prescription benefits for about 50 million people. Generation Health will analyze 17 drugs that have accompanying diagnostic tests indicating how well the drug would work in an individual, including those for cancer, heart disease, and HIV, and determine which tests CVS should offer to its customers next year. “Now there is the opportunity to bring genetics to all the people CVS Caremark serves, and that is significant,” says Raju Kucherlapati, a geneticist at Harvard Medical School in Boston.

A number of genetic tests can influence treatment decisions. For instance, a test might suggest which drug or how much of a drug a patient should take. But only a handful of such tests are commonly used. That’s in large part thanks to economics. In an informal poll at a conference on personalized medicine last week at Harvard Medical School, attendees identified “lack of reimbursement” as the major barrier preventing the adoption of personalized medicine. “From the payers perspective, there is overall skepticism of the clinical efficacy and cost-effectiveness,” said Jerel Davis, a consultant at McKinsey and Company, speaking at the conference. From the providers perspective, not only will they get no reimbursement for the tests, they might even lose income, because the tests might indicate that some procedures should be avoided, Davis said.

In some cases, tests can reduce costs by reducing prescriptions or procedures that are unlikely to help an individual patient. But they can also increase costs.Sir Michael Rawlins, chairman of the National Institute for Health and Clinical Excellence in the U.K., said at the conference that it’s cheaper to give all patients undergoing a specific surgery the blood thinner heparin than it is to do genetic testing to determine who is most at risk of blood clotting. On the other hand, he said, the breast-cancer drug herceptin, which is most effective in patients with a high concentration of a protein called human epidermal growth factor receptor 2, is only cost-effective if physicians can identify the small percentage of patients most likely to benefit from it. This type of testing is now routinely done in breast cancer.

The data needed to decide whether a particular test falls into the latter category is complex, and sometimes controversial. For example, the U.S. Food and Drug Administration changed the label of the blood thinner warfarin in 2007 to note that two specific genetic variations affect a patient’s sensitivity to the drug. However, there has been huge disagreement among physicians, insurers, and others over whether genetic testing improves outcomes and is more cost-effective than traditional methods of monitoring warfarin response.

Enter Generation Health. “This is a rapidly growing field that is largely disorganized from a payer perspective,” says Lofberg. He compares his company’s approach to that of a pharmacy benefit manager, which largely took over the prescription business 15 years ago, negotiating cheaper drug prices and developing tools to reduce medication errors and inappropriate prescriptions from physicians. The company will use data from research studies and insurance claims to create models for how to put the tests into use, and then offer those services to payers, such as CVS Caremark, insurers, and employers.

Taken together, CVS Caremark and the two other behemoths in the pharmacy benefits field, Express Scripts and Medco Health Solutions, cover about three-quarters of the nation’s prescription market. They face such tight competition that they are continually looking for new services, such as access to genetic testing. Medco has already waded into this territory, funding studies on genetic testing for warfarin and the breast-cancer drug tamoxifen–some women have a genetic variation that makes them metabolize the drug into its active form less effectively. Medco has already instituted a software program for some of its members that automatically highlights these tests to pharmacists when a patient puts in a prescription for warfarin or tamoxifen.

Generation Health is building a preferred-provider network of labs that perform the different tests. It’s also developing patient and physician educational materials. “Today, if you google a genetic test, there is no simple way for a physician or patient to figure out how to use it,” says Lofberg.

One test Generation Health will explore–and one that is likely to be of great interest–is for the anti-clotting drug Plavix, often prescribed to people given heart stents. To work, the drug must be metabolized into its active form. A study published earlier this year in the New England Journal of Medicine found that about 30 percent of Caucasians have a poorly functioning variant of the gene for a drug-metabolizing enzyme and thus are more likely to suffer heart attacks and die after surgery. “It’s a common drug with a big effect in a significant part of the population, meaning it can have a large clinical effect,” says Marc Sabatine, a physician at Brigham and Women’s Hospital who ran the study.

20091125-1

The-Scientist.com, November 23, 2009, by Victoria Stern

The paper:

K. Lage et al., “A human phenome-interactome network of protein complexes implicated in genetic disorders,” Nat Biotech, 25: 309-316, 2007. (Cited in 86 papers)

The finding:

Søren Brunak at the Technical University of Denmark and Kasper Lage, now at the Broad Institute in Boston, developed a computational method to predict which proteins most likely cause a particular disease. By correlating known disease-causing proteins with specific disease phenotypes, the team generated a model that predicts whether similar proteins might also be behind the same diseases. The researchers found new genes linked to ailments such as inflammatory bowel disease and ovarian cancer.

The advantage:

Instead of randomly sequencing or manually searching for genes, this computational method is a more efficient way to look for individual disease-linked genes that were previously unnoticed or might have otherwise been missed, Brunak says.

The relevance:

This study was one of the first to use computational methods to predict a gene-phenotype relationship. “This paper is very interesting and inspiring. It is one of several quantitative ways to infer gene-disease association,” says Guanghui Hu, a computational biologist at GlaxoSmithKline in King of Prussia, Pa, who was not involved in the research. In 2008, Brunak’s group used the same method to identify new genes likely associated with Parkinson’s disease, cardiomyopathies, and muscular dystrophy syndromes (Proc Natl Acad Sci, 105:20870-75, 2008).

The limitations:

The method will likely miss key associations, and return some false positives, Hu notes. “[The study’s] real impact remains to be seen.”

Diseases for which the computational method has helped uncover new genes

Amyotrophic lateral sclerosis (ALS), inflammatory bowel disease, epithelial ovarian cancer, retinitis pigmentosa, Alzheimer disease, type 2 diabetes, coronary heart disease

News Author: Lisa Nainggolan
CME Author: Charles P. Vega, MD 

 

Medscape.com, November 24, 2009 – A new study has shown that walking speed over 6 m in older people is predictive of cardiovascular mortality, with those in the slowest tertile three times more likely to suffer CV death over five years than those who walked faster. Dr Julien Dumurgier (INSERM, Paris, France) and colleagues say this kind of walking test could be part of a general clinical assessment of those aged over 65; they report their findings online November 10, 2009 in BMJ.

“We found that old persons who walk slowly have an increased risk of death, in particular cardiovascular death; it’s an easy message,” second author, epidemiologist Dr Alexis Elbaz (INSERM), told Medscape. “This shows us the very important role of trying to maintain good fitness in older persons,” he added.

Geriatricians Drs Rowan H Harwood (Queen’s Medical Center, Nottingham, UK) and Simon P Conroy (Leicester Royal Infirmary, Leicester, UK) are the authors of an editorial accompanying the study. Harwood told Medscape that the study was “technically well done,” if not new information.

Nevertheless, he says, what the French group has done, “nicely, is that they show a strong relationship” between slow walking speed and cardiovascular death. “People have looked at vascular events before and they have looked at vascular mortality, but they haven’t put it in the context of all the other sorts of mortality, and they haven’t pulled mortality apart in the way that this group did.”

No Association Between Walking Speed and Cancer Mortality

In their linked prospective cohort study, Dumurgier and colleagues recruited 3208 men and women living in the community in Dijon between 1999 and 2001 aged 65 or older who were participating in the Three-City study. They were followed for an average of 5.1 years.

The main outcome measures were mortality overall and according to the main cause of death, by tertiles of baseline walking speed, adjusted for several potential confounders.

Elbaz explained that walking speed was measured by asking participants to walk at their usual speed and then asking them to walk, over 6 m down a corridor, at their maximum pace without running. Although chronometers were used in this study, this measure could also be simply performed in a doctor’s office using a watch or timer, to obtain walking speed in meters per second, he noted.

During follow-up, 209 participants died (99 from cancer, 59 from cardiovascular disease, 51 from other causes); those in the lowest third of baseline walking speed had a 44% increased risk of death (hazard ratio 1.44), compared with the upper tertiles.

Analyses for specific causes of death showed that those with a low walking speed had about a threefold increased risk of cardiovascular death (HR 2.92) compared with participants who walked faster. There was no association between walking speed and cancer mortality (HR 1.03), however.

Walking Speed: An Objective Measure of Physical Fitness

Elbaz said that assessment of walking speed is simple and can be performed easily in a routine clinical setting, “in fact, some geriatricians already do this kind of thing, following the measure over time, seeing if it remains stable, etc,” he noted.

However, he cautioned that walking speed should not be used in isolation to identify people at high risk of cardiovascular death but rather “in the context of a global assessment.” And he noted that the participants studied by his group were community-dwelling, well-functioning older people in fairly good health who were able to come by themselves to the study center. Assessment of older, frailer individuals “is more complicated,” he admitted.

Harwood said: “I would like to caution against being too simplistic. One of the things this study is telling us is how fit someone is, which is a reflection of how much exercise they do, and we know exercise is good for us.” And he says that in an observational study, “you can never completely adjust for all confounders, so residual confounding is always a problem.”

But he agrees that measuring walking speed in this way is, at least, “an objective measure of physical performance, and it’s far more accurate than asking a person how much regular exercise they do. It’s a bit like the difference between asking a person how much they eat and weighing them.”

References

  1. Dumurgier J, Elbaz A, Ducimetiere P, et al. Slow walking speed and cardiovascular death in well functioning older adults: prospective cohort study. BMJ 2009; DOI: 10.1136/bmj.b4460. Available at: http://www.bmj.com.
  2. Harwood RH and Conroy SP. Slow walking speed in elderly people. BMJ 2009; DOI:10.1136/bmj.b4236. Available at: http://www.bmj.com.