Redefining Disease, Genes and All

Correction Appended

Duchenne muscular dystrophy may not seem to have much in common with heart attacks. One is a rare inherited disease that primarily strikes boys. The other is a common cause of death in both men and women. To Atul J. Butte, they are surprisingly similar.

Dr. Butte, an assistant professor of medicine at Stanford, is among a growing band of researchers trying to redefine how diseases are classified — by looking not at their symptoms or physiological measurements, but at their genetic underpinnings. It turns out that a similar set of genes is active in boys with Duchenne and adults who have heart attacks.

The research is already starting to change nosology, as the field of disease classification is known. Seemingly dissimilar diseases are being lumped together. What were thought to be single diseases are being split into separate ailments. Just as they once mapped the human genome, scientists are trying to map the “diseasome,” the collection of all diseases and the genes associated with them.

“We are now in a unique position in the history of medicine to define human disease precisely, uniquely and unequivocally,” three scientists wrote of the new approach last year in the journal Molecular Systems Biology. Such research aims to do more than just satisfy some basic intellectual urge to organize and categorize. It also promises to improve treatments and public health.

Scientists are finding that two tumors that arise in the same part of the body and look the same on a pathologist’s slide might be quite different in terms of what is occurring at the gene and protein level. Certain breast cancers are already being treated differently from others because of genetic markers like estrogen receptor and Her2, and also more complicated patterns of genetic activity.

“In the not too distant future, we will think about these diseases based on the molecular pathways that are aberrant, rather than the anatomical origin of the tumor,” said Dr. Todd Golub, director of the cancer program at the Broad Institute in Cambridge, Mass.

The reclassification may also help find drugs. “There are 40 drugs to treat heart attacks, but none to treat muscular dystrophy,” Dr. Butte said. If the 6

diseases are similar in some molecular pathways, perhaps the heart attack drugs should be tested against muscular dystrophy.

Dr. Golub and colleagues at the Broad Institute have developed a “Connectivity Map,” which profiles drugs by the genes they activate as a way to find new uses for existing drugs.

The research will also improve understanding of the causes of disease and of the functions of particular genes. For instance, two genes have recently been found to influence the risk of both diabetes and prostate cancer.

“I’m shaking my head with disbelief that two genes would pop up in these two diseases that have absolutely nothing in common,” said Dr. Francis S. Collins, the director of the National Human Genome Research Institute. He said another gene, cyclin-dependent kinase inhibitor 2A, seemed to be involved in cancer, diabetes and heart disease.

A consistent way to classify diseases is also essential for tracking public health and detecting epidemics. The World Health Organization takes pains to periodically revise its International Classification of Diseases, which is used, among other ways, to tally the causes of death throughout the world. The classification is also the basis of the ICD-9 codes used for medical billing in the United States.

The first international classification, in the 1850s, had about 140 categories of disease, according to Dr. Christopher G. Chute, chairman of biomedical informatics at the Mayo Clinic. The 10th edition, in 1993, had 12,000 categories, said Dr. Chute, chairman of the committee developing the 11th version, due in 2015.

The increase stems mainly from better knowledge and diagnostic techniques that allow diseases to be distinguished from one another. For most of human history, diseases were named and classified by symptoms, which was all people could observe.

Linnaeus, the 18th-century Swedish scientist known for categorizing creatures into genus and species, also developed a taxonomy of disease. He had 11 classes — painful disease, motor diseases, blemishes and so on — that were further broken down into orders and species. But not knowing about viruses, for instance, he classified rabies as a mental disease, Dr. Chute said.

In the 19th century, a big shift occurred. Doctors began learning how to peer inside the body. And diseases began to be classified by their anatomic or physiological features. 7

The stethoscope let doctors realize that what had been thought of as 17 conditions — like coughing up blood and shortness of breath — could all be different symptoms of the same disease, tuberculosis.

“The advent of the stethoscope made it possible to unify tuberculosis,” said Dr. Jacalyn Duffin, a professor of the history of medicine at Queen's University in Ontario.

The shift from symptoms to anatomical measurements had big implications for patients, said Dr. Duffin, who is also a hematologist.

“Up until the 18th century, you had to feel sick to be sick,” she said. But now people can be considered sick based on measurements like high blood pressure without feeling ill at all.

Indeed, Dr. Duffin said, people who feel sick nowadays “don’t get to have a disease unless the doctor can find something” and instead might be told that it’s all in their head. Doctors argue, for instance, about whether fibromyalgia or chronic fatigue syndrome, which have no obvious anatomical causes, are really diseases.

Genes might allow the study of diseases at a finer level than even physiological tests. Genes are the instructions for the production of proteins, which interact in complex ways to carry out functions in the body. Disruptions in these molecular pathways can cause disease.

“It gives you a direct connection to what the root causes are,” said Dr. David Altshuler, a professor of medicine and genetics at Harvard and Massachusetts General Hospital, and a researcher at the Broad Institute. “That is different from listening to a stethoscope.”

Some of the earliest work has until now been with inherited diseases caused by mutations in a single gene. Diseases have been subdivided by the type of mutation.Hemophilia was divided into hemophilia A and B, caused by mutations in different genes for different clotting factors. And what was once considered a mild form of hemophilia was later identified as a variant of a different clotting disorder, von Willebrand disease, caused by mutations in a different gene and requiring a different clotting factor as treatment.

Diseases are being lumped, as well as split. Researchers at Johns Hopkins reported in the April issue of Nature Genetics that two rare syndromes with different symptoms might represent a continuum of one disease. One syndrome, Meckel-Gruber, is tied to neural defects and death in babies. The other, Bardet-Biedl, is marked by vision loss, obesity, diabetes and extra fingers and toes. 8

The techniques are being applied to diseases for which the genetic cause is not as clearly known and which might be a result of multiple genes.

Dr. Butte uses data from gene chips that measure which genes are active, or expressed, in a cell. Amid thousands of studies using such chips, many compared the gene activity patterns in diseased tissue with that of healthy tissue.

Much of the raw data from such studies are deposited in a database. So Dr. Butte can gather data on gene activity for scores of diseases without leaving his desk. He then performs statistical analyses to map diseases based on similarities in their patterns of gene activity.

Other scientists use data on which genes appear to cause disease or contribute to the risk of contracting it.

Using such data, Marc Vidal, a biologist at Harvard, and Albert-Laszlo Barabasi, now a physicist at Northeastern University, created a map of what they called the “diseasome” that was published last year in The Proceedings of the National Academy of Sciences.

Diseases were represented by circles, or nodes, and linked to other diseases by lines that represent genes they have in common — something like the charts linking actors to one another (and ultimately to Kevin Bacon) based on the movies they appeared in together.

The number of genes associated with diseases is expanding rapidly because of so-called whole genome association studies. In these studies, gene chips are used to look for differences between the genomes of people with a disease and those without.

Multiple techniques can be combined. In a paper published online in Nature in March, scientists at Merck reconstructed the network of genes involved in obesity.

One area that might benefit from genetic disease classification is psychiatry. Because of the difficulty of measuring the brain, psychiatric diagnoses are still mainly based on symptoms. The Diagnostic and Statistical Manual of Mental Disorders contains descriptions of conditions as diverse as acute stress disorder and voyeurism.

Scientists have found that certain genes appear to be associated with both schizophrenia and bipolar disorder. Those links, and the fact that some drugs work for both diseases, have prompted a debate over whether they are truly distinct disorders. “The way we categorize these into two 9

separate entities is almost certainly not correct,” said Dr. Wade H. Berrettini, a professor of psychiatry at the University of Pennsylvania.

But Dr. Kenneth S. Kendler, a professor of psychiatry and human genetics at Virginia Commonwealth University, said that even if the two diseases shared genes, the diseases remained distinct. Schizophrenia is marked by hallucinations and impaired social functioning, and bipolar disorder by mood swings.

“It’s extremely naïve to think that psychiatric illnesses will collapse into categories defined by a gene,” he said. “Each gene at most has a quite modest effect on the illness.”

Some experts say that such limitations may hold true for other diseases, as well, and that genetics will not be able to unequivocally define and distinguish diseases. “We shouldn’t expect, nor will we get, this decisive clarity,” said Fiona A. Miller, associate professor of health policy, management and evaluation at the University of Toronto.

She and others said genetic classification could bring its own ambiguities. Newborns are now often screened for cystic fibrosis with the idea that they can be treated early to help avoid complications. But some infants with a mutation in the gene responsible for the disease are unlikely ever to have symptoms. Do they have the disease?

“We don’t know what to call these infants,” said Dr. Frank J. Accurso, a professor of pediatrics at the University of Colorado. “We don’t even have a good language for it yet.”

Still, Dr. Butte said nosology based on genes would one day make today’s classifications look as quaint as ones from 100 years ago look now. One category in the 1909 listing of the causes of death, for instance, was “visitation of God.”

“Imagine how they are going to be laughing at us,” he said. “Not 100 years from now, but even 50 or 20 years from now.”

This article has been revised to reflect the following correction:

Correction: May 9, 2008

An article on Tuesday about changes in the way diseases are classified 10

misspelled the university where Dr. Jacalyn Duffin works as a professor of the history of medicine. It is Queen’s University, not Queens.