Sleuthing Out a Diagnosis for Medical Mysteries

Solving a difficult diagnosis is far from elementary—even if your name is Dr. Watson.

Unusual congenital disorders, particularly unexplained neurological and cognitive problems, chronic health conditions that don’t fit any one category, and deadly stealth predispositions lurking like time bombs ready to set off heart disease, cancer and other lethal disorder now may be easier to track down with the help of clues from genome sequencing.

“Sometimes we can look at the patient and the history to make an educated guess, and then test the most likely gene for a condition we suspect. However, in many cases it’s not that straightforward. Not that much is visible to suggest a diagnosis, and the symptoms are so nonspecific that there is nothing to give you a starting point,” Bruce Korf, MD, PhD said.

A nationally recognized leader in human genetics and neurofibromatosis, Korf is Chair of the UAB Department of Genetics. He and his colleagues are often called on to search for clues to the genetic basis of baffling diseases.

“When we don’t have evidence to suspect a specific gene is involved, it’s faster and may be more cost effective to take a broader approach and sequence the exome,” Korf said.

The exome, or protein encoding regions of the genome, are the most likely “scene of the crime” where mutations causing genetic disorders tend to occur. Sequencing the exome rather than the entire genome, is often more efficient and generates a more manageable quantity of information.

“This still produces an incredible volume of data to sort through,” Korf said. “There are around three billion base pairs in any one person’s genome. Working with that amount of information often tests the limits of current information technology. Rather than trying to upload a data set, we may have to move it on a hard drive.

“You may be looking at up to three million variants. How do you know which are involved? We begin by comparing them to databases of known variants whose effects have already been described. Though databases aren’t as robust as we’d like them to be, or anywhere near what they will be in a few years, they do help us filter out the known benign mutations.”

This brings down the numbers, but there may still be hundreds or thousands of possibilities.

“If the nature of the disease suggests that it’s more likely to be inherited as a recessive disorder than a dominant trait, we start with variants that show up in both genes. If the patient is a child and neither parent has a variant, it’s a new mutation and could be significant.”

When symptoms show up in other relatives who have the mutation, that can also be evidence a variant is likely to be relevant. In recessive disorders, a variant that shows up in both parents may be a strong indication that it warrants closer examination.

“An infant had a seizure disorder that didn’t fit into any other diagnosis. There seemed to be a problem with how the cells were handling ions and blood chemistry. We suspected a gene and found both parents had a copy of the same variant, which gave us a strong indication it might be involved,” Korf said.

The relative severity of the effects of known variants can also provide clues.

“Though a patient may have mutations that are more benign, when we’re looking for the cause of a disorder with more severe effects, we focus on the mutations that are known to have a greater impact. When we have a short list of ten to twenty suspects, we compare them to databases looking for reports and descriptions of conditions associated with those variants,” Korf said.

“Sometimes when you see the most likely possibilities, it makes sense and you say, ‘of course.’ But it’s often something that wouldn’t normally have come to mind if you hadn’t done the work. The patient’s symptoms may be atypical, or the condition can be so rare, it may take more work to get a clearer picture. In some cases, you may the first to identify a variant with effects that haven’t been previously described, and you may be adding new information to what we know about genetic disorders.”

In experienced hands, gene sequencing can be a powerful tool in getting to the root causes of hard to diagnose diseases. Though clinicians will want to become familiar with the benefits it can offer patients and when it is indicated, it is more likely to be a tool they will be using through referral rather than directly.

“Some cardiologists are ordering genetic testing for variants that are known to be related to heart conditions such as specific rhythm disorders, but when a patient needs sequencing to determine whether a genetic factor is involved in an unknown disorder, it’s probably best to get someone with experience on the case,” Korf said.

“A specialist in genetics can make an informed judgment on which tests to order. Experience in sequencing provides a background for making the best educated guess on where to start and how to make sense of the results,” Korf said. “The data you get back is usually not black and white. You’re likely to have multiple variants, and it takes judgment to identify which are involved.”

Genetic counseling is another area where experience is important

“When you sequence a genome or exome, you may discover unexpected incidental findings that could be significant to the patient’s health, and you need to know what to do with that information,” Korf said.

Does the patient want to know he has a gene for a disorder he may develop later in life, or would he prefer not knowing so he can live his life without that worry ever-present? Consent forms for testing should have an opt-out provision for patients who prefer not to know. However there are 56 gene variants that are generally considered important to disclose if they are discovered in incidental findings.

“These are reported variants that are known or predicted to be pathogenic that we can do something about—either through close monitoring, medication or surgery,” Korf said. “Examples are hereditary breast and ovarian cancer, and hypertropic cardiomyopathy that can make exercise lethal. Patients who want to have children may want to know if they are carrying the gene for cystic fibrosis, Tay Sacks or hemophilia.”

When the answer is yes, experienced genetic counselors are prepared to provide the facts and ease undue fears.

The cost of sequencing a patient’s genes, particularly the exome, has decreased dramatically and is continuing to become more affordable. It is also a key element in the trend toward personalized medicine. Gene testing can predict which blood thinner is likely to be most effective in a specific patient, and what dosage is likely to be optimum. The patient’s genetic code can also be a predictor of likely response to certain cancer drugs and other medications. The day may not be far off when referring to a patient’s genome before prescribing medication is a routine part of medical practice.

Genetics and genomics are evolving rapidly. Recent developments making news include a prenatal test of the mothers’ blood that can pick up fetal cells and screen for abnormalities in chromosomes 13, 18 and 21. This test may prevent the need for amniocentesis to screen for Down’s syndrome and certain other congenital disorders.

“We’re already seeing screening for more genetic disorders at birth where early intervention and surveillance can make a life-saving difference,” Korf said. “We’re likely to see much more robust screening of newborns in the future. “

Will we be able to predict whether a newborn will develop type 2 diabetes or a similar disorder?

“The jury is still out on some of that,” Korf said. “However, the mystery of what causes many chronic complex disorders will be solved. Understanding more about the basis of diseases could suggest avenues of treatment. When we know why they happen, we’ll be in a better position to solve the next mystery—how to treat them.