Acta Oto-Laryngologica ISSN: 0001-6489 (Print) 1651-2251 (Online) Journal homepage: http://www.tandfonline.com/loi/ioto20 Factors affecting hearing deterioration in vestibular schwannoma patients treated with gamma knife radiosurgery: the Asan Medical Center experience Marn Joon Park, Hong Ju Park, Jong Woo Chung, Do Heui Lee, Young Hyun Cho, Young Jun Choi & Joong Ho Ahn To cite this article: Marn Joon Park, Hong Ju Park, Jong Woo Chung, Do Heui Lee, Young Hyun Cho, Young Jun Choi & Joong Ho Ahn (2017): Factors affecting hearing deterioration in vestibular schwannoma patients treated with gamma knife radiosurgery: the Asan Medical Center experience, Acta Oto-Laryngologica, DOI: 10.1080/00016489.2017.1386800 To link to this article: http://dx.doi.org/10.1080/00016489.2017.1386800 View supplementary material Published online: 25 Oct 2017. Submit your article to this journal Article views: 5 View related articles View Crossmark data Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ioto20 Download by: [University of Florida] Date: 27 October 2017, At: 22:56 ACTA OTO-LARYNGOLOGICA, 2017 https://doi.org/10.1080/00016489.2017.1386800 RESEARCH ARTICLE Factors affecting hearing deterioration in vestibular schwannoma patients treated with gamma knife radiosurgery: the Asan Medical Center experience Marn Joon Parka, Hong Ju Parka, Jong Woo Chunga, Do Heui Leeb, Young Hyun Chob, Young Jun Choic and Joong Ho Ahna a Downloaded by [University of Florida] at 22:56 27 October 2017 Department of Otorhinolaryngology Head & Neck Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; bDepartment of Neurosurgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; cDepartment of Radiology and Research Institute of Radiology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea ABSTRACT ARTICLE HISTORY Objectives: To investigate the changes in hearing and to determine factors predicting hearing deterioration in patients with vestibular schwannoma (VS) who undergo gamma knife radiosurgery (GKRS). Design: A retrospective review of medical records in patients diagnosed with VS and initially treated with GKRS at a tertiary care medical center between 1995 and 2015 was performed. Tumor factors (location, volume), parameters related to irradiation to the tumor and cochlea, and distance between the tumor and cochlea were reviewed. Results: Fifty-six patients were included in the final analysis with a mean observation period following GKRS as 24.4 ± 27.8 months. Prior to GKRS, the average pure tone threshold at 500, 1k, 2k, and 4k Hz (PTA4) was 51.0 ± 29.7 dB HL. After GKRS, the mean PTA4 was 71.6 ± 33.3 dB HL. Significant independent odds ratios for hearing deterioration were 8.5 for extracanalicular tumors, 18.8 for more than 10 shots in GKRS, and 12.2 for a distance between the tumor center and cochlea modiolus less than 20 mm. Conclusions: A significant hearing deterioration was shown in 2 years after GKRS. Tumor location, number of radiation shots, and distance between the tumor and cochlea affected hearing level after GKRS. Received 26 July 2017 Revised 12 September 2017 Accepted 17 September 2017 Introduction Vestibular schwannoma (VS) is a slow-growing benign tumor mostly originating from the eighth cranial nerve that accounts for 6–9% of all brain tumors . Due to compression of the cranial nerves and nucleus, symptoms in VS patients depend on tumor location. The treatment options for VS include observation, microsurgical resection, and gamma knife radiosurgery (GKRS). Because less than 25% of patients with VS experience significant hearing decline, some physicians prefer to perform regular serial imaging follow-up, only considering an intervention upon symptomatic development or tumor growth . However, tumor manipulation may lead to complications such as further hearing deterioration, vertigo, or facial palsy. GKRS was first used in 1969 by Leksell et al. . Since then, advances in stereotactic radiosurgery and fractional irradiation technology have led to a 90% tumor control rate, as well as functional preservation of the cranial nerves. Elliott et al.  reported an almost 50% rate of hearing preservation in patients treated with GKRS. Golfinos et al.  particularly stressed the merits of GKRS in patients with small-sized VS (less than 2.8 cm), and Mousavi et al.  KEYWORDS Vestibular schwannoma; acoustic neuroma; gamma knife stereotactic radiosurgery; hearing preservation; hearing deterioration argued that there was a better chance of hearing preservation in patients whose hearing is initially preserved. Despite the benefits of GKRS, the auditory nerve function is the least preserved of that of all cranial nerves in VS patients managed by GKRS. A decline in hearing was observed in 30–35% of patients, whereas over 98% of patients with VS managed by GKRS showed preservation of both the facial and trigeminal nerves . Therefore, many recent studies have focused on hearing preservation in VS after GKRS. However, it is still unclear which clinical factors account for the hearing deterioration and to what extent. Therefore, we investigated the changes in hearing by frequency and prognostic factors for hearing deterioration in patients with VS who underwent GKRS. Methods Study subjects This retrospective review of medical records was performed with patients confirmed with VS at a tertiary care medical center in South Korea between January 1995 and December 2015. Among the 993 patients who were diagnosed with VS in those 10 years, patients who underwent GKRS without CONTACT Joong Ho Ahn firstname.lastname@example.org Department of Otolaryngology, Asan Medical Center, University of Ulsan College of Medicine, 88 Olympic-ro 43-gil, Songpa-gu, Seoul 05505, Republic of Korea Supplemental data for this article can be accessed here. ß 2017 Acta Oto-Laryngologica AB (Ltd) Downloaded by [University of Florida] at 22:56 27 October 2017 2 M. J. PARK ET AL. Figure 1. Study design and patient recruitment. An illustrated description of the study design. Between 1995 and 2015, patients confirmed to have VS on TBMRI who underwent GKRS were retrospectively reviewed. In total, 56 patients were recruited for the final statistical analysis. TBMRI: gadolinium-enhanced temporal bone magnetic resonance imaging; GKRS: gamma knife radiosurgery; NF-II: neurofibromatosis type II; PTA: pure tone audiometry; SA: speech audiometry; VS: vestibular schwannoma. any prior treatment (e.g. microsurgical resection) for VS were included. Patients who did not undergo pure tone audiometry (PTA) with speech audiometry (SA) both prior to and after GKRS at least once were excluded. Additionally, patients with bilateral VS and neurofibromatosis type II (NF-II) were excluded. Patients with a follow-up period less than 6 months following GKRS were also excluded. In total, 56 patients were included in the final analysis (Figure 1). This study was approved by institutional review board (IRB) of Asan Medical Center IRB (approval no. 2016-1007). Gadolinium-enhanced temporal bone magnetic resonance imaging All 56 patients had temporal bone MRI (TBMRI) obtained by using a 3T magnetic resonance imaging system (Achieva; Philips Healthcare, Best, The Netherlands) with an 8-channel head coil. Coronal T1- and T2-weighted turbo spinecho, axial three-dimensional (3D) T2-weighted, pre-contrast and post-contrast 3D fluid-attenuated inversion recovery (FLAIR), and post-contrast T1-weighted imaging were performed. Gadolinium-DOTA (Dotarem; Guerbet, Paris, France) at 0.2 mmol/kg of body weight was used for the contrast enhancement. Board-certified radiologists confirmed the diagnosis of VS. Hearing assessments: pure tone audiometry and speech audiometry Initial hearing was assessed with PTA and SA. For PTA, with the masking of the contralateral ear, warble-tone thresholds were measured at 250, 500, 1000, 2000, 3000, 4000, and 8000 Hz and then averaged (PTA4). For SA, word discrimination scores (WDSs) at 40 and 60 dB SPL in quiet were measured using the Korean standard word lists for adults . Patients’ initial hearing levels were stratified using the American Academy of Otolaryngology – Head and Neck Surgery (AAO-HNS) classification  (Supplementary Table S1). GKRS and follow-up In all cases, GKRS was performed by applying a Leksell Gamma knife (B. and C. Elekta, Stockholm, Sweden) device in the following order. First, a Leksell stereotactic coordinate frame was applied on the recipients’ vertex. Then, a gadolinium contrast magnetic resonance T1-weighted image with 2-mm slice reconstruction on the axial plane was taken. Afterward, the GKRS was designed using GammaPlan (Elekta) instrument software. The prescribed dose at the tumor margin was set at 12.2 ± 0.5 Gy (minimum, 10.0 Gy; maximum, 13.0 Gy) (Table 1), giving a 50% median isodose line. Leksell GammaPlanV software (version 10.2.1, Stockholm, Sweden) was used to accurately calculate the amount of irradiation received at each point and a distance–dose curve for the irradiation dosage was subsequently drawn (Figure 2). After GKRS, all patients were admitted for 1 day. Contrast-enhanced head computed tomography (CT) was conducted immediately after GKRS to rule out any immediate GKRS-related complications. After discharge, the patients were asked to attend a 1-week postoperative visit and a yearly follow-up to undergo TBMRI using the above protocol. By applying the criteria suggested by Kanzaki et al.  in 2003, the response of VS to GKRS was categorized as regression, stable, and progression. Patients were asked to undergo a baseline hearing assessment within 1 month before the administration of GKRS. However, due to the retrospective nature of the study, only a small number of patients had their hearing evaluated after GKRS, in an inconsistent manner. R Measured variables Patient sex and age at diagnosis were collected. Changes in clinical symptoms (tinnitus, facial palsy, vertigo, and facial ACTA OTO-LARYNGOLOGICA Table 1. Demographic and clinical characteristics of the study patients (n ¼ 56). Downloaded by [University of Florida] at 22:56 27 October 2017 Characteristics Age at diagnosis Sex (male/female), n (%) Tumor characteristics Lateralization (L/R), n (%) Intracanalicular/extracanalicular, n (%) Tumor volume (cm3) LAD (mm) Long–short-axis ratioa Distance parameters Center-to-center distance (mm)b Minimal distance (mm)c Irradiation parameters related to GKRS Prescribed dose (Gy) Maximal dose (Gy) Minimal dose (Gy) Mean dose (Gy) Cochlear dose (Gy) Number of shots Observation period following GKRS (months) Tumor response to GKRSd Regression/stable/progression, n (%) Immediate complications following GKRS, n (%) Hydrocephalus Limb weakness Mean ± SD (Min–Max) 50.2 ± 12.0 (16–71) 24/32 (42.9/57.1) 30/26 15/41 2.07 ± 2.36 17.2 ± 6.9 1.60 ± 0.58 (53.6/46.4) (26.8/73.2) (0.01–9.10) (4.0–30.8) (1.0–4.4) 11.4 ± 6.0 (4.0–60.7) 3.3 ± 2.6 (1.0–13.3) 12.2 ± 0.5 24.1 ± 1.5 9.8 ± 1.5 17.0 ± 1.1 4.6 ± 1.8 8.5 ± 6.0 24.4 ± 27.8 (10.0–13.0) (20.0–26.2) (6.2–14.8) (14.3–19.1) (2.1–8.9) (1–26) (6.1–137.2) 24/26/6 (42.9/46.4/10.7) 2 (3.6) 1 (1.8) GKRS: gamma knife radiosurgery; L: left; LAD: long-axis diameter; R: right; SD: standard deviation. a Long-axis diameter divided by short-axis diameter. b Distance between the tumor center and cochlea modiolus (mm). c Minimal distance between the tumor and cochlea modiolus (mm). d Tumor response following GKRS was categorized by applying the generally accepted criteria . 3 numbness) and audiometric data before and after GKRS were also reviewed. The time interval after GKRS was regarded as the ‘observation period’, and the final state of the tumor was assessed. On the obtained TBMRI, the location of the tumor was identified according to its presence on the contrast-enhanced T1-weighted image: VS limited to the intra-auditory canal (IAC) was defined as ‘intracanalicular’ (Figure 3(A)), whereas VS partially overlapping between the IAC and cerebellopontine angle or located predominantly at the cerebellopontine angle was defined as ‘extracanalicular’ (Figure 3(B)). The long- and short-axis diameters were measured, and the long–short-axis ratio (the long-axis diameter divided by the short-axis diameter) was calculated for each tumor. On T2-weighted imaging, the spatial relationship between the cochlea and the tumor was analyzed, and the following two distance parameters were defined (Figure 4): center-to-center distance, the distance between the tumor center and the cochlea modiolus; minimal distance, the minimal distance between the tumor and the cochlea modiolus. For the variables associated with GKRS, the parameters reflecting the dose of irradiation were defined as follows: prescribed dose, the ‘intended dose’ at the 50% median isodose line, drawn along the border of the tumor; mean dose, Figure 2. Dose of irradiation to the tumor, cochlea, and adjacent structures during GKRS in VS patients. An explanatory model for the higher irradiation dose to the cochlea that elevates the risk of hearing deterioration according to the tumor location in GKRS planning. The upper plots show the percentage deposition of the irradiation dose in relation to the relative distance in the GKRS while the lower images show magnified T2-weighted TBMRI of two VS patients. A shows a rightsided intracanalicular VS in a 34-year-old male, whereas B shows a right-sided extracanalicular VS in a 44-year-old female. By using Leksell GammaPlanV (version 10.2.1), the expected irradiation discrimination margin (yellow line; light-gray line in black & white printing) is designed to be almost parallel to the border of the tumor (red line; dark-gray line in black & white printing). Although the minimal and central distances between the cochlea (arrowheads) and the expected margin of irradiation are much shorter in intracanalicular VS (A) than in extracanalicular VS (B), note that a similar dose of irradiation is administered to the cochlea in the two patients. Additionally, it is notable that the slope of the tangent is much more precipitous in an intracanalicular tumor (16.23) than in an extracanalicular tumor (5.06) at the point where 50% of the maximum irradiation is expected. This model is a possible explanation of the poorer hearing in patients with extracanalicular VS following GKRS. R Downloaded by [University of Florida] at 22:56 27 October 2017 4 M. J. PARK ET AL. Figure 3. Definitions of intracanalicular and extracanalicular VSs. Gd-enhanced T1-weighted images of TBMRI of two patients diagnosed with VS. Image A shows an intracanalicular VS (arrow) located exclusively in the right internal auditory canal. Image B shows an extracanalicular VS (arrow) extruding out to the cerebellopontine angle. Table 2. Changes in clinical manifestations following GKRS (n ¼ 56). Cranial nerve function b PTA4 (dB HL) Word discrimination score (%)b Initial hearing, n (%) Serviceable/nonserviceable Tinnitus, n (%) Facial palsy, n (%) Vertigo, n (%) Facial numbness, n (%) Initial (prior to GKRS) Final (following GKRS) p Valuea 51 ± 30 (3–100) 55 ± 42 (0–100) 72 ± 33 (5–120) 39 ± 39 (0–100) <.001 .001 28/28 (50/50) 28 (50.0) 1 (1.8) 18 (32.1) 1 (1.8) 13/43 (23.2/76.8) 32 (57.1) 6 (10.7) 28 (50) 3 (5.4) .003 .449 .113 .054 .618 GKRS: gamma knife radiosurgery; PTA4: average pure tone threshold at 500, 1000, 2000, and 4000 Hz; SD: standard deviation. a Comparison between the initial and final states: p values were calculated using a paired t-test with Bonferroni correction, chi-square test, or Fisher’s exact test. Bold text indicates p < .05. b Values are presented as mean ± SD (Min–Max). Figure 4. Definition of the distance parameters: spatial relationship between the cochlea and the tumor. Magnified TBMRI of a 53-year-old female patient diagnosed with left extracanalicular-type VS. A T2-weighted image shows the inner ear structure as well as the tumor (arrowhead, basal turn of the left cochlea; asterisk, left lateral semicircular canal). Distance parameters reflecting the spatial relationship between the cochlea and the tumor were measured. Arrow A indicates the minimal distance, defined as the length between the cochlea modiolus and the tumor; Arrow B indicates the center-to-center distance, defined as the length between the cochlea modiolus and the center of the tumor. the average dose of irradiation received at the 50% median isodose line; maximal dose, the maximal amount of irradiation received along the 50% median isodose line; minimal dose, the minimal amount of irradiation received along the 50% median isodose line; cochlear dose, the average irradiation dosage at four points at the cochlea border. The number of fractioned beams of irradiation given to the tumor to achieve the target dose of irradiation was defined as the ‘number of shots’, which was determined by the physician (D.H.L.). Statistical analysis A paired t-test with Bonferroni correction and a chi-square test or Fisher’s exact test were used to compare the changes in the clinical values and proportions between the initial and final states (Table 2 and Figure 5). Student’s t-test and chi-square/Fisher’s exact test were used for the comparison of the mean values and proportions between the preserved and deteriorated hearing groups (Table 3). To compare the clinical variables between serviceable and nonserviceable hearing (Supplementary Table S2) and between intra- and extracanalicular VS (Table 6), Student’s t-test with Levene’s test and chi-square/Fisher’s exact test were used to assess the equality of variances. One-way independent analysis of variance (ANOVA) was used to evaluate the significance differences in the level of changes associated with each frequency before and after GKRS (Figure 6). To calculate the odds ratio (OR) for hearing deterioration for each clinical factor, binary logistic regression analysis was used. Multivariate binary logistic regression analysis with backward elimination was performed with variables found to be significant (p < .05) in univariate analysis. The estimated ORs and 95% confidence intervals (CIs) were calculated as well (Tables 4 and 5). Two-sided p values less than .05 were considered significant. All statistical analyses were performed using IBM SPSS software version 23.0 (IBM, Armonk, NY). ACTA OTO-LARYNGOLOGICA 5 Downloaded by [University of Florida] at 22:56 27 October 2017 and the mean tumor volume was 2.07 cm3. The mean distance from the central portion of the tumor to the center of the cochlea modiolus was 11.4 mm and the mean minimal distance between the tumor margin and the modiolus was 3.3 mm. Although a mean of 12.2 Gy of irradiation was designed to be given at the tumor margin, a mean value of 17.0 Gy was actually irradiated at the tumor margin when calculated retrospectively. The mean irradiation dose received at the cochlea was determined to be 4.6 Gy, with a minimum of 2.1 Gy and a maximum of 8.9 Gy. The average number of shots given was 8.5, ranging from 1 to 26. The mean observation period after GKRS was 24.4 months. Upon follow-up, 24 patients (42.9%) had tumor regression, 26 (46.4%) had a stable tumor, and six (10.7%) had tumor progression despite GKRS. Four patients (7.1%) had additional microsurgical resection of the tumor for uncontrolled growth despite GKRS. Immediately after GKRS, one patient (1.8%) complained of mild weakness of the lower limb but showed spontaneous recovery within 1 month with conservative management. Hydrocephalus was evident in two patients (3.6%) in contrast-enhanced head CT following GKRS. One patient had spontaneous shrinkage of the ventricle upon close follow-up, whereas another patient had persistent hydrocephalus with no symptoms. This patient underwent a ventricleperitoneal shunt insertion for hydrocephalus management. Changes in the clinical outcome after GKRS Following GKRS, a significant increase was seen in the PTA4 and a decrease in the WDS, with an increased proportion of patients with nonserviceable hearing (both p < .05) (Table 2). In comparison, no significant changes were observed in the other clinical manifestations. There was a significant increase in the pure tone thresholds at each frequency and a decrease in the WDS in all 56 patients and in patients with serviceable hearing (Figure 5). Similarly, in the nonserviceable hearing patient group, significant changes were seen in every register, except 250 Hz and WDS. There were no differences between the level of changes in each frequency (Figure 6). Factors associated with hearing deterioration Figure 5. Changes in hearing following GKRS. Changes in hearing from before GKRS to the final hearing in terms of frequencies and word discrimination scores. In all 56 patients (above) and in the serviceable hearing group (middle), all seven frequencies showed a significant increase in the pure tone threshold (left) in addition to a significant decrease in word discrimination scores (right). In the nonserviceable hearing group (below), the word discrimination score (right) and the pure tone threshold (left) showed no significant increase with 250 Hz but the pure tone threshold level significantly increased from 500 to 8000 Hz. p < .05 is indicated with an asterisk (). Results Patient characteristics This study included 56 patients – 24 men (42.9%) and 32 women (57.1%) – with a median age at diagnosis of 50.2 years. Patient characteristics are summarized in Table 1. Among the 56 VSs, 15 (26.8%) were intracanalicular tumors, The center-to-center distance was higher in the preserved hearing group in all patients and in the serviceable hearing group (both p < .05 (Table 3). The dose of irradiation at the cochlea was higher in the group with hearing deterioration in all patients and in patients with serviceable hearing (both p < .05). Although the number of shots given was significantly higher in the hearing deterioration group only in nonserviceable hearing patients (p < .05), weak significance was shown for a higher number of shots in all 56 patients in the hearing deterioration group (p < .1). A higher proportion of extracanalicular tumors and a lower long–short-axis ratio were seen in patients with serviceable hearing in the hearing deterioration group (all p < .05). Univariate analyses showed that center-to-center distance (>20 mm), cochlear irradiation dose (>5 Gy), number of 6 M. J. PARK ET AL. Table 3. Comparison of clinical characteristics according to hearing deteriorationa following GKRS. All patients (n ¼ 56) Downloaded by [University of Florida] at 22:56 27 October 2017 Preserved (n ¼ 28) Deteriorated (n ¼ 28) Age, y 51.0 ± 10.3 49.5 ± 13.7 Sex (M/F), n (%) 10/18 (36/64) 14/14 (50/50) 58.3 ± 32.1 43.6 ± 25.5 Initial PTA4 (DB HL) Initial WDS (%) 55.6 ± 40.7 55.2 ± 43.8 Distance parameters 14.2 ± 3.8 10.9 ± 4.4 Center-to-center (mm)b 3.3 ± 2.8 3.3 ± 2.5 Minimal distance (mm)c Radiation parameters Prescribed dose (Gy) 12.2 ± 0.4 12.1 ± 0.7 Maximal dose (Gy) 23.8 ± 1.6 24.3 ± 1.3 Minimal dose (Gy) 9.9 ± 1.7 9.8 ± 1.4 Mean dose (Gy) 17.1 ± 0.9 16.9 ± 1.2 Cochlear dose (Gy) 3.7 ± 1.3 5.4 ± 1.8 Number of shots 6.9 ± 5.3 9.9 ± 6.4 Tumor characteristics 1.71 ± 2.28 2.43 ± 2.42 Tumor volume (cm3) Intra/extracanalicular, n (%) 11/17 (39/61) 4/24 (14/86) LAD (mm) 15.8 ± 7.3 18.7 ± 6.2 1.6 ± 0.4 1.6 ± 0.7 Long–short-axis ratiob Tumor response to GKRSf Regression/stable/ 8/17/3 (29/61/10) 17/8/3 (61/29/10) progression, n (%) Serviceable hearing (n ¼ 28) Nonserviceable hearing (n ¼ 28) p valueb Preserved (n ¼ 13) Deteriorated (n ¼ 15) p valueb .645 .280 .064 .976 45.3 ± 9.7 5/8 (38/62) 33.5 ± 16.9 88.7 ± 13.2 53.1 ± 7.8 7/8 (47/53) 27.2 ± 15.3 92.5 ± 16.5 .047 .963 15.5 ± 4.6 4.0 ± 3.7 .922 .149 .978 .636 .015 .076 .257 .068 .123 .843 .050 Preserved (n ¼ 15) Deteriorated (n ¼ 13) p valueb .025 .662 .307 .514 55.3 ± 9.0 5/10 (33/67) 80.5 ± 24.9 24.0 ± 30.4 45.3 ± 17.6 7/6 (54/46) 62.2 ± 21.8 12.8 ± 18.4 .064 .274 .051 .262 9.8 ± 4.1 3.9 ± 2.9 .032 .887 11.6 ± 5.7 2.4 ± 1.3 11.1 ± 4.7 3.0 ± 2.1 .810 .367 12.1 ± 0.4 23.8 ± 1.6 9.7 ± 2.3 17.1 ± 0.9 3.2 ± 0.8 6.8 ± 5.3 11.9 ± 0.7 23.9 ± 1.4 9.6 ± 1.1 16.5 ± 1.0 4.5 ± 1.1 8.4 ± 4.6 .362 .764 .910 .137 .037 .424 12.2 ± 0.4 23.8 ± 1.7 10.0 ± 1.1 17.0 ± 1.0 4.6 ± 1.5 7.3 ± 3.6 12.4 ± 0.5 24.8 ± 1.0 10.1 ± 1.6 17.3 ± 1.2 6.2 ± 1.9 11.1 ± 4.2 .298 .075 .883 .453 .197 .017 1.41 ± 1.93 7/6 (54/46) 14.5 ± 7.9 1.7 ± 0.5 2.47 ± 2.59 2/13 (13/87) 18.0 ± 6.5 1.4 ± 0.3 .244 .042 .204 .046 2.05 ± 2.59 4/11 (27/73) 16.8 ± 7.2 1.5 ± 0.3 2.21 ± 2.26 2/11 (15/85) 19.2 ± 5.9 1.9 ± 1.0 .862 .655 .346 .112 3/9/1 (23/29/8) 9/5/1 (60/33/7) .093 5/8/2 (33/53/13) 7/4/2 (54/31/15) .473 F: female; GKRS: gamma knife radiosurgery; LAD: long-axis diameter; M: male; PTA4: average pure tone threshold at 500, 1000, 2000, and 4000 Hz; WDS: word discrimination score. Values indicate the mean ± standard deviation. a Definition of deteriorated hearing: patient with DPTA4 greater than or equal to 15 dB HL (final PTA4 following GKRS – PTA4 prior to GKRS); preserved hearing: patient with DPTA4 less than 15 dB HL. b For the comparison of the mean values between the preserved hearing group and the deteriorated hearing group, a Student's t-test was used. To compare the differences in the proportion, a chi-square test or a Fisher's exact test was used. Bold text indicates p < .05. c Distance between the tumor center and the cochlea modiolus (mm). d Minimal distance between the tumor and the cochlea modiolus (mm). e Long-axis diameter divided by short-axis diameter. f Tumor response following GKRS was categorized by applying the generally accepted criteria . Discussion Figure 6. Changes in pure tone thresholds at each frequency (n ¼ 56). In all 56 patients, ~PTA (final PTA – pre-GKRS PTA) at each of the seven frequencies (250, 500, 1000, 2000, 3000, 4000, 8000 Hz) were calculated and are shown in Figure 4. ~PTAs at all seven frequencies were compared using a one-way independent ANOVA test, which showed no significant differences between the frequencies (p ¼ .703). shots given (>10), tumor location (extracanalicular tumors), and tumor volume (>0.6 cm3) were significantly associated with hearing deterioration (DPTA4 15 dB HL) (Table 4). Subsequent multivariate analyses showed that a center-tocenter distance less than 20 mm, more than 10 shots, and extracanalicular tumors were independent predictive factors for hearing deterioration in all 56 patients and in patients with serviceable hearing (all p < .05) (Table 5). The present study shows the degree of hearing changes and the factors related to hearing deterioration following GKRS in 56 patients diagnosed with VS. During the mean observation period of 2 years, we found a significant decrease in hearing at all frequencies and in word discrimination. In addition, tumor location, tumor volume, number of shots, amount of irradiation received at the level of the cochlea, and distance between the tumor and cochlea modiolus were related to hearing deterioration. An extracanalicular VS, more than 10 shots, and a distance between the tumor center and cochlea modiolus less than 20 mm were independent risk factors for hearing deterioration in all 56 patients, with a higher OR evident in patients who presented with serviceable hearing. Therefore, this study reveals the changes in the final hearing outcomes of VS patients treated with GKRS and allows final hearing to be predicted from initial clinical variables. The exact mechanism and factors contributing to hearing decline in patients with untreated VS are not yet fully understood. In patients with NF-II, the tumor itself was shown to secrete neurotoxins, and a higher concentration of secreted toxin was related to more aggressive hearing decline, regardless of tumor size, and volume . However, in patients with unilateral and sporadic VS, it is unclear whether these toxins are related to advanced hearing decline. Our study shows that none of these clinical factors were associated with a poorer degree of hearing in patients with ACTA OTO-LARYNGOLOGICA 7 Table 4. Univariate analysis of factors associated with hearing deteriorationa in VS patients following GKRS. All patients (n ¼ 56) OR (95% CI) Downloaded by [University of Florida] at 22:56 27 October 2017 Age >60 y Distance parameters Center-to-center distance <20 mm Minimal distance <2 mm Radiation parameters Maximal dose >24 Gy Minimal dose >10 Gy Mean dose >17 Gy Cochlear dose >5 Gy Number of shots >10 shots Tumor characteristics Extra-/intra-canalicular Volume >0.6 cm3 Long–short-axis ratio <1.5 Tumor response to GKRSc Regression Stable Progression Serviceable hearing (n ¼ 28) p value b OR (95% CI) Nonserviceable hearing (n ¼ 28) p value b OR (95% CI) p valueb 1.62 (0.59–8.54) .399 6.00 (0.60–60.16) .128 0.98 (0.22–4.34) .978 9.00 (1.03–78.94) 1.56 (0.24–10.14) .047 .641 7.58 (1.20–48.00) 1.95 (0.27–13.98) .031 .506 3.00 (0.27–33.09) 1.19 (0.55–10.87) .370 .608 1.67 2.46 1.71 6.64 8.00 (0.53–5.28) (0.67–9.03) (0.46–6.39) (1.48–14.54) (2.19-29.25) .385 .174 .428 .009 .002 1.13 1.46 2.22 6.29 8.25 .885 .665 .379 .047 .024 3.43 1.91 2.33 6.40 7.58 (0.65–18.22) (0.33–11.01) (0.40–13.61) (1.18–34.61) (1.20–48.00) .148 .472 .346 .031 .031 11.50 (3.24–40.86) 5.31 (1.57–17.97) 2.06 (0.71–5.98) <.001 .024 .184 21.68 (3.02–155.36) 6.50 (1.60–50.69) 3.20 (0.68–15.07) <.001 .013 .141 2.00 (0.30–13.27) 1.67 (0.31–9.01) 1.68 (0.32–8.76) .473 .553 .538 1 (reference) 0.27 (0.08–0.85) 0.50 (0.08–3.06) .084 .069 .453 1 (reference) 0.19 (0.03–1.02) 0.33 (0.02–7.14) .152 .052 .482 1 (reference) 0.36 (0.07–1.88) 0.71 (0.07–6.92) .473 .224 .772 (0.23–5.54) (0.26–8.05) (0.38–13.18) (1.02–38.65) (1.33–51.26) CI: confidence interval; GKRS: gamma knife radiosurgery; OR: odds ratio. a Deterioration of hearing defined as DPTA4 greater than or equal to 15 dB HL (final PTA4 following GKRS – PTA4 prior to GKRS). b Univariate binary logistic regression analysis was used to calculate the OR. Bold text indicates p < .05. c Tumor response following GKRS was categorized by applying the generally accepted criteria . Table 5. Multivariate analysis of factors associated with hearing deteriorationa in VS patients following GKRS. All patients (n ¼ 56) Distance parameter Center-to-center distance <20 mm Radiation parameters Cochlear dose >5 Gy Number of shots >10 Tumor characteristics Extra-/intra-canalicular Volume >0.6 cm3 Serviceable hearing (n ¼ 28) Nonserviceable hearing (n ¼ 28) OR (95% CI) p value OR (95% CI) p value OR (95% CI) p valueb 12.23 (2.31–254.59) .015 17.20 (1.44–167.07) .019 7.42 (0.38–146.58) .188 3.55 (0.69–18.26) 18.75 (1.58–223.08) .129 .032 1.06 (0.06–4.50) 37.02 (1.36–305.61) .197 .020 12.03 (0.78–186.11) 14.68 (0.53–406.00) .075 .113 8.54 (1.29–56.34) 2.30 (0.34–15.49) .026 .391 28.25 (1.78–783.97) 17.66 (0.71–43.93) .026 .080 3.97 (0.10–156.64) 3.35 (0.09–120.56) .462 .509 b b CI: confidence interval; GKRS: gamma knife radiosurgery; OR: odds ratio. a Deterioration of hearing defined as DPTA4 greater than or equal to 15 dB HL (final PTA4 following GKRS – PTA4 prior to GKRS). b Multivariate binary logistic regression analysis was used to calculate the OR. Bold text indicates p < .05. unilateral VS (Supplementary Table S2), indicating the need for future studies in VS patients who are managed with surveillance alone. During recent decades, the success rate for the management of VS with GKRS has been increasing and there has been a decrease in the dysfunction rate of the cranial nerves V, VII, and VII and in postoperative complications, such as hydrocephalus. Our study found a controlled tumor (regression or stable tumor) in 50 patients (89.3%) who underwent GKRS (Table 1), similar to the outcomes of recently published articles . The ratio of patients with serviceable hearing to patients with nonserviceable hearing significantly changed from 1:1 to 1:3.3. However, although there was an increase in the percentage, no significant changes were observed in the proportion of patients with tinnitus, facial palsy, vertigo, and facial numbness. When compared with other articles, our rate of vertigo and tinnitus development was higher . Our results have also counted the number of patients with ‘transient’ dysfunction of the lower cranial nerve, in addition to patients with ‘permanent’ dysfunction, which explains the higher rate of lower cranial nerve dysfunction following GKRS in our study compared with previous reports. Once the deterioration in hearing began after GKRS, none of the patients showed hearing, facial palsy, and facial numbness recovery. In contrast, three patients (5.3%) with tinnitus and 11 patients (19.6%) with vertigo achieved a full recovery upon conservative management during the follow-up period (data not shown). The immediate post-GKRS complication rate in our study was less than 5%. Altogether, our 10-year results show similar rates of tumor control and complications to previous studies. Many previous attempts have been made to ascertain what truly determines the overall hearing following GKRS. Kano et al.  and Han et al.  reported that older patients (>60 years) with poorer initial hearing were more likely to have hearing deterioration following GKRS. Hayden et al.  argued that a poorer hearing outcome was associated with a larger tumor volume, but this finding was not in accordance with those of other studies. The irradiation dosage delivered to the tumor was also found to be an adverse factor when it comes to hearing preservation in the treatment of VS. Foote et al.  believed that a mean dose greater than 12 Gy administered to the tumor margin was associated with a poorer hearing outcome. Yoma et al.  measured the amount of irradiation at the cochlea in a similar fashion to our study, showing that a cochlear dosage 8 M. J. PARK ET AL. Table 6. Comparison of clinical parameters between intracanalicular and extracanalicular VS (n ¼ 56). 3 Tumor volume (cm ) Long–short-axis ratio Number of shots Cochlear dose (Gy) Downloaded by [University of Florida] at 22:56 27 October 2017 a Intracanalicular Extracanalicular p valuea 0.20 ± 0.02 2.0 ± 0.2 3.5 ± 0.5 3.4 ± 1.3 2.76 ± 0.38 1.5 ± 0.1 10.1 ± 1.0 5.1 ± 1.7 <.001 .002 <.001 .034 p value calculated by using Student's t-test. Bold text indicates p < .05. larger than 4.2 Gy was significantly associated with hearing deterioration after GKRS. Massager et al.  claimed that patients with VS with a small intracanalicular volume were more likely to have their cochlea exposed to radiation during GKRS and suggested that the cochlear dose should be reduced to less than 4.2 Gy, if possible. In 2013, Linsky et al. published an article on the received radiation dose and the tolerance of suborgans in the human temporal bone in patients with VS treated with GKRS . The basal turn near the modiolus and the inferior part of the cochlea were the most susceptible, with a received dose greater than 12 Gy in 10–15% of cases. Although our results did not show increased risk in the older age group and with a higher dose of irradiation at the tumor, in addition to the differences in the degree of initial hearing with hearing deterioration, we did find that a cochlear dose exceeding 5 Gy, tumor volume over 600 cm3, extracanalicular volume, and a center-to-center distance more than 20 mm were significantly associated with hearing deterioration. Combining these previous findings with our results, we strongly suggest that the irradiation dose to the cochlea, particularly the modiolus, is the most important determining factor for hearing preservation in VS patients undergoing GKRS. In addition, we identified significant differences in clinical parameters in intra-versus extracanalicular VS (Table 6). Intracanalicular VSs showed a smaller tumor volume and higher long–short-axis ratio (indicative of a more cylinder-like shape), reflecting the tumors’ predominant location in the IAC. In addition, significant differences were seen in the number of shots given and the irradiation dose to the cochlea between intra- and extracanalicular VS. To account for these findings, an explanatory model is presented in Figure 2. The dose of irradiation received at each point follows an almost normal distribution pattern, with the maximal dose at the center of the tumor and 50% of the maximum dose (mean dose) at the border of the tumor, as designed during GKRS planning with the obtained MR image. This explains why a longer center-to-center distance is associated with poorer hearing preservation, because the cochlea will receive a higher dose of irradiation when the cochlea is closer to the tumor. Using patients A and B in Figure 2 as examples, the 50% iso-radiation dosage at the tumor margin (yellow) and the maximal dose received at the center of the tumor were almost identical in the two patients. In addition, the dose received at the point of the cochlea modiolus was almost similar in the two cases, even though patient B shows a longer distance between the cochlea and the tumor. The slope of each tangent was calculated at the point of the 50% maximal dose given, with a higher slope of 16.23 in patient A (intracanalicular, cylinder-shaped, and lower tumor volume) and a slope of 5.06 in patient B (extracanalicular, round-shaped, and larger tumor volume). It can be acknowledged that the lower the slope of a tangent, the lower the likelihood that the targeting tissue be accurately discriminated and the organs preserved. Thus, we conclude that the tumor location and the distance between the cochlea and the tumor are the two most important factors determining VS patients’ hearing outcome following GKRS. Regarding hearing preservation and rehabilitation, the tumor should be managed more prudently in patients with initial serviceable hearing (class A and B). Clinicians should be careful when planning interventions, and a more precise effort should be made when GKRS is designed in these patients. In addition, many audiologists have reported successful hearing rehabilitation in patients with moderate-degree sensorineural hearing loss (average pure tone thresholds >50 dB HL and WDS >50%) with various hearing aids . Because patients classified as class C (PTA4 > 50 dB HL and WDS >50%) could be candidates for hearing rehabilitation with hearing aids, an effort should also be made in class C hearing patients. Interestingly, we have also discovered that stable tumors showed the lowest likelihood of hearing loss, followed by progressive tumors, despite GKRS, although the significance was weak in patients with serviceable hearing. We believe that these findings could be due to excess irradiation that enabled the complete eradication of the tumor in patients with tumor regression and the effect of tumor re-growth in patients with progressive tumors. However, due to the low power of the study, this hypothesis needs to be proven in future studies. Our study has some limitations due to its retrospective nature: (1) the time to hearing deterioration following GKRS was not reviewed, (2) there were no protocols for a uniform hearing assessment in all patients, (3) although most of the patients underwent the baseline hearing assessment prior to GKRS, only a small percentage of the patients had their hearing assessed following GKRS, (4) there was a marked difference in the follow-up period among the patients, with a standard deviation of 27.8 months, and (5) despite effort to include GKRS recipients from the last 10 years to minimize the time period bias, we acknowledge that two different gamma knife irradiation and planning devices were applied. So as to not regard any early hearing deterioration as the ‘final hearing’, we have excluded patients with less than a 6month follow-up. Although the data were not presented, we observed no significant additional changes in patients’ hearing after the 6-month follow-up period. Disclosure statement The authors have no conflicts of interest to disclose. References  Yamakami I, Uchino Y, Kobayashi E, et al. Conservative management, gamma-knife radiosurgery, and microsurgery for ACTA OTO-LARYNGOLOGICA      Downloaded by [University of Florida] at 22:56 27 October 2017     acoustic neurinomas: a systematic review of outcome and risk of three therapeutic options. Neurol Res. 2003;25:682–690. Jethanamest D, Rivera AM, Ji H, et al. Conservative management of vestibular schwannoma: predictors of growth and hearing. Laryngoscope. 2015;125:2163–2168. Leksell L. A note on the treatment of acoustic tumours. Acta Chir Scand. 1971;137:763–765. Elliott A, Hebb AL, Walling S, et al. Hearing preservation in vestibular schwannoma management. Am J Otolaryngol. 2015;36:526–534. Golfinos JG, Hill TC, Rokosh R, et al. A matched cohort comparison of clinical outcomes following microsurgical resection or stereotactic radiosurgery for patients with small- and medium-sized vestibular schwannomas. J Neurosurg. 2016;125: 1472–1482. Mousavi SH, Kano H, Faraji AH, et al. Hearing preservation up to 3 years after gamma knife radiosurgery for Gardner– Robertson class I patients with vestibular Schwannomas. Neurosurgery. 2015;76:584–590. discussion 90-1. Noren G. Long-term complications following gamma knife radiosurgery of vestibular schwannomas. Stereotact Funct Neurosurg. 1998;70(Suppl1):65–73. Lee J. Standardization of Korean speech audiometry. Audiol Speech Res. 2016;12:S7–S9. Committee on hearing and equilibrium guidelines for the evaluation of hearing preservation in acoustic neuroma (vestibular schwannoma): committee on hearing and equilibrium. Otolaryngol Head Neck Surg. 1995;113:179–180. Kanzaki J, Tos M, Sanna M, et al. New and modified reporting systems from the consensus meeting on systems for reporting results in vestibular schwannoma. Otol Neurotol. 2003;24: 642–648. discussion 8-9.           9 Asthagiri AR, Vasquez RA, Butman JA, et al. Mechanisms of hearing loss in neurofibromatosis type 2. PLoS One. 2012;7:e46132. Kano H, Kondziolka D, Khan A, et al. Predictors of hearing preservation after stereotactic radiosurgery for acoustic neuroma: clinical article. J Neurosurg. 2013;119(Suppl):863–873. Flickinger JC, Kondziolka D, Niranjan A, et al. Results of acoustic neuroma radiosurgery: an analysis of 5 years’ experience using current methods. J Neurosurg. 2013;119(Suppl):1–6. Han JH, Kim DG, Chung HT, et al. Hearing preservation in patients with unilateral vestibular schwannoma who undergo stereotactic radiosurgery: reinterpretation of the auditory brainstem response. Cancer. 2012;118:5441–5447. Hayden Gephart MG, Hansasuta A, Balise RR, et al. Cochlea radiation dose correlates with hearing loss after stereotactic radiosurgery of vestibular schwannoma. World Neurosurg. 2013;80:359–363. Foote KD, Friedman WA, Buatti JM, et al. Analysis of risk factors associated with radiosurgery for vestibular schwannoma. J Neurosurg. 2001;95:440–449. Yomo S, Carron R, Thomassin JM, et al. Longitudinal analysis of hearing before and after radiosurgery for vestibular schwannoma. J Neurosurg. 2012;117:877–885. Massager N, Nissim O, Delbrouck C, etet al. Irradiation of cochlear structures during vestibular schwannoma radiosurgery and associated hearing outcome. J Neurosurg. 2013;119(Suppl): 733–739. Linskey ME, Johnstone PA, O’Leary M, et al. Radiation exposure of normal temporal bone structures during stereotactically guided gamma knife surgery for vestibular schwannomas. J Neurosurg. 2013;119(Suppl):800–806. Aazh H, Moore BC. Audiological rehabilitation for facilitating hearing aid use: a review. J Am Acad Audiol. 2017;28: 248–260.