Exploring the Connexin Connection Connexin 26 mutations are the most common cause of congenital sensorineural hearing loss, responsible for at least 20 percent of all genetic hearing loss and 10 percent of all childhood hearing loss. Pediatric otolaryngologist Margaret Kenna, an expert in congenital and childhood-onset hearing loss, recently led an online chat on Cx26. The Leader listened in. Overheard
Overheard  |   December 01, 2014
Exploring the Connexin Connection
Author Notes
  • Margaret Kenna, MD, MPH, is on the pediatric otolaryngology faculty at Boston Children’s Hospital, where she directed the pediatric cochlear implant program for eight years and has been director of the department’s clinical research since 2003. Her research focuses on congenital and childhood-onset hearing loss. Margaret.Kenna@childrens.harvard.edu
    Margaret Kenna, MD, MPH, is on the pediatric otolaryngology faculty at Boston Children’s Hospital, where she directed the pediatric cochlear implant program for eight years and has been director of the department’s clinical research since 2003. Her research focuses on congenital and childhood-onset hearing loss. Margaret.Kenna@childrens.harvard.edu×
Article Information
Hearing Disorders / Special Populations / Genetic & Congenital Disorders / Overheard
Overheard   |   December 01, 2014
Exploring the Connexin Connection
The ASHA Leader, December 2014, Vol. 19, online only. doi:10.1044/leader.OV.19122014.np
The ASHA Leader, December 2014, Vol. 19, online only. doi:10.1044/leader.OV.19122014.np
Bridget Gonzalez: I believe that you said in your presentation that Connexin 26 protein is not found in hair cells. Later, I believe you said that people with the Connexin 26 V84L variant did very well with a CI. Throughout the webinar I have heard presenters say that people with Connexin 26 losses do well with CIs. Do you believe that this is due to the fact that the loss is not in the hair cells? Is it now standard for CI children candidates to be tested for Cx26/30 (GJB2/6) genes before implantation? Thank you.
Margaret Kenna: It is unclear whether the patients who have Cx26 as the cause of their hearing loss do well because they are generally otherwise healthy and cognitively normal, or whether there is something about the presence in the supporting cells, but not primarily the hair cells, or whether there are other factors. Also, Cx26 patients generally have grossly normal inner ear anatomy.
Vaibhavi Dharkar: In your lecture you talked about the Connexin 26 expression in cochlea, liver, bladder and brain. Can this explain the association of renal disease in association with hearing loss (syndromes like BOR)?
Kenna: Yes, it can; that is, if the gene, in the case of BOR (which is mainly the EYA1 gene) abnormal expression in the ear, the kidneys, etc., can explain the presentation. What is interesting about many of these genes, such as the EYA1 gene, is that they can obviously affect development as well. What is quite unclear is why, over time, the hearing may get worse in Cx26-related hearing loss. The gene expression is already abnormal to begin with, so either less and less protein is produced, or the work of the supporting cells—where Cx26 is expressed—just goes away over time, affecting the hair cells and ultimately the hearing.
Deborah Hayes: I am interested in the association between Connexin 26 and vestibular function/dysfunction. From your own research, can you please provide more information about this issue, especially in children?
Kenna: In the original Cx26 studies from the late 80s and early 90s, the primary presentations discussed were bilateral severe-to-profound SNHL. Over time, however, it has become very clear that there is a much wider phenotype. Vestibular function certainly can be affected; how commonly that happens or whether it gets worse, like the hearing loss can, is less clear.
We are doing a study right now trying to begin to get some answers to this question, but I have definitely seen children present with both hearing loss and vestibular dysfunction who have biallelic Cx26 mutations. In terms of expression of Cx26, it has mainly been studied in the cochlea; it is now clear that it needs to have much more of an assessment in the vestibular system. Our patients with Cx26 mutations tend to walk on time, unlike the Usher1 patients, but either present with dizziness or more subtle, but measurable, vestibular dysfunction.
Kristen Kowalski: Regarding audiologic phenotype of biallelic mutations … you mention that hearing loss is usually congenital, though normal hearing is detected at birth in 4 to 8 percent—how so? Could this be based on how newborn hearing screening is performed—that is, OAE versus ABR etc.—or is that a rate of true normal hearing at birth that turns into a progressive hearing loss over time?
Kenna: Yes, some of the Cx26 patients can present with “normal” hearing. So, either it is truly normal, by ABR or OAE, or the child passes a newborn hearing screen with possible mild hearing loss that becomes more apparent with time. We have seen it present both ways.
One of the theories about why [an infant has] normal hearing to start with, then progression, is probably the same as why they progress in general—abnormal protein from the mutations. About 40 percent of the Cx26 biallelic patients present with bilateral severe-to-profound SNHL; the others present with measurable hearing, so it is this group of patients who have the possibility of getting worse. With regard to other aspects of the phenotype, we occasionally see patients with mixed hearing losses, and fluctuating losses. In those patients, I always worry about other additional causes, like CMV or EVA, but we rarely find additional causes of the hearing loss.
Sandra Oba: You mentioned there may be a Connexin 26 heterozygous carrier advantage—can you give a bit more information about this?
Kenna: There is a lot that we don’t know about the carriers; in some cases they seem to be more likely to have hearing loss, but in others they seem to maintain normal hearing. For example, the carriers of single mutations, whose children have biallelic mutations with hearing loss, often have excellent hearing. It is also possible that single mutations may be modifiers for other genes or other epigenetic factors, but as I also mentioned, in large groups of patients with SNHL, there seem to be more-than-expected numbers of patients with single Cx26 mutations compared to normal-hearing cohorts.
There have also been some reports of a “heterozygous” advantage, postulated because of the relatively high carrier rate of Cx26 mutations in many populations. Several studies have suggested an increased thickness of the epidermis in carriers, potentially making the subjects less susceptible to infection.
Donna Ettinger: You state that some GJB2 hearing loss patients have other potential causes of SNHL. Does this mean that when the genetic tests show the GJB2 hearing loss, the parents should be advised to keep looking for other causes?
Kenna: A wonderful question and one that I wrestle with all the time. In looking at large cohorts of hearing loss patients, things like congenital CMV, Cx26 and EVA are common, and occasionally occur together. When we see a new patient with hearing loss, I worry about stopping the evaluation without testing for everything, but also know that testing for everything in everyone is usually overkill. So, in general, if I find a likely cause that makes sense, and fits the clinical picture, then I will generally not do further testing unless the hearing loss begins to act differently than expected for the current diagnosis.
For example, a patient with two very clear Cx26 mutations who starts to have night vision issues: The chance they also have Usher Syndrome is low statistically, but I would worry. The one thing we do, even if you think there is a clear cause, is test all newborns with hearing loss for CMV, because it is potentially treatable and common. For example, I have several patients with CMV and other things, such as Cx26, CHARGE and EVA.
Terry Sullivan: I work with a woman whose son was born with hearing loss and has what appear to be “cafe au lait” spots and is undergoing genetic testing. I usually think of neurofibromatosis when I hear of those symptoms, but is there anything else?
Kenna: You are correct, cafe au lait lesions make you think about NF, especially NF1. In theory, NF1 is not supposed to have a higher rate of hearing loss than non-NF1 populations. However, we are beginning to notice that many of our NF1 children also have hearing loss, but no evidence of acoustic neuromas, so again, it could be our referral base, or there could be something to it. However, if I see a child with NF1 and hearing loss, I do look for other causes. In the handful of children we follow with NF1 and hearing loss, the hearing loss does not always stay stable, but again it’s not clear why that is. If you end up getting imaging on these children you must be ready to see and deal with other findings—optic gliomas, etc., which you won’t find in most other children.
Ellen Mastman: What is the oldest age at which you have noticed hearing loss begin in cases where Cx26 is found?
Kenna: The oldest I have found was about 8 years, and that was a sudden-onset, rapidly progressive SNHL, so a somewhat atypical presentation with somewhat uncommon mutations. I think it is certainly possible that much milder phenotypes go unnoticed for long periods of times, so they are not discovered until older ages.
Zarin Mehta: I worked with a family that had four children and all but the youngest had bilateral sensorineural hearing loss. The severity of the hearing loss was in the mild-to-moderately severe range and rather similar for all three children. No family history of hearing loss except age-related (I worked with the children’s grandfather, too). They did not have a mutation for Connexin 26. The parents were not interested in further testing, but for such a family, what other mutations might you advise them to look for—especially if the children want to know their risk of passing on the mutation to their offspring?
Kenna: It certainly sounds recessive, although I guess possibly it could be mitochondrial, especially if the mom had any degree of hearing loss. In our patient population, after Cx26 and SLC26A4 (who mainly have EVA), the Usher genes, as a group, are the next most common recessive genes we find. So I guess I would be concerned about Usher 2, especially USH2A, the most common Usher 2 gene. But there are so many other genes, with widely varying frequency depending on the population you are studying. We have recently found children with mutations in otoancorin, and stereocilin, for example, and of course there are many still-undiscovered genes.
Submit a Comment
Submit A Comment
Comment Title

This feature is available to Subscribers Only
Sign In or Create an Account ×
December 2014
Volume 19, Issue 12