HONOLULU — As the number of genes discovered for epilepsy has mushroomed, the implications for genetic testing have become more, not less, complex, according to Jeffrey Noebels, MD, PhD. Recent findings about ion channel gene mutations have shown that while some discoveries can lead quickly to better treatment, others are just as quickly making easy generalizations about the genetics of epilepsy a thing of the past.
“We have been able to identify over 120 genes, many of them for ion channels,” said Dr. Noebels, professor of neurology, neuroscience, and molecular genetics at Baylor University in Houston, TX, here at a plenary session at the AAN annual meeting. “But how can we put this to work in the clinic? There is an enormous rush for direct-to-consumer genetic testing, but we still have little insight into what the results mean,” he cautioned.
Gene tests can be ranked according to their level of utility. So-called “level one” tests are those for gene variants that, if present, “would trigger a change in management.” Second-level tests are those whose results might lead to preventative lifestyle changes or family planning decisions. “For third-level genes, we wouldn't order the test, either because there is only a small increased risk from a positive result, or we don't know what the results means.” Ion channel genes encompass the entire range, Dr. Noebels explained.
ONE CLEAR SUCCESS
The voltage-gated potassium channel gene, KCNQ1, “may be the first level-one ion channel gene worth testing for in epilepsy planning.” The gene, he said, causes a “dual phenotype” in epilepsy and cardiac arrhythmia, and is linked to sudden unexplained death in epilepsy (SUDEP).
SUDEP causes up to 18 percent of all early deaths in epilepsy, making it a more significant cause of death than accidents or suicides in this population. Until recently, there was no marker for it, and no knowledge of what the “molecular target” might be for the syndrome.
The story connecting heart arrhythmia, epilepsy, and SUDEP to the KCNQ1 gene began in 1995, with the linking of the gene to one form of long QT syndrome (LQTS), and its identification as a cardiac potassium channel gene. In 2009, researchers showed that about a third of patients with that form of LQTS had seizures, suggesting that some cases of SUDEP might be caused by the cardiac effects of the gene.
To test this hypothesis, Dr. Noebels began a series of experiments with mice bearing mutations in the KCNQ1 gene. He found that the gene is expressed in the brain, and that the mice had both a long QT syndrome and spontaneous seizures. “That established that one gene, when co-expressed in heart and brain, could cause arrhythmia in both of them,” he said. “Therefore there was a parsimonious explanation for this syndrome.”
Further study revealed the gene was most heavily expressed in the dorsal motor nucleus of the vagus, and that cortical discharges “were often triggered in lockstep with brief cardiac asystole.” With prolonged video monitoring in the mice, a SUDEP event was observed. It began with bradycardia, which lasted an hour, followed by asystole and then death.
But while the channel gene mutation is clearly linked to serious risks in epilepsy, “the problem with bringing this to the clinic is long QT syndrome ascertainment,” Dr. Noebels said. It is far less common than idiopathic epilepsy, “and only a small fraction of people with epilepsy would likely have this.” And electrocardiograms are not currently standard for patients presenting in the United States with epilepsy, he noted. “We are hoping in the epilepsy community that we can turn this into better practice of epilepsy management, and look early on for a risk factor which is a really serious one.”
Dr. Noebels envisions routine EKG screening for appropriate patients in the epilepsy clinic. A patient with an abnormality would then undergo genetic testing and, if positive, would be referred to a cardiologist. Treatment might include beta-blockers or implantation of a cardiac pacemaker.
There are well-established guidelines for LQT management, “so the question is, why not look for it, and why not treat it?” [Clinical guidelines for treating LQT syndrome, which were developed in 2006 by the American Heart Association, American College of Cardiology, and the European Society of Cardiology, are available at: http://bit.ly/l0WHix.]
A COMPLICATED PICTURE
But while the implications of a positive KCNQ1 gene test should trigger a change in epilepsy management, there are many other genes for which the implications are not at all clear.
“In ancient times — about 25 years ago — people didn't believe you could mutate sodium channels and survive,” Dr. Noebels said. “The belief was that sodium channels were either opened or closed. But in fact we found that the situation is more complicated,” with a number of mutations that “prevent the door from shutting all the way,” making the cell more excitable. “There is probably a large gradient of such changes,” especially given that there are 14 different genes for various subunits of the channel, and different forms are expressed to different extents on different cell types. “Mutation in any one of them can have a very specific effect in the brain,” Dr. Noebels said.
“We know now that mutations can change the channel in a huge number of ways,” including the number of channels produced, the conductance or ion selectivity of the individual channel, and in how channels interact with one another, Dr. Noebels said. Effects of missense mutations differ from frame-shift mutations, and the effects of a mutation in the pore site will differ from those in a cytoplasmic loop, which protrudes into the cell and may contain regulatory regions. “Understanding what a mutation does to a channel is going to be important.”
“The question is, knowing there is this much variability in ion channel gene mutations, what is our ability to return a useful answer to the patient? At the moment, we don't know.”
While some attempts have been made to develop a unified theory linking the type of mutation to the severity of effect, exceptions abound. “Some people have mutations and there is absolutely nothing wrong with them at all,” Dr. Noebels said.
Within the same family, the same mutation may cause epilepsy in one member while another remains an unaffected carrier. “So even within a family it is a poor predictor. There must be other genes in those members who are protected,” genes that are undiscovered as of yet, he explained.
Dr. Noebels's current work is the simultaneous genetic characterization of 237 ion channels in 350 patients, “to see what the landscape of ion channel mutations was. We found that everyone has many rare variants,” with large numbers of them in unaffected controls as well. “Even if you find a mutation, that doesn't prove it is the cause.”
Dr. Noebels concluded: “We are all excited by the new genomics, but we will need a lot more than a single gene to fully understand the ion channelopathy mutations, and to use this information in a wise way.”
“I think one of the critical issues for genetic testing is to have an understanding of what it can and can't tell us,” agreed Timothy Pedley, MD, professor of neurology at Columbia University Medical Center. “It's clear that many of these genetic mutations are far more complicated in terms of their effects than has been appreciated. One thing that comes through very clearly is that you can have a mutation that causes a disease, and you can have the same mutation and be asymptomatic. One important corollary is that unsupervised genetic testing has lots of problems with it, even when you are dealing with a clinically important gene. To do that without the context of genetic counseling is to my mind a mistaken way of doing it.”