Vestibular schwannoma is a benign lesion derived from the neurilemma sheath and accounts for approximately 6% of newly diagnosed intracranial tumors.1 Microsurgery and stereotactic radiotherapy are the most widely used treatment modalities for vestibular schwannoma. Patient quality of life after surgery for vestibular schwannoma has been emphasised recently. In addition to sensorineural hearing loss, tinnitus is one of the most common symptoms in vestibular schwannoma patients,2-4 and bothersome tinnitus can occur after treatment in some cases. Many theories have been proposed to explain the origin of tinnitus, including tinnitus associated with vestibular schwannoma. Some authors suggest that tinnitus originates in peripheral organs such as the cochlea or cochlear nerve, whereas others propose a central cause such as the cochlear nucleus, brainstem, and cerebral cortex. Some patients still suffer from tinnitus even after tumor removal and vestibulocochlear nerve section, which supports the theory that tinnitus is likely a phantom symptom of central origin.5-7

Some authors state that even if the vestibulocochlear nerve is preserved, surgical trauma during treatment of vestibular schwannoma may cause postoperative tinnitus.8 Ionized radiation used to treat vestibular schwannoma can also cause cochlea and cochlear nerve toxicity9 and may aggravate tinnitus postoperatively, regardless of hearing preservation.

In this study, we investigated changes in tinnitus in patients who underwent translabyrinthine microsurgery (TLM) and gamma knife radiosurgery (GKS). Factors related to tinnitus, such as pretreatment hearing and tumor volume, were also evaluated.

We evaluated outcomes of 78 vestibular schwannoma patients who underwent TLM (n=54) or GKS (n=24) and regularly followed at Severance Hospital between January 2009 and December 2012. Among them, 64 patients completed a tinnitus questionnaire. Microsurgery was performed at the Department of Otolaryngology by one senior surgeon via the translabyrinthine approach and tumor was completely or near totally removed in all patients. GKS was performed at the Department of Neurosurgery by the other one senior surgeon. After excluding 12 patients in TLM group and 6 patiens in GKS group who did not have pre- or postoperative tinnitus, 27 patients who underwent surgical removal via TLM and 19 patients who underwent GKS were finally enrolled. In the TLM group, the cochlear nerve was sectioned in all cases and resulted in deafness of operated ear. GKS was performed using the 201 source Cobalt-60 Leksell gamma unit. The tumor margin was covered by the 50% isodose line, and 12-13 Gy was delivered to the margin.

We evaluated preoperative pure tone audiometry, and mean hearing levels were expressed as the average of hearing thresholds at 500, 1000, 2000 and 3000 Hz (4-tone average). Preoperative tumor volume on magnetic resonance imaging was measured using a three-dimensional reconstruction program (Aquarius iNtuition, TeraRecon) (Fig. 1). Tinnitus severity was evaluated using the tinnitus handicap inventory (Supplementary Fig. 1, only online) and visual analogue scale pre- and one month postoperatively. Visual analogue scale was used for loudness (1-10), awareness (1-10), and annoyance (1-10). Change in postoperative tinnitus was defined as change in tinnitus handicap inventory scores of more than 20% and 10 scores positively or negatively, otherwise postoperative tinnitus was defined as unchanged. This study was approved by the Institutional Review Board of Severance Hospital, Yonsei University Health System (4-2012-0910).

Results were analysed using Fisher's exact test, paired t-test, and Pearson correlation test as appropriate. The level of significance was set at p value less than 0.05. Statistical analyses were performed using SPSS v16.0 (SPSS Inc., Chicago, IL, USA).

We evaluated 46 patients including 11 males and 35 females with a mean age of 45.9±11.3 years at treatment. The vestibulocochlear nerve was definitely cut during TLM, but anatomically preserved in the GKS group. Forty three patients had tinnitus on the side with tumor and three patients had bilateral tinnitus. Mean tumor volume was 5.05±8.68 cm³ in the TLM group and 4.37±6.91 cm³ in the GKS group (Table 1).

The ratios of patients with change in tinnitus, based on the tinnitus handicap inventory, are shown in Fig. 2. There was a higher rate of improved patients following TLM than GKS (p=0.016). In fact, the GKS group showed a strong tendency of tinnitus aggravation (p<0.001) (Fig. 2). This tendency was also seen in visual analogue scale (Supplementary Fig. 2, only online).

The mean tinnitus handicap inventory score was significantly reduced from 51.25±24.45 preoperatively to 31.7±30.33 postoperatively in the TLM group (p=0.006). Mean score of visual analogue scale also showed significant decrease in all items of loudness, awareness and annoyance (p<0.05) (Fig. 3). In GKS group, mean score of tinnitus handicap inventory was significantly increased from 15.29±21.47 preoperatively to 41.2±2.58 postoperatively (p<0.001). Mean score of visual analogue scale was significantly increased in items of loudness and annoyance (p<0.05), but not awareness (p=0.061) (Fig. 4).

We also evaluated factors related to tinnitus such as tumor volume and preoperative residual hearing using Pearson's correlation test. Preoperative tinnitus handicap inventory scores did not correlate with preoperative tumor volume (p>0.05). Postoperative tinnitus handicap inventory score and tinnitus handicap inventory change (difference between pre- and postoperative tinnitus handicap inventory scores) did not correlate with preoperative tumor volume in both the TLM and GKS groups (p>0.05). Furthermore, we compared preoperative pure tone audiometry and tinnitus handicap inventory scores to investigate the relationship between preoperative hearing function and postoperative tinnitus handicap inventory score, and found no significant correlation in both groups (p>0.05).

There are various results of changes in tinnitus after microsurgery. Kameda, et al.10 reported that tinnitus disappeared in 25.2%, improved in 33.3%, remained unchanged in 31.6%, and worsened in 9.9% of patients after a retrosigmoid approach. There was no difference in tinnitus incidence based on whether or not the vestibulocochlear nerve was resected during surgery. Schaller, et al.11 suggested that sudden drop of blood pressure after hearing preserving surgery can be a risk factor for developing tinnitus after microsurgery. Another study reported reduction in tinnitus severity in 45% of patients who underwent the middle cranial fossa approach.12 Harcourt, et al.13 investigated change in tinnitus after TLM and reported that patients who had probable or definite cochlear nerve section had significantly lower postoperative tinnitus severity. In addition, it was found that cochlear neurectomy improved tinnitus control rates compared to vestibular nerve section alone.14 In our study, 52% (14/27) of patients who underwent TLM with vestibulocochlear nerve section showed improved tinnitus.

We have unpublished results on eight patients who underwent tumor removal via middle cranial fossa approach (MCFA) in our hospital at the same period and by the same surgeon. In the middle cranial fossa microsurgery group, cochlear nerve and hearing can be preserved after tumor removal. Tinnitus handicap inventory and visual analogue scale scores were also evaluated in this group. The result of MCFA group is shown in Supplementary Fig. 3 (only online). MCFA group showed a similar rate of tinnitus improvement to the TLM group; however, there was no significant changes in mean tinnitus handicap inventory or visual analogue scale scores (p>0.05). They were excluded in the present study, because the number of MCFA group patients is too small and this technique was applied only to patients who had relatively smaller size of the tumor and a better hearing compared to GKS group and TLM group. Future research in MCFA would help us prove the relationship between cochlear nerve continuity and tinnitus in the surgical treatment.

Previous reports on patients who underwent GKS showed that change in tinnitus occurred only in a few patients (4%),15,16 whereas other studies showed reduced tinnitus in about 45% of patients treated with GKS.17,18 In our study, patients who underwent GKS showed worsened tinnitus in 74%, based on tinnitus handicap inventory score.

Although various results of changes in tinnitus after microsurgery and GKS have been reported, our study showed better tinnitus control rate in the TLM than GKS group. Moller19 suggested that the cause of tinnitus in vestibular schwannoma is peripheral nerve injury, whereas Lockwood, et al.20 proposed that neural network malfunction due to the cerebral neural plasticity can be the cause of tinnitus. It is highly likely that a peripheral cause such as cochlea or cochlear nerve degeneration or firing has an important role in addition to a central cause in generating tinnitus in vestibular schwannoma patients. Therefore, vestibulocochlear nerve integrity is an important factor for tinnitus control after treatment; and cochlear nerve section may have benefit in postoperative tinnitus improvement. In some cases, secondary changes in the brainstem or post-brainstem central nervous system might act like phantom areas in generating tinnitus.5,21-23 In these patients, vestibulocochlear nerve section may be less effective or not helpful in reducing tinnitus after operation. After GKS, tumor swelling may occur prior to shrinkage, and damage to the peripheral organs can result from radiation exposure. Direct radiation-induced cochlear nerve toxicity is a causative factor for hearing loss and may result in worsening tinnitus.9,24 Although controversial our results suggest that TLM with vestibulocochlear neurectomy may relieve or prevent tinnitus; even though GKS may induce damage to the cochlea and cochlear nerve by direct injury or swelling and cause worsening of tinnitus symptoms in a large proportion of patients.

There are many reports of tinnitus severity related to tumor size. Kohno, et al.25 reported that the smaller tumour group show a significantly worse prognosis of tinnitus, and others claim that tinnitus is common when tumours are small or larger than 4.5 cm.26 However, our study did not find any relationship between tumor volume and tinnitus handicap inventory scores or differences in pre- and postoperative tinnitus handicap inventory scores in both groups.

In studying hearing function, Kanzaki, et al.27 found that tinnitus present in 78.6% of patients preoperatively increased to 89.3% postoperatively when hearing preservation was attempted. Kohno, et al.25 found that the group with better preoperative hearing showed higher rate of worsening tinnitus. In our study, all patients in the TLM group sacrificed hearing, and there was no relationship between preoperative pure tone audiometry and pre- or postoperative tinnitus handicap inventory scores or change in tinnitus handicap inventory scores. In the GKS group, we also did not find any relationship between preoperative pure tone audiometry and pre- or postoperative tinnitus handicap inventory scores or change of tinnitus handicap inventory scores.

In cases where hearing preservation is not intended, cutting the vestibulocochlear nerve may help reduce the development of tinnitus in vestibular schwannoma treatment.