Benefits of Active Middle Ear Implants in Mixed Hearing Loss: Stapes Versus Round Window

Transcription

1 The Laryngoscope VC 2016 The American Laryngological, Rhinological and Otological Society, Inc. Benefits of Active Middle Ear Implants in Mixed Hearing Loss: Stapes Versus Round Window Jeon Mi Lee, MD; Jinsei Jung, MD, PhD; In Seok Moon, MD, PhD; Sung Huhn Kim, MD, PhD; Jae Young Choi, MD, PhD Objectives/Hypothesis: We compared the audiologic benefits of active middle ear implants with those of passive middle ear implants with hearing aids in mixed hearing loss, and also compared the outcomes of stapes vibroplasty with those of round window vibroplasty. Study Design: Retrospective chart review. Methods: Thirty-four patients with mixed hearing loss due to chronic otitis media were treated with a middle ear implant. Of these, 15 were treated with a passive middle ear implant (conventional ossiculoplasty with a partial ossicular replacement prosthesis), nine with an active middle ear implant coupling to the stapes, and 10 with an active middle ear implant coupling to the round window. Patients underwent pure-tone/free-field audiograms and speech discrimination tests before surgery and 6 months after surgery, and the results of these tests were compared. Results: The active middle ear implant resulted in better outcomes than the passive middle ear implant with hearing aids at mid to high frequencies (P <.05). Patients who received either a stapes vibroplasty or a round window vibroplasty showed comparable hearing gain except at 8,000 Hz (48.9 db vs db, P <.05). Patients who received a stapes vibroplasty showed an improvement even in bone conduction at 1,000 Hz and 2,000 Hz (both P <.05). Conclusions: Active middle ear implantation could be a better option than treatment with passive middle ear implants with hearing aids for achieving rehabilitation in patients with mixed hearing loss. Vibroplasty via either oval window or round window stimulation shares similar good results. Key Words: Mixed hearing loss, middle ear implant, Vibrant Soundbridge, vibroplasty, couplers. Level of Evidence: 4 Laryngoscope, 127: , 2017 INTRODUCTION Active middle ear implants (AMEI) are an alternative treatment option for various forms of hearing loss. The Vibrant Soundbridge (VSB) (MED-EL, Innsbruck, Austria) was initially developed for patients with sensorineural hearing loss and functions by coupling the floating mass transducer (FMT) to the long process of the incus. In 2006, Colletti et al. suggested placing the FMT on the round window (RW) and obtained positive results. 1 Since then, the use of the VSB has expanded to many patients who are affected by mixed hearing loss who could not benefit from conventional hearing aids (HAs), usually due to a previous surgery. Before the introduction of VSB, passive middle ear implantation, by way of a conventional ossiculoplasty, was usually performed to restore hearing. Whereas a conventional ossiculoplasty only reduces the air-bone gap (ABG), AMEI From the Department of Otorhinolaryngology, Yonsei University College of Medicine, Seoul, Republic of Korea. Editor s Note: This Manuscript was accepted for publication July 14, The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Jae Young Choi, MD, Department of Otorhinolaryngology, Yonsei University College of Medicine, Severance Hospital, Yonsei University Health System, 50 Yonsei-ro, Seodaemun-gu, Seoul , Republic of Korea. jychoi@yuhs.ac DOI: /lary in patients with mixed hearing loss shows promising results. 1 4 Many techniques to apply AMEI to patients with mixed hearing loss were introduced, and each technique has its advantages. When attaching the FMT to the stapes, it seems more intuitive to fix the FMT in a position with simple crimping. In case of the RW application of the VSB, the vibrational energy is directly transferred into the inner ear, bypassing the outer and middle ear 5 ; thus, the chronically disabled middle ear condition can be overcome. These techniques have structural and surgical advantages, and also good hearing gains. Published studies demonstrated hearing gain of 30 to 35 db for coupling to the stapes and 30 to 55 db for coupling to the RW. 4 There were a few articles reporting outcomes of vibroplasty in various techniques, but no article reported a direct comparison of stapes vibroplaty and RW vibroplasty in a single institute. We evaluated the audiologic benefits of active middle ear implantation over conventional ossiculoplasty with HAs, and also compared the outcomes of stapes vibroplasty and RW vibroplasty. MATERIALS AND METHODS Patients A total of 34 subjects with mixed hearing loss were included in this study. The patients were between 20 and 73 years of 1435

2 Fig. 1. The surgical views of stapes vibroplasty (A) and round window vibroplasty (B). (A) A clip coupler floating mass transducer (FMT) was attached to the stapes head and reinforced with cartilage. (B) Position of the round window (RW) soft coupler FMT assembly, RW, and the harvested cartilage. Harvested cartilage was placed behind the FMT to improve contact to the RW membrane and to optimize stabilization of the FMT (B). The black arrow indicates harvested cartilage, and the white arrow indicates the RW soft coupler. [Color figure can be viewed in the online issue, which is available at age (mean years), 41% of patients were men, and 59% of patients were women. All of the patients had chronically disabled middle ears, and some of them had previous ear surgery; 15 patients underwent canal wall up mastoidectomy, eight patients underwent canal wall down mastoidectomy, and four patients had myringoplasty operation. Procedures were performed based on the patient s middle ear conditions, which were evaluated with preoperative physical examination and/or an image study. When the middle ear space was well aerated, a conventional ossiculoplasty was performed, and if not, vibroplasty was considered. All of the ossiculoplasty patients were fitted with conventional HAs in the same ear after the operation. The types of HA were variable, ranging from open fit, receiver-in-the-ear types, closed fit, and complete-in-the-ear types, depending on each patient s preferences and inner ear condition. When the suprastructure of stapes was intact, and the mobility of the footplate was mobile, stapes vibroplasty was performed. Otherwise, RW vibroplasty was performed. In this study, 15 patients received ossicular reconstructions using a partial ossicular replacement prosthesis (PORP) (ossicular replacement prosthesis; Heinz Kurz GmbH, Dusslingen, Germany), nine received stapes vibroplasty, and 10 received RW vibroplasty between October 2011 and October 2015 at our tertiary hospital center. All implanted devices were the Vibrant Soundbridge (VSB) (MED-EL, Innsbruck, Austria) with couplers, and the speech processors were Amade (MED-EL). All procedures were in accordance with the Declaration of Helsinki of The study was approved by the institutional review board of the Severance Hospital in Seoul, Korea ( ). All of the subjects provided informed consent to participate in the study. Surgical Procedure In the PORP group, the surgery was performed under local anesthesia via a transcanal approach. None of the patients had a perforated tympanic membrane (TM). The transmeatal flap was elevated, and an appropriate length was measured for the PORP to be between the TM and the stapes head. Tragal cartilage was inserted between the TM and the PORP to prevent extrusion of the prosthesis. In the stapes vibroplasty group, a tympanomastoidectomy and a posterior tympanotomy were simultaneously performed in seven patients; however, two patients had already undergone a previous mastoidectomy before implantation. The incudal buttress was removed to expose the stapes. These patients surgical history and intraoperative status were varied, but generally, the incudostapedial joints of the patients were not properly functioning, and the stapes structure and mobility was intact. A clip coupler was used to fix the floating mass transducer (FMT) to the stapes head (Fig. 1A). Harvested cartilage was inserted between the TM and the FMT to prevent extrusion of the prosthesis. An implant bed was drilled into the occipitotemporal bone to fix the implant housing with bone-anchored sutures. RW vibroplasty was performed using a similar procedure. Eight patients had already undergone a previous mastoidectomy, mostly canal wall down, and the remaining two patients had undergone a tympanomastoidectomy and a posterior tympanotomy simultaneously. Six patients were classified with ossiclemissing status, one patient exhibited stapedial fixation, and one had an ossicular anomaly and a stapedial fixation, and two patients with fractured or thinned stapes were observed. The RW niche was identified, and granulation around the RW was cleared to expose the intact RW membrane. An FMT with a RW soft coupler was positioned in the RW, and harvested cartilage was placed in the opposite side of RW membrane to improve contact with the RW membrane and to optimize stabilization of the FMT (Fig. 1B). One patient did receive the RW soft coupler initially; however, this patient subsequently underwent a revision surgery due to granulation tissue formation between the FMT and middle ear structures, and reimplantation was performed with a RW soft 1436

3 TABLE I. Characteristics of the Patients. Patient No. Gender/Age, yr Ear Surgery History Surgical Finding PORP group 1 F/49 CWDM Incus missing 2 M/55 CWUM Incus missing 3 F/54 CWDM Incus missing 4 M/65 CWUM Incus missing 5 F/62 CWUM Incus missing 6 M/71 CWUM Incus missing 7 M/64 CWUM Incus missing 8 F/71 CWUM Incus missing 9 M/59 CWUM Incus missing 10 M/68 CWUM Incus missing 11 M/51 CWUM Incus missing 12 F/72 CWUM Incus missing 13 M/72 CWUM IS joint erosion 14 F/54 CWUM IS joint erosion 15 F/20 None IS joint erosion Stapes vibroplasty group 1 F/70 Myringoplasty Incus missing 2 M/67 None Incus missing 3 M/56 CWDM Incus missing 4 M/47 None Incus missing 5 F/65 Myringoplasty IS joint erosion 6 F/64 Myringoplasty IS joint erosion 7 M/52 Myringoplasty IS joint erosion 8 F/65 None IS joint fixation 9 F/73 CWUM Intact IS joint but stapes head erosion RW vibroplasty group 1 F/63 CWUM All ossicle missing 2 M/31 None All ossicle missing 3 F/56 CWDM All ossicle missing 4 F/66 CWDM All ossicle missing 5 F/52 CWDM All ossicle missing 6 M/47 CWUM All ossicle missing 7 F/56 CWDM Stapedial fixation 8 F/60 None Stapedial fixation 9 F/25 None Stapes anomaly with stapedial fixation 10 F/68 CWDM, ossiculoplasty Thinned and fractured stapes CWDM 5 canal wall down mastoidectomy; CWUM 5 canal wall up mastoidectomy; F 5 female; IS 5 incudostapedial; M 5 male; PORP 5 partial ossicular replacement prosthesis; RW 5 round window. coupler attached. After removing the granulation tissue and performing a reimplantation with a coupler, hearing gain was stabilized and showed improvement. An implant bed was created using the same methods as above. Patients information is listed in Table I. Audiologic Assessment All patients underwent pure-tone audiograms (PTA) before and 6 months after surgery. Pure-tone air-conduction (AC) (250 Hz to 8,000 Hz) and bone-conduction (BC) (250 Hz to 4,000 Hz) thresholds were measured with clinical audiometers in a doublewalled audio booth. The average hearing threshold was defined as the mean value of the measurements taken at frequencies of 500 Hz, 1 khz, 2 khz, and 4 khz. A word recognition test was performed to acquire the word recognition score (WRS). The WRS was measured at the most comfortable hearing level using 50 monosyllabic Korean words that are heard during everyday life. A word recognition test was performed before and after implantation, except for one patient who received stapes vibroplasty, because she was a foreigner and did not understand Korean. Patients were reassessed 6 months after the surgery; the assessments consisted of a free-field PTA and a word discrimination score. Functional hearing gain (FHG) was determined as the difference between unaided and aided free-field audiometry. Statistical Analysis Statistical analysis was performed with SPSS software for Windows version 21 (IBM, Armonk, NY). The results of 1437

4 no statistical significance was seen. Meanwhile, stapes vibroplasty showed better hearing gain at high frequencies over 3 khz, and the FHG of the stapes vibroplasty group was significantly better than that of the RW vibroplasty group at 8 khz (48.9 db vs. 31 db, P <.05). Because the BC thresholds remained stable whereas the AC hearing gain was remarkable, ABGs were meaningfully decreased in all three groups (P <.01). The AC hearing gain in the two vibroplasty groups was so prominent that even a reversal of AC and BC, especially at around 2 khz, was present in both vibroplasty groups (Fig. 3A-2,A-3). WRS was also shown to have improved from 58.0% % to 68.9% % in the PORP-with-HA group, from 61.0% % to 66.3% % in the stapes vibroplasty group, and from 54.4% % to 73.0% % in the RW vibroplasty group, though no statistical significance was observed in any group. Fig. 2. Preoperative air- and bone-conduction threshold in patients of three groups: partial ossicular replacement prosthesis (PORP) with hearing aid group (n 5 15), stapes vibroplasty group (n 5 9), and round window (RW) vibroplasty group (n 5 10). Preoperative air- and bone-conduction thresholds and the air-bone gaps showed no significant differences from each other (analysis of variance, P >.05). Results are presented as the mean 6 standard error of the difference. VSB 5 Vibrant Soundbridge. multiple experiments are presented as the mean 6 standard deviation. Statistical analysis was performed using analysis of variance (ANOVA), paired t-test, and independent t test. Values of P <.05 were considered statistically significant. RESULTS Preoperative AC and BC thresholds and ABG of three groups are displayed in Figure 2. These data showed no differences from each other (ANOVA, P >.05). FHG The average AC threshold was improved from db to db (P <.01) in the PORP group without HAs and to db (P <.01) in those with HAs, from db to db (P <.01) in the stapes vibroplasty group, and from db to db (P <.01) in the RW vibroplasty group. When we compared the functional gain of each group, the average hearing gain was not statistically different (ANOVA, P 5.16), though a difference of more than 10 db resulted from the PORP-with-HA group and the RW vibroplasty group (28.5 db vs db). We compared the FHG of each group by frequency range (Fig. 3). All three groups showed comparable hearing gain at less than 1 khz; however, in the PORP-with- HA group, the hearing gain was greatest at 1 khz and was less improved at higher frequencies. Both the stapes and RW vibroplasty group showed meaningfully better hearing gains than the PORP-with-HA group at over 1 khz (P <.05). The stapes vibroplasty group tended to result in less hearing gain at under 3 khz, low to mid frequencies, and more hearing gain at mid to high frequencies than that of the RW vibroplasty group, though Improvement of Bone Conduction In the PORP-with-HA group, the average BC threshold did not change ( db to db). The RW vibroplasty group also showed a stable BC threshold of db and db pre- and postoperatively. Interestingly, in the stapes vibroplasty group, an average BC threshold was slightly improved from db to db, and it was more significant at 1 khz and 2 khz (both P <.05) (Fig. 4). DISCUSSION Conventional passive ossiculoplasty improves hearing by restoring the continuity of the ossicular chain, but a complete closure of the ABG is mostly hard to achieve. We also observed a limited hearing gain of only 10 db in this study. Even with conventional amplification, aided AC could hardly overcome BC, and the limitation was more prominent at mid to high frequencies. In contrast, both stapes vibroplasty and RW vibroplasty overcame the ABG and provided significant hearing gain. Hearing gain was outstanding at high frequency rather than at low frequency, and the VSB-aided threshold was even better than the original BC threshold at 1 khz, 2 khz, and 3 khz. However, amplified frequencies are different, which is thought to be due to the mass loading effect given by the FMT. A loaded mass would result in a stiffening of the ossicular ligaments, increasing the tension on the joints, and changing the movement pattern of the ossicular chain. A human temporal bone model was studied to examine the mass loading effects on ossicles, and it was revealed that middle ear function did not change, but the displacement of the stapes footplate decreased in response to the increased masses, 6 and the effect was prominent at high frequencies. 7 Various experimental studies 6,8,9 consistently demonstrated no significant changes below 0.9 khz, regardless of the weight of the mass. No experimental study to reveal this phenomenon has been reported; however, it is thought to be due to the resonance frequency of the stapes vibration, which is around

5 Fig. 3. Audiometric results after middle ear implants in patients treated with partial ossicular replacement prosthesis (PORP) with hearing aid (HA) (A-1, n 5 15), stapes vibroplasty (A-2, n 5 9), and round window (RW) vibroplasty (A-3, n 5 10). Postoperative air-conduction (AC) hearing threshold became better in the three groups at whole frequencies (paired t test, both P <.01). (B) Functional hearing gains as frequencies in the three groups. Vibroplasty groups showed significantly better hearing gain at higher than 1 khz than the PORP with HA group. Among vibroplasty groups, the stapes vibroplasty group showed statistically better hearing gain at 8 khz than the RW vibroplasty group (P <.05). The stapes vibroplasty group tended to result in less hearing gain at low to mid frequencies and more hearing gain at mid to high frequencies than that of the RW vibroplasty group, though no statistical significance was seen. Results are presented as the mean 6 standard error of the difference. *P <.05. BC 5 bone conduction; VSB 5 Vibrant Soundbridge. khz. 9 Further study for the mass loading effect on low frequencies is needed. Stapes vibroplasty and RW vibroplasty both resulted in good hearing results. Wever and Lawrence first proposed that the reverse drive via the RW membrane was equivalent to the forward drive, 10 though cochlear stimulation from the RW is not a normal sound transmission pathway. Colletti et al. proved its beneficial outcome by placing the FMT on the RW. 1 Recently, the basilar membrane response to forward and reverse stimulation was measured using an animal model, and the response was also similar in nature. 11 Average hearing gain in stapes vibroplasty and RW vibroplasty, which is a forward and reverse stimulation, did not differ significantly in our study in accordance with the recent report. However, RW vibroplasty tended to show better hearing gain at low to mid frequencies, whereas stapes vibroplasty tended to show better hearing gain at high frequencies, especially at 8 khz (Fig. 3B). These differences are consistent with middle ear physiology. The presence of the footplate increases system stiffness, thus enhancing high frequencies and limiting low-frequency transmission. 12 Moreover, placing the FMT on the stapes results in increased system stiffness. An increased hearing gain at 8 khz and a decreased hearing gain at 250 Hz in the stapes vibroplasty group were observed in our study, though differences measured at 250 Hz were not statistically meaningful. These differences were not influenced by the preoperative ACs, which were identical before the implantation as frequencies, and the 1439

6 Fig. 4. Preoperative (Pre-op.) and postoperative (Post-op.) bone-conduction (BC) threshold in treated patients. (A-1) Audiometric results in patients with partial ossicular replacement prosthesis (PORP) with hearing aid (HA) (n 5 15). BC threshold remained stable. (A-2) Audiometric results in patients with stapes vibroplasty (n 5 9). An average BC threshold was not statistically different before and after surgery. (A-3) Audiometric results in patients with round window (RW) vibroplasty (n 5 10). BC threshold remained stable. (B) Functional hearing gains of bone conduction at frequencies between 250 and 4,000 Hz in the three groups. When analyzed at every frequency, there are remarkable hearing gains of bone conduction at 1 and 2 khz (both P <.05) in the stapes vibroplasty group. Results are presented as the mean 6 standard error of the difference. *P <.05. differences clearly demonstrated the characteristics of each restoration method. Another interesting finding of our study is the change in bone conduction thresholds in the stapes vibroplasty group; significant changes were found at 1 khz and 2 khz (Fig. 4B). There are three primary components of BC hearing, as suggested by Tonndorf, which are the distortional, inertial-ossicular, and external canal-osseotympanic components. 13 The distortional component includes the distortion and stimulation of the otic capsule caused by vibrational sound forces. The hearing gain from this component does not differ between groups. Inertial-ossicular and external canal-osseotympanic components might be involved in the hearing gain seen in the stapes vibroplasty group. Stapes movement occurs when a vibrational sound force is applied onto the skull bones along the axis of the 1440 ossicle, and hearing gain from this movement is called the inertial-ossicular component. The inertial-ossicular component is not applicable to the RW vibroplasty group due to the lack of a stapes. Though the stapes moves equally in both the PORP and stapes vibroplasty groups, a difference in weight (stapes, 3 g; FMT, 25 g) would distribute a different degree of force to the cochlea. The external canalosseotympanic component is also responsible for hearing gain from BC. It is contributed to by the external auditory canal. When a vibrator is placed on the skull, the bony external auditory canal also vibrates and creates a sound. Though most vibrations escape from the unoccluded ear, a portion strikes the tympanic membrane and reduces the BC threshold. Therefore, when a sound is applied to the skull, it is not only transmitted via bone but also equally via the external auditory canal and the ossicular chain.

7 With a reconstructed ossicular chain, an intact tympanic membrane and normal stapes with weights provides hearing gain from BC in the stapes vibroplasty group. Our study demonstrates the efficiency of VSB over ossiculoplasty with HA. A conventional ossiculoplasty showed significant benefits after the operation, especially at around 1 khz, by restoring the ossicular chain and middle ear pressure. Additional amplification via HA could also provide meaningful hearing gain, though such hearing gain was limited to low frequencies. However, an active middle ear implant could provide better FHG than ossiculoplasty with or without HAs, and the hearing gain was outstanding at mid to high frequencies, which are essential for conversation. The patients with VSB were free from feedback or occlusion effects and had better cosmetic outcomes. Also, there is a significant chance for the retraction of the middle ear cavity and the extrusion of the prosthesis in the middle ear cavity in chronically disabled middle ears, 14 whereas no extrusion of the VSB has been reported yet. Moreover, access via the oval window or the RW both showed similar useful outcomes, so that surgery could be performed more easily, regardless of the patient s previous surgical history or the presence of any anatomical disruption. However, all retrospective studies have their limitations. Patients were fitted with different types of HAs with different durations, and whether they had the best fitting HA could not be guaranteed. Patients with a stapes vibroplasty and/or a RW vibroplasty were not in identical conditions during the preoperative period. All patients in the RW vibroplasty group were not capable of receiving a stapes vibroplasty. The stapes vibroplasty was only performed when the suprastructure of stapes was intact and the mobility of the footplate was mobile. Otherwise, the RW vibroplasty was performed. Existence of a stapes would have an effect on hearing, though its mechanism and the degree of the effect are unknown. A 6-month follow-up period was sufficient to compare the techniques; however, for a better comparison, a longer-term follow-up period is needed. Cavity problems, extrusion of the devices, or stapedial fixation can occur anytime in chronically disabled middle ears, and migration of the FMT from the RW niche can develop due to imperfect coupling or stabilization of the FMT. More candidates and further studies are needed to obtain more accurate results. CONCLUSION The use of vibroplasty in patients with mixed hearing loss resulted in positive hearing outcomes when compared to conventional ossiculoplasty with HAs. Based on the data from this study, stapes vibroplasty and RW vibroplasty both provided equivalent benefits, and surgeons are more able to complete coupler vibroplasty in instances of distorted anatomy. The VSB could provide a safe and effective option for cases of conductive and mixed hearing loss. Acknowledgments The authors thank Dong-Su Jang, MFA, Medical Illustrator, Medical Research Support Section, Yonsei University College of Medicine, Seoul, Korea, for his help with the illustrations. BIBLIOGRAPHY 1. Colletti V, Soli SD, Carner M, Colletti L. Treatment of mixed hearing losses via implantation of a vibratory transducer on the round window. Int J Audiol 2006;45: Luers JC, Huttenbrink KB. Vibrant Soundbridge rehabilitation of conductive and mixed hearing loss. Otolaryngol Clin North Am 2014;47: de Abajo J, Sanhueza I, Giron L, Manrique M. Experience with the active middle ear implant in patients with moderate-to-severe mixed hearing loss: indications and results. Otol Neurotol 2013;34: Luers JC, Huttenbrink KB, Zahnert T, Bornitz M, Beutner D. Vibroplasty for mixed and conductive hearing loss. Otol Neurotol 2013;34: Baumgartner WD, Boheim K, Hagen R, et al. The Vibrant Soundbridge for conductive and mixed hearing losses: European multicenter study results. Adv Otorhinolaryngol 2010;69: Gan RZ, Dyer RK, Wood MW, Dormer KJ. Mass loading on the ossicles and middle ear function. Ann Otol Rhinol Laryngol 2001;110: Needham AJ, Jiang D, Bibas A, Jeronimidis G, O Connor AF. The effects of mass loading the ossicles with a floating mass transducer on middle ear transfer function. Otol Neurotol 2005;26: Nishihara S, Aritomo H, Goode RL. Effect of changes in mass on middle ear function. Otolaryngol Head Neck Surg 1993;109: Nishihara S, Goode RL. Experimental study of the acoustic properties of incus replacement prostheses in a human temporal bone model. Am J Otol 1994;15: Wever EG, Lawrence M. Sound conduction in the cochlea. Trans Am Otol Soc 1952;40: Chen Y, Guan X, Zhang T, Gan RZ. Measurement of basilar membrane motion during round window stimulation in guinea pigs. J Assoc Res Otolaryngol 2014;15: Canale A, Dagna F, Cassandro C, et al. Oval and round window vibroplasty: a comparison of hearing results, risks and failures. Eur Arch Otorhinolaryngol 2014;271: Tonndorf J. A new concept of bone conduction. Arch Otolaryngol 1968;87: Blake DM, Tomovic S, Jyung RW. Extrusion of hydroxyapatite ossicular prosthesis. Ear Nose Throat J 2013;92:490,