Narrative review
Potential benefits from cochlear implantation of children with unilateral hearing loss Paul J Boyd Consultant with Cochlear Europe Ltd Objectives/methods: The aim of this discussion paper is to review several issues relevant to the viability of cochlear implantation of children with severe-profound unilateral hearing loss (UHL) and to discuss to what extent published findings on these issues can predict likely benefits from implantation in this population. Results: Several key issues are apparent from the recent literature: (i) UHL results in significant educational and psycho-social difficulties, but these are not universal in pre-lingual cases and may not be apparent for several years after birth, (ii) conventional treatments (contralateral routing of signal aids or bone-anchored hearing aids) provide limited benefit in the majority of sensorineural cases, (iii) early published outcomes from implantation of a limited number of children with acquired UHL suggest benefits similar to those observed in postlingually deafened adults, (iv) unilateral auditory deprivation results in poorer outcomes from delayed implantation of children with congenital losses, and (v) a large proportion of cases of severeprofound sensorineural UHL are associated with structural abnormalities of the cochlea or VIII nerve, such that not all children with UHL may be suitable for cochlear implantation. Conclusions: Children with acquired UHL are likely to gain similar positive benefits from cochlear implantation as those recently reported in adults (improved localization and better speech understanding in specific noise conditions). However, implantation of children with prelingual UHL is currently problematic as the impact of UHL may not become apparent until the child enters full-time education, by which time outcomes from cochlear implantation may be sub-optimal due to auditory deprivation. Development of appropriate candidacy criteria is important but challenging as criteria may need to be based on real-world hearing difficulties as well as audiological measures. Keywords: Cochlear implant, Unilateral hearing loss, Single-sided deafness, Clinical management, Hearing impaired children
Introduction In recent years, a number of reports have been published on outcomes of cochlear implantation of adults with severe-profound unilateral hearing loss (UHL), also often termed ‘single-sided deafness’ (SSD). Initially, these were performed in the expectation of reducing intractable tinnitus, but it soon emerged that significant listening benefits are also often obtained, including improved speech understanding in noise, presumably relating to partial restoration of binaural localization abilities (Punte et al., 2011). More recently, several series of adults have been implanted where anticipation of hearing benefits has been the principal motivation for implantation, rather than treatment of tinnitus, with positive reported outcomes and high levels of device use (Vlastarakos et al., 2014).
Correspondence to: Paul J Boyd, BMI Thornbury Hospital, 312 Fulwood Road, Sheffield S10 3BR, UK. Email:
[email protected] © W. S. Maney & Son Ltd 2015 DOI 10.1179/1754762814Y.0000000100
The success of these studies has prompted some clinicians to consider whether implantation of children with UHL might also be beneficial and, indeed, there have already been a small number of children implanted on a highly selected and individual basis (e.g. Cadieux et al., 2013). The real-life consequences of UHL in children have been reported to be considerable, notably with reduced performance in educational settings (Kuppler et al., 2013), so that it is possible that benefits from implantation of these children might actually be greater than those obtained with adults. There are, however, important differences among some of the characteristics of UHL in children and adults that might influence potential benefits from implantation. One important issue is that the quality of the auditory percept produced by a cochlear implant (CI) probably needs to be relatively good in order to provide a useful and acceptable addition to the normal hearing in the opposite ear. From clinical experiences of implantation of subjects with bilateral hearing loss and bilateral implantation, it is clear
Cochlear Implants International
2015
VOL.
16
NO.
3
121
Boyd
Potential benefits from cochlear implantation of children with UHL
that this is more likely with early implantation of congenital losses (relating to neural plasticity) or with implantation of acquired hearing losses with short duration of deafness (Van Deun et al., 2010). The large majority of implanted adults with UHL have had relatively short duration acquired hearing losses, such that the auditory percept from the CI was probably relatively good. In children, however, there is a lower incidence of acquired hearing loss, and many congenital unilateral losses are associated with structural abnormalities and co-morbidities such as cognitive impairment, so that only a subset of children with UHL may potentially benefit from cochlear implantation. Furthermore, there are significant ethical and psycho-social factors that must be considered with children. The aim of this discussion paper, therefore, is to review and consider the various issues that might have a bearing on the viability and possible outcomes of cochlear implantation of children with UHL. This includes an overview of reported prevalence/aetiology as well as the consequences and impact of UHL in children, and a review of outcomes of conventional treatments and of existing publications on implantation of adults and a small number of highly selected children with UHL. Reviewing this information, suggestions are made regarding likely benefits from implantation of children with various types of UHL, including conclusions relevant to prospective studies. NB: Many recent publications have used the term ‘single-sided deafness’ (SSD) to describe subjects with UHL who have been considered for implantation with a CI or bone-anchored hearing aid (BAHA). However, as the term ‘deafness’ implies a profound or total hearing loss, then ‘SSD’ is not really the ideal term for cases where there is some functional hearing, even though most subjects in such publications have had relatively severe hearing losses. For this reason, UHL is used as the term of preference throughout this discussion paper, although SSD is also used in places when quoting published literature.
Prevalence/aetiology of UHL in children In the following discussion, we are primarily interested in the characteristics of hearing losses which might potentially benefit from cochlear implantation. The typical characteristics of UHL where a CI might be considered would be a severe-profound sensorineural hearing loss in the poorer ear (i.e. unaidable using conventional hearing aids) and with normal or nearnormal hearing in the good side. In this situation, candidates would fall outside the standard criteria for implantation in most counties, including the UK. Speech and language would be expected to develop more or less normally, but such individuals would
122
Cochlear Implants International
2015
VOL.
16
NO.
3
have difficulty localizing sounds and understanding speech in noisy listening environments, which might compromise educational and social performance. The impact of UHL in children is discussed in more detail in the following section. In addition to UHL, certain aetiologies may result in markedly asymmetric hearing loss (AHL), where both ears have a significant hearing loss, but with only one ear within standard CI criteria. Many of the issues discussed below are applicable to such cases as well as to those with purely unilateral losses. At a recent symposium on the use of bone-anchored hearing devices as a treatment for SSD and UHL (Cochlear, 2013), it was proposed that the following definitions for the hearing in the better ear should be used to clearly distinguish between subjects in these two groups: SSD better ear ≤30 dB HL to 4 kHz inclusively, AHL better ear ≤60 dB HL to 4 kHz inclusively and >30 dB HL at one or more frequencies to 4 kHz inclusively and with an interaural asymmetry of >30 dB (four-frequency average). Hearing loss can be conveniently categorized as congenital (including conditions where the hearing loss develops postnatally) and acquired. Certain aetiologies involve structural malformations and others are associated with cognitive or other concomitant disabilities. Therefore, when evaluating the potential of cochlear implantation, we need to consider that the target population is heterogeneous in terms of aetiology and that some hearing losses may be more amenable to certain types of intervention than others. However, a significant difficulty in estimating the numbers of individuals with UHL who might be suitable for a CI is that many published reports on prevalence do not provide details on the degree of hearing loss (only those with severe or profound losses would be potential candidates). The widespread introduction of neonatal hearing screening over the last 15 years or so has generated a number of reports on the overall incidence of congenital hearing loss, though often aetiological details are lacking. The incidence of permanent bilateral hearing loss (defined as ≥40 dB HL in the better ear) is generally accepted to be around 1 in 1000, where some 20–25% are profound losses (Prieve et al., 2000; Kennedy et al., 2006; Uus and Bamford, 2006). Uus and Bamford (2006) provided the audiological basis of such cases from a national screening programme in the UK, reporting 7% with a conductive or mixed loss, 10% with auditory dys-synchrony, and 83% sensorineural. Aetiology of neonatal hearing loss varies to some degree according to the genetic profile and immunological status of the population, but generally profound losses have genetic origin in over half of cases (Nance et al., 2006), many others being due to perinatal infections or complications
Boyd
linked to prematurity. The most common genetic causes involve mutations affecting the gap junction protein connexin 26 (Pandya et al., 2003), while a relatively common example of perinatal infection causing deafness is congenital cytomegalovirus (CMV), though the majority of cases are asymptomatic at birth (Yamamoto et al., 2011). The incidence of neonatal UHL has been less widely reported, but appears to be of a similar order of magnitude to that of bilateral impairment. Prieve et al. (2000) estimated prevalence of 0.83/1000 where hearing loss was defined as auditory brainstem thresholds of >20 dB HL (1–4 kHz), and Uus and Bamford (2008) reported a prevalence of 0.64/1000 for UHL of moderate or greater degree. Prevalence of hearing loss increases in older children, and this appears to be particularly true for unilateral losses. Reported rates inevitably vary considerably, however, according to the audiological criteria employed as well as the age of the study populations. Geographical variations also exist due to variations in genetic and immunological status of different populations. Cone-Wesson et al. (2000) reported on visual reinforcement audiometry outcomes used to assess hearing status of 2995 infants at 8–12 months corrected age. When cases with middle ear effusion were excluded, prevalence of hearing loss of at least mild degree was 1 and 0.9% for bilateral and unilateral losses, respectively. Vartiainen and Karjalainen (1998), on the other hand, assessed the prevalence of UHL (>25 dB HL) in Finnish children up to age 10 as only 1.7 per 1000 live births, of which 35% were profound losses and 15% severe. Lundeen (1991) reported a prevalence of 1.9 and 0.73% for unilateral and bilateral hearing losses of >25 dB HL, respectively, in school-aged children, and Bess et al. (1998) reported a 3% prevalence of sensorineural UHL in schoolaged children (3rd, 6th, and 9th US school grades). The highest reported overall prevalence rate is 14% in a series of adolescents by Shargorodsky et al. (2010), though this included ‘slight’ losses with thresholds of >15 dB HL. Ross et al. (2010) demonstrated that prevalence figures depend on the definition of UHL, which varies among studies. Their study analysed data from a national populationbased, cross-sectional survey in the US (data from 1988 to 1994) to identify prevalence of UHL in children aged 6–19 years applying a range of commonly used audiological criteria (various combinations of thresholds and frequencies). Resultant prevalence rates were between 3.0 and 6.3%, though about 20% of the children failed tympanometry testing and so may have had conductive losses. The increased prevalence of UHL in older children is due to underlying congenital pathology as well as a wide range of factors producing acquired hearing loss.
Potential benefits from cochlear implantation of children with UHL
Among congenital factors, CMV infection (90% of which is asymptomatic at birth) and enlarged vestibular aqueduct were reported to be major causes of lateronset deafness in children by Nance et al. (2006). Yamamoto et al. (2011) found a 10% incidence of sensorineural deafness in a group of 102 children aged 12–84 months who were identified with CMV infection in the neonatal period. Of these, 50% were unilateral (all moderate to severe losses). Auditory neuropathy spectrum disorder (ANSD) is unilateral in about 25% of cases (Teagle et al., 2010; Liu et al., 2012). Dedhia et al. (2013) investigated a series of 78 children who passed neonatal screening but were subsequently diagnosed with a hearing loss during childhood; 47% were severe or profound and 26% were unilateral. Aetiology was genetic in 17%, anatomic abnormality in 14%, acquired perinatal in 12%, and unknown in 54%. While high proportions of unknown aetiology are reported in many studies, Dodson et al. (2012) found that 59% of 34 subjects with UHL (2 months to 36 years of age) had a family history of hearing loss, which was of unilateral loss in 45% of cases, suggesting a genetic origin. Thus, later acquired losses appear to be less common in children than those with congenital origins, the most commonly reported causes of acquired UHL being viral infections, meningitis, and head trauma (Tharpe and Sladen, 2008). With technical refinements and increased access to imaging techniques, it has become evident that a large proportion of severe-profound UHLs are associated with abnormal cochlear and internal auditory canal (IAC) anatomy. Song et al. (2009) found that 29% of 322 children with unilateral profound hearing loss had abnormal computerized tomography (CT) scans (mostly incomplete partition type-II, narrow IAC, and enlarged vestibular aqueduct). Masuda et al. (2013) reported extremely high prevalence (67%) of cochlear and IAC malformations in 69 Japanese children with UHL aged 0–15 years (32% with IAC anomaly) when middle ear conditions/abnormalities were excluded, and similar prevalence (57%) was reported in a Korean study by Yi et al. (2013). The latter study measured diameters of IAC and other cochlear structures in 51 children with SSD (aged 0–20 years) using highresolution CT. Bony cochlear nerve canal (BCNC) diameter of 70 dB HL. Levi et al. (2013) also found that most cases of hypoplastic cochlear nerves were associated with unilateral, usually severe-profound, hearing loss. Clemmens et al. (2013) assessed 134 children identified with sensorineural UHL using magnetic resonance imaging (MRI) and CT. Cochlear nerve deficiency (CND) was present in 26% overall, but increased to 48% in
Cochlear Implants International
2015
VOL.
16
NO.
3
123
Boyd
Potential benefits from cochlear implantation of children with UHL
cases of severe-profound deafness, and 100% in infants (n = 10) and was strongly predicted by narrow (