Hearing preservation in vestibular schwannoma management

YAJOT-01556; No of Pages 9 AMER IC AN JOURNAL OF OT OLA RYNGOLOGY– H E A D A N D NE CK M E D ICI N E AN D S U RGE RY X X (2 0 1 5) XXX – XXX

Available online at www.sciencedirect.com

ScienceDirect www.elsevier.com/locate/amjoto

Original Contribution

Hearing preservation in vestibular schwannoma management☆,☆☆,★,★★ Anna Elliott, BSc, MD a, 1 , Andrea L.O. Hebb, MSc, PhD, RN b, 2 , Simon Walling, MBChB, FRCSC b, 2 , David P. Morris, MBBS, FRCS (Eng), FRCS (ORL-HNS) a, 1 , Manohar Bance, MB, MSc, FRCS(C) a,⁎, 1 a

Maritime Lateral Skull Base Clinic, Division Otolaryngology, Dalhousie University 3rd Floor Dickson Building, QEII VG site, 1278 Tower Road, Halifax, NS b Maritime Lateral Skull Base Clinic, Division of Neurosurgery, Dalhousie University, QEII Halifax Infirmary Site, 1796 Summer Street, Halifax, NS

ARTI CLE I NFO

A BS TRACT

Article history:

Objective: To compare hearing preservation between stereotactic radiotherapy (SRT) and

Received 11 February 2015

conservative treatment of patients with unilateral vestibular schwannoma. Data sources: Retrospective case series comparing hearing outcomes of patients with a unilateral vestibular schwannoma managed conservatively or with stereotactic radiotherapy in a tertiary care academic centre. Review methods: Patients: Tumor database patients with American Academy of Otolaryngology Head and Neck Surgery Class A or B hearing at the onset of study. Stereotactic radiotherapy patients were predominantly those who failed conservative management. Interventions: Audiometric pure tone averages and speech discrimination scores as well as patient demographics, tumor location, size and growth were extracted. Main outcome measures: Hearing outcome measures were: 1) Hearing Preservation, i.e. no drop from Class A/B to Class C/D hearing, 2) Hearing Survival of Class A/B hearing in months, 3) Audiometric Pure Tone Averages, Difference between post-treatment and pre-treatment, and 4) Speech Discrimination Score Difference (pre-treatment − post treatment). Survival analysis and non-parametric tests were used for hearing outcome measures, with multiple covariates tested. Results: Overall, serviceable hearing preservation among the 123 patients was 51%. The median hearing survival time was 46 months (mean 59 months). The Pure Tone averages and Speech Discrimination score differences were 16 dB and 82% respectively over a median follow-up time of 43 months. No significant difference was found between the

Abbreviations: AAOHNS, American Academy of Otolaryngology Head and Neck Surgery; IAC, Internal auditory canal; CPA, Cerebellopontine angle; PTA, Pure Tone Average; SDS, Speech Discrimination Score; Cox PH, Cox Proportional Hazards Regression. ☆ Funding support: Dalhousie Medical Research Fund. ☆☆ Data presented at the Canadian Society of Otolaryngology–Head and Neck Surgery 66th Annual Meeting, May 20–22, 2012, Fairmont Royal York Hotel, Toronto, Ontario, Canada. ★ Levels of evidence: Level 2c Outcomes research. Level 4 Case–control studies. ★★ Conflicts of Interest: No conflicts of interest or financial disclosures. ⁎ Corresponding author at: Division of Otolaryngology-Head and Neck Surgery, Room 3184, 3rd Floor Dickson Building, VG site, QEII Health Sciences Centre, 1278 Tower Road, Halifax, NS B3H 2Y9. Tel.: + 1 902 473 5975; fax: +1 902 473 4345. E-mail address: [email protected] (M. Bance). 1 Tel.: + 1 902 473 5975; fax: +1 902 425 9572. 2 Tel.: + 1 902 473 4824; fax: +1 902 425 4176. http://dx.doi.org/10.1016/j.amjoto.2015.02.016 0196-0709/© 2015 Elsevier Inc. All rights reserved.

Please cite this article as: Elliott A, et al, Hearing preservation in vestibular schwannoma management, Am J Otolaryngol–Head and Neck Med and Surg (2015), http://dx.doi.org/10.1016/j.amjoto.2015.02.016

2

AMER ICA N JOURNAL OF OT OLA RYNGOLOGY– H E A D A N D N E CK M EDI CI N E AN D S U RGE RY X X (2 0 1 5) XXX – XXX

conservative and SRT groups in any hearing outcome. Class A patients showed far better hearing survival than Class B patients. Conclusions: No significant difference was demonstrated as to measures of hearing outcomes between stereotactic radiotherapy and conservative management. Excellent existing hearing appears to be the best predictor of long term hearing survival in the cohort studied. © 2015 Elsevier Inc. All rights reserved.

1.

Introduction

Vestibular schwannomas (VS) (or acoustic neuromas (AN)) are benign brain tumors arising from the myelin-forming Schwann cell of the 8th cranial nerve (vestibulocochlear nerve). The prevalence of VS is approximately 1 in 1000 persons, with increased incidence due to incidental finding through MRI imaging [1]. VS can cause auditory and vestibular symptoms such as hearing loss, vertigo, and tinnitus; facial symptoms such as paralysis and numbness; and with increased growth life threatening sequelae such as brainstem compression and hydrocephalus. Clinical management of VS includes conservative management or “wait and scan”, surgery, and/or stereotactic radiotherapy [2]. These treatment strategies vary in their hearing preservation rates. Conservative management entails carefully following the tumor's growth with serial MRI scans [3]. In many clinics, most tumors are followed by this approach and do not require direct intervention. However, hearing tends to decline with time, even in the absence of tumor growth [4]. Surgical resection is another treatment option. However, surgical resection with hearing preservation, while feasible in very small tumors, has a high likelihood of hearing loss in larger tumors; indeed it is a certainty if the translabyrinthine surgical route is chosen [5]. Stereotactic radiotherapy (SRT) is an alternate treatment modality which uses multiple beams of radiation converging onto a single target over several sessions in order to deliver a high dose of therapy, while limiting exposure of healthy surrounding tissue to damaging radiation. SRT has been reported to have a success rate of over 95% in controlling growth of VS [6,7]. However, it is currently unclear whether SRT stabilizes hearing in treated individuals, or actually accelerates hearing loss. At the Maritime Lateral Skull Base Clinic (formerly the Atlantic Lateral Skull Base Clinic) we employ a patient-centered model with all options explained to the patient; which has resulted in most tumors being followed conservatively. However, if a tumor is growing (> 2 mm/MRI scan), and is under 25 mm in diameter, SRT is often considered as the next treatment option. There is, however, a lack of literature focusing on the hearing outcomes of SRT versus conservative wait and watch treatment. The objective of our study was to compare hearing preservation and hearing survival times of patients undergoing either conservative or SRT treatments in the management of unilateral acoustic neuromas. This is an important question, as if SRT is strongly protective of hearing, it could be offered as a first line treatment option, superseding conservative management. However, if SRT accelerates hearing loss, it might be reasonable to wait longer in the face of mild–moderate tumor growth before offering this modality. Lastly, if SRT has no appreciable effect on hearing, than it would be a reasonable option for patients with

small tumors, demonstrating growth on imaging. This finding would affect treatment strategies for patients under the care of the Maritime Lateral Skull Base clinic and other treatment centers involved with the management of VS. We recognize that there is a difference in the two populations in that the SRT group is comprised mostly of growing tumors, whereas the conservative group represents stable tumors. The difference is unavoidable as we do not advocate for SRT in the majority of cases without evidence of tumor growth. In some respects this is encouraging as the sample is comprised of patients where the most reasonable option was SRT, and no change or preservation of hearing would both be favorable outcomes. Surgical patients were not included.

2.

Materials and methods

This retrospective case series compared patients with VS who were managed conservatively to patients who either were primarily managed with SRT or who failed conservative management and went on to have SRT. Patient details were extracted, with prior informed consent signed, from the Maritime Lateral Skull Base clinic database. Ethics approval for this study was obtained from our local Research Ethics Board.

2.1.

Management protocol

At our center, conservative management predominated for patients with VS using the “wait-and-scan” approach. This included serial MRI scanning and audiometric testing (conservative management) as outlined previously [8]. However, if a tumor was large (> 25 mm), or was growing (>2 mm/MRI), then active treatments such as surgical or radiotherapy interventions were considered. The present study focused only on the conservative and SRT treatment modalities. For patients who received SRT, the procedure was performed using a standardized linear accelerator for delivery of stereotactic radiation. Patients received either 3125 cGy in 5 fractions, or in one case 6250 cGy in 25 fractions.

2.2.

Patient characteristics

The case series was drawn from a database of 405 patients with VS followed between July 1996 and July 2011 by the Maritime Lateral Skull Base clinic. Excluded were: surgical management (n = 87), having had SRT at another center (n = 16), lack of follow up (e.g. <2 audiograms) or insufficient data (n = 46), additional brain tumors or uncertain findings (n = 17), neurofibromatosis type II (n = 7) or non-serviceable hearing at onset (AAOHNS class C or D) (n = 109) [9,10]. Lack of adequate hearing was the major exclusion criterion. The statistical analyses were performed on 123 patients, 98 of whom underwent

Please cite this article as: Elliott A, et al, Hearing preservation in vestibular schwannoma management, Am J Otolaryngol–Head and Neck Med and Surg (2015), http://dx.doi.org/10.1016/j.amjoto.2015.02.016

AMER IC AN JOURNAL OF OT OLA RYNGOLOGY– H E A D A N D NE CK M E D ICI N E AN D S U RGE RY X X (2 0 1 5) XXX – XXX

Table 1 – Patient characteristics in the case series (n = 123). Sex Male n (%) 62 Female n (%) 61 Age (years) Median (Range) 55.0 Tumor location IAC n (%) 48 CPA n (%) 75 Tumor size IAC n (%) 48 Small n (%) 59 Medium n (%) 16 Large n (%) 0 Tumor growth status Stable n (%) 87 Growing n (%) 30 Shrinking n (%) 6 AAOHNS hearing class at onset on tumor side A n (%) 48 B n (%) 73 Decreased hearing on both tumor and contralateral sides No n (%) 120 Yes n (%) 2 Missing n (%) 1 Follow-up in months Median (Range) 43 Treatment modality Conservative n (%) 98 SRT n (%) 25

(50%) (50%) (16–85) (39%) (61%) (39%) (48%) (13%)

(71%) (24%) (5%) (41%) (59%) (98%) (2%)

(80%) (20%)

conservative therapy and 25 whom underwent SRT. In total, there were 62 male and 61 female participants with a median age of intervention of 55 years (range 16–85) and a median follow up time of 43 months (range 3–131). For further descriptive statistics, see Table 1.

Hearing assessment

The audiometric protocol has been described previously [8]. Briefly, patient hearing levels were determined by audiometric testing based on speech discrimination scores (SDS) and puretone averages (PTAs) of 0.5, 1, 2, and 3 kHz. The American Academy of Otolaryngology–Head and Neck Surgery [9,10] hearing classification system was used to categorize both pre- and post-treatment hearing levels, with only patients having serviceable hearing (class A or B) at the onset included in the analyses.

2.4.

tumor size. For cerebellopontine angle (CPA) tumors, size was calculated as the square root of the product of the maximum tumor diameter parallel to the petrous ridge and the perpendicular linear measurement of the extracanalicular portion of the tumor. If the tumor was purely intracanalicular (IAC), a linear measurement was taken parallel to the portion extending the length of the internal auditory canal. Furthermore tumors were classified into size categories: IAC, small (<15 mm), medium (15–30 mm), and large (>30 mm) as previously documented [4]. Tumor growth was defined as an increase of >2 mm on 3–12 month interval MRIs, and tumors were further classified as growing or stable/shrinking [8].

2.5.

Outcome measures

The four outcome measures of interest were: 1. Hearing preservation: hearing status at the end of follow-up with preservation defined as retention of serviceable hearing class A/B, and loss defined as a drop from class A/B to C/D. 2. Hearing survival in months (time spent in class A/B). 3. PTA difference (post-treatment PTA − pre-treatment PTA). 4. SDS difference (pre-treatment SDS − post treatment SDS).

(3–131)

AAOHNS: American Academy of Otolaryngology Head and Neck Surgery IAC: Internal auditory canal CPA: Cerebellopontine angle PTA: Pure Tone Average SDS: Speech Discrimination Score SRT: Stereotactic Radio Therapy

2.3.

3

Tumor characteristics

Tumor characteristics were based on location, size, and growth. Tumor size was measured using a protocol previously outlined [9,10]. One author (A.E.) reviewed all MRI images and calculated

Considered as independent variables were: demographic characteristics (gender, age); tumor characteristics (location, size class, growth); and, possible hearing predictors (AAOHNS hearing class, contralateral ear hearing class, pre-SRT (including contralateral ear) hearing decline, and length of follow up time).

2.6.

Statistical analysis

All data were summarized using descriptive statistics. Univariate comparisons were conducted between each of the outcomes of interest and each of the independent variables. Univariate non-parametric tests for binary dependent variables included Chi-square, Fisher’s exact, and Kolmogorov–Smirnoff, while for continuous dependent variables Kruskal Wallis and linear regression were used. Multivariate models were conducted to predict treatment modality and all hearing outcomes. Multivariate tests included logistic regression for binary dependent variables and multiple regression for continuous dependent variables. Survival analysis for hearing preservation in months was conducted comparing Kaplan–Meier survival curves with log rank Mantel–Cox tests, and Cox Proportional Hazards Regression was used to analyze covariates within the model. Results were considered significant at an alpha of 0.05. The statistical software used was SPSS 15.0 (SPSS, Chicago, IL).

3.

Results

3.1.

Patient characteristics

Table 1 shows the characteristics of the 123 patients in the case series. The patients were equally distributed between male and female with an overall median age of 55 years. The median follow-up time was 43 months. With regards to tumor characteristics, 39% of tumors were IAC and the remainder were primarily small CPA tumors. With regards to hearing assessment at the admission into the case series, 41% of patients were class A, and the remainder were class B.

Please cite this article as: Elliott A, et al, Hearing preservation in vestibular schwannoma management, Am J Otolaryngol–Head and Neck Med and Surg (2015), http://dx.doi.org/10.1016/j.amjoto.2015.02.016

4

AMER ICA N JOURNAL OF OT OLA RYNGOLOGY– H E A D A N D N E CK M EDI CI N E AN D S U RGE RY X X (2 0 1 5) XXX – XXX

Table 2 – Predictors of stereotactic radio therapy (SRT) versus conservative treatment.

Sex Male Female Age

n (%) n (%) median years (range)

b c d

Multivariate model b Logistic Regression

p-value

O.R.

p-value

(48%) (52%) (30–73)

0.826 0.219

1 1.11 1.03

0.891 0.388

0 25

(0%) (100%)

<0.001

n/a

(49%) (44%) (7%)

0 16 9

(0%) (64%) (36%)

<0.001

0.00 0.17 1

(88%) (12%)

7 18

(28%) (72%)

<0.001

1 39.82

<0.001

(43%) (57%)

8 17

(32%) (68%)

0.369

1 1.89

0.324

SRT (n = 25)

50 48 55.0

(51%) (49%) (16–85)

12 13 59.0

(49%) (51%)

Tumor location IAC n (%) 48 CPA n (%) 50 Tumor size c IAC n (%) 48 Small n (%) 43 Medium n (%) 7 Tumor growth status d Not growing n (%) 86 Growing n (%) 12 AAOHNS hearing class at onset on tumor side A n (%) 42 B n (%) 56 a

Univariate comparison a

Conservative Treatment (n = 98)

n/a 0.101

Chi-square, Fisher’s exact, or Kolmogorov–Smirnoff was used, as appropriate, for univariate comparisons. The full multivariate model excludes “tumor location” as this information is included in “tumor size”. The sample included no case of “large” tumor size. The group “not growing” includes “stable” (n = 80) and “shrinking” (n = 6).

3.2.

based on their growth, or size (i.e. being at a size at which further growth was felt could preclude SRT in the future).

Predictors of treatment modality

Approximately 80% of patients in the case series underwent conservative treatment, while 20% underwent SRT. Table 2 shows that the two treatment groups did not differ with regards to patient demographics (sex and age) and hearing class at the outset. Tumor characteristics of location, size, and growth status were all highly significant on univariate comparisons. However, the multivariate model revealed only tumor growth status to be an independent predictor of treatment modality (p < 0.001), which is not surprising as these tumors were chosen for SRT

3.3.

Hearing outcomes

The median follow up time was not significantly different between the two groups (42.5 months and 44 months in Conservative and SRT respectively, p = 0.416). Table 3 shows the four outcomes of interest for the entire case series. Overall hearing preservation was 51%, with hearing being preserved for a median of 46 months. The PTA increased by 16 dB and the SDS dropped by 82%. Table 3 also

Table 3 – Hearing outcomes according to treatment modality. Overall (n = 123)

Hearing status on tumor side at the end of follow-up Preserved n (%) 63 (51%) Lost n (%) 60 (49%) Hearing survival in months Median (Range) 46 (0–130) PTA difference (last − first) (dB) Mean (SD) 16.4 ±17.6 SDS difference (first − last) (%) Mean (SD) 82.0 ±20.4 a

By treatment modality

Univariate comparison between treatment modalities a p-value

Conservative Treatment (n = 98)

SRT (n = 25)

53 45

(54%) (46%)

10 15

(40%) (60%)

0.264

48

(0–130)

35

(9–72)

0.276

14.2

±17.3

25.0

±16.6

0.001

82.0

±21.6

81.8

±15.4

0.975

Fisher’s exact or Kolmogorov–Smirnoff test was used, as appropriate, for univariate comparisons.

Please cite this article as: Elliott A, et al, Hearing preservation in vestibular schwannoma management, Am J Otolaryngol–Head and Neck Med and Surg (2015), http://dx.doi.org/10.1016/j.amjoto.2015.02.016

AMER IC AN JOURNAL OF OT OLA RYNGOLOGY– H E A D A N D NE CK M E D ICI N E AN D S U RGE RY X X (2 0 1 5) XXX – XXX

shows a univariate comparison of Conservative versus SRT treatment groups with regards to the four outcomes of interest. PTA difference was found to be the only outcome that differed significantly between the two treatment modalities (14.2 and 25.0 in the Conservative and SRT groups respective, p = 0.001). It must be remembered that the PTA drop measure differs from hearing preservation measure, as dropping from class A/B to C/D can occur with a large PTA drop, or a more modest one.

5

As an additional point of interest, we noted that of the 25 patients who underwent SRT, 18 (72%) had stable hearing prior to the initiation of treatment. This means that the SRT population was not pre-selected for already having a decline in hearing trajectory. Furthermore, following the course of the study, of all patients who had a decline in hearing on the tumor side, only 2 had subsequent decline in hearing on the contralateral side, with no difference between the distribution in the treatment groups (p = 1.000). Hence,

Table 4 – Predictors of the outcome “hearing preservation”. Defined by AAOHNS classification at the end of follow up

Sex Male Female Age

n (%) n (%) median years (range)

Hearing Preserved (n = 63)

34 29 53.0

Hearing Loss (n = 60)

(54%) (46%) (18–85)

28 32 57.0

(47%) (53%) (16–78)

Univariate comparisona

Multivariate modelb Logistic Regression

p-value

O.R.

p-value

0.418 0.327

0.54 1 1.03

0.192 0.159

Tumor location IAC n (%) 29 (46%) 19 (32%) CPA n (%) 34 (55%) 41 (68%) 0.103 n/a n/a 0.989 Tumor sizec IAC n (%) 29 (46%) 19 (32%) 1.08 Small n (%) 27 (43%) 32 (53%) 1.11 Medium n (%) 7 (1%) 9 (15%) 0.261 1 Tumor growth statusd Not growing n (%) 52 (83%) 41 (68%) 0.37 Growing n (%) 11 (17%) 19 (32%) 0.067 1 0.113 AAOHNS hearing class at onset on tumor side A n (%) 39 (62%) 11 (18%) 0.08 B n (%) 24 (38%) 49 (82%) <0.001 1 <0.001 Treatment modality 1 Conservative n (%) 53 (84%) 45 (75%) 0.806 SRT n (%) 10 (16%) 15 (25%) 0.209 1.20 Follow-up in months Median (Range) 27 (3–130) 48(11–131) 0.011 1.03 <0.001 a Chi-square, Fisher’s exact, or Kolmogorov–Smirnoff was used, as appropriate, for univariate comparisons. bMultiple regression was used for multivariate comparisons. cThe sample included no case of “large” tumor size. “PTA difference” (last – first audiogram) and “SDS difference” (first-last audiogram).

PTA difference univariate comparisona

PTA Multivariate modelb Multiple Regression

SDS difference univariate comparisona

p-value

p-value

p-value

Sex

0.243 0.433

Age

0.034 0.916

0.634 Tumor location

0.477 n/a

0.232 Tumor size class

0.164 0.770

0.230 Tumor growth status

0.146 0.011

0.057 AAOHNShearingclassatonsetontumorside

0.022 0.924

0.421 Treatment modality

<0.001 0.213

0.013 Follow-up in months

0.304 <0.001

0.024 a

0.190 b

Kruskal Wallis or simple linear regression test was used for univariate comparisons. Multiple regression was used for multivariate comparisons.

Please cite this article as: Elliott A, et al, Hearing preservation in vestibular schwannoma management, Am J Otolaryngol–Head and Neck Med and Surg (2015), http://dx.doi.org/10.1016/j.amjoto.2015.02.016

6

AMER ICA N JOURNAL OF OT OLA RYNGOLOGY– H E A D A N D N E CK M EDI CI N E AN D S U RGE RY X X (2 0 1 5) XXX – XXX

the natural rates of hearing loss with time or presbyacusis or preexisting systemic factors were not likely involved. This indicates that both pre-SRT hearing status and contralateral hearing status had little effect on the hearing status of tumor-side ear.

3.4.

Predictors of the outcome: hearing preservation

Table 4 identifies the characteristics predictive of hearing preservation. Both univariate comparisons and the multivariate model revealed that only hearing class at outset (OR 0.08, p < 0.001) and follow-up time (OR 1.03, p < 0.001) were significant predictors of hearing preservation at the end of follow up.

3.5. Predictors of the outcomes: PTA difference and SDS difference Table 4 identifies the characteristics predictive of PTA difference and SDS difference. Univariate comparisons revealed that treatment modality (p = 0.013) and follow-up time (p = 0.024) were significant predictors of PTA difference. The SRT group tended to have a larger PTA drop. However, the multivariate model revealed only tumor growth status (p = 0.011) and follow-up time (p = <0.001) to be independent predictors of PTA difference and thus a worse hearing outcome. In contrast, univariate comparisons revealed that sex (p = 0.034), tumor growth status (p = 0.022), and hearing class at outset (p < 0.001) were all significantly associated with SDS difference and thus a worse hearing outcome. There was no multivariate analysis for SDS as this variable did not meet the specifications for any multivariate non-parametric tests for continuous variables.

3.6.

Predictors of the outcome: hearing survival

The median hearing preservation survival time for the case series was 46 months (Fig. 1). Using Cox PH regression, the model revealed that only hearing class at outset predicted hearing survival (OR 0.261, p < 0.001) and with all covariates taken into account in the model, the median hearing survival time was 25 months (Fig. 2). There was no significant difference between the hearing survival curves of conservative versus SRT patients (median survival 48 and 35 months

for conservative and SRT groups respectively, p = 0.276) (Fig. 3). Similarly, when creating a model stratified by intervention and analyzed with Cox PH, hearing class at outset was still the only predictive factor of survival (OR 0.240, p < 0.001). Thus, the hearing class at outset variable was explored further and analysis of the survival curves using both log rank and Cox PH revealed that better hearing at outset (class A) conferred a significant survival advantage over class B (median survival 71 and 28 months for class A and B respectively, p < 0.001) (Fig. 4). Furthermore, the type of intervention (conservative versus SRT) had no effect on this finding (p = 0.414).

4.

Discussion

The purpose of this study was to compare hearing outcomes between patients with vestibular schwannomas who either were treated conservatively or underwent stereotactic radiotherapy. We were particularly interested in determining if patients who underwent SRT had a worse hearing prognosis. This is a retrospective case series analysis, and limited to clinical history, clinical outcomes as captured in our database, and patient reports. At our center, we manage most VS conservatively. The decision to intervene with SRT is based on tumor growth (p < 0.001). This treatment algorithm stems from the finding that approximately 75% of VS do not grow, and no definite predictors of growth have been identified [1]. Thus, the “wait and scan” approach attempts to avoid the morbidity associated with interventions until it is deemed that these interventions are truly necessary [3,4]. The major finding of this study was that we failed to demonstrate a significant difference in the outcome measures of hearing preservation, hearing survival and speech discrimination score between conservative versus SRT managed patients. Although SRT patients had a higher Pure Tone Average difference of approximately 10 dB compared to conservative patients, treatment method (SRT versus conservative) was not

Fig. 1 – Survival curve showing the probability of experiencing serviceable hearing loss according to AAOHNS classification in all patients (n = 123). Median survival time is 46 months. Please cite this article as: Elliott A, et al, Hearing preservation in vestibular schwannoma management, Am J Otolaryngol–Head and Neck Med and Surg (2015), http://dx.doi.org/10.1016/j.amjoto.2015.02.016

AMER IC AN JOURNAL OF OT OLA RYNGOLOGY– H E A D A N D NE CK M E D ICI N E AN D S U RGE RY X X (2 0 1 5) XXX – XXX

7

Fig. 2 – Survival curve showing the probability of experiencing serviceable hearing loss within a Cox Proportional Hazards Regression model taking into account covariates (treatment modality, gender, age, tumor size class and growth status, and AAOHNS class at outset). The only covariate that significantly contributes to the model is AAOHNS hearing class at outset (p < 0.001). Median survival time is 25 months.

predictive of this difference in the context of a multivariate analysis, suggesting it to hold little clinical significance. Overall, serviceable hearing preservation was 51% which is consistent with reports in the literature. However, studies citing hearing preservation ranging from 43% to 74% have not analyzed both SRT and conservative management in any comparative manner [2,3,7,8,11,12]. Overall, the variable “hearing preservation” is a better outcome measure than

solely PTA or SDS. Hearing preservation is based on AAOHNS classification, which in turn is comprised of a grading of SDS and PTA. This combination seems to avoid the seemingly spurious predictors, (i.e. such as gender in our case), of these variables when they stand alone. The AAOHNS system is the commonly accepted measure of hearing preservation; however, there is debate over the value of this pass/fail system and the definition of “serviceable” hearing [8]. There

Fig. 3 – Kaplan Meier survival curves showing the probability of experiencing hearing loss according to AAOHNS classification in patients who were followed with observation (n = 98) or who received SRT (n = 25). Survival curves show no significant long-term differences between the two treatment groups (p = 0.276). Median survival times are 48 and 35 months for the conservative and SRT groups, respectively. Please cite this article as: Elliott A, et al, Hearing preservation in vestibular schwannoma management, Am J Otolaryngol–Head and Neck Med and Surg (2015), http://dx.doi.org/10.1016/j.amjoto.2015.02.016

8

AMER ICA N JOURNAL OF OT OLA RYNGOLOGY– H E A D A N D N E CK M EDI CI N E AN D S U RGE RY X X (2 0 1 5) XXX – XXX

Fig. 4 – Kaplan Meier survival curves showing the probability of experiencing hearing loss according to AAOHNS classification in patients who had the initial hearing class A (n = 48) or B (n = 73). Survival curves show significant long-term hearing survival advantage to those with superior hearing status at outset (p < 0.001). Median survival times are 71 and 28 months for the Class A and Class B hearing groups, respectively.

are multiple conflicting reports on the ideal measure of hearing and how it should be standardized [13–15]. The major and probably the most important finding of this study was that patients who had better hearing at the outset (AAOHNS class A versus B), were more likely to retain serviceable hearing, and for longer periods of time, irrespective of treatment. Patients who had class A had a 62% hearing preservation rate and conferred a median of 46 additional months compared to the 38% hearing preservation rate of those who had class B hearing at the onset. Furthermore, the class A hearing group tended to stabilize after about 50 months, whereas the class B group continued to deteriorate. This finding was regardless of treatment modality and in accordance with a study by Stangerup et al. [16] which followed 932 conservatively managed patients and found that the best predictor of serviceable hearing at the end of follow-up was an initial excellent hearing status. Follow-up time was also found to be a predictor of serviceable hearing preservation. The shorter the follow-up time, the more likely a patient was to be classified as having hearing preservation (p < 0.001). While the overall hearing survival rate was 51%, the survival curves shown indicate that this could be partly due to lack of longer term follow up. Despite our relatively long average follow up time of 43 months, comparable or longer than many studies, those patients followed for longer continued to have deterioration of hearing [17]. Loss of follow-up creates a bias against those with gradual hearing loss [3]. This issue has been documented by others with the majority of studies reporting a follow-up of <5 years [15]. Surprisingly, a review by Sughrue et al. [12] found that there was no correlation between length of follow-up and hearing preservation. Hearing class at onset and follow up time were the only major and minor predictors, respectively, of hearing preservation.

Other studies have found additional predictors, albeit inconsistently and often contradicted by other studies including age [18], tumor growth [4,12], and cochlea radiation dosage [7,11]. The meaning of class A/B hearing versus C/D in real life functionality can be debated, as our previous sound field hearing comprehension studies [19] in groups of VS patients with different classes of hearing have not shown a clear demarcation at the cut-off points for the classes. However, most would agree that class C/D hearing is substantially worse than class A/B, if only because the lower speech discrimination in class C/D cannot be improved with hearing aids.

5.

Conclusion

Our findings are relevant to our clinical center, but may also inform clinical practice elsewhere. The strength of our study is grounded in the real world of clinical practice utilizing clinically relevant inclusion and exclusion criteria to build a strong case series. With a rare tumor and uncertain prognostic markers, this study can inform best-practice in the management of VS. Secondly, the analysis used for hearing survival was original, as we were not able to identify any other studies using Cox Proportional Hazards Regression to model hearing survival. Hearing is a sense with loss having a serious impact on the quality of life of patients with VS. One of the goals of our center is to retain serviceable hearing while maintaining tumor control. This study failed to demonstrate a difference between the conservative and SRT treatment modalities with regards to hearing outcomes. While it must be recognized that this is a retrospective analysis of clinical cases, without the rigorous controls of experimental design (i.e. radiation doses to the cochlea), one can take comfort in the decision to

Please cite this article as: Elliott A, et al, Hearing preservation in vestibular schwannoma management, Am J Otolaryngol–Head and Neck Med and Surg (2015), http://dx.doi.org/10.1016/j.amjoto.2015.02.016

AMER IC AN JOURNAL OF OT OLA RYNGOLOGY– H E A D A N D NE CK M E D ICI N E AN D S U RGE RY X X (2 0 1 5) XXX – XXX

proceed with SRT, that there is no absolutes in the loss and/or acceleration of loss of hearing. Furthermore, the strongest predictor of maintenance of serviceable hearing is excellent hearing (class A) at the onset of treatment. These results could have a significant effect on how patients with vestibular schwannoma are managed and the discussions intricate to client-centered care.

Acknowledgments This work was made possible from a Dalhousie University Medical Research Foundation Music-in-Medicine summer research grant awarded to the principal author.

REFERENCES

[1] Nikolopoulos TP, Fortnum H, O’Donoghue G, et al. Acoustic neuromas growth: a systematic review of the evidence. Otol Neurotol 2010;31:478–85. [2] Smouha EE, Yoo M, Mohr K, et al. Conservative management of acoustic neuroma: a meta-analysis and proposed treatment algorithm. Laryngoscope 2005;115:450–4. [3] Martin TP, Tzifa K, Kowalski C, et al. Conservative versus primary surgical treatment of acoustic neuromas: a comparison of rates of facial nerve and hearing preservation. Clin Otolaryngol 2008;33:228–35. [4] Hajioff D, Raut VV, Walsh RM, et al. Conservative management of vestibular schwannomas: third review of a 10-year prospective study. Clin Otolaryngol 2008;33:255–9. [5] Nickele CM, Akture E, Gubbels SP, et al. A stepwise illustration of the translabyrinthine approach to a large cystic vestibular schwannoma. Neurosurg Focus 2012;33:E11. [6] Sweeney P, Yajnik S, Hartsell W, et al. Stereotactic radiotherapy for vestibular schwannoma. Otolaryngol Clin North Am 2009; 42:655–63. [7] Yang I, Aranda D, Han SJ, et al. Hearing preservation after stereotactic radiosurgery for vestibular schwannoma: a systematic review. J Clin Neurosci 2009;16:742–7.

9

[8] Pennings RJE, Morris DP, Clarke L, et al. Natural history of hearing deterioration in intracanalicular vestibular schwannoma. Neurosurgery 2011;68:68–77. [9] AAOHNS. American Academy of Otolaryngology–Head and Neck Foundation, Inc. Committee on Hearing and Equilibrium guidelines for the diagnosis and evaluation of therapy in Menière's disease. Otolaryngol Head Neck Surg 1995;113:181–5. [10] Monsell EM, Balkany TA, Gates GA, et al. Committee on Hearing and Equilibrium guidelines for the evaluation of hearing preservation in acoustic neuromas (vestibular schwannoma). Otolaryngol Head Neck Surg 1995:179–80. [11] Thomas C, Di Maio S, Ma R, et al. Hearing preservation following fractionated stereotactic radiotherapy for vestibular schwannomas: prognostic implications of cochlear dose. J Neurosurg 2007;107:917–26. [12] Sughrue ME, Yang I, Aranda D, et al. The natural history of untreated sporadic vestibular schwannomas: a comprehensive review of hearing outcomes. J Neurosurg 2010;112:163–7. [13] 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;24:642–9. [14] Timmer FCA, Hanssens PEJ, van Haren AEP, et al. Gamma knife radiosurgery for vestibular schwannomas: results of hearing preservation in relation to cochlear radiation dose. Laryngoscope 2009;119:1076–81. [15] Bassim MK, Berliner KI, Fisher LM, et al. Radiation therapy for the treatment of vestibular schwannoma: a critical evaluation of the literature. Otol Neurotol 2010;31:567–73. [16] Stangerup SE, Tos M, Homsen JT, et al. Hearing outcomes of vestibular schwannoma patients managed with ‘wait and scan’: predictive value of hearing level at diagnosis. J Laryngol Otol 2010;124:490–4. [17] Van de Langenberg R, de Bondt BJ, Nelemans PJ, et al. Predictors of columetric growth and auditory deterioration in vestibular schwannomas followed in a wait and scan policy. Otol Neurotol 2011;32:338–44. [18] Tamura M, Yoma S, Arkha Y, et al. Hearing preservation after gamma knife radiosurgery for vestibular schwannomas presenting with high-level hearing. Neurosurgery 2009;64:289–96. [19] Lee A, Clarke L, Bance M. Vestibular schwannoma: how much residual hearing is useful? J Otolaryngol Head Neck Surg 2008;37:399–410.

Please cite this article as: Elliott A, et al, Hearing preservation in vestibular schwannoma management, Am J Otolaryngol–Head and Neck Med and Surg (2015), http://dx.doi.org/10.1016/j.amjoto.2015.02.016