Significance of cochlear dose in the radiosurgical treatment of vestibular schwannoma: controversies and unanswered questions.

RESEARCH—HUMAN—CLINICAL STUDIES RESEARCH—HUMAN—CLINICAL STUDIES

TOPIC

Significance of Cochlear Dose in the Radiosurgical Treatment of Vestibular Schwannoma: Controversies and Unanswered Questions BACKGROUND: Cochlear dose has been identified as a potentially modifiable contributor to hearing loss after stereotactic radiosurgery (SRS) for vestibular schwannoma (VS). OBJECTIVE: To evaluate the association between computed tomography-based volumetric cochlear dose and loss of serviceable hearing after SRS, to assess intraobserver and interobserver reliability when determining modiolar point dose with the use of magnetic resonance imaging and computed tomography, and to discuss the clinical significance of the cochlear dose with regard to radiosurgical planning strategy. METHODS: Patients with serviceable pretreatment hearing who underwent SRS for sporadic VS between the use of Gamma Knife Perfexion were studied. Univariate and multivariate associations with the primary outcome of time to nonserviceable hearing were evaluated. RESULTS: A total of 105 patients underwent SRS for VS during the study period, and 59 (56%) met study criteria and were analyzed. Twenty-one subjects (36%) developed nonserviceable hearing at a mean of 2.2 years after SRS (SD, 1.0 years; median, 2.1 years; range 0.6-3.8 years). On univariate analysis, pretreatment pure tone average, speech discrimination score, American Academy of Otolaryngology-Head and Neck Surgery hearing class, marginal dose, and mean dose to the cochlear volume were statistically significantly associated with time to nonserviceable hearing. However, after adjustment for baseline differences, only pretreatment pure tone average was statistically significantly associated with time to nonserviceable hearing in a multivariable model. CONCLUSION: Cochlear dose is one of many variables associated with hearing preservation after SRS for VS. Until further studies demonstrate durable tumor arrest with reduced dose protocols, routine tumor dose planning should not be modified to limit cochlear dose at the expense of tumor control.

Jeffrey T. Jacob, MD* Matthew L. Carlson, MD‡ Terry K. Schiefer, MD* Bruce E. Pollock, MD*§ Colin L. Driscoll, MD*‡ Michael J. Link, MD*‡ *Department of Neurologic Surgery, ‡Department of Otolaryngology-Head and Neck Surgery, and §Department of Radiation Oncology, Mayo Clinic College of Medicine, Rochester, Minnesota Correspondence: Michael J. Link, MD, Department of Neurologic Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905. E-mail: [email protected] Received, August 24, 2013. Accepted, January 8, 2014. Published Online, January 28, 2014. Copyright © 2014 by the Congress of Neurological Surgeons.

KEY WORDS: Acoustic neuroma, Cerebellopontine angle, Cochlear dose, Gamma Knife radiosurgery, Hearing preservation, Stereotactic radiosurgery, Vestibular schwannoma Neurosurgery 74:466–474, 2014

DOI: 10.1227/NEU.0000000000000299

H

earing preservation after stereotactic radiosurgery (SRS) for vestibular schwannoma (VS) remains a major challenge. Recent studies have shown young patient age, small tumor size, excellent pretreatment hearing, and low radiation dose to the cochlea to be favorably associated with hearing outcomes.1-9 ABBREVIATIONS: AAO-HNS, American Academy of Otolaryngology-Head and Neck Surgery; CI, confidence interval; PTA, pure tone average; SDS, speech discrimination score; SRS, stereotactic radiosurgery; VS, vestibular schwannoma

466 | VOLUME 74 | NUMBER 5 | MAY 2014

www.neurosurgery-online.com

Although SRS aims to deliver a highly conformal dose of radiation to the tumor volume with a steep radiation falloff gradient, inner ear structures adjacent to the irradiated target may be exposed to intolerable doses of radiation, particularly when there is significant tumor extension involving the lateral internal auditory canal.6,10 Several studies have evaluated the relationship between modiolar point dose and hearing outcomes; however, few have examined the significance of the dose volume to the inner ear.2,6,8,11,12 In the present study, we evaluate radiation dose received by the cochlear volume among a cohort


Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

COCHLEAR VOLUME AND HEARING PRESERVATION AFTER SRS

TABLE 1. Summary of Clinical Features and Outcomes for 59 Study Patients With Vestibular Schwannomaa Feature Baseline patient data Age, y 58.9 6 10.1 Sex, n (%) Female 32 Male 27 Tumor laterality, n (%) Left 27 Right 32 CPA tumor size, mm 7.1 6 5.3 AAO-HNS tumor size classification, n (%) 1 = Intracanalicular 15 2 = CPA, 0 to , 10 mm 27 3 = CPA, 10 to , 20 mm 17 CSF fundal cap, mm 1.6 6 1.9 83.3 6 12.9 Cochlear volume, mm3 Radiosurgical planning parameters Volume treated, mm3 1276 6 1167 Cochlear volume maximal dose, Gy 11.8 6 3.4 Cochlear volume mean dose, Gy 4.9 6 1.2 Point dose to the modiolus, Gy 6.5 6 1.8 Isocenters, n 5.7 6 2.6 Dose to the tumor margin, Gy 12 38 13 21 Audiometric data Pretreatment PTA, dB 32.3 6 12.7 Pretreatment speech 88 6 14 discrimination, % Pretreatment hearing class, n (%) A 26 B 33 Posttreatment hearing class, n (%) A 10 B 28 C 6 D 15

(59; 33-79) (54) (46) (46) (54) (6.7; 0.0-16.9)

(25) (46) (29) (1.0; 0-8.9) (83.8; 51.9-107.3) (909; 97-5600) (11.4; 3.7-19.2) (4.9; 1.9-8.3) (6.2; 2.7-11.7) (5; 1-15) (64) (36) (32.5; 3.8-50.0) (90; 50-100)

(44) (56) (17) (47) (10) (25)

a

AAO-HNS American Academy of Otolaryngology-Head and Neck Surgery; CPA, cerebellopontine angle; CSF, cerebrospinal fluid; PTA, pure tone average. Values are mean 6 SD (median; range) when appropriate.

of patients with pre-treatment serviceable hearing (American Academy of Otolaryngology—Head and Neck Surgery [AAOHNS] class A or B; Gardner-Robertson class 1 or 2). Volumetric cochlear dose analysis was performed with high-resolution computed tomography (CT) and thin-slice magnetic resonance imaging (MRI), and all patients were treated with the Leksell Gamma Knife Perfexion system (Elekta Instruments, Norcross, Georgia). Associations between patient variables, tumor characteristics, radiosurgical planning parameters, and posttreatment hearing capacity were explored. Additionally, intraobserver and interobserver variability analyses were performed comparing modiolar point dose calculations between clinicians using CT

NEUROSURGERY

FIGURE 1. Kaplan-Meier graph showing estimated rates of serviceable hearing among 59 patients with unilateral vestibular schwannoma treated with stereotactic radiosurgery.

and MRI. Finally, we review the clinical significance of cochlear dose: whether it should influence radiosurgical treatment planning or merely serve in hearing outcome prognostication to assist in pretreatment patient counseling.

METHODS Patient Selection, Clinical Features, and Outcomes Studied After Institutional Review Board approval (11-00799), a prospectively maintained clinical database including all patients who underwent SRS for VS from October 2007 to March 2011 was reviewed. Initial data collection included pretreatment and posttreatment audiometric data (pure tone average [PTA], speech discrimination score [SDS], AAO-HNS hearing class, length of audiometric follow-up, SRS planning parameters, tumor size, volume treated, neurofibromatosis type 2 status, history of prior tumor treatment, and basic demographic information). Pretreatment tumor size, location, and hearing capacity were reported according to the 1995 AAO-HNS guidelines.13 Patients with nonserviceable pretreatment hearing (AAO-HNS class C or D; Gardner-Robertson class 3, 4, or 5), inadequate or missing pretreatment audiometric data, neurofibromatosis type 2, history of prior tumor treatment, and age , 18 years were excluded. After treatment, patients underwent clinical, audiometric, and radiological follow-up every 6 months for the first year and annually thereafter.

Radiosurgical Planning Strategy All patients were evaluated by a neurosurgeon-neurotologist team, and a thorough pretreatment neurological examination with audiometric analysis was performed. Subjects were treated with the Leksell Gamma Knife Perfexion system. A Leksell model G stereotactic head frame (Elekta, Inc) was applied with local anesthesia. Pretreatment noncontrast high-resolution CT and postgadolinium axial spoiled gradient recalled acquisition MRI was performed to obtain 1-mm-thick slices. Images were exported to a computer workstation for dose planning with Leksell GammaPlan software (Elekta Instruments). Three-dimensional volumes

VOLUME 74 | NUMBER 5 | MAY 2014 | 467


JACOB ET AL

TABLE 2. Univariate Associations With Time to Nonserviceable Hearinga Feature Baseline patient data Age, y Sex Female Male Tumor laterality Left Right CPA tumor size CSF fundal cap 0 . 0 Cochlear volume Pretreatment audiometric data Pretreatment PTA Pretreatment PTA , 30 $ 30 Pretreatment speech discrimination Pretreatment speech discrimination $ 70 , 70 Pretreatment hearing class A B Radiosurgical planning parameters Volume treated Tumor maximum 24 26 Dose to the tumor margin 12 13 Cochlear volume maximal dose Cochlear volume mean dose Cochlear volume mean dose , 5 $ 5 Point dose to the modiolus Number of isocenters

Odds Ratio (95% CI) 1.56 (0.98-2.48)b

P Value

c Index

.06 .20

0.060 0.521

.74

0.597

.29 .90

0.606 0.518

.22

0.578

1.0 (reference) 0.53 (0.21-1.39) 1.0 (reference) 0.86 (0.36-2.06) 0.96 (0.88-1.04)d 1.0 (reference) 0.94 (0.39-2.28) 0.82 (0.60-1.12)b 2.41 (1.48-3.92)b 1.0 (reference) 7.91 (1.81-34.62) 0.60 (0.44-0.81)b

, .001 .006

0.742 0.669

, .001 .009

0.681 0.566

.006

0.668

.64 .09

0.539 0.576

.01

0.572

.11 .049 .006

0.643 0.684 0.693

.08 .76

0.656 0.516

1.0 (reference) 0.24 (0.08-0.70) 1.0 (reference) 5.65 (1.64-19.53) 0.90 (0.59-1.38)c 1.0 (reference) 3.54 (1.14-11.05) 1.0 2.08 1.11 1.37

(reference) (0.87-4.99) (0.98-1.26)d (1.00-1.87)d

1.0 4.12 1.20 0.97

(reference) (1.51-11.23) (0.98-1.47)d (0.82-1.16)d

a

CI, confidence interval; CPA, cerebellopontine angle; CSF, cerebrospinal fluid; PTA, pure tone average. Odds ratio represents a 10-unit increase in the feature listed. c Odds ratio represents a 1000-unit increase in the feature listed. d Odds ratio represents a 1-unit increase in the feature listed. b

of the cochlea were generated from the temporal bone CT, and the mean and maximal doses to the cochlear volume were obtained for each patient.

Statistical Methods Continuous features were summarized with means, standard deviations, medians, and ranges; categorical features were summarized with frequency counts and percentages. Time to nonserviceable hearing (posttreatment AAO-HNS class C or D) was estimated with the Kaplan-Meier method. Duration of follow-up was determined from the


time of treatment to the date of the first audiogram indicating nonserviceable hearing or the date of the most recent audiogram for patients still retaining AAO-HNS class A or B hearing. Univariate and multivariate associations with time to nonserviceable hearing were evaluated with Cox proportional hazard regression models and summarized with hazard ratios and 95% confidence intervals (CIs). All continuous variables that were found to be statistically significantly associated with time to nonserviceable hearing on univariate analysis (P , .05) were further interrogated to determine critical cutoff points that best discriminated between patients who retained serviceable hearing at last follow-up and those who




developed AAO-HNS class C or D hearing. Additionally, the prognostic ability of a feature included in a Cox model was evaluated with the c (for concordance) index. The interpretation of a c index is the same as the interpretation of the area under a receiver-operating characteristic curve. A c index can range from 0.5 to 1.0, with higher values indicating greater prognostic ability. Modiolar point dose measurements between 2 independent observers using CT and MRI were plotted. Intraobserver and interobserver agreement between modiolar point dose measurements was statistically assessed with the Lin concordance correlation coefficient. This coefficient ranges from 0.0 to 1.0, with higher values indicating a greater level of agreement. Statistical analyses were performed with the SAS software package (SAS Institute, Cary, North Carolina). All tests were 2 sided, and values of P , .05 were considered statistically significant.

RESULTS Between October 2007 and March 2011, 105 patients underwent SRS for VS, and 59 (56%) met the study criteria and were included. Patient characteristics, SRS parameters, and cochlear dose-volume calculations are summarized in Table 1. Forty-six patients were excluded for 1 or more of the following reasons: 28 patients had inadequate audiometric follow-up, 21 had nonserviceable hearing before SRS, 3 patients had neurofibromatosis type 2, and 1 patient was , 18 years of age. The mean duration of audiometric follow-up after SRS was 25.2 months (median, 25.0 months; range, 6.0-46.0 months). Twentyone subjects (36%) developed nonserviceable hearing at a mean of 2.2 years after SRS (SD, 1.0 years; median, 2.1 years; range, 0.63.8 years). Among patients with serviceable hearing, the mean duration of follow-up was 2.0 years (SD, 1.0 years; median, 2.0 years; range, 0.5-4.1 years). Estimated serviceable hearing rates at 1, 2, and 3 years after SRS were 96% (95% CI, 92-100; number still at risk, 51), 83% (95% CI, 73-95; number still at risk, 33), and 57% (95% CI, 42-77; number still at risk, 13), respectively (Figure 1). No patient demonstrated improvement in hearing capacity after treatment. The overall tumor control rate was 95% (56 of 59 patients) during the follow-up period, and no patient developed facial or trigeminal nerve dysfunction. Univariate associations with time to nonserviceable hearing are summarized in Table 2. Pretreatment PTA, pretreatment SDS, pretreatment AAO-HNS hearing class (A vs B), tumor marginal dose (12 vs 13 Gy), and mean dose to the cochlear volume were statistically significantly associated with time to nonserviceable hearing. In evaluations for critical cutoff points, the estimated rates of serviceable hearing at 1, 2, and 3 years after SRS were 100% (95% CI, 100-100; number still at risk, 25), 96% (95% CI, 88-100; number still at risk, 16), and 90% (95% CI, 77-100; number still at risk, 7), respectively, for patients with pretreatment AAO-HNS class A hearing compared with 94% (95% CI, 85-100; number still at risk, 26), 73% (95% CI, 58-93; number still at risk, 17), and 36% (95% CI, 20-65; number still at risk, 6), respectively, for patients with AAO-HNS class B hearing (P = .006; Figure 2). The estimated rates of serviceable hearing at 1, 2, and 3 years after SRS were 100% (95% CI, 100100; number still at risk, 21), 95% (95% CI, 86-100; number

NEUROSURGERY

FIGURE 2. Kaplan-Meier graph showing estimated rates of serviceable hearing comparing patients with pre-treatment American Academy of OtolaryngologyHead and Neck Surgery (AAO-HNS) hearing class A vs class B. SRS, stereotactic radiosurgery.

still at risk, 14), and 95% (95% CI, 86-100; number still at risk, 7), respectively, for patients with pretreatment PTA , 30 dB compared with 94% (95% CI, 87-100; number still at risk, 30), 76% (95% CI, 62-94; number still at risk, 19), and 36% (95% CI, 20-65; number still at risk, 6), respectively, for patients with pretreatment PTA $ 30 dB (P = .006; Figure 3). The estimated rates of serviceable hearing (95% CI; number still at risk) at 1, 2, and 3 years after SRS were 96% (95% CI, 91-100; number still at risk, 45), 84% (95% CI, 74-96; number still at risk, 25), and 68% (95% CI, 53-87; number still at risk, 13), respectively, for patients with pretreatment SDS $ 70% (P = .009; Figure 4). In comparison, the estimated percentages of patients with serviceable hearing at 1 and 2 years after SRS were 100% (95% CI, 100100; number still at risk, 6) and 80% (95% CI, 52-100; number still at risk, 4), respectively, for patients with pretreatment

FIGURE 3. Kaplan-Meier graph showing estimated rates of serviceable hearing comparing patients with pretreatment pure tone average (PTA) , 30- vs $ 30dB hearing loss. SRS, stereotactic radiosurgery.



JACOB ET AL

FIGURE 4. Kaplan-Meier graph showing estimated rates of serviceable hearing comparing patients with pretreatment speech discrimination score (SDS) $ 70% and , 70%. SRS, stereotactic radiosurgery.

FIGURE 6. Kaplan-Meier graph showing estimated rates of serviceable hearing comparing patients receiving a mean dose with the cochlear volume of , 5 vs $ 5 Gy.

SDS , 70%; the rate had fallen to 0% and there were no patients left at risk at 3 years after treatment. The estimated rates of serviceable hearing at 1, 2, and 3 years following SRS were 94% (95% CI, 87-100; number still at risk, 32), 88% (95% CI, 78-100; number still at risk, 20), and 76% (95% CI, 60-97; number still at risk, 9), respectively, for patients receiving 12 Gy to the tumor margin compared with 100% (95% CI, 100-100; number still at risk, 19), 77% (95% CI, 60-100; number still at risk, 13), and 35% (95% CI, 17-72; number still at risk, 4), respectively, for patients receiving 13 Gy (P = .02; Figure 5). Finally, the estimated percentages of patients with serviceable hearing at 1, 2, and 3 years after treatment were 100% (95% CI, 100-100; number still at risk, 27), 96% (95% CI, 88-100; number still at risk, 19), and 76% (95% CI, 57-100; number still at risk, 10), respectively, for subjects receiving a mean dose of , 5 Gy to the cochlear volume compared with 93% (95% CI,

83-100; number still at risk, 24), 71% (95% CI, 55-92; number still at risk, 14), and 37% (95% CI, 18-73; number still at risk, 3), respectively, for patients who received a cochlear mean dose of $ 5 Gy (P = .006; Figure 6). A multivariable model to predict posttreatment nonserviceable hearing is summarized in Table 3. After adjustment for pretreatment PTA, no other studied feature was statistically significantly associated with time to nonserviceable hearing. Specifically, each 10-dB increase in pretreatment PTA was associated with a 2.4-fold increased risk of nonserviceable hearing (hazard ratio 2.41; P = , .001). Finally, intraobserver and interobserver variability comparing CT- and MRI-based modiolar point dose measurements was determined with the Lin concordance correlation coefficient. Modiolar point dose measurements using CT showed very strong interobserver agreement compared with only moderate agreement using MRI (Figure 7A and 7B). Additionally, there was only moderate correlation between point doses measured using MRI and CT by the same observer (Figure 7C and 7D).

DISCUSSION Over the last several decades, there has been growing interest in identifying prognostic factors associated with hearing outcome in patients with VS. Several studies have found tumor size,4 patient

TABLE 3. Multivariable Associations With Time to Nonserviceable Hearinga Feature Pretreatment PTA FIGURE 5. Kaplan-Meier graph showing estimated rates of serviceable hearing comparing patients receiving a tumor marginal dose of 12 vs 13 Gy.


a b

Odds Ratio (95% CI) 2.47 (1.48-4.14)

b

P Value ,.001

CI, confidence interval; PTA, pure tone average. Odds ratio represents a 10-unit increase in the feature listed.




FIGURE 7. Plots comparing interobserver calculations of modiolar point dose using (A) magnetic resonance imaging (MRI) and (B) computed tomography (CT). C and D, intraobserver calculations of modiolar point dose comparing MRI- and CT-based measurements. The Lin concordance correlation coefficient (CCC) ranges from 0.0 to 1.0, with higher values indicating a greater level of agreement.

age,2,4 and pretreatment hearing status3 to be among factors that influence hearing decline after SRS. Recently, cochlear dose has become a “hot topic” within the radiosurgical literature, receiving significant attention as a critical prognostic factor for hearing preservation.1-9,14 Most studies have used MRI-based modiolar point dose or MRI-based dose-volume histograms and have determined critical cutoff points between 3 and 5.3 Gy to be statistically significantly associated with loss of serviceable hearing after SRS.2,4,6,9 Although MRI-based modalities appear to be most widely used, in the present study, we found significantly greater agreement between observers using CT compared with MRI when determining modiolar point dose values (Figure 7). Therefore, assigning a cochlear dose value based on a single pixel is problematic because there are many reasonable candidate pixels and each may have significantly different values (Figure 8). Understanding that CT affords superior resolution of bony cochlear anatomy and that assigning

NEUROSURGERY

a dose value based on a single pixel using MRI is problematic, we began using CT-based volumetric cochlear dose analysis in 2007 to more accurately investigate how cochlear dose influences hearing outcome. Clinical and histopathological studies have suggested that multiple anatomic subunits of the cochlea, beyond the modiolus, may be susceptible to radiation injury.15 Temporal bone specimens procured from patients receiving external beam radiation and results of animal studies have demonstrated that the stria vascularis, outer hair cells, and spiral ganglion cells may be most vulnerable to the effects of ionizing radiation.15,16 Although modiolar point dose may provide a gross estimation of radiation dose to the basal spiral ganglion cells within the modiolus, this method fails to directly assess the outer wall of the cochlea, containing the stria vascularis and the organ of Corti. Furthermore, it is known that poor bone contrast, coupled with the heterogeneity and nonlinearity of magnetic fields, can result in



JACOB ET AL

FIGURE 8. Modiolar point dose calculations can vary significantly, depending on the imaging modality used and the exact placement of the cursor as demonstrated. A, with the use of 1-mm-thick spoiled gradient recalled magnetic resonance imaging and (B) 1-mm-thick computed tomography of the temporal bones, there is a significant difference in point dose by moving the cursor only 1 pixel away. Each cursor placement was felt to reasonably represent point dose to the modiolus of the cochlea.

distortions of MRI ranging from 0.2 to 5.0 mm, limiting its usefulness in defining cochlear volumes.17,18 Poetker et al17 demonstrated a mean image shift of 1.92 mm in 23 patients undergoing SRS for the treatment of VS using MRI. They concluded that this translated to almost 17% of the treatment volume missing the tumor and irradiating surrounding healthy tissues, which can make it difficult to extrapolate or interpret the relative doses to the cochlea or its relationship to hearing outcome. There are very few studies evaluating cochlear dose using CT-based volume-dose analysis.6,8,9,11,12 Vastly discrepant hearing outcomes reporting and heterogeneity in cochlear dose analysis make the data on the relationship of cochlear dose and hearing outcome after SRS difficult to interpret. Massager et al6 were the first to perform dose-volume analysis using 3-dimensional volumes of the cochlea based on thin-cut CT. They found a high degree of correlation between the dose of radiation delivered to the intracanalicular part of the tumor and to the cochlea. Patients receiving a mean cochlear dose , 4 Gy, which also correlated with extent of tumor penetrating the internal auditory canal, were more likely to maintain hearing after treatment. Although the authors could not differentiate whether hearing loss occurred as a result of a higher radiation dose to the cochlea or secondary to tumor filling a large part of the canal, these 2 parameters appeared to be linked. Analyzing MRI-based radiosurgical plans, Kano et al4 determined that a radiation dose of 4.2 Gy to the modiolus was the critical threshold for hearing preservation. Brown et al2 generated dose-volume histograms for the cochlear volumes based on T2-weighted MRI and found a strong correlation between the percentage of the volume of cochlea receiving . 5.3 Gy and hearing outcomes. Similarly, in


the present study, we found that patients receiving , 5 Gy to the cochlear volume were more likely to retain serviceable hearing at last follow-up on univariate analysis. In a recent report, Baschnagel et al9 advocated limiting the mean cochlear dose to , 3 Gy in patients with serviceable hearing to maximize the prospects of hearing preservation. It is notable that all patients in their series received a mean cochlear dose , 5 Gy and only 3 patients had a mean dose . 4 Gy, much lower than in the present and previous studies. Some centers have advocated fractionated stereotactic radiotherapy in an attempt to minimize radiation-induced tissue complications and to improve functional outcomes relative to single-session SRS.19 However, when comparing SRS with a variety of fractionation schemes, studies have failed to show improved hearing preservation rates.20-23 Recently, Rasmussen et al24 compared fractionated stereotactic radiotherapy for VS with an untreated control group and found that fractionated stereotactic radiotherapy, in fact, accelerated the progression of hearing loss. Hayden-Gephart et al25 analyzed their experience using a hypofractionated SRS scheme (18 Gy/3 fractions) and correlated hearing outcome with cochlea volume and cochlear dose. Similar to studies using single-fraction SRS, a direct correlation was noted between increased cochlear dose and the risk of hearing loss. Curiously, they noted that larger cochlear volumes were associated with a lower risk of hearing loss, suggesting that these patients may have greater “reserve,” protecting them from adverse radiation effects. In contrast, we did not find any meaningful difference in the mean cochlear volume in those with retained serviceable hearing (83.5 mm3) compared with those with loss of useful hearing (83.4 mm3).




In the present study, we identified several features that were statistically significantly associated with time to nonserviceable hearing on univariate analysis. Namely, patients with better pretreatment hearing (PTA, SDS, and AAO-HNS hearing class) and those receiving lower radiation dose parameters (marginal dose and dose to the cochlear volume) were more likely to maintain useful hearing at the last audiometric follow-up. Importantly, point dose to the modiolus neared statistical significance with a value of P = .08; however, mean dose to the cochlear volume carried a stronger statistical association. After adjustment for baseline differences, only pretreatment PTA was statistically significantly associated with time to nonserviceable hearing with a multivariable model. In our experience, among non-VS tumors of the temporal bone such as cerebellopontine angle meningiomas extending into the internal auditory canal, post-SRS sensorineural hearing loss is a rare event, despite the cochlea receiving radiation doses in excess of 10 Gy.26 This would suggest that there are other more significant tumor-related factors beyond cochlear dose that ultimately dictate hearing outcomes. Collectively, the vast majority of literature suggests that, if followed up long enough, most patients harboring VS trend toward accelerated hearing loss in the tumor ear compared with the “control” ear, regardless of management.27-29 Many of the factors that influence hearing outcome with 1 treatment strategy equally influence other strategies. For example, patients with excellent pretreatment hearing, smaller tumors, or tumors with a greater distance from the fundus are more likely to have good long-term hearing whether managed with SRS, microsurgery, or observation. Thus, even when optimally managed, long-term hearing outcomes are strongly dictated by the tumor itself rather than treatment. Although reducing the radiation dose to the cochlea may play some role in improving hearing outcomes after SRS, this may necessitate reducing the margin dose or intentionally undertreating the lateral portion in many cases. Both of these strategies would theoretically run the risk of reducing long-term tumor control; therefore, we would not advocate those strategies.

CONCLUSION Whether reliable long-term functional hearing preservation is the “holy grail” or “unicorn” of VS management remains to be seen. Fortunately, the literature has shown conservative management with serial imaging, microsurgery, and radiosurgery to be excellent treatment options in select patients. In those who ultimately require active treatment, radiosurgery provides a safe, noninvasive means toward excellent tumor control with minimal morbidity. Mean dose to the cochlear volume is one of many variables that influence hearing outcomes after SRS for VS. Even with the most conformal SRS devices available and specific “cochlea-shielding” strategies, there is still very little that can be done to limit dose to the cochlea when the tumor extends to the fundus of the internal auditory canal. Without data to support effective tumor control

NEUROSURGERY

with significantly lowered radiation doses, we caution against undertreating the tumor in the distal fundus or further reducing the marginal prescription dose to achieve lower cochlear doses. Disclosure Internal departmental funding was used without commercial sponsorship or support. Institutional Review Board approval: 11-00799. The authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article.

REFERENCES 1. Link MJ, Pollock BE. Chasing the holy grail of vestibular schwannoma management. World Neurosurg. 2013;80(3-4):276-278. 2. Brown M, Ruckenstein M, Bigelow D, et al. Predictors of hearing loss after gamma knife radiosurgery for vestibular schwannomas: age, cochlear dose, and tumor coverage. Neurosurgery. 2011;69(3):605-613; discussion 613-614. 3. Hasegawa T, Kida Y, Kato T, Iizuka H, Yamamoto T. Factors associated with hearing preservation after Gamma Knife surgery for vestibular schwannomas in patients who retain serviceable hearing. J Neurosurg. 2011;115(6):1078-1086. 4. Kano H, Kondziolka D, Khan A, Flickinger JC, Lunsford LD. Predictors of hearing preservation after stereotactic radiosurgery for acoustic neuroma. J Neurosurg. 2009;111(4):863-873. 5. Lasak JM, Klish D, Kryzer TC, Hearn C, Gorecki JP, Rine GP. Gamma knife radiosurgery for vestibular schwannoma: early hearing outcomes and evaluation of the cochlear dose. Otol Neurotol. 2008;29(8):1179-1186. 6. Massager N, Nissim O, Delbrouck C, et al. Irradiation of cochlear structures during vestibular schwannoma radiosurgery and associated hearing outcome. J Neurosurg. 2007;107(4):733-739. 7. Timmer FC, Hanssens PE, van Haren AE, et al. Gamma Knife radiosurgery for vestibular schwannomas: results of hearing preservation in relation to the cochlear radiation dose. Laryngoscope. 2009;119(6):1076-1081. 8. Wackym PA, Runge-Samuelson CL, Nash JJ, et al. Gamma Knife surgery of vestibular schwannomas: volumetric dosimetry correlations to hearing loss suggest stria vascularis devascularization as the mechanism of early hearing loss. Otol Neurotol. 2010;31(9):1480-1487. 9. Baschnagel AM, Chen PY, Bojrab D, et al. Hearing preservation in patients with vestibular schwannoma treated with Gamma Knife surgery. J Neurosurg. 2013;118 (3):571-578. 10. Linskey ME, Johnstone PA, O’Leary M, Goetsch S. Radiation exposure of normal temporal bone structures during stereotactically guided Gamma Knife surgery for vestibular schwannomas. J Neurosurg. 2003;98(4):800-806. 11. Tamura M, Carron R, Yomo S, et al. Hearing preservation after Gamma Knife radiosurgery for vestibular schwannomas presenting with high-level hearing. Neurosurgery. 2009;64(2):289-296; discussion 296. 12. Hayden-Gephart MG, Hansasuta A, Balise RR, et al. Cochlea radiation dose correlates with hearing loss following stereotactic radiosurgery of vestibular schwannoma. World Neurosurg. 2013;80(3-4):359-363. 13. Committee on hearing and equilibrium guidelines for the evaluation of hearing preservation in acoustic neuroma (vestibular schwannoma): American Academy of Otolaryngology-Head and neck Surgery Foundation, INC. Otolaryngol Head Neck Surg. 1995;113(3):179-180. 14. Yomo S, Tamura M, Carron R, Porcheron D, Régis J. A quantitative comparison of radiosurgical treatment parameters in vestibular schwannomas: the Leksell Gamma Knife Perfexion versus model 4C. Acta Neurochir (Wien). 2010;152(1): 47-55. 15. Linskey ME, Johnstone PA. Radiation tolerance of normal temporal bone structures: implications for Gamma Knife stereotactic radiosurgery. Int J Radiat Oncol Biol Phys. 2003;57(1):196-200. 16. Hoistad DL, Ondrey FG, Mutlu C, Schachern PA, Paparella MM, Adams GL. Histopathology of human temporal bone after cis-platinum, radiation, or both. Otolaryngol Head Neck Surg. 1998;118(6):825-832. 17. Poetker DM, Jursinic PA, Runge-Samuelson CL, Wackym PA. Distortion of magnetic resonance images used in Gamma Knife radiosurgery treatment planning: implications for acoustic neuroma outcomes. Otol Neurotol. 2005;26 (6):1220-1228.



JACOB ET AL

18. Prott FJ, Haverkamp U, Eich H, et al. Effect of distortions and asymmetry in MR images on radiotherapeutic treatment planning. Int J Cancer. 2000;90(1):46-50. 19. Hansasuta A, Choi CY, Gibbs IC, et al. Multisession stereotactic radiosurgery for vestibular schwannomas: single-institution experience with 383 cases. Neurosurgery. 2011;69(6):1200-1209. 20. Collen C, Ampe B, Gevaert T, et al. Single fraction versus fractionated linac-based stereotactic radiotherapy for vestibular schwannoma: a single-institution experience. Int J Radiat Oncol Biol Phys. 2011;81(4):e503-e509. 21. Henzel M, Hamm K, Sitter H, et al. Comparison of stereotactic radiosurgery and fractionated stereotactic radiotherapy of acoustic neurinomas according to 3-D tumor volume shrinkage and quality of life. Strahlenther Onkol. 2009;185(9):567-573. 22. Kopp C, Fauser C, Müller A, et al. Stereotactic fractionated radiotherapy and LINAC radiosurgery in the treatment of vestibular schwannoma: report about both stereotactic methods from a single institution. Int J Radiat Oncol Biol Phys. 2011;80(5):1485-1491. 23. Meijer OW, Vandertop WP, Baayen JC, Slotman BJ. Single-fraction vs. fractionated linac-based stereotactic radiosurgery for vestibular schwannoma: a single-institution study. Int J Radiat Oncol Biol Phys. 2003;56(5):1390-1396. 24. Rasmussen R, Claesson M, Stangerup SE, et al. Fractionated stereotactic radiotherapy of vestibular schwannomas accelerates hearing loss. Int J Radiat Oncol Biol Phys. 2012;83(5):e607-e611. 25. 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(3-4):359-363. 26. Pollock BE, Link MJ, Foote RL, Stafford SL, Brown PD, Schomberg PJ. Radiosurgery as primary management for meningiomas extending into the internal auditory canal. Stereotact Funct Neurosurg. 2004;82(2-3):98-103.


27. Stangerup SE, Thomsen J, Tos M, Cayé-Thomasen P. Long-term hearing preservation in vestibular schwannoma. Otol Neurotol. 2010;31(2):271-275. 28. Shelton C, Hitselberger WE, House WF, Brackmann DE. Hearing preservation after acoustic tumor removal: long-term results. Laryngoscope. 1990;100(2 pt 1): 115-119. 29. Carlson ML, Jacob JT, Pollock BE, et al. Long-term hearing outcomes following stereotactic radiosurgery for vestibular schwannoma: patterns of hearing loss and variables influencing audiometric decline. J Neurosurg. 2013; 118(3):579-587.

COMMENT

T

his is a relevant and important contribution to the increasing body of work on hearing preservation in relation to radiotherapy of vestibular schwannomas. The results are conflicting with previous reports, which documents that there are no simple truths between radiotherapy and preservation of hearing for patients with a solitary vestibular schwannoma. It is evident that a considerable proportion of patients lose useful hearing by radiotherapy, but the specific factors involved seem variable, judged by available publications. Differences between patient cohorts and treatment regimens are obvious suspects concerning these variable findings. As an example, tumors growing until treatment (and growth rate) may be associated with poorer results. Per Caye-Thomasen Copenhagen, Denmark



Cochlear obliteration after translabyrinthine vestibular schwannoma surgery.

Unanswered Questions.

Non surgical treatment of vestibular schwannoma.

Baltimore's unanswered questions.

Rejection--unanswered questions.

The myxomycetes--some problems and unanswered questions.

Fiber: unanswered questions.

Unanswered questions in metal chelation.

E-Learning: some unanswered questions.

The apnea-PaCO2 relationship--unanswered questions?

HPV vaccination: unanswered questions remain.

Unanswered questions about prazosin hydrochloride.

Nine unanswered questions about cytokinesis.

radiosurgical management of vestibular schwannomas.

Unanswered questions about the periodic health examination.

Polyp Resection - Controversial Practices and Unanswered Questions.

Aging and depression: some unanswered questions.

Eccentric exercise: many questions unanswered.

Some unanswered questions about vaccination.

Proteomic analysis of vestibular schwannoma: conflicting role of apoptosis on the pathophysiology of sporadic vestibular schwannoma.

Some unanswered questions in thyroid-related ophthalmopathy.

Electrical vestibular stimulation after vestibular deafferentation and in vestibular schwannoma.

Unanswered questions on access from the margins.

The map of dizziness in vestibular schwannoma.