Otology & Neurotology29:214Y220 Ó 2008, Otology & Neurotology, Inc. Revision Cochlear Implant Surgery in Children *Robert D. Cullen, *Jose N. Fayad, *William M. Luxford, and †Craig A. Buchman *Children`s Auditory Research and Evaluation Center, House Ear Institute, Los Angeles, California; and ÞWP Biggers Carolina Children`s Communicative Disorders Program, Department of Otolaryngology/Head and Neck Surgery, University of North CarolinaYChapel Hill, Chapel Hill, North Carolina, U.S.A.Objective: To determine the incidence of revision cochlear magnet dislodgement requiring revision surgery occurred in 2%implant (CI) surgery in children and the indications for revision (n = 2). Head trauma was associated with 41% of the hardsurgery and to examine the pattern of events that lead to revi- failure cases (n = 20). Device analyses revealed identifiablesion CI surgery. abnormalities in most of both hard and soft failure cases. InStudy Design: Retrospective case review. most patients, auditory performance equaled or surpassed theSetting: Two tertiary pediatric CI centers. best preoperative performance by 6 to 12 months after revision.Patients: Pediatric CI patients who underwent revision surgery Conclusion: Revision CI surgery is common among pediatricrelated to their CI. CI recipients. Hard failure is the most common reason forMain Outcome Measures: Reasons for revision, surgical out- undertaking revision surgery, and this mode of failure is fre-comes, complications, performance, and device analyses were quently associated with preceding head trauma. Patients andsought. parents should be counseled that performance is expected toResults: Nine hundred fifty-two pediatric CI operations were equal or surpass the child`s best level of performance beforeperformed between 1991 and 2005. Ninety-three patients revision surgery, although this may take some time, and excep-underwent 107 (11.2%) revision operations. Hard device failure tions do exist. Key Words: Cochlear implantVPediatricVoccurred in 46% (n = 49); soft failure occurred in 15% (n = 16); Revision.medical/surgical causes were responsible for 37% (n = 40); and Otol Neurotol 29:214Y220, 2008. Cochlear implantation has provided auditory (re)habi- the indications for revision surgery and to examine thelitation to children for more than 2 decades. With any pattern of events that lead to revision CI surgery.surgically implanted medical device, there is an inherentrisk that the device will necessitate revision surgery forthe patient. This holds true for cochlear implants (CIs) as MATERIALS AND METHODSwell. The pattern of events that lead to revision surgery A collaborative and retrospective review of revision CI sur-can be broken down into the following categories: device gery was performed for children followed at 2 large pediatric CIfailure, suspected device malfunction, medical or surgi- programs: the WP Biggers Carolina Children`s Communicativecal problems, and the desire for technological upgrades. Disorders Program in the Department of Otolaryngology/Head Revision surgery accounts for approximately 5 to 10% and Neck Surgery at the University of North Carolina, Chapelof adult CI operations in many centers (1Y6). The reasons Hill, NC, and the Children`s Auditory Research and Evaluationfor revision fall into the categories listed above. Perfor- Center at the House Ear Institute, Los Angeles, CA. Nine hun-mance has been shown to be good after revision surgery, dred fifty-two pediatric CI operations were performed from 1991 to 2005 at these 2 institutions. The records of thoseincluding those that require replacement of the CI. patients who had undergone revision surgery were reviewed. There is less data available regarding revision CI sur- Reasons for revision, surgical outcomes, complications, perfor-gery in children. The purpose of this study was to deter- mance, and device analyses were sought for this subset ofmine the incidence of revision CI surgery in children and patients. The reason for revision was categorized into 5 groups. Device failure (also called a hard failure) was defined as com- plete failure of the CI. A patient with a device failure received Address correspondence and reprint requests to Craig A. Buchman, no auditory input from the device, and evaluation confirmedM.D., GO412 Neurosciences Hospital, UNC-Chapel Hill, Chapel Hill, that no connection could be made with the device by externalNC 27599-7600; E-mail: email@example.com means. Suspected device malfunction (also called a soft failure) 214 Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
REVISION COCHLEAR IMPLANT SURGERY IN CHILDREN 215 TABLE 2. Signs and symptoms associated with hard failure (n = 49) n % Symptoms Sudden loss of sound percepts 49 100 Signs No lock 43 88 No percepts only 2 4 Fluctuating impedances 1 2 Loss of electrodes 1 2 Open circuits 1 2 Trauma 20 41FIG. 1. Indication for revision surgery. Preceding signs or symptoms Atypical tinnitus 2 4 Intermittent function 3 6 Decrement in performance 2 4was defined as a clinical failure of the CI, where the device Shocking sensation 1 2itself may have still been providing some auditory input; how-ever, unacceptable aversive symptoms or a decrement in per-formance drove the family or child to undertake revisionsurgery. A small subset of patients underwent revision surgery patients experienced aversive symptoms (atypical tinni-for magnet displacement. Patients or parents wishing to have tus, pain or shocking in the region of the internal receiverrevision surgery to take advantage of new technology, despite stimulator, or intermittent performance) or a decrementhaving a functioning CI, were categorized as a technological in performance before the complete failure of the device.upgrade. The remainder of patients required surgery for medical Medical/surgical causes were responsible for 37%or surgical complications directly related to their CI. Medical or (n = 40) of revision operations (Table 3). Wound andsurgical problems not related to the CI such as cases of uncom- flap complications were the cause in 50% (n = 20) of theplicated otitis media, nonsurgical mastoiditis, or mastoiditismanaged with the device left undisturbed were not considered procedures. Six patients underwent a single procedure foras a part of this study. wound- and flap-related complications. Five patients required multiple procedures, accounting for the remain- ing 14 procedures. Two patients requiring multiple revi- RESULTS sions ultimately had device explantation with re- implantation on the opposite side. Electrode-associated Ninety-three pediatric CI patients underwent a total of complications occurred in 14 patients (35%), 7 of whom107 revision operations (Fig. 1). This comprises 11.2% had an inner ear malformation or ossification of theof all pediatric CI operations performed at the 2 study cochlea due to meningitis. Five patients who underwentcenters. Of the 922 patients who underwent a primary CI CI surgery at outside institutions had the electrode arraysurgery at 1 of the 2 study centers, 77 (8.4%) had revi- placed outside of the cochlea. Cholesteatoma occurred insion surgery during the study period. Demographic data 5 other patients, requiring subsequent revision surgery.were available for 84 of the 93 patients (Table 1). Another patient who underwent revision surgery was initially classified as a soft failure because of the pre- Reason for Revision sence of a functional device with limited auditory Hard failure was the most common reason for revision input. The patient was later diagnosed as having anCI surgery, accounting for 46% (n = 49) of revision absent VIII cranial nerve. This was reclassified as a med-operations. Forty-one percent (n = 20) of these patients ical cause for revision surgery, as the diagnosis was initi-had a preceding history of head trauma. Other associated ally missed on preoperative magnetic resonance imagingsigns or symptoms are shown in Table 2. Of note, several (7,8). Soft failure occurred in 15% (n = 16) of revisions. The signs and symptoms leading to the diagnosis and subse- TABLE 1. Demographics quent reimplantation are listed in Table 4. The most Age at implantation Time to revision, common reasons were aversive auditory or nonauditory (mo) (mo) symptoms and decline in performance. Abnormal in situCategory Average Range Average RangeMedical/surgical 52.5 T 38 9.5Y181.8 26.8 T 26 1.5Y96.4 TABLE 3. Medical/surgical causes for revision (n = 40) (n = 27) n %Hard failure (n = 43) 54.2 T 44 11.7Y205.6 38.5 T 25 1.7Y104.2Soft failure (n = 14) 69.1 T 70 18.2Y210.9 42.9 T 33.4 4.5Y104 Flap/wound 20 50Total (n = 84) 54.7 T 44.3 9.5Y210.9 31.5 T 25 1.5Y104.2 Electrode 14 35 Extrusion 6 Complete demographic data including date of birth, date of initial Extracochlear placement 5implantation, and date of revision implantation were available on 84 Unable to replace 2of 91 patients with revision surgery because of the listed causes. Time to Kinked in cochlea 1revision refers to the time between the initial and revision operations. Cholesteatoma/tympanic membrane 5 12.5Only times to initial revisions are included if patients had more than 1 Absent auditory nerve 1 2.5revision. Otology & Neurotology, Vol. 29, No. 2, 2008 Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
216 R. D. CULLEN ET AL. TABLE 4. Signs and symptoms of soft failures (n = 16) the identical tests scored at 6- and 12-month postoperative n % intervals. As a variety of open- and closed-set tests of auditory performance are performed at both centers, aAversive symptoms meaningful statistical analysis of any 1 test was pre- Auditory 13 81% cluded. The following tests were included: Early Speech Intermittent/poor performance 5 Atypical tinnitus 3 Perception Test, Glendonald Auditory Screening Pro- Increased loudness 1 cedure, Ling Six Sound Test, Lexical Neighborhood Nonauditory 7 44% Test, and the Phonetically Balanced Kindergarten Pain 4 Test. Figure 4, A and B, demonstrates the preoperative Headache 3Signs and postoperative percentage scores at the 6- (n = 28 Abnormal testing 11 88% patients) and 12-month (n = 21 patients) postoperative Abnormal impedances 4 periods, respectively. The absolute values of the scores Loss of electrodes 3 vary widely because these graphs are inclusive of all admi- Intermittent lock 2 Increased thresholds 2 nistered tests, both open and closed set. What can be gath- Performance issue 4 ered from the data is that most patients ultimately reach or Facial stimulation 0 exceed their prerevision performance level. In addition, 12-month scores are better than 6-month scores. However, 5 of the 21 children with 12-month data had not yet returned to their best prerevision level of performance.device testing, although not universal, was found in mostpatients (Table 4). DISCUSSION Two patients underwent revision surgery to replacemagnets that had become dislodged from the internal Cochlear implantation has become a very commonreceiver stimulator. No patients in this series underwent procedure among children throughout the world. As CIrevision surgery solely for a technology upgrade. surgery necessitates implantation of an electrical stimu- lator with complex circuitry in the confines of the Device Analysis cochlea, middle ear, mastoid and subcutaneous region, For patients who underwent explantation for device- revision surgery should probably be anticipated at somerelated issues (hard failure and soft failure), the devices point throughout the child`s life span. The present studywere returned to the manufacturer for a complete device used data from 2 large CI centers that have accounted foranalysis. Device analyses were available for review in 49 nearly 1,000 implanted children during a 14-year period.cases (hard failure, n = 37/49; soft failure, n = 12/16). One hundred seven (11.2%) revision operations wereThe results of these analyses are summarized in Table 5. required in 93 children during this period. Thus, revision surgery is common among pediatric CI users. The dis- Manufacturer Differences tribution of pediatric revision CI cases within the cate- Three manufacturers have had CIs approved for gories of medical/surgical, hard failure, and soft failureimplantation within the United States: Cochlear Corpora- are similar to those reported in a recent study (5). Astion (Englewood, CO, USA), Advanced Bionics Cor- expected, both device-related (n = 67; 62.6%) and med-poration (Sylmar, CA, USA), and MedEl Corporation ical/surgical reasons (n = 40; 37.4%) for revision were(Innsbruck, Austria). An analysis was performed to eval- observed. Irrespective of cause, revision surgery wasuate the incidence of revision surgery for each manufac- undertaken at a wide variety of times after initial im-turer, as well as for major device lines within each plantation (Table 1). Medical/surgical revisions gener-manufacturer. Advanced Bionics Corporation devices ally occurred earlier (mean, 26.8 T 26.1 mo; range,where subcategorized into the Clarion 1.2 (C1.2), Clar-ion CII (CII), and HiResolution 90 K (90 K). CochlearCorporation devices include the Nucleus 22 (N22), TABLE 5. Device analyses for device failures and suspectedNucleus 24 (N24), and Nucleus Freedom (Freedom). device malfunctionMedEl Corporation produced only the Combi40+ duringthe study period. The incidence of revision for each man- Device analysis Hard failure Soft failureufacturer and device model is shown in Table 6, along (n = 49/65) (n = 37/49) (n = 12/16)with the total number of devices implanted. Figures 2 Case fracture 14 0and 3 show the manufacturer-specific and device-specific Loss of hermeticity 12 2incidence of revision surgery by indication, respectively. Moisture 3 0 Broken substrate 1 1 Electrode fatigue 1 3 Postrevision Auditory Performance Circuit failure 1 1 Audiologic performance data were sought on all Contamination 0 1patients. A comparison of preoperative (best perfor- Electrode short 0 1mance before a decline or absence of function) and post- Inconclusive 1 0 No failure mode identified 4 3operative performance was possible when the patient hadOtology & Neurotology, Vol. 29, No. 2, 2008 Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
REVISION COCHLEAR IMPLANT SURGERY IN CHILDREN 217 TABLE 6. Incidence of revision by device trauma may clearly lead to some of these findings,Device Total Revisions % maximizing device durability in the manufacturing pro- cess should be of paramount importance. CI manufac-Advanced Bionics Corporation 364 63 17.3 turers have attempted to address these issues by using Clarion 1.2 235 52 22.1 a variety of more durable materials, changing device Clarion CII 54 5 9.3 HiResolution 90K 75 6 8.0 design when appropriate, as well as improving theCochlear Corporation 541 38 7.0 manufacturing process. Surely, these changes are Nucleus 22 178 24 13.5 what account for some of the observed differences in Nucleus 24 316 13 3.8 revision rates among the various device models and Nucleus Freedom 47 1 2.1MedEl Corporation manufacturers. Combi 40+ 41 5 12.2 Medical and surgical reasons accounted for 37% of the revision operations. Fifty percent of these procedures were performed for skin flap or wound complications.1.5Y96.4 mo) than those related to the device (mean, Of note is that most of these complications occurred39.5 T 27.0 mo; range, 1.7Y104.2 mo). early in the series. This is supported by the fact that The most common reason for revision surgery among most of these revisions occurred in patients with olderpediatric CI users is hard device failure, accounting for version devices (Fig. 3). By learning from these mis-46% of the revision operations in this group. Of these takes, skin incisions and flap designs have been devel-children, 41% had a preceding history of trauma. In some oped to help avoid these problems. Smaller skin incisionscases, a direct blow to the head was immediately fol- with smaller skin flaps have reduced wound complica-lowed by device failure, suggesting a causal relationship. tions. In the pediatric patient, the device should sit deepIn others, the history of trauma was remote and only to the periosteum of the cranium and not in the directelicited by careful history. Because the device did not subcutaneous layer. Ample room between the internalalways malfunction immediately, the role of trauma in housing of the device and the external, behind-the-earthe root cause of device failure in these cases is less clear. processor, is required to prevent unwanted magnet inter-Prevention of mild head trauma and associated device actions and pressure necrosis on the skin. A seat for theimpact in the pediatric population is difficult, if not CI is drilled into the calvarium to prevent migration ofimpossible. The CI centers in this study discourage par- the device. However, the use of Btacking^ sutures orticipation in contact sports where direct device impacts other materials to prevent the implant from migratingare probable. In the end, families are left to weigh the does not seem to be necessary.risks and benefits of participation in these and other Otitis media remains very common among children. Insports for their child. children with CIs, the consequences of recurrent or Regarding the cause of device failure, Bcase fracture^ chronic otitis media and eustachian tube dysfunctionand Bloss of hermeticity/case leak^ remain the most com- are noteworthy (9). In the present study, cholesteatoma,mon findings among devices that stop working. Although tympanic membrane retraction, and serous otitis mediaFIG. 2. Manufacturer-specific incidence of revision surgery by indication expressed as a proportion of the total number of each vendor`simplants represented in the population. *Includes 6 patients with unknown device manufacturer, 1 of which had revision surgery. Otology & Neurotology, Vol. 29, No. 2, 2008 Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
218 R. D. CULLEN ET AL.FIG. 3. Device-specific incidence of revision surgery by indication expressed as a proportion of the total number of each modelrepresented in the population. C1.2 indicates Clarion 1.2; CII, Clarion CII; Freedom, Nucleus Freedom; N22, Nucleus 22; N24, Nucleus24; 90K, HiResolution 90K.were all encountered. The management of cholesteatoma within the cochlea until the new device is in place andin the same ear as a CI is complex and must be indivi- immediately ready for insertion.dualized. When tympanic membrane retraction occurs, it Suspected device malfunction, or soft failure, is a rela-must be monitored closely to prevent electrode extrusion. tively newly recognized entity that has been wellThe use of cartilage graft tympanoplasty for the preven- described in adults (12). This type of failure occurstion of further retraction and electrode extrusion should when a clinical failure of the device occurs while thebe considered in select cases where the electrode comes device continues to provide auditory perceptions. Clini-in contact with the tympanic membrane. Closure of the cal failure, from a patient-oriented perspective, includesear canal may also be considered (10,11). progressive decline in function or adverse stimuli related Extracochlear placement of electrodes was observed in to the implant. Adverse stimuli may be auditory such as5 patients referred from outside institutions. Proper tech- intermittent function, decrement in performance, atypicalnique and definitive identification of the round window tinnitus, or increasing loudness. Nonauditory symptomsbefore making the cochleostomy can avoid this compli- such as pain or headache are particularly alarming. Acation. The appropriate evaluation of the CI candidate thorough discussion of the diagnosis and managementshould likewise reduce the likelihood of implanting a of these patients has been well described (13).patient with an absent VIII cranial nerve (7,8). Identification of children with a CI soft failure can be Electrode extrusion complicated several cases. The very challenging and presumably impossible in someprecise cause behind extrusion is not completely under- cases. For example, 8 children in the present study thatstood. However, it is feasible that the Bmemory^ of a were reimplanted after a hard device failure had preced-straight electrode may exert force against the inner wall ing soft failure symptoms. These symptoms preceded theof the cochlea, causing the electrode to back out of the hard failure of the device by weeks or even years in somecochlea. This is supported by the fact that all 6 cases of cases. This underscores the difficulty in evaluating theelectrode extrusion occurred in Nucleus 22 devices, subtle signs and symptoms of soft failures in this agewhich have a straight intracochlear electrode array. The group. A high index of suspicion is needed in all casesadvent of Bmodiolar-hugging^ precurved electrodes may when a child presents with aversive symptoms or anprevent this problem. unexplained decrement in performance. The manufac- Two revision cases were complicated by the fact that a turer can also assist in the evaluation of these suspectnew electrode could not be reinserted into the cochleost- devices before intervention by providing advanced test-omy. This occurred when an electrode array was com- ing using diagnostic tools not routinely available in thepletely removed from the cochlea early in the procedure. commercial software. Imaging is also undertaken. In theOn attempting to reinsert the electrode, the tract was end, testing results may be unrevealing. In these cases,unable to be reestablished. Because of this experience, we usually recommend ipsilateral device removal andthe electrode is now severed at the facial recess and left replacement when aversive symptoms are intolerable.Otology & Neurotology, Vol. 29, No. 2, 2008 Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
REVISION COCHLEAR IMPLANT SURGERY IN CHILDREN 219FIG. 4. Auditory performance at 6 months (A) and 1 year (B) after revision. The line (unity) represents the point at which preoperative andpostoperative scores are equal. Points above the line indicate that performance improved at 6 months after revision. Points below the lineindicate a decrement in performance. The following tests were included: Early Speech Perception Test, Glendonald Auditory ScreeningProcedure, Ling Six Sound Test, Lexical Neighborhood Test, and the Phonetically Balanced Kindergarten Test.Contralateral device placement is considered when a par- Postrevision Performancetial decrement in performance is evident in the absence of In general, most patients maintained or improved onthe previously mentioned aversive stimuli to avoid the their best preoperative level of performance level,removal of a functional device. Obviously, these deci- although exceptions were observed. Presumably, differ-sions must be individualized for the patient. ences in electrode position or the use of a different device In the present study, the analysis in 9 of 12 devices and/or programming strategy may delay the return to theremoved for soft failure reasons revealed an identifiable preoperative level of function. In 1 child with a commoncause for failure. Despite having abnormal impedances, cavity, revision surgery resulted in a substantial drop inloss of electrodes, and intermittency issues, some of performance that never returned to the preoperative level.these devices were considered to be working within spe- This important outcome should prompt professionals tocification by the manufacturer before removal. This pro- use caution when counseling parents and children regard-vides additional support for a clinical diagnosis of device ing expectations after device replacement. Specifically,malfunction. Certainly, continued evaluation by the man- parents and children should be counseled that perfor-ufacturer is essential to improve our evaluation process mance after revision surgery is likely to be equal to orof these patients and devices. better than the best prerevision performance; however, Otology & Neurotology, Vol. 29, No. 2, 2008 Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
220 R. D. CULLEN ET AL.attaining this level of performance may take several 3. Miyamoto RT, Svirsky MA, Myres WA, et al. Cochlear implantmonths to more than a year. In addition, not all patients reimplantation. Am J Otol 1997;18:S60Y1. 4. Balkany TJ, Hodges AV, Gomez-Marin O, et al. Cochlear reim-will attain this goal by 1 year after revision. plantation. Laryngoscope 1999;109:351Y5. 5. Lassig AA, Zwolan TA, Telian SA. Cochlear implant failures and CONCLUSION revision. Otol Neurotol 2005;26:624Y34. 6. Alexiades G, Roland JT Jr, Fishman AJ, et al. Cochlear reimplan- tation: surgical techniques and functional results. Laryngoscope Revision CI surgery is common among pediatric CI 2001;111:1608Y13.recipients. Hard failure is the most common reason for 7. Buchman CA, Roush PA, Teagle HF, et al. Auditory neuropathyundertaking revision surgery, and this mode of failure is characteristics in children with cochlear nerve deficiency. Ear Hearfrequently associated with preceding head trauma. 2006;27:399Y408. 8. Adunka OF, Roush PA, Teagle HF, et al. Internal auditory canalPatients and parents should be counseled that, for most morphology in children with cochlear nerve deficiency. Otol Neu-individuals, performance after revision surgery can be rotol 2006;27:793Y801.expected to equal or surpass the child`s best level of 9. Fayad JN, Tabaee A, Micheletto JN, et al. Cochlear implantation inperformance before revision surgery, although this may children with otitis media. Laryngoscope 2003;113:1224Y7.take some time, and exceptions do exist. 10. El-Kashlan HK, Arts HA, Telian SA. Cochlear implantation in chronic suppurative otitis media. Otol Neurotol 2002;23:53Y5. 11. El-Kashlan HK, Arts HA, Telian SA. External auditory canal clo- REFERENCES sure in cochlear implant surgery. Otol Neurotol 2003;24:404Y8. 12. Buchman CA, Higgins CA, Cullen R, et al. Revision cochlear 1. Parisier SC, Chute PM, Weiss MH, et al. Results of cochlear implant surgery in adult patients with suspected device malfunc- implant reinsertion. Laryngoscope 1991;101:1013Y5. tion. Otol Neurotol 2004;25:504Y10; discussion 10. 2. Fayad JN, Baino T, Parisier SC. Revision cochlear implant surgery: 13. Balkany TJ, Hodges AV, Buchman CA, et al. Cochlear implant soft causes and outcome. Otolaryngol Head Neck Surg 2004;131: failures consensus development conference statement. Otol Neuro- 429Y32. tol 2005;26:815Y8.Otology & Neurotology, Vol. 29, No. 2, 2008 Copyright @ 2008 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.