20091106-4

Body-wide survey also finds unique mix of bacteria for each person
 

GoogleNews.com, USNews.com, November 5, 2009, by Peter West — A new “atlas” of bacterial life on and within the human body has uncovered the fact that your palms, feet and forearms are a veritable United Nations of germs.

The scientists, using sophisticated gene-sequencing technologies, pinpointed specific bacterial types and where they like to thrive on and within the body. They found a wide variability of bacteria, depending on the spot on the body. They also found that bacterial colonies differed person to person, with each individual carrying his or her own “personalized” assortment of microorganisms.

“We’ve always known that there are microbes on us and in us,” said study author Noah Fierer, an assistant professor of microbial ecology at the University of Colorado at Boulder. “But we weren’t always able to isolate them and differentiate them from each other. With these new techniques, we can.”

“Each of us is really an archipelago of distinct habitats, at least as far as bacteria are concerned,” added senior study author Robert Knight, an assistant professor of chemistry and biochemistry and computer science at the University of Colorado. “It’s truly amazing how different the sites within the same body are, and how different the corresponding sites on different people are.”

The team’s work is being published Nov. 5 in the online version of Science. The study builds on earlier research, including a 2008 study that found that women had a greater diversity of bacteria on their palms than men.

The mapping project focused on seven to nine men and women, who were examined four times each over a three-month period. Researchers swabbed 27 different sites, searching for bacteria in virtually every nook and cranny, from hair to ear wax, mouth to nostril and trunk and legs.

So, where are the most popular places for bacteria to hang out? They seem to like the gut (no surprise there), forearms, palms, index fingers, the backs of the knees and soles of the feet, according to the study. At least that is where some of the most diverse and thriving colonies take root.

Some bacterial communities appeared to strongly prefer one body spot over another. For example, when the scientists transplanted microbial groupings from the forearm to the forehead, the germs failed to thrive. But doing the reverse — transplanting communities from forehead to forearm — didn’t seem to impede bacterial growth.

Researchers focused on bacterial diversity rather than bacterial quantity. There are thousands of bacterial types, but certain ones seemed prominent in the study, said Fierer. Four bacterial groups stood out — Actinobacteria, Firmicutes, Proteobacteria and Bacteroidetes — and they were relatively stable over time. Individual bacterial types included the Streptococcus and Staphylococcus bacteria.

In some ways, the mapping efforts spurs more questions than it answers. Researchers are not sure why the forearms and other body parts attract so much microbial diversity. Also, it is not clear whether the different strains compete with each other or live in a kind of bacterial harmony.

Researchers also want to know whether people are born with these bacterial colonies or develop them after birth, through external contact with their mothers and during the delivery process. Even more fascinating is why each person has his or her own blend of bacteria — much like a fingerprint. Perhaps diet, climate, locale or a combination of variables are responsible for each person’s unique bacterial make-up, said Fierer.

Most important, researchers want to know how the colonies impact health and whether they can be harnessed in some way to treat disease. In the future, for example, microbial mapping could lead to routine screening of patients as a way to record a baseline of their normal microbial communities. Any later deviation could be a sign of disease.

“Our ultimate goal is to devise strategies for personalized medicine, based not on the human genome, where we are all 99.9 percent identical, but on the human microbiome, where we are 80 to 90 percent different from one another,” added Knight.

20091106-5

20091106-3

Colorado.edu, November 5, 2009  —  A University of Colorado at Boulder team has developed the first atlas of bacterial diversity across the human body, charting wide variations in microbe populations that live in different regions of the body and which aid us in physiological functions that contribute to our health.

The study showed humans carry “personalized” communities of bacteria around that vary widely from our foreheads and feet to our noses and navels, said CU-Boulder’s Rob Knight, senior author on the paper published in the Nov. 6 issue of Science Express. The researchers found unexpectedly wide variability in bacterial communities from person to person in the study, which included nine healthy volunteers and which targeted 27 specific sites on the body.

“This is the most complete view we have yet of the microbial side of ourselves, one that our group and others will be adding to over the coming years,” said Knight an assistant professor in CU-Boulder’s chemistry and biochemistry department. “The goal is to find out what is normal for a healthy person, which will provide a baseline for further studies to look at people with diseased states. One of the biggest surprises was how much variation there was from person to person in a healthy group of subjects.”

Co-authors on the Science Express study, the online version of Science magazine, included CU-Boulder’s Elizabeth Costello, Christian Lauber, Micah Hamady and Noah Fierer, as well as Jeffrey Gordon from the Washington University School of Medicine in St. Louis.

There are an estimated 100 trillion microbes residing on and within each human being that are thought to collectively endow us with the essential traits we rely on for a variety of functions, including the proper development of our immune systems, efficient digestion of key foods and resistance to invasion by lurking microbial pathogens.

The CU-Boulder team looked high and low, analyzing microbial communities in places such as hair on the head, ear canals, nostrils, mouth, lower intestine, and 18 different skin sites ranging from foreheads and armpits, forearms, palms, index fingers, navels, the back of the knees and the soles of the feet. The team used the latest generation of massively parallel DNA sequencers and new computational tools developed at CU-Boulder.

The study subjects were sampled four times each over a three-month period, typically after showering an hour or two earlier. Microbial DNA was then isolated directly from swabs used for sampling each body site, eliminating the standard culturing step. Specific bacterial RNA genes present in the DNA were then amplified using a technique known as PCR and the genes were then sequenced with high-capacity DNA sequencers, said Knight.

The specific bacterial RNA genes amplified from each sample, which were obtained from each body site of each individual, were “tagged” during the PCR step with a sample-specific DNA barcode developed by Knight’s group. This allowed the team to pool hundreds of samples together prior to a single sequencing “run,” reducing the cost and increasing the speed of the work.

Specific skin sites, as well as hair, nostril and ear canal sites, had the highest levels of variability within individuals over time and were roughly on a par with the human lower intestine, according to the study. The highest diversity skin sites were the forearms, palm, index finger, back of the knee and sole of the foot. The armpits and soles of the feet showed some similarities, perhaps because they are dark and moist environments, said Fierer.

The mouth cavity showed the least variation in diversity both within individuals and between people, according to the study. The team also found the skin “head group” — which included forehead, external nose, external ear and hair — was dominated by one type of bacteria, while sites on the trunk and legs were dominated by a different group.

“We have an immense number of questions to answer,” said Fierer, an assistant professor in CU-Boulder’s ecology and evolutionary biology department who was a co-author on the study. “Why do healthy people have such different microbial communities? Do we each have distinct microbial signatures at birth, or do they evolve as we age? And how much do they matter? We just don’t know yet.”

Costello, the first author on the paper who recently accepted a postdoctoral position at Stanford University, likened the analysis of human bacterial communities to charting the growth of newborns. “Just as babies are tracked for weight and height as they grow to see where they fall in relation to normal ranges, we’d like to be able to find out if there are normal ranges of microbial communities for humans that could be tracked over time.”

In an intriguing microbial community “transplant” experiment, the team disinfected the forearms and foreheads of some test subjects, then “inoculated” both sites with bacterial communities harvested from the tongue. The tongue bacteria persisted longer on the forearms than foreheads, suggesting some bacterial populations more strongly prefer sebaceous, or oily sites.

“As some others have speculated, it may be that drier areas of the skin like forearms make generally more hospitable landing pads for bacteria,” Costello said. The team did not find any significant difference in how easily a person’s forehead or forearm could be colonized by his or her own “transplanted” microbes as opposed to those of other people.

“These patterns suggest that the search for microbial factors associated with disease, although difficult to ascertain due to the high intrinsic levels of variability among healthy individuals, may be achieved using broad profiling techniques such as those employed here,” the authors wrote in Science Express.

Previous microbial studies of healthy individuals have generally focused on individual body habitats including the lower intestine, skin and mouth. The new study builds on a 2008 CU study on hand bacteria indicating that while more than 4,200 species of bacteria resided on 102 human hands, only about five species were shared by all 51 participants. The 2008 study also showed women had a greater diversity of bacteria on their palms than men.

Knight, also a faculty member in CU-Boulder’s computer science department and who is a member of the university’s Colorado Initiative in Molecular Biotechnology, said understanding the variation in human microbial communities holds promise for future clinical research.

“If we can better understand this variation, we may be able to begin searching for genetic biomarkers for disease,” he said. The CU-Boulder researchers said it might someday be possible to identify sites on the human body that would be amenable to microbial community transplants with either natural or engineered microbial systems that would be beneficial to the health of the host.

The research was funded by the Howard Hughes Medical Institute, the National Institutes of Health, the Bill and Melinda Gates Foundation and the Crohn’s and Colitis Foundation of America.

Contact

Rob Knight, 303-492-1984
Rob.Knight@colorado.edu
Noah Fierer, 303-492-5615
Noah.Fierer@colorado.edu
Elizabeth Costello
costelle@colorado.edu
Jim Scott, 303-4923114 

20091106-2

Arul Chinnaiyan, M.D., Ph.D., from the University of Michigan Comprehensive Cancer Center. Credit: University of Michigan Health System

Researchers at the University of Michigan Comprehensive Cancer Center have discovered what leads to two genes fusing together, a phenomenon that has been shown to cause prostate cancer to develop.

GoogleNews.com, Physorg.com, November 5, 2009  —  The study found that pieces of chromosome relocate near each other after exposure to the hormone androgen. This sets the scene for the gene fusion to occur. The finding is reported online Oct. 29 in Science Express.

“This work shows the origin of how the gene fusion is actually created and perhaps the origin of prostate cancer itself. This is a triggering event for the genesis of prostate cancer,” says study author Arul Chinnaiyan, M.D., Ph.D., director of the Michigan Center for Translational Pathology and S.P. Hicks Professor of Pathology at the U-M Medical School. Chinnaiyan is also a Howard Hughes Medical Institute investigator.

Chinnaiyan and his team identified in 2005 a prostate-specific gene called TMPRSS2 that fuses with the gene ERG, which is known to play a role in prostate cancer. Their earlier research has shown that this gene fusion acts as an “on switch” to trigger prostate cancer. In the current study, the researchers focused on what causes the gene fusion to occur.

The researchers took prostate cancer cells that did not reflect the gene fusion but that were sensitive to androgen, a male hormone known to play a role in some prostate cancers. They exposed the cells to androgen and found that two pieces of chromosome that are normally far apart are relocated near each other.

Next, the researchers applied radiation to the androgen-stimulated cells. This stress or insult to the cells – designed to induce chromosomal breaks – led to the gene fusion occurring.

“We thought the gene fusions occurred as a chance event, but it’s not. Chromosomes can actually be induced in three-dimensional space to be close to each other. Then when an insult to the DNA occurs, the fusion happens,” says lead study author Ram-Shankar Mani, Ph.D., a research fellow in pathology at the U-M Medical School.

The researchers believe the findings could have implications for gene fusions that occur in other cancer types. By understanding how gene fusions occur, the researchers suggest that screening tools or prevention strategies could potentially be developed.

AARP, Key Retirees’ Lobby, To Endorse Democrats’ Health Bill

Nov 05, 2009

AARP, the influential retirees’ lobby, is expected to endorse House Democrats’ plan to overhaul the health system today, the Associated Press reports, citing unnamed officials. “Backing the 10-year, $1.2 trillion House bill is a tricky move for AARP. Many retirees are concerned about cuts in Medicare payments to medical providers, which will be used to finance an expansion of health insurance coverage to millions of working families who now lack it. Also, AARP says its membership is about evenly divided among Democrats, Republicans and independents, meaning its endorsement in today’s highly politicized atmosphere could anger many members. Floor votes on the House bill could come as early as this weekend”
ABC News reports that “Speaker [Nancy] Pelosi and her leadership team have been lobbying the lobby for weeks,” adding that Democratic sources had confirmed the AARP would endorse the bill. One Democrat called it a “big victory”

This is part of Kaiser Health News’ Daily Report – a summary of health policy coverage from more than 300 news organizations. The full summary of the day’s news can be found here and you can sign up for e-mail subscriptions to the Daily Report here. In addition, our staff of reporters and correspondents file original stories each day, which you can find on our home page.

Medscape.com, November 5, 2009, by Roxanne Nelson – For decades, researchers have observed an apparent association between low serum cholesterol levels and higher overall cancer incidence and mortality. However, 2 new papers published online November 3 in Cancer Epidemiology, Biomarkers & Prevention appear to have put that concern to rest and shed new insight into the role that cholesterol plays in cancer.

Some experts have attributed the association between low cholesterol levels and increased cancer risk/mortality to reverse causation – undiagnosed cancer causing a reduction in cholesterol levels. The first of the new studies adds evidence to the hypothesis that lower levels of serum cholesterol might be a marker of existing malignancy, not a causal factor. That study results also found that high-density-lipoprotein (HDL) cholesterol was modestly but significantly associated with an overall decreased cancer risk.

The results from this analysis should help dispel any lingering doubts that low cholesterol can cause cancer.

“The results from this analysis should help dispel any lingering doubts that low cholesterol can cause cancer,” said Eric Jacobs, PhD, strategic director of pharmacoepidemiology at the American Cancer Society, during a press briefing that was held to discuss the results of these 2 papers.

“It also raises an interesting question about whether levels of HDL cholesterol might lower the risk for some cancers,” added Dr. Jacobs, who coauthored an accompanying editorial. “This is a new and exciting question but we need to do a great deal more research before we have any clear answers.”

The second study found that men with low circulating cholesterol levels have a reduced risk for high-grade prostate cancer. The authors observed that men with low levels of total cholesterol (<200 mg/dL) had a lower risk for prostate cancer with a Gleason score of 8 to 10 than men with high cholesterol (odds ratio [OR], 0.41; 95% confidence interval [CI], 0.22 – 0.77). Overall, participants with serum cholesterol levels below 200 mg/dL had a 59% reduction in their risk for high-grade disease.

The second study raises a different question: Does having low total cholesterol reduce the risk for very high-grade prostate cancer? “There is increasing evidence that this may be true, but evidence from human studies is still limited,” said Dr. Jacobs. “Again, more research is needed before we have clear answers to this question.”

Dr. Jacobs notes that the results from the 2 studies provide 1 answer and raise 2 new issues. Results from the first study clearly show that low total cholesterol is unlikely to increase the risk for cancer and, at the same time, raise a concern about the potential role of high HDL cholesterol in reducing risk for cancer.

Results from the second study raise concern about the association between low total cholesterol and a reduced risk for high-grade disease. “Analyses to replicate the association between low total cholesterol and reduced risk of high-grade prostate cancer are well justified,” according to the editorial, which is coauthored by Susan M. Gapstur, PhD, MPH, also from the American Cancer Society. “Analyses of associations of cholesterol levels and use of cholesterol-lowering drugs with prostate cancer progression among men with localized low-grade prostate cancer could also clarify the role of cholesterol in prostate carcinogenesis.”

ATBC Study

In the first study, researchers from the National Cancer Institute (NCI) and the National Institute for Health and Welfare in Helsinki, Finland, examined the relation between serum total and HDL cholesterol and the risk for overall and site-specific cancers among participants enrolled in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) Study. The cohort consisted of 29,093 male smokers residing in Finland. The trial period concluded on April 30, 1993, but follow-up continued until diagnosis, death, or March 31, 2003.

Fasting serum total and HDL cholesterol were assayed at baseline, and a total of 7545 incident cancers were identified during follow-up, which was as long as 18 years.

“Several studies have attempted to exclude reverse causality by looking at time trends, but most [were not] large enough or had insufficient length of follow-up to provide more definitive information on this point,” said study author Demetrius Albanes, MD, a senior investigator at NCI, during the briefing. “At the same time, very few studies mentioned HDL cholesterol and any relationship between HDL cholesterol and overall cancer risk.”

“With this background in mind, we set out to enhance our scientific understanding of the cholesterol/cancer relationship,” he added.

Cancer Marker Rather Than Cause

Dr. Albanes and colleagues found that, in a multivariate model, higher serum total cholesterol was associated with a lower overall cancer incidence. Patients in the highest quintile for serum total cholesterol concentration had lower overall risk for cancer than those in the lowest quintile (>276.7 vs <203.9 mg/dL; relative risk [RR], 0.85; 95% confidence interval [CI], 0.79 – 0.91; P < .001).

“Cancers of the lung, kidney, and liver contributed substantially to this finding,” Dr. Albanes said. “However, this relationship disappeared when we excluded cases diagnosed within the first half of the follow-up period or within 9 years of when their baseline blood sample was collected.”

“This finding supports the idea that the low serum cholesterol levels that we detected as a possible risk factor may actually have been the result of undiagnosed cancers,” he added. “In addition, we observed a greater decline in total serum cholesterol from baseline to 3 years, specifically among the cases that were diagnosed in the early half of the observation, as opposed to the latter portion.”

The researchers also observed that higher HDL cholesterol levels were associated with a decreased risk for cancer (>55.3 vs <36.2 mg/dL; RR for the highest vs the lowest quintile, 0.89; 95% CI, 0.83 – 0.97; P = .01). This inverse association of HDL cholesterol was evident for cancers of the lung, prostate, liver, and hematopoietic system.

This is the first study to show a significant relationship between higher HDL and lower risk for all cancer combined.

This is the “first study to show a significant relationship between higher HDL and lower risk for all cancer combined, with an 11% lower risk for the highest category of HDL,” Dr. Albanes said.

In contrast to the findings for total cholesterol, the exclusion of cases in the early years of follow-up made the HDL association stronger rather than weaker, he emphasized. “Thus, the 11% lower risk we had earlier became a 14% to 15% lower risk in the highest HDL cholesterol category.”

Prostate Prevention Trial

In the second study, Elizabeth Platz, ScD, MPH, codirector of the Cancer Prevention and Control Program at the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University in Baltimore, Maryland, and colleagues evaluated the association between low serum cholesterol and prostate cancer risk. The cohort in this prospective study consisted of 5586 men, 55 years and older, who were participants in the Prostate Prevention Trial and who had been randomized to the placebo group of that study between 1993 and 1996.

Unlike previous studies, the authors note, the number of high-grade cases in the placebo group of this trial was sufficiently large to allow the estimation of the association between cholesterol levels and disease with a Gleason score of 8 to 10.

In a previous study, explained Dr. Platz, they had observed that men who were using statin drugs had a lower risk for advanced prostate cancer and, with longer-term follow-up, had a lower risk for high-grade disease, but there was no association between statin use and overall prostate cancer risk.

Association With Gleason Score of 8 to 10

A total of 1251 cases of prostate cancer were identified: 993 cases with a Gleason score of 2 to 6, 199 cases with a Gleason score of 7, and 59 cases with a Gleason score of 8 to 10. Although an association was observed between low cholesterol levels and a Gleason score of 8 to 10, none was observed for overall prostate cancer risk (OR, 0.97; 95% CI, 0.85 – 1.11). In addition, no association was observed for disease with a Gleason score of 2 to 6 (OR, 1.03; 95% CI, 0.89 – 1.18) or with a Gleason score of 7 (OR, 0.93; 95% CI, 0.69-1.24).

“Our results confirm the prior work that we did and are compatible with the findings we had for statins and prostate cancer,” said Dr. Platz.

The next move might be to look not just at total cholesterol but also the relation with high HDL and low low-density-lipoprotein cholesterol, she said. Another step might be to “evaluate whether cholesterol lowering, rather than low cholesterol as the usual state, would explain this relationship,” she added.

Both studies were funded by the National Institutes of Health. The authors and editorialists have disclosed no relevant financial relationships.

Cancer Epidemiol Biomarkers Prev. 2009;18:2805-2806, 2807-2813, 2814-2821.

Physorg.com, October/November 2009  —  Chronic sleep deprivation in a mouse model of Alzheimer’s disease makes Alzheimer’s brain plaques appear earlier and more often, researchers at Washington University School of Medicine in St. Louis report online this week in Science Express.

 They also found that orexin, a protein that helps regulate the sleep cycle, appears to be directly involved in the increase.

Neurodegenerative disorders like Alzheimer’s disease and Parkinson’s disease often disrupt sleep. The new findings are some of the first indications that sleep loss could play a role in the genesis of such disorders.

“Orexin or compounds it interacts with may become new drug targets for treatment of Alzheimer’s disease,” says senior author David M. Holtzman, M.D., the Andrew and Gretchen Jones Professor and chair of the Department of Neurology at the School of Medicine and neurologist-in-chief at Barnes-Jewish Hospital. “The results also suggest that we may need to prioritize treating sleep disorders not only for their many acute effects but also for potential long-term impacts on brain health.”

Holtzman’s laboratory uses a technique called in vivo microdialysis to monitor levels of amyloid beta in the brains of mice genetically engineered as a model of Alzheimer’s disease. Amyloid beta is a protein fragment that is the principal component of Alzheimer’s plaques.

Jae-Eun Kang, Ph.D., a post-doctoral fellow in Holtzman’s lab, noticed that brain amyloid beta levels in mice rose and fell in association with sleep and wakefulness, increasing in the night, when mice are mostly awake, and decreasing during the day, when they are mostly asleep.

A separate study of amyloid beta levels in human cerebrospinal fluid led by Randall Bateman, M.D., assistant professor of neurology and a neurologist at Barnes-Jewish Hospital, also showed that amyloid beta levels were generally higher when subjects were awake and lower when they slept.

To confirm the link, Kang learned to use electroencephalography (EEG) on the mice at the Sleep and Circadian Neurobiology Laboratory at Stanford University with researchers Seiji Nishino, M.D., Ph.D., and Nobuhiro Fujiki, M.D., Ph.D. The EEG readings let researchers more definitively determine when mice were asleep or awake and validated the connection: Mice that stayed awake longer had higher amyloid beta levels.

20091106-1

Keeping patients on the heart-lung machine improved survival, study found

GoogleNews.com, HealthDay.com, November 5, 2009, by Ed Edelson  —  Longer-term outcomes for people who had coronary bypass surgery “off-pump,” meaning without the use of a heart-lung machine, were worse than for those undergoing the conventional procedure, a major study finds.

One year after surgery, about one in 10 patients getting the off-pump procedure had died, suffered major complications, had heart attacks or required repeat bypasses, compared to 7.4 percent of those who underwent operations using heart-lung machines, researchers report in the Nov. 5 issue of the New England Journal of Medicine.

And the hoped-for advantage of off-pump bypass in other areas, such as less effect on mental function, did not show up, said study co-author Dr. Frederick L. Grover, chair of surgery at the University of Colorado, Denver.

“We thought they would end up doing better neurologically, or that other organs would do better, or they would get out of the hospital faster — and that did not pan out,” Grover said.

Follow-up scans also showed that fewer of the off-pump grafts remained open after a year (82.6 percent vs. 87.8 percent). The percentage of grafted blood vessels that became narrower and less able to supply the heart with blood was also higher in the off-pump group, the researchers reported.

The study included more than 2,200 adults, almost all of them men, who had bypass surgeries at 18 Veterans Administration medical centers across the country. The research was needed because initial enthusiasm for the off-pump procedure has since been “tempered by concern” about the long-term effectiveness of the operation, the report’s authors explained.

About 20 percent of bypass operations in the United States are done off-pump, Grover estimated.

There still is a role for off-pump bypass surgery, he said, based on individual surgeon’s expertise and patient characteristics, but the study indicates that the procedure’s role will be limited.

“Some surgeons around the country do off-pump bypass and they are very accomplished at it,” he said. “Most haven’t done the long-term follow-up that we have done, but they have pretty good results, and I would doubt they would change their practice. But those who haven’t taken it on are less likely to take it on.”

Grover said he would continue to do off-pump bypass surgery for some patients — those who are elderly, who have major disabilities or have suffered previous strokes. “I would do it for Jehovah’s Witnesses, who can’t take [donated] blood,” he said.

Another expert agreed. Off-pump bypass surgery will continue to be done, said Dr. Eric David Peterson, a professor of medicine at Duke University, even though the newly reported study, which he noted is larger than any done before, does not show the expected benefits.

“Surgeons who use it would say the results are about the same, and in their hands it would be an attractive alternative,” Peterson said. “But for the physician who has not yet mastered it, this [off-pump surgery] would not be necessary because conventional means are at least as good, if not better.”

The verdict in the study for bypass surgery in general was favorable, Peterson said. “Both methods look remarkably good, with very little effect on neurocognitive function,” he said.

The study does not disprove the possible benefits of off-pump surgery for older, sicker people, Peterson said. The men who had bypass operations in the study were “a bit younger than average, and their risks are a little lower than what you see in the general bypass population,” he noted.

“Those populations thought to be at most risk for complications probably were not enrolled in this study,” Peterson said. “Off-pump bypass may have a role there.”

More information

The basics of coronary bypass surgery are explained at the U.S. National Library of Medicine.
SOURCES: Frederick L. Grover, M.D., professor and chair, surgery, University of Colorado, Denver; Eric David Peterson, M.D., professor, medicine, Duke University, Durham, N.C.; Nov. 5, 2009, New England Journal of Medicine