Panel Discussion

Toward Epidemiologic Genetic Research in Genetics of Congenital Hearing Impairment: Progress to Date - Robert Morell, Ph.D., Senior Staff Fellow, NIDCD, NIH, moderator.

The beginning afternoon session consisted of the following six presentations.

1) The Changing Landscape of Early Hearing Detection and Intervention Programs (EHDI)- Terese Finitzo, Ph.D., University of Texas/Callier Center, Hearing Health Institute.

2) Unique sub-populations of "the Deaf" who fail ABRs, but pass otoacoustic emissions - Charles Berlin, Ph.D., Kresge Hearing Research Laboratory of the South, Louisiana State University.

3) Connexin-26 Gene Mutation Detection Using Guthrie Cards from Deaf Children Identified Through the Rhode Island Audiology Screening - Jeffrey Milunsky, M.D., Clinical Genetics Center for Human Genetics, Boston University School of Medicine.

4) Genetic Studies of Hearing Impairment in Special Education Groups - Bronya Keats, Ph.D., Center for Molecular and Human Genetics, Louisiana State University Medical Center.

5) Hearing Phenotype in DFNB1/GJB2/Connexin-26 - Edward Cohn, M.D., Boystown National Research Hospital.

6) Genetics of hearing loss: how common is it? - Arti Pandya, M.D., M.B.A., Department of Human Genetics and Pediatrics, Medical College of Virginia.

7) Panel Discussion: Potential Strategies for Epidemiologic/Population-Based Research in the Genetics of Congenital Hearing Impairment: How do we get from here to there?

Dr. Finitzo reviewed four areas relating to EHDI programs: the state of the nation, the state of the states, the state of the information, and strategies for collaboration between EHDI systems and genetic research.

The Newborn and Infant Hearing Screening and Intervention Act of 1999, was introduced on March 18, 1999. The house version of the bill, HB 1193, was introduced by Rep. James Walsh of New York, who introduced the first hearing screening bill in the early 1990s. The Walsh bill currently has over 80 co-sponsors. The senate version, SB 956, was introduced by senators Snow, Harkin, and Frist. Of the $25 million dollars originally proposed, $13 million would be used for grants to states to design and implement model newborn and infant hearing screening, evaluation, and intervention programs, and another $12 million would be used to support technical assistance and data collection through the CDC, and research through the National Institutes of Health (NIH). At last count, only some $6-10 million dollars remains in the proposed legislation.

The second national activity is the President's FY 2000 Initiative, which will provide a total of $4 million, divided among 45 to 50 grants. These grants will:

"Develop and expand statewide universal newborn hearing screening programs;

"Link screening programs to intervention within the community service system;

"Monitor the impact of early detection and intervention on child, family and systems; and

"Provide technical assistance to states.

In 1992, only Rhode Island and Hawaii had universal newborn screening programs. Although the numbers change daily, Dr. Finitzo said that currently, 19 states have enacted legislation mandating universal newborn hearing screening programs. Eight more have legislation pending, three states' bills are awaiting their Governors' signatures, and four more states have pilot or model projects in progress. If all the bills are signed, Texas and California will be screening 20% of all newborns in the nation.

Dr. Finitzo reviewed how well the programs are actually working, using Texas, New York and Rhode Island as examples.

The Sounds of Texas Project, is a public/private partnership involving the Department of Health, the Hearing Health Institute, the OZ Corporation, the University of Texas Callier Center, and the Texas Commission for the Deaf and Hard of Hearing. Funding has been provided by the Meadows Foundation and Title V. The goal of the project was to screen 85% of the births by the year 2000. The project was begun in FY 1997, and by mid-1998, it was evident that 85% of the births in Texas would not be screened on a voluntary basis. Texas is second in land area to Alaska, with vast distances among people and facilities. Texas is second in births to California, with over 330,000 births in 1997. Forty-eight percent of all births are Medicaid births. There are vast differences in the amount of births per hospital, with some having only five births per year, and with others such as Parkland Hospital, which has more births per year than the states of Rhode Island or Maine. The state also is racially and ethnically diverse, and shares a long border with Mexico, which creates very complex public health issues, and greatly complicates tracking efforts.

Between January of 1994 and December of 1998, there were 116,598 births in 26 hospitals. Of those births, 112,702 (97%) were screened during hospital admission, and 97% of those passed the birth admissions screening. Since each hospital had its own screening mechanism, the Birth Screening Performance Index (BSPI) was used as a measure of comparison. The percent of the babies the hospital was able to screen on birth admission was multiplied by the percent who passed the screening. Overall the BSPI on the 26 hospitals was 93%. Of the babies screened, 3,987 required follow-up, with 2,916 (73%), returning for follow-up. Through 1997, 2.06/1000 babies were identified with permanent hearing loss. The 1998 follow-up is still in progress.

The New York Project included 72,000 babies, with 97% being screened on birth admission. Ninety-six percent passed the screening, 0.39% refused the screening, and 72% returned for follow-up. Both Texas and New York began to distinguish those children who failed the screening from those who were never screened, and found that if a baby was missed on birth admission, the likelihood of that child returning for follow-up was 18-30%. Of those children returning for outpatient rescreening, 79% passed the screen, and 74% were referred for ABR. In 1995-96, 1.96/1000 babies were identified with hearing loss.

Between January of 1993 and December 1996, there were 53,121 births in Rhode Island. Universal screening is mandated by the state, and 99% of the babies born in those years were screened. Ninety percent of the babies screened passed, with a BSPI of 89%. Follow-up was needed by 5,397 children, and 85% of those children did return. One hundred and eleven children, or 2/1000, were identified as having permanent hearing loss.

How information flows and how information varies, will affect the collaboration that can take place to research the genetics of hearing loss. EHDI data goes from the hospitals, to the state health department, to the CDC database, or to the Bureau of Maternal and Child Health (MCH). Most data become more restrictive as they move up the system. Not only are the most data available at the hospital level, but the information tends to vary the most at that level as well. Different hospitals keep different types of data. At the hospital level, the data available may include: the baby's name, mother's name, doctor's name, test results, technology used, the screener's name, content information, etc. At the state level some information may not be included, such as the screener's name. Similarly, at the CDC and MCH levels, the hospital name, or other such data, may not be required. Of course, what is asked for at the MCH and CDC levels, must be collected at the other levels.

The Bureau of MCH has asked for two pieces of information: the number of births in the state; and the number of births screened. CDC has asked for additional information, such as the number of birthing hospitals in the state, how many hospitals have universal screening programs, how many live births do these data account for, how many babies are screened before discharge and one month of age, how many are referred for diagnostic evaluation, how many received audiologic evaluation by three months, how many have permanent congenital hearing loss, whether it is unilateral or bilateral, the type of hearing loss, the median and average age of diagnosis, the minimum and maximum age of diagnosis, and which children received intervention by six months.

EHDI Information Systems, at least in Texas, have three major roles:

"to take care of the baby - get the care needed and no more care than needed;

"to assure a quality program; and

"to assist with public health and population demographics.

Lessons from the Sounds of Texas Project, include:

"the system is only as strong as the weakest link - a triple redundant system was needed;

"quality and accountability are essential - have to prove that the programs were of high quality;

"confidentiality is a major issue - found that parts of the data can be made public while other parts are kept private.

Tracking systems must include information on phenotype, genotype and case level association of genotype and phenotype.

Strategies for collaboration include: the EHDI Information System; the Immunization and Tracking Registry; and the Metabolic Database.

The EHDI system could provide a way to assist research into the genetics of hearing loss.

"The comment was made that using the term "failing" a hearing screening is too negative, and makes the child a failure in the eyes of the parents.

"The Rhode Island universal screening data provide much information and shows that earlier adoption of hearing impaired children takes place because of screening.

"In Texas, parents are informed of the screening process and its consequences, both in writing and verbally.

"If a child fails the screen, the hospital tries to rescreen the child later, with 99% of children passing the rescreen.

"Children are referred to hospitals.

"In Texas, 60% of the children identified with hearing loss were from at-risk situations, and 40% had no risk factors. If a child is at-risk, s/he is rescreened at six months, even if s/he passed the initial screen.

Return to Top

2) Unique sub-populations of "the Deaf" who fail ABRs, but pass otoacoustic emissions - Charles Berlin, Ph.D., Kresge Hearing Research Laboratory of the South, Louisiana State University.

Dr. Berlin showed an animated illustration of what happens inside the auditory systems, to demonstrate why there are special sub-populations of the deaf who do not meet our expectations. The Organ of Corti is spiral shaped, and moves in a particular way that can be separated into three parts by various tests. In the Organ of Corti, the outer hair cells, which move, are connected to the tectorial membrane. The hair cells fire only when moving in a particular direction. One test studies the motion and presence of the outer hair cell, another studies the motion and presence of the inner hair cell and nerve fiber together, and the third studies the presence or absence of the hair cell in the aggregate. Some people appear to be deaf, but have perfectly normal outer or inner cells. Discriminating between and managing the cells becomes a very important issue in terms of early identification. Dr. Berlin refers to the people as unique sub-populations of the deaf, although they have been recently named auditory neuropathy patients, a name Dr. Berlin thinks poorly chosen, because they may not have true pathology in their nerves all the time.

The Kresge Laboratory is a tertiary care facility, which does not screen, nor run screening programs. The goals of the facility are: to recognize and support parents feelings and uncertainties toward deafness; help parents raise a literate taxpayer; be a messenger and source for guidance for parents; and help parents network and meet other parents and other deaf people.

These children often have in their histories:

"mild jaundice at birth, often considered sub-clinical;

"other blood dyscrasias, which necessitate transfusions;

"no apparent reason for deafness and no high-risk markers;

"siblings with "deafness", some of whom have had cochlear implants*;

"passing an otoacoustic emissions screen and appearing to be deaf later by ABR.

*These children have normal outer hair cells, so it would seem counter-intuitive that they should have cochlear implants, yet for parents who want oral/auditory children, that is probably the only way for them to acquire auditory skills.

Some of these children actually get better and seem "cured of their sensory deafness", some get worse and lose their normal hair cell function, and some stay the same. Still others develop peripheral neuropathies, e.g., CMT or similar demyelinating diseases.

Dr. Berlin related an example of a genetic + environmental case, in which a pair of identical twins, both had hyperbilirubinemia. One child has gradually gotten worse and lost his cochlear microphonic, the electrical response of the hair cells, and has also lost his otoacoustic emissions. The other child retains his otoacoustic emissions, and shows a nearly normal audiogram in developing speech and language. Neither child has an auditory brain stem response that can be measured.

In another example, a child was raised in cultural deafness. He gradually became more responsive to sound, and ultimately, by age 12, shows a normal audiogram. He cannot get services in the schools, because he is no longer deaf, and yet, he still does not have an auditory brain stem response. He does have normal otoacoustic emissions. The simplest screen for this condition is otoacoustic emissions, plus middle ear muscle reflexes. An ABR is not critical.

How can what's happening in these children's auditory systems be followed?

"Some patients get better, and some others show progressive loss of cochlear microphonic and emissions.

"None seem to do well with hearing aids in terms of learning language auditorally.

"Some show remarkable progress after cochlear implantation, becoming quite auditory and verbal, if progress here is met while developing auditory/oral systems.

"Some develop peripheral neuropathies in other systems.

This is sometimes a genetic disorder -- there are entire families that have been recorded. Putative genes have been localized for two types of auditory neuropathy. There may actually be a genetic hypersensitivity to clinically low levels of bilirubin in people who recover.

(Dr. Berlin played a recording of what speech sounds like to these children.) This condition does not respond to ordinary amplification of hearing aids. These may be the children forced to wear hearing aids, when no auditory device can help. The question is, "why does a cochlear implant help them?" and the answer is, "we aren't sure". Dr. Berlin played a recording of how "normal" hearing impairment sounds. (Animations of cochlear functions are available at: www.neurophys.wisc.edu.)

The ABR is not a direct hearing test. If a screening system purports to measure hearing, one does not do that with and ABR. It is best understood as a test of Neural Synchrony in response to a brief acoustic stimulus. Patients can have a totally absent ABR, but have some portions of the inner ear that are normal. Their voluntary behavioral pure tone audiograms can range anywhere from normal to absent, depending upon their ability to resolve brief signals in noise.

Similarly, Otoacoustic Emissions are not direct tests of hearing, but they are sounds which come from the outer hair cells of the cochlea. They are not measurable unless there are normal:

"Endocochlear potentials (the cochlear battery) and stria vascularis (the probable source of homeostasis for the cochlea).

"Outer hair cells and tectorial membrane whose interactive motion combine to make the sounds themselves.

"Middle ear structures to allow the sound out of the ear.

We can now split the organ of Corti into three parts for testing.

"The presence of Normal Otoacoustic Emissions shows that the outer hair cells and the cochlear battery, as well as the tectorial membrane and middle ear are working normally.

"The presence of Cochlear Microphonics reflect the presence of at least outer, and perhaps, some inner hair cells, and confirms their electrical integrity.

"Wave I of the ABR reflects neural synchrony between the inner hair cell and its neural elements.

In contrast, the kinds of patients picked up more easily, have hair cell deafness in the organ of Corti, and they show little or no ABRs, no emissions, and they have loss of inner and outer hair cells.

It is important that we see the difference between a normal latency intensity function. Dr. Berlin shows a graph of an inverting cochlear microphonic, deliberately done to show CM separates cochlear microphonic from the action potential. Dr. Berlin showed another graph of an ABR that was misread as coming from someone with normal hearing. It shows a five-wave complex. It was then misread as coming from someone with a central hearing loss. In fact, it is a cochlear microphonic. When you invert the polarity of the click, you get a complete inversion. This is the only way we know to easily separate out auditory neuropathy patients from true synchronous discharge patients. The cochlear microphonic disappears over time, in some of these patients.

Properly classifying and managing children who fail ABR, but pass Otoacoustic Emission Screenings is crucial. These children have been labeled "auditory neuropathy patients", but may have either axonal, synaptic, or even pre-synaptic disorders. Therefore, the following clinical suggestions were offered.

Clinical suggestions for preventing screening errors:

"Automated ABRs should use alternating polarity or MLS types of stimuli to insure that AN patients are referred for further testing. If the old ABR systems are used, and only use one polarity click, these children will pass both ABR and otoacoustic emissions.

"However, the final diagnostic ABR test should use one positive and one negative polarity transient.

"Children who fail ABRs should also have typanometry, reflexes, and otoacoustic emissions diagnostic tests.

Clinical suggestions to aid in identification in older children:

"Do emissions test at least once on every new patient, especially before hearing aids, cochlear implants, and general diagnostic intake. We have seen normal emissions in long-standing "deaf ears" and in unilateral hearing losses.

"Do ABRs with one positive and one negative polarity click to separate CM from AP.

"Do MEMR on every patient at least once. A paradoxically absent MEMR is neither normal nor always due to absent stapedius muscle. A non-acoustic reflex test will help rule out the latter.

The Middle Ear Muscle Reflex has been ignored for too long. None of our auditory neuropathy patients show normal MEMR, and therefore it becomes a very good first line screening test.

The Kresge Laboratory is currently screening hundreds of children in schools for the deaf, and has already found 10 of these children, out of 500. There were likely many more who began their lives with normal emissions, but lost them for one reason or another, including the natural progression of the condition. The incidence and prevalence of auditory neuropathy is hard to determine, because once they become three to five years old, the emissions often disappear in these patients. Also, the forced use of hearing aids, which destroys Otoacoustic emissions, contributes to this condition as well. It is important to identify these patients, because they can't learn by auditory/oral methods. When families insist on hearing aids, we recommend that hearing aids be tried in only one ear, so as not to damage the total hearing of the child. If the otoacoustic emissions disappear, they will disappear in both ears, and we are not responsible for making the condition worse.

The Laboratory is consulting on 200 such children in the U.S. and Australia. There are many more of these cases than there are of acoustic tumors. Some children with genetic histories have yet to show spontaneous improvement, whereas children with blood dyscrasias often improve their speech awareness and language abilities, but their ABRs often remain absent.

Dr. Berlin thanked the parents who have stayed in touch with the researchers for their continuing assistance.

"Concern was expressed about misinformation in these cases. Some cases may result in withholding appropriate services. All of those 10 cases have electrically [inaudible] ABR, because we do not implant any ear before a pre-operative electrically [inaudible] ABR test. Dr. Berlin replied that he understood these concerns. He stated that his group had a particular bias toward getting the family into acute speech and acute language, because it is English. Because it is English, if the children grow into normal sensitivity, they are already in an English mode as a primary language. Then, if they get worse, acute language does not compete with sign language -- they are mutually supportive. So by placing the parents "in the middle of the chess board", they can go with English, American Sign Language, etc. The only service Dr. Berlin said he would recommend withholding is bilateral hearing aids. "I recommend that if a family wants hearing aids, they try only one. They should then watch the child's progress over 12-14 months and if they do not see direct acceleration of auditory sensitivity and awareness, then we go back with the parents and negotiate what they would like." Dr. Berlin related the story of how they came to use implants in these type of cases. There were two children in the family who went to an oral school for the deaf, despite the advise that it would probably not do much good. The children were raised in acute speech and did extremely well in language acquisition. When they went to the school, they had the best language comprehension in the school, but the worst speech production. The principal agreed to put a hearing aid on one ear and monitor the emissions in both ears, finding that the emission held up in the ear without the aid, and disappeared in the ear with the aid. The parents finally implanted the oldest of the two children, who did so well, that the second child was quickly implanted. "We would not have recommended an implant for that child based on our understanding of physiology. But what was lacking was neural synchrony, and the Mom directing the show, picked that up and taught us all a lesson. We now have 12 patients we are following."

"Dr. Berlin was asked whether he thought the 12 patients would develop Charcot-Marie-Tooth disease? The CMT patients acquire their deafness much later in life. They develop normally, but their demyelinating disease seems to strike their auditory nerve for reasons not understood. He said he had not seen any of his newborn patients develop neurologic disease later on, if their history is genetic.

Return to Top

3) Connexin-26 Gene Mutation Detection Using Guthrie Cards from Deaf Children Identified Through the Rhode Island Audiology Screening - Jeffrey Milunsky, M.D., Clinical Genetics Center for Human Genetics, Boston University School of Medicine.

Dr. Milunsky began his presentation with an update on the preliminary data from the pilot Rhode Island study he is conducting.

There are an estimated 80,00-160,000 profoundly deaf children in the U.S.. The estimates vary widely, but with pre-lingual sensorineural deafness the estimates are between 2-3/1000. This is approximately 4% of people under 45 years of age. At least half of all deafness is inherited, with some studies showing as much as 62% being inherited. Of these inherited hearing losses, 70% ares non-syndromic, and 30% are associated with additional abnormalities, such as Waardenburg's syndrome, Alport's disease, Usher's, etc.

Dr. Milunsky established a deafness center at Boston to try to provide more information to the hard of hearing community, and to make a more formal genetic assessment of people within the community to search for any abnormalities associated with syndromes. The majority of the clinics patients are non-Caucasian. Most of the research on Connexin-26 has been in Caucasians and, therefore, the ramifications of its effects on minorities is not well understood.

There are more than 300 syndromes associated with neurosensory and conductive hearing loss. There are mutations at greater than 40 loci causing non-syndromic deafness. DFNA refers to the dominant loci, with 25 loci identified. DFNB recessive loci, also around 25 identified, and DFN refers to the x-linked loci, with around five identified.

The inheritance of pre-lingual non-syndromic hearing impairment, the majority of which is autosomal receptive, is dominant in 25%, has x-links in 1.5%, and mitochondrial disorders with sensitivity to ototoxic medications, and diabetes also play a role.

Both the NIH and the Joint Committee on Infant Hearing Screening are recommending universal newborn hearing screening. In the past, however, studies looked only at high-risk infants. It is now clear that 50% of children with hearing loss would be missed if only high-risk children are screened. Of the children not screened at birth, it takes 2.5 years to diagnose a hearing problem. Many studies have shown the benefits of early diagnosis, such as the use of hearing aids before six months to achieve age appropriate language development.

Rhode Island established universal screening in 1993. There are approximately 14,000 births per year, with 99.9% of newborns screened. Two percent of the babies fail the initial screening, and 90% return for repeat screening. In the last five years, 32 children per year have been identified with unilateral or bilateral deafness. Fifteen to 18 of the children have bilateral deafness and have been fitted with amplification by two months of age.

Dr. Milunsky reviewed the Connexin-26 gene, saying it was located on Chromosome 13, and produces a gap junction protein that allows cell to cell diffusion of small inorganic molecules, including potassium, involved in auditory transduction. The gap junctions are found between the outer cells and the supporting cells of the inner ear.

Dr. Milunsky then discussed the functions of the molecular laboratory. The Connexin-26 gene is a simple gene that has only two exons, only one of which has been shown to produce the protein. The one coty exon has one very common mutation, 35delG. Literature now suggests that up to 90% of all mutations in the Connexin gene maybe this 35delG. The deletion of the single G, causes a frame shift mutation, leading to a premature termination in an abnormal protein.

This common mutation has been looked at extensively in the Caucasian population. In the Ashkenazi Jewish population, a second mutation, 167delT, is also common. The carrier frequency has been approximated at 1/28. All the data confirms that the 35delG mutation is the most common mutation causing sensorineural deafness, with a carrier frequency of 1/31 in Caucasians (mainly from Mediterranean data). This would account for 10% of all childhood hearing loss. In comparison with syndromic hearing loss, Waardenburg's accounts for 3% of hearing loss.

The objectives of the Rhode Island pilot study were:

"To determine the efficacy and utility of Connexin-26 analysis from DNA isolated from the Guthrie Newborn Screening Cards;

"To define a major cause of the pre-lingual non-syndromic deafness in those requiring hearing aids (a sub-population of those requiring hearing aids);

"To identify families with an affected child in order to provide accurate genetic counseling, prenatal diagnosis for future pregnancies if wanted, and protection of carriers in the family.

A very rigorous informed consent process was conducted with all of the parents. The study went back five years, and genetic counseling was provided.

Preliminary data identified 46 children as non-syndromic within the last five years, and with no other known cause of hearing loss. The first problem in the study was encountered in obtaining the Guthrie cards. Only 45 cards were available, as one had been lost from the state lab. Also, it took seven months to get the cards, which is a major impediment.

From the 45 cards, the following were identified:

"3 children as 35delG homozygotes;

"3 as 35delG heterozygotes*;

"34 tested negative for the 35delG mutation.

"2 cards contained insufficient DNA for sampling;

"3 other cards were too contaminated to analyze

* Two of the three were Ashkenazi Jews, and further tests are being conducted on them.

Results indicated that 15% of the individuals have at least one Connexin-26 mutation. But, 11% of the Guthrie cards did not work. Cheek swab data is being obtained from the five children with unusable Guthrie cards.

The benefits of Connexin-26 gene mutation detection are:

"there is a precise diagnosis of those affected, defining the cause of their hearing impairment;

"allows for the provision of accurate genetic counseling;

"provides prenatal diagnosis for future children;

"provides carrier determination.

Of course, prenatal diagnosis is a difficult issue. Dr. Milunsky worked with one deaf couple, of which one had the Connexin-26 mutation in the homozygous state, the other had the Connexin-26 mutation in the heterozygous state, who wanted deaf children. They are still discussing whether they would want prenatal diagnosis to identify and keep a deaf fetus. He also related a story of a couple with achondroplasia, who wanted children with the condition, and, therefore, aborted a normal fetus.

"Dr. Milunsky said he did not receive many requests for prenatal testing.

"When asked whether he advocated the use of Guthrie cards or cheek swabs, Dr. Milunsky replied cheek swabs. He said the best approach was to identify the child after birth and perform a follow-up screen. Then in one to two months, obtain a cheek swab (with consent). This procedure reduces public health costs.

Return to Top

4) Genetic Studies of Hearing Impairment in Special Education Groups - Bronya Keats, Ph.D., Center for Molecular and Human Genetics, Louisiana State University Medical Center.

Dr. Keats discussed a study just begun at the LSU Medical Center, on the genetics of hearing impairment in special education groups, i.e., schools for the deaf.

The goals of the study are:

"To estimate the frequency of 35delG homozygotes in schools for the deaf.

"To estimate the frequency of the 35delG allele in different ethnic groups.

From the schools for the deaf, 563 cases have been identified as follows: Florida - 31; Upstate New York - 132; Long Island, New York - 96; and Louisiana - 304.

First the principal is contacted to see whether there is interest in participating in the study. If there is, then information is sent which includes consent forms, and a questionnaire for the parents. If the parents allow their child to participate, then the school is visited.

Data collected includes:

"audiograms, otoacoustic emissions

"age of onset of hearing impairment

"family history of hearing impairment

"other anomalies, e.g., vision, kidney, thyroid, heart

"environmental insults, e.g., meningitis, CMV, ototoxic drugs

"ethnic background and gender

"management - hearing aid, cochlear implant

"cheek swab

Of the 563 children identified, their deafness was attributed as follows:

Congenital deafness....147

infection....................13

unknown genetic........134

35delG.......................20

35delG/?.....................16

other heterozygote.......14

other homozygote..........8

Samples have been collected from all over the world, including: Africa; South America; Australia; North America (European-American, French Acadian, Mvskoke); and Asia. Many native groups have been included. The study group has plans to sequence the Connexin-26 gene, in order to estimate the frequency of the various alleles in each of these population groups.

Preliminary conclusions from the study of schools for the deaf include:

"The frequency of 35delG homozygotes among children with congenital deafness is about 15%.

"GJB2 mutations may explain as much a 43% of congenital deafness.

There are 400 more samples to analyze, for a total sample size of over 1,000.

"Dr. Keats was asked how informed consent was obtained from other populations with different views on such things. She replied that all of the samples were anonymous, and that many of the samples were taken many years ago. Now there is a large group of people working on a diversity project, and they have people of that culture, who speak the indigenous language, explain the project and consent to the participants.

"Dr. Keats was asked how generalizable this would be to the public schools, and how different the phenotype was from children in public schools. Dr. Keats replied that most children in schools for the deaf were profoundly hearing impaired from birth.

"Dr. Keats was asked whether the were large enough family groups for any of these individuals who were heterozygous to one mutation, but not the other mutation, or for whom you are not certain what the other mutation might be, to demonstrate the DFNB1 linkage? She answered that they did not have large enough families.

Return to Top

5) Hearing Phenotype in DFNB1/GJB2/Connexin-26 - Edward Cohn, M.D., Boystown National Research Hospital.

Dr. Cohn's group is interested in the DFNB1 phenotype, and has studied it using several methods, such as:

" conducting retrospective hearing analyses of classical DFNB1 families;

" conducting retrospective hearing evaluations of patients with sporadic DFNB1; and

" performing a Connexin-26 evaluation of newborns failing diagnostic ABR.

Approximately 50% of hearing loss is genetic, with the majority being recessive.

DFNB1 is part of the recessive non-syndromic group. From 1993-999, the most common form of NSRHL, DFNB1, has been localized to chromosome 13 Q11-12. The GJB2 gene was found to have mutations which segregate with the hearing loss phenotype of DFNB1. This has allowed research groups to characterize the mutations in the GJB2 gene and the hearing phenotype resulting from these mutations.

A connexin is a membrane protein that can oligomerize in a hexameric structure to form a connexon. Connexeons from adjacent cells interact to form complete intercellular channels that become clustered in specialized membrane regions, the gap junction. In other words, they are the proteins that allow cells to communicate.

Dr. Cohn showed a slide of the Connexin26 protein structure, with various mutations sites placed within the structure. Six of these sites can form a connexon. Each molecule will form together in a six-sided structure to form a hemichannel. That hemichannel extends from inside the cell to extra cellularly. If on the other side of it is another connexon that can talk with it, then it allows the cells to act as an organized tissue. The hypothesis is that these connexons help the cochlea recirculate potassium.

In the classical families, classical recessive hearing loss was defined as families in which two or more siblings have hearing loss, and the parents do not have hearing loss.

In the retrospective study, three major genotypes of DFNB1 were found. The most common, 35delG homozygous form, has about one third profound hearing loss. In another third, it accounts for severe-profound hearing loss. The remaining third varies from moderate to severe, mild to moderate. Only six patients were homozygous 167, who seem to have a worse hearing loss, with five of the six in the profound or severe to profound range. The small number causes concern over the statistical reliability of the figure. Additional research will confirm or refute that number. In the compound heterozygotes, there is a mutation on both alleles, but they are different mutations. Forty percent were profoundly deaf.

Hearing Data

Five out of 22 35delG patients had progression. Progression was defined as 1db per year with a MTTA of 4, using 500,000, 2,000, and 4,000. Serial data was available on 60% of the patients.

Two-thirds of these patients are in the profound or severe-profound range, but the homozygotes 35delG mutations will have a mild-to moderate hearing loss, even though they are in the same family as those with profound hearing loss. That fact has brought up the concept of modifying genes.

The importance of these data follows:

"The hearing phenotype is defined.

"One-third to one-half of the children with DFNB1 will have profound hearing loss.

"Progression will occur in a significant number of these children. Perhaps more so in the heterozygotes.

"DFNB1 is non-syndromic. Therefore, cx26 determination becomes a critical diagnostic crossroads eliminating the need to R/O Usher syndrome, Pendred's syndrome, Jervell Lange Nielsen syndrome, and the expensive tests supporting these diagnoses, i.e., electroretinograms, opthamology consultations, vestibular function tests, CT inner ear, perchlorate washout, EKG.

If a cx26 mutation is found, then the following are not necessary:

"a CT of the inner ear;

"an ophthalmologic consultation;

"ERG (Usher syndrome);

"EKG;

"perchlorate washout.

There appears to be utility in screening newborns who fail diagnostic ABR. This allows accurate genetic counseling, more careful audiologic and educational observation to detect early those children failing to learn language with traditional amplification.

The study also looked at people with sporadic hearing loss. Having a cloned gene allows you to test individuals. The frequency of DFNB1 in individuals with sporadic hearing loss varies from 10%, reported by Lench in the UK, to 37% in the southern Italian population, reported by Estivil. Wells has reported a 31% biallelic finding in individuals with sporadic hearing loss.

Dr. Cohn provided an overview of the project.

"Twenty-five percent of hearing loss followed with serial audiograms will have DFNB1.

"The distribution of hearing loss will be similar to data previously reported with recessive families. Hearing will be variable, with one-third or more having profound hearing loss, and one-third having severe-profound hearing loss.

"Some DFNB1 children with serial audiograms will show significant progression of hearing loss.

"Progression will be greater in those with compound heterozygous mutation and those with asymmetry as defined by Cohn et. al., in the March 1999 issue of Pediatrics.

"Heterozygote effects will be seen on sophisticated OAE algorithms.

"Determine the frequency of DFNB1 in patients with sporadic hearing loss.

"Determine the spectrum of hearing loss, asymmetry and progression over time to help inform physicians, audiologists, educators, and genetic counselors about the natural history of DFNB1 in sporadic cases.

"Study children who have had serial audiograms at BTNRH.

"Begin with the audiology advocate list of children with sensorineural hearing loss.

"Recruit through parents of children ages => 10, or adults who have been followed by an audiology advocate with serial audiograms. Recruit all persons with hearing loss who are not multiply handicapped.

"Review the serial audiograms of those with cx26 mutation to determine distribution of hearing loss severity, degree of progression of hearing loss, and percentage of hearing loss secondary to cx26 mutation.

"Analyze medical and hearing data on patients with biallelic mutation.

"Convert study to prospective study at time of data review to continue observations.

"Outline background of NSRHL, DFNB1/GJB2/connexin26 relevant history and goals of research project and recruitment techniques.

"Complete standard IRB application including creation of consent form, which will allow blood draw, DNA extraction, test for mutations causing hearing loss, and review of medical records of those who are positive for cx26 mutations.

After approval was granted, Dr. Cohn's group obtained a list of children over 10 years of age who were followed by audiology advocates. They also obtained names and addresses of parents and those who are adults. Recruitment letters were mailed that included informed consent forms, and a phone number potential participants could call for additional information. Once people agreed to participate, an appointment was made to draw a blood sample. DNA was extracted from the blood sample and analyzed for cx26 mutation. Audiometric data of those with DFNB1 was reviewed as well. The data was then placed on a secure database, and analyzed for degree of hearing loss, progression, asymmetry, and success with amplification.

Preliminary results include:

"31 sporadic patients

"3 biallelic DFNB1 (35delG) -10%

"2 heterozygous mutations -6%

"Hearing loss

"1 profound symmetrical PTA 101

"1 moderate-severe asymmetric PTA 78

"1 moderate-severe symmetric PTA 62 (slightly progressive .9db/year/1.5 year=7.8 db/year)

Newborn Study

"20% of newborns with hearing loss will have DFNB1 mutation in the cx26 gene.

"One-third or more of these patients will have profound hearing loss and another third will have severe-profound hearing loss.

"Cx26 will prove a valuable marker to help busy newborn screening programs prioritize follow-up.

"Determine the frequency of DFNB1 in newborns who fail diagnostic ABR.

"Determine the spectrum of hearing loss, asymmetry and progression over time to establish the natural history of DFNB1 in the newborn infant population. This will help inform physicians, audiologists, educators, and genetic counselors about the natural history of DFNB1 in sporadic newborn cases.

"Study children who have failed diagnostic ABR.

"Inform parents of children who fail the diagnostic ABR that 10-37% of non-syndromic recessive hearing loss is caused by cx26 mutation and that this test can be performed on a swab of the baby's cheek. A diagnosis of DFNB1 is made if biallelic cx26 mutations are found.

"Distribute brochure regarding cx26 hearing loss to parents whose child has failed the diagnostic ABR (distributed by the audiologist).

"Obtain informed consent for buccal smear.

"Obtain buccal smear and send to BTNRH in kit provided.

"Extract DNA.

"Inform genetic counselor of results - biallelic mutation, heterozygous mutation, no mutation.

"8 samples for newborns failing ABR;

"1 positive for homozygous 35delG

"7/8 or 87.5% successful DNA extraction from buccal smears.

It is important to continue to understand and define the natural history of DFNB1, one of the most common causes of hearing loss in childhood. We also need to define what happens to these children as they grow older, and what happens to them as they enter late adulthood.

"Dr. Cohn said only one of the sporadic children studied showed distibular symptoms, and he would have to review the findings.

"A positive test for cx26 can mean many things in terms of hearing phenotype. Fifty percent of the families with one child with cx26 mutation had similar hearing losses in the family, and 50% had diverse hearing losses within the family.

Return to Top

6) Genetics of hearing loss: how common is it? - Arti Pandya, M.D., M.B.A., Department of Human Genetics and Pediatrics, Medical College of Virginia.

Hearing loss is an economically and socially important cause of human morbidity. Although this has not changed within the past decade, the rapid and dramatic advances in this field have changed the perception of the magnitude of this problem, recently. The purpose of this conference is to develop and explore strategies for population-based study of genetics of congenital hearing impairment, and to recognize the ethical implications this type of research could have. The question that needs to be addressed is, "how common is genetic hearing loss"?

Existing estimates are based on segregation analysis on very large samples of family data, and suggests that 50% of profound deafness is genetic in etiology, with 30% being syndromic, and 60-70% being from non-syndromic causes. Also, 70-80% of non-syndromic deafness is accounted for by recessive transmission, 15-20% by dominant transmission, with 1-20% being mitochondrial, depending on the ethnic background and the part of the world from which the person comes.

Several genes for syndromic and non-syndromic deafness have been characterized that make up the genetic architecture of deafness. The availability of molecular information at a rapid pace is exciting, but also poses several questions, such as: "how prevalent is each of these conditions in the population", "can any of these disorders be prevented, if there is a need for it", and, "if so, what are the ethical implications of this approach"?

The role of mitochondria in deafness was limited to its association with systemic neuromuscular disorders. However, more recently two major mitochondrial changes have been recognized, one in the 12S ribosomal gene, which is a single point mutation, and a single point mutation in a precursor for the transfer RNA for hearing, which result in non-syndromic deafness in several ethnic populations. There is also a resurgence in diabetes mellitus and deafness, because of a single point mutation, and a fourth Usher syndrome has been identified. Aminoglycoside ototoxicity is a very common cause of preventable, but irreversible, hearing loss, which is often characterized by matrilineal inheritance. Usually, there is a history of exposure to streptomycin, causing a homoplastic change in ribosyl RNA gene in the mitochondria. More recently, investigators have noticed sporadic cases with aminoglycoside ototoxicity without any other matrilineal family history.

As for DFNB1, there are 30 different mutations that have been identified, and there is a prevalence of ethnic diversity with respect to mutation heterogeneity, as seen at the 167delT mutation in Ashkenazi Jewish population.

Knowledge that Connexin26 mutations account for a large proportion of genetic deafness has lead to several studies. The genotype/phenotype correlation suggests that there is immense phenotypic heterogeneity in this group. The studies have also shown that homozygotes for the delG mutation, although are more likely to have severe-profound hearing loss, have immense variability, including people with mild hearing loss that may not usually be picked up. Progression in this category is still questionable. Homozygotes for the 167delT mutation are more likely to have severe-profound deafness, but the reason is still elusive. More interestingly, it is the compound heterozygotes who had severe-profound deafness with some degree of progression that was not noticed in the homozygote delG mutations. There are several studies ongoing, and there are several questions yet to be answered.

In order to enhance the understanding of the phenotypic variability relative to phenotype/genotype correlation, but to estimate the total frequency of genetic deafness in the population, Dr. Pandya's group is proposing a prospective nationwide study which would establish a DNA repository of samples to be obtained from a large, annual, nationwide, population-based survey conducted by the research institute at Gallaudet University.

The target is to obtain samples from 2,000 probands from both multiplex and simplex families. The survey will also use a sequential screening strategy in which all the samples will first be screened for testing at the cx26 locus, and then at the two mitochondrial mutations. A screening of negative samples would then follow for mutations in other known or potential genes. A subset with syndromic deafness would also be screened for genetic mutations. The process of this endeavor is too large to be done in a single lab, and so anonymous samples would be available to external investigators for testing of candidate genes. The power of this method is immense, because it can be used to map loci identified in mice, and obviate the need for collecting large families and doing linkage analysis. Once the genetic changes have been identified, it would be interesting to see the gene/gene interactions.

Dr. Pandya presented preliminary results of a cx26 screening of 259 subjects. Two hundred of the samples were sequenced, and 59 were analyzed only for 35delG mutation. Connexin deficit accounted for 25% of the U.S. probands, and at least 77% were identified as 35delG mutation. There were five previously reported mutations in the samples, and two new mutations. One was an R32C, which occurs in combination with an already described deletion. The most common pathologic allele in the Mongolian samples was a novel double mutation that occurs in sets, one of this the ralene [?] to isobucene [?] polymorphism described at codon 27 along with the glycine to glutamic acid change at position 114. We did not detect a single delG homozygote in the Mongolian population, which is also true when screening samples from China and Japan. This suggests that cx26 mutations, especially the delG mutation, although common in the Caucasian population, is not the explanation in some of the Asian populations.

Dr. Pandya reviewed the various methods of cx26 testing. Since not all affected individuals are homozygotes for the delG 30 or 35 mutation, an extensive analysis of cx26 will be needed in a high proportion of cases.

Dr. Pandya presented two pedigrees to illustrate the advantages and disadvantages of the testing she is planning to use for general populations. She presented two families she prefers to call simplex to sporadic, because there is no family history of deafness. One child was six months old, the other 18 months old. One baby was identified at newborn screening, the other was not detected until 18 months, although he had profound hearing loss. Both received cochlear implants. Both children were delG 30 homozygous, and the parents were carriers. In one case, the mother decided not to have prenatal testing, but to wait until the baby was born and use the cord blood for testing. Dr. Pandya raised the question of whether these children were really sporadic cases, or are the simplex cases being reclassified as chance, isolated familial cases. Once you call them sporadic, you are saying they are not genetic, and that is confusing to consumers.

Another case consists of two families with a deaf by deaf non-complementary mating, which suggests that all children would have hearing loss, and there would be a mutation in the cx26 gene. However, when the first family was screened a delG 30 mutation wasfound on one allele, but even after sequencing, the mutation could not be detected on the second allele. However, there was an unrelated deaf individual with a deaf child, which indicates this had to be a cx mutation, because of the prevalence. The second case was a classic delG 30 homozygous family, deaf by deaf non-complimentary mating. But if only screening for the delG mutation had been used, appropriate answers would not have been available.

The last case consists of two children, ages 10 and 8, with severe hearing loss. However, there was a history of three generations of unilateral hearing loss on the paternal side, with some suggestion of noise exposure and gun use. Both children were homozygous of the delG mutation, and the parents were heterozygous. We are not sure whether there are two independent etiologies.

Rationale for genetic testing:

"identifies the genetic nature of sporadic cases;

"helps in offering counseling;

"identifies candidates for early cochlear implants (in Europe they start implants as early as 12 months of age);

"can be cost efficient to devise a protocol starting with the simplest and easiest tests;

"helps provide an assessment of the progressive phenotype;

"avoiding aminoglycosides after obtaining the appropriate family history, and doing testing would have an impact.

"Dr. Pandya was asked whether the sole purpose of genetic testing was to identify candidates for cochlear implantation, and whether other options were offered. Dr. Pandya replied that cochlear implants were not the only option, but if they were the desired option, then the earlier they were identified and the implant took place, the better the outcome.

"Dr. Pandya was asked why Mongolia was chosen and how the IRB could be understood by a different culture. She replied that Mongolia was a happen stance, in that one of the researchers was in Mongolia for a different reason, and upon talking with researchers there, found a great enthusiasm for participating in this study. The gathering of data in Mongolia is handled by a local subcontractor. The IRB and informed consent rules were explained in full and that contractor then explains them to the subjects. Since the contractor is of that culture and speaks that language, we have to assume he knows that population better than do we.

"It was pointed out that the reason for calling the cases sporadic rather than simplex, was so that everyone would be discussing the same thing.

Return to Top

Dr. Khoury began the discussion by saying CDC needs some strategies for population-based research which it and its partners can support financially and otherwise. Dr. Khoury said he wanted to highlight three things that had come up in the discussion and presentations.

1)When talking about population-based, rather than clinically based strategies, the question of what to do with the gene after it's been discovered, arises.

2)The distinction between research and testing must be kept in mind. The point of a population-based research agenda is to gather information that can to be used for testing purposes.

3)When referring to causes of any condition, genetic and non-genetic components get separated. Increasingly, genetic means a few single gene conditions with high penetrance. Non-genetic components include infectious and environmental things, but the genetic components of those things are not considered. But, the entire picture is genetic and environmental together.

Also, whatever is done, must be done under an acceptable, ethical agenda. That may be harder to adhere to in a large study.

Dr. Burke said that what is being learned is that there is not a clear one-to-one relationship between a particular genotype and a particular phenotype. There is a complex relationship between genetics contributing to hearing impairment and the actual outcome of hearing impairment. It also seems clear that there are a number of environmental factors, and much is to be learned about the way in which different genes contributing to hearing impairment interact with one another, and how environmental factors interact with genotypes to produce a certain phenotype. That is the essential question that needs to be addressed in research. As research goals are considered, there needs to be clarity about two different potential goals that are not mutually exclusive. The first is, research should go forward to understand how hearing impaired phenotypes come to be -- to do the types of population-based research that allow a rigorous look at that interaction, and understand the pathophysiology. Also, what options can be offered to reduce the likelihood of a hearing impaired phenotype.

Another goal is whether the question really being asked is, "how can a genetic test be used as a tool"? Understanding how such tests are used as a clinical tool, really can't be addressed without addressing that underlying issue of the correlation between genotype and phenotype. Once that is answered, research needs to be conducted to look whether a genetic test can be used to determine management of the condition. An effectiveness research agenda needs to be pursued.

A third issue is, "what should the role be for mutations of this kind, as tools for testing and diagnosis".

Dr. Arnos said CDC should be commended for including members of the deaf community. This is the first time that has happened at a scientific meeting, and it needs to be done more.

Dr. Morton said the genotype/phenotype issue is one that is not new to the genetics community. Chromosome abnormalities are often detected in prenatal diagnoses, for which the parents cannot be given an accurate prognosis, which is what they want. One of the goals should be to understand the science behind the genotype, and to give all sides of the information as accurately as possible.

Dr. Morell emphasized genotypes and phenotypes, but reminded everyone of another aspect of the phenotype question. Almost all the genes found so far, are expressed in other tissue as well. A better understanding is needed of what the effects of those mutations are in other tissue in order to provide better health care to that phenotype.

Dr. Boyle asked the group to discuss the prevalence rates of the genotype in the population, including what needs to be answered.

"Longitudinal studies of cx26 progression are needed, including what happens throughout life, and what are the environmental factors.

"The existing databases containing serial audiograms need to be preserved, retrospective to DFNB1, 2, 3, 4, 45.

"A better understanding of hearing impairment is needed to prevent and abate its occurrence. An agenda has to be identified that deals with providing optimal opportunities and training for hearing impaired people.

"Consented prospective studies will tell more than retrospective studies, although retrospective is usually more cost effective in generating hypotheses.

"We need to design studies that will give us the best of prospective and retrospective methods. Much can be learned from case/control studies. Penetrance estimates can be developed from 100 case/controls. We need to capitalize on all the existing programs, and see how we can add on to them to get the information needed.

Inclusion of the deaf in IRBs and deaf research.

"It was asked whether institutions would be open to including deaf people, or their families, on IRBs, to ensure that research has the ethical awareness of it's real life impact on the deaf? Responses:

"Brigham and Women's Hospital already has lay persons on the IRB. It is a tremendous time commitment, and meets on many different topics. Anyone volunteering would be looked on favorably.

"University of Washington has lay members, as well, and would welcome the deaf.

"The University of Michigan sees a problem. By including the deaf on IRBs concerning the deaf, the door is opened to including people with a specific condition on each IRB that concerns that condition. An IRB must be representative of the general population.

"Emory always includes public participants, and would be amenable to having someone from whatever the community might be. Geneticists are already working with many populations and groups. Communities have to come and be advocates.

"A bigger problem arises, in that the deaf community is not monolithic. There are varying opinions. There could be a covert stacking of the committee to stop important research that particular group doesn't like.

"Some states are using IRBs to oversee newborn screening programs because IRBs have a history of well thought through, inclusive behavior.

"The deaf community is fractured, how can more support from the deaf community be encouraged?

"Programs need to be established to train and support deaf scientists to work on deaf research. The problem is finding the numbers of deaf people to go to medical school.

"Parents are usually turned-off by a strong bias towards one method of management over another. A CDC bulletin board is needed that explores all types of management.

"The deaf community is concerned about involvement. Gallaudet wants to set up a certification training program to talk with people about deaf issues. Research has a way of becoming policy, e.g., in vitro fertilization is happening now and is out of control. If we don't want that to happen with deaf issues, we need to set up an advisory board.

"In order to do proper population studies, large numbers of deaf people are needed to get statistical power. This gives the deaf the opportunity to participate in the oversight committees, and to provide leadership and input into what the deaf community will embrace.

"Meetings such as this one are a good way to influence the research agenda.

"There needs to be an advisory board for genetic research on deafness. There is no grassroots effort by researchers. People can't attend these meetings if they don't know they are occurring. The average person doesn't read scientific journals, in which these meetings are announced.

"Some members of the deaf community see attempts to prevent and abate deafness as genocide. They say they are a unique community with its own language and culture, and by trying to ensure that future children not be born deaf, or by providing cochlear implants for those who are, researchers/scientists are essentially ensuring the death of the deaf community.

Other comments include:

"One participant stated that most people in the room have a bias that hearing is better. Deaf people were referred to as having limited communication options, yet the deaf are fluent in two languages. Also, cochlear implants are not the answer. Parents are not given true informed consent when they are told about cochlear implants. The medical route is too strongly emphasized. The role of public health is to improve the quality of life through better access to communication. The hearing should learn sign language.

"Informed consent language should be developed that satisfies members of the various communities.

"There was disagreement on whether audiologists were the gatekeepers, and whether they should be trained to offer genetic counseling. One person thought they should, and another said they should not. That as newborn screening changes, a medical home for every child is being pushed, and counseling should occur at the pediatrician's. It was agreed that whomever the gatekeepers are, they should have wide choices available. A multidisciplinary team, representing many fields, should present all options to the parents. Without this approach, you do not have informed consent.

"Research and testing are different.

"It was asked whether any state laws mandate what takes place after identification of the child. The reply was that legislation varies, for example, Missouri pays for hearing aids. No one knew of a study of state requirements.

"It was pointed out that not every child would be screened. Often a pediatrician is blamed when a child is not diagnosed early on, but if the physician does not know what to look for, or to whom and when to refer a child, it is not really the physician's fault. Professional education is needed to remedy this situation.

"In Colorado, physicians have created an advisory board for screening.

"There is a gap in what scientists know and what the consumer knows. Much of what is done is misinterpreted by the public. If would help if a sample study could be shown, and a family followed through the process, to help the public understand what they are reading about research. Dr. Khoury commented that the speaker was giving too much credit to scientists.

"There are many reasons for early diagnosis of genetic causes of hearing impairment. Some genetic mutations affect more than just hearing, they may affect the heart or other organs, so early detection and treatment benefits the child's entire body.

"We must keep in mind that most congenitally deaf children are born to hearing parents. Their wishes and how they deal with the child's deafness are other issues to take into consideration.

Dr. Boyle thanked everyone for their participation. She said CDC needed to take a few steps back and think about the research agenda for the hearing impaired, and how to involve stakeholders. She said a report of the meeting would be available within a few weeks.

The meeting was adjourned.