The Cautious Optimism of Biomarker Research
“Diseases are complicated because people are complicated.”
Although eight years have passed since the sequencing of the human genome, John Harley, MD, PhD, is not surprised the achievement has yet to produce sweeping advances in medical diagnostics and treatment.
New technical capabilities in genomic sequencing—including faster processing speeds and lower costs—are unlocking clues for detecting and treating deadly cancers and diseases like Alzheimer’s. Still, the work is tedious and researchers are finding that new payloads of data often reveal more layers of complexity instead of breakthrough cures.
“Diseases are complicated because people are complicated,” explains Harley, director of the Center for Autoimmune Genomics and Etiology at Cincinnati Children’s and the Division of Rheumatology. “We now have enormous haystacks of data and the trick is how we find nuggets of information that are useful for understanding, preventing and curing disease.”
Mining the nuggets
Among the nuggets researchers want to mine are new diagnostic biomarkers – the gene mutations, specific proteins or other signals in the body that reveal the presence or likelihood of disease. The goal is to find faster, more accurate technologies for diagnosis and treatment. It can be a scientific game of chance, Harley says.
Sometimes a gene gives only tiny clues to a disease process. Other times it unlocks a vault of new answers. The clues may be hiding behind a gene variation not yet detected. Then there is the issue of whether even useful genetic information applies to a broad population of people or is highly individualized.
Take the disease Harley studies – systemic lupus erythematosus. A mainly adult disease, when it hits during childhood, it puts children at risk of brain inflammation, kidney damage, arthritis, skin lesions and more. About a year ago, researchers studying the disease were looking at roughly 35 associated genes. At last check they were up to 50.
So far, most of the associated genes in lupus each has a small effect, making it difficult to find reliable genetic signatures that distinguish sick from healthy people. Even so, the information gives researchers biological insights and new ways to look at the disease.
“Progress may be hard to come by, but that should not change what we come to work every day to do,” Harley says. “We do what we can. You may study a large number of genes that don’t have a dramatic impact, and then one study comes along that gives us the answer we need and helps 26 million people. That justifies it all.”
Advancing the biomarker
Researchers have known about biomarkers for more than a century – the telltale biological signs in blood, urine, hair and breath that indicate if a person’s blood glucose is too high or the body has been invaded by llness-causing pathogens.
Still, most of the diagnostics used today are retrospective, Harley points out. The tests tend to catch problems after they have already started, or in some cases after it is too late to help.
This is where the hope – and challenge – of advancing new and meaningful biomarker technologies rests: finding disease at its earliest stages, making treatment precise and effective or preventing an illness from happening in the first place.
Not so far off
Gregory Grabowski, MD, director of Human Genetics at Cincinnati Children’s, sees a day when it may be possible to stop a disease before it takes root, soon after birth. The early foundations for achieving this ultimate form of personalized and preventive medicine are under development, although the goal remains in the realm of science fiction – for now.
Nevertheless, the efforts are producing results. Cincinnati Children’s has been on the leading edge of gene chip technology, which emerged a few years ago as a bold and less invasive way to test children for specific gene mutations associated with hearing loss. The technology has enhanced diagnoses for a number of diseases, Grabowski says.
A decade of research, and installing faster computing and high-throughput analysis technologies, has allowed Cincinnati Children’s to build large and growing databases of normal and disease-related variations in genetic makeup. A patient’s clinical sample can be tested against the databases to get an idea of which part of the genome is affected in a particular disease.
“We are doing roughly 3,000 of these tests a year and it’s been highly useful,” Grabowski explains. “It’s led to a refinement of diagnosis by telling us what parts of the genome might be duplicated or deleted in individuals, or if it involves longer or shorter pieces of DNA.”
Narrowing the field
A major hurdle to moving the technology forward is the sheer size of the human genome and the computing and analytic power needed to process data so it is useful. The human genome has 3 billion base pairs.
At a recent meeting of the American College of Medical Genetics, where conquering this mountain of information was the chief topic, Grabowski says the main message to clinicians was to focus analyses exclusively on exomes. Exomes, which encode proteins, make up only 1 percent of the human genome.
“That alone is daunting,” Grabowski says. “From the 3 billion pairs you are still left with 3 million pairs that you have to sequence and analyze to see if they are normal, abnormal or associated with disease.”
Still, the effort is worth it, Grabowski says, pointing to the rare blood disorder Fanconi anemia. Sixteen gene mutations are associated with the disease. Finding out which gene is involved once required a painstaking process of trial and error. Scientists used a reprogrammed virus, inserting a correct copy of each gene into blood cells to see which one fixed the disorder. It’s now possible at Cincinnati Children’s to take a sample of DNA and use a computerized robot to sequence all 16 genes with the sample.
“This is powerful technology that is starting to displace a lot of established approaches we have in molecular diagnostics,” Grabowski says. “My guess is that one day this approach will become common practice around the country.”