Audiovestibular Consequences of Congenital Cytomegalovirus Infection: Greater Vulnerability of the Vestibular Part of the Inner Ear

Vestibular Impairment and Postural Development in Children With Bilateral Profound Hearing Loss

Importance

Children with profound hearing loss (HL) and vestibular impairment have worse cochlear implant outcomes compared with those without vestibular impairment. However, the decision for cochlear implantation is rarely based on vestibular function assessment as a complement to audiologic testing.

Objectives

To identify the prevalence of vestibular impairment according to HL origin and to assess the association between vestibular impairment and delayed posturomotor development in children with profound HL.

Design, Setting, and Participants

This cohort study was conducted in a pediatric referral center for cochlear implantation in Paris, France, using medical records data on HL origin, vestibular assessment, and ages of developmental milestone achievement. The cohort included children with profound HL (loss >90 dB HL) who completed vestibular assessment prior to cochlear implantation between January 1, 2009, and December 31, 2019. Data analyses were conducted between January and June 2023.

Main Outcomes and Measures

The primary outcome was prevalence of vestibular impairment according to HL origin. Children were classified into 3 groups according to their responses to vestibular testing: normal vestibular function (NVF), partially impaired vestibular function (PVF), and complete bilateral vestibular loss (CBVL). Generalized logit models were performed to evaluate the association between vestibular impairment and causes of HL as well as posturomotor development delay.

Results

A total of 592 children were included (308 males [52.0%]; mean [SD] age, 38 [34] months). In children with documented HL origin (n = 266), 45.1% (120) had HL with genetic origin, 50.0% of which were syndromic (mainly Usher and Waardenburg syndromes) and 50.0% were nonsyndromic (mainly associated with connexin 26). Among patients with infectious HL origin (n = 74), 70.3% (52) had cytomegalovirus (CMV) infection. Vestibular impairment was found in 44.4% (263 of 592) of the children; it was mostly symmetrical in 88.9% (526) and was CBVL in 5.7% (34) of the cases. Vestibular impairment was present in 78.3% (47) of children with genetic syndromic HL (56.7% [34] with PVF; 21.7% [13] with CBVL) and in 69.2% (36) of children with CMV infection (57.7% [30] with PVF; 11.5% [6] with CBVL). Genetic syndromic HL origin was found to be more often associated with both PVF and CBVL than other HL causes. The odds of having delays in 4 developmental milestones (head holding, sitting, standing with support, and independent walking) were higher in both PVF and CBVL (eg, head-holding odds ratios: 2.55 and 4.79) compared with NVF, and the age of achieving these milestones was higher in CBVL than PVF (eg, head holding: 7.33 vs 4.03 years; P < .001). All 4 developmental milestones were associated with the degree of vestibular impairment.

Conclusions and Relevance

This cohort study found that among children with profound HL, vestibular impairment was prevalent, varied according to HL origin, and associated with posturomotor development; while all developmental milestones were associated with vestibular impairment severity, not all HL causes were associated with vestibular impairment severity. Children with profound HL may benefit from complete vestibular assessment before cochlear implantation, which would support early and adapted management, such as physical therapy for CBVL and cochlear implantation strategy.

Vestibular Follow-up Program for Congenital Cytomegalovirus Based on 6 Years of Longitudinal Data Collection

OBJECTIVES

Congenital cytomegalovirus (cCMV), the leading nongenetic cause of pediatric sensorineural hearing loss, can also affect vestibular function. Literature findings suggest clinical presentation of vestibular loss in cCMV to be as variable as the hearing loss. Still, probably due to the considerable additional burden it entails for both patients and diagnostic centers, longitudinal vestibular follow-up in cCMV is not well-established in clinical practice. Therefore, this study aims to propose an evidence-based vestibular follow-up program with proper balance between its feasibility and sensitivity.

DESIGN

In this longitudinal cohort study, 185 cCMV-patients (mean age 3.2 years, SD 1.6 years, range 0.5-6.7 years) were included. Vestibular follow-up data were obtained through lateral video head impulse test (vHIT) and cervical vestibular evoked myogenic potential (cVEMP) evaluations around the ages of 6 months, 1 year, and 2 years. Around 3 and 4.5 years of age, data from vertical vHIT and ocular vestibular evoked myogenic potentials (oVEMP) were also collected.

RESULTS

At birth, 55.1% (102/185) of patients were asymptomatic and 44.9% (83/185) were symptomatic. The mean duration of follow-up for all patients was 20.8 (SD 16.3) months (mean number of follow-up assessments: 3.2, SD 1.5). Vestibular loss occurred at some point during follow-up in 16.8% (31/185) of all patients. Six percent (10/164) of patients with normal vestibular function at first assessment developed delayed-onset vestibular loss; 80.0% (8/10) of these within the first 2 years of life. Vestibular deterioration was reported both in patients who had been treated with postnatal antiviral therapy and untreated patients. At final evaluation, both the semicircular and the otolith system were impaired in the majority of vestibular-impaired ears (29/36, 80.6%). Dysfunctions limited to the semicircular system or the otolith system were reported in 4 (4/36, 11.1%) and 3 (3/36, 8.3%) ears, respectively. The occurrence of vestibular loss was highest in patients with first trimester seroconversion (16/59, 27.1%) or with an unknown timing of seroconversion (13/71, 18.3%), patients with sensorineural hearing loss (16/31, 51.6%), and patients with periventricular cysts on magnetic resonance imaging (MRI) (7/11, 63.6%).

CONCLUSIONS

Longitudinal vestibular follow-up, most intensively during the first 2 years of life, is recommended in cCMV-patients with vestibular risk factors (first trimester or unknown timing of seroconversion; sensorineural hearing loss; periventricular cysts on MRI). If those risk factors can be ruled out, a single evaluation early in life (around 6 months of age) might be sufficient. Both semicircular and otolith system evaluation should be part of the follow-up program, as partial losses were reported.

Is CMV PCR of inner ear fluid during cochlear implantation a way to diagnose CMV-related hearing loss?

To evaluate the diagnostic performance of cytomegalovirus (CMV) polymerase chain reaction (PCR) on inner ear fluid collected during cochlear implantation and to assess its interest in current practice. This monocentric prospective study included consecutive children presenting with severe to profound sensorineural hearing loss (SNHL) who were candidates for unilateral and/or bilateral cochlear implantation. The etiology of the SNHL was determined before cochlear implantation when possible. During the surgery, drop-like samples of inner ear fluid and blood were collected. CMV PCR was then performed on both samples. Between January 2017 and September 2021, 113 children with severe to profound SNHL underwent cochlear implantation with inner ear fluid collection. Among these children, 77 of them presented with a known cause of SNHL (68%) and 36 of them had an unknown cause of SNHL at the time of surgery (32%). Sensitivity and specificity of the CMV PCR on inner ear fluid were 60% (95% CI: [49-71]) and 98% (95% CI: [96-100]), respectively. Positive and negative predictive values were 90% (95% CI: [83-97]) and 92% (95% CI: [86-98]), respectively. A sensitivity analysis according to age at cochlear implantation showed a decrease with age.

CONCLUSION

Sampling of inner ear fluid during cochlear implant surgery is an interesting, simple and safe way to diagnose CMV-related hearing loss, especially when the diagnosis of congenital infection can no longer be confirmed. However, the sensitivity decreases with age.

TRIAL REGISTRATION

NCT04724265 What is Known: • Congenital cytomegalovirus infection is the leading infectious cause of neurological disabilities and sensorineural hearing loss in children. In the absence of systematic screening at birth, many cCMV infections go undetected and are often undiagnosed despite the development of sensorineural sequelae.  • Nearly 40% of indications for cochlear implantation are of unknown etiology.

WHAT IS NEW

• Performing CMV PCR on inner ear fluid at the time of cochlear implantation is a safe way with high diagnostic performance (PPV = 90%, NPV = 92%) to detect a CMV-related hearing loss. • This sample may be interesting in cases of unknown cause of hearing loss in order to identify undiagnosed cCMV infections.

Predicting Early Vestibular and Motor Function in Congenital Cytomegalovirus Infection

OBJECTIVES

Congenital cytomegalovirus (cCMV) can affect vestibular function, which is an important cornerstone for early motor development. This study aims to identify risk factors for early vestibular dysfunction with severe repercussions on the motor outcome.

METHODS

This prospective cohort study included 169 cCMV-patients with complete vestibular assessment (lateral video Head Impulse Test and cervical Vestibular Evoked Myogenic Potentials) before the age of 18 months (mean 8.9, standard deviation 3.27 months). Motor results using the Alberta Infant Motor Scale were collected in 152 of these patients. Logistic and linear regression models were applied to identify risk factors for the vestibular and motor outcomes, respectively.

RESULTS

The odds of developing early vestibular dysfunction were 6 times higher in patients presenting with hearing loss at birth compared to those born with normal hearing (p = .002). Within the latter group, significant predictors for vestibular dysfunction were (delayed-onset) hearing impairment at the time of vestibular testing (p = .003) and the presence of periventricular cysts on magnetic resonance imaging (p = .005). Remarkably, none of the patients infected during the third trimester of pregnancy (n = 14) developed early vestibular dysfunction. On average, vestibular-impaired patients had a z-score on the Alberta Infant Motor Scale that was 1.17 points lower than patients without vestibular deficit (p < .001).

CONCLUSION

Early vestibular loss can have a significant adverse effect on motor development. Hearing and cranial imaging findings could facilitate the widespread implementation of a (targeted) vestibular assessment approach in the cCMV-population.

LEVEL OF EVIDENCE

3 Laryngoscope, 2022.