Predictive Modeling for EVA Syndrome

UH ENT team creates a new tool to calculate the risk and rate of hearing loss in patients with enlarged vestibular aqueduct syndrome

Innovations in Ear, Nose & Throat | Summer 2021

Nauman Manzoor, MD

Todd Otteson, MD

Enlarged vestibular aqueduct (EVA) syndrome is a rare malformation of the inner ear that is commonly linked to sensorineural hearing loss in children. Because the disease progresses differently for each patient, it’s incredibly difficult for otolaryngologists and audiologists to anticipate outcomes to provide proactive treatment.

“Some patients can have stable hearing, while others fluctuate or lose hearing, and recovery from such hearing loss is unpredictable” says Nauman Manzoor, MD, an otologist-neurotologist at University Hospitals Ear, Nose & Throat Institute. “A high number of patients eventually decline to the point that they need implantable devices such as a cochlear implant for hearing rehabilitation.”

But the latest innovations may soon eliminate a lot of that uncertainty, Dr. Manzoor says. He is part of a team at UH ENT Institute that’s pioneering a new risk prediction tool to help providers better predict an EVA patient’s risk of hearing loss.

“This is a disorder where a lot of what we do is reactive — we intervene when someone loses hearing,” he says. “We can now provide exact probability ranges for hearing loss at one year, three years and even a decade later.”

Putting data to work

The UH team developed their EVA nomogram — a risk prediction tool that uses data modeling to approximate the likelihood of a clinical event — by analyzing existing patient data, including imaging and audiometric factors. Fortunately, they didn’t have to go far to find what they needed.

In 2016, faculty at UH Rainbow Babies & Children’s Hospital in Cleveland initiated a project to address the lack of EVA patient data available to providers. That team, led by Todd Otteson, MD, MPH, division chief of pediatric otolaryngology at UH Rainbow Babies & Children's Hospitals, designed and spearheaded a national EVA patient registry based at UH.

UH faculty and residents, including Dr. Manzoor, quickly began using the registry data to publish new findings about the disorder. This year, he had another idea: What if the team could leverage all that registry data — drawn from more than 150 patients worldwide — to develop a prognostic tool to determine which EVA patients are at the greatest risk of hearing loss?

“We’ve never had a tool like this before,” he explains. “I’ve seen risk prediction used in other areas, but we don’t have a similar example in EVA. I knew that we had this data, but it needed to be looked at from a different perspective.”

Dr. Manzoor partnered with Dr. Otteson and team of researchers to build the tool, using the national registry data and referencing previously published studies that show how different factors correlate with EVA hearing loss. They also enlisted help from Case Western Reserve University School of Medicine student Nathan Farrokhian, who applied his knowledge of complex analytics and machine learning to design the nomogram where clinicians will input patient data to be analyzed.

Through their research, the team discovered three factors that were very significant in predicting risk of further hearing loss with EVA: gender, the size of the abnormality in the inner ear and the patient’s hearing level at the time of diagnosis. Previous studies had examined these factors in isolation; but there was never an effort to pull the findings together in one place, Dr. Manzoor explains.

“We didn’t have an equation to plug everything into,” he says.

Using readily available data allowed the team to complete the EVA nomogram in just three months and submit their results to a peer-reviewed journal. Now they’re working to validate their tool externally, collaborating with other institutions with higher numbers of EVA patients. Collecting more evidence will help the team roll out a more formalized version to clinics across the country, Dr. Manzoor says.

“We’re getting closer to the point where at the time of initial consultation, we can input the information into an iPad app and get a prediction of whether a particular patient is going to lose hearing or not, and at what rate,” he says.

Proactive EVA care

By successfully predicting EVA hearing loss, the UH team hopes to give ENTs and audiologists a broader perspective to improve how they approach treatment. At the top of the list? Alleviating anxiety for patients and parents.

“When a pediatric patient comes to our clinic, the parents are very anxious because their child has this rare inner ear condition,” Dr. Manzoor says. “Some of these patients totally lose hearing through their lifespan and they’ll need intervention such as cochlear implantation and hearing aid.”

Having that knowledge upfront allows care teams to proactively recommend interventions. If a patient is at high risk of hearing loss, clinicians may increase the frequency of hearing surveillance and/or treat the patient with steroids proactively to help maintain hearing. They may also recommend that the patient avoid contact sports, which can cause head trauma that may accelerate hearing loss in EVA patients.

“If you have someone who is flagged immediately for a high risk for hearing loss, you and the entire practice’s approach to that child will be different than if their hearing is likely to be stable,” Dr. Manzoor says. “We can use that data in our counseling, as well as patient and parent education.”

Just as important, clinicians can better prepare patients facing moderate to severe hearing loss by introducing hearing rehab options — including cochlear implantation and hearing aids — earlier in the treatment process.

“With the modern rehabilitative tools we have for hearing loss, there won’t be a period in these kids’ lives where they would have significant untreated hearing loss,” he says.

To reach the UH Ear, Nose & Throat Institute, call 216-868-8943.

Contributing Expert:

Nauman Manzoor, MD
Otologist-neurotologist
UH Ear, Nose & Throat Institute
Assistant Professor
Case Western Reserve University School of Medicine