Progression of hearing loss after LINAC-based stereotactic radiotherapy for vestibular schwannoma is associated with cochlear dose, not with pre-treatment hearing level

BACKGROUND: Although stereotactic radiotherapy (SRT) for vestibular schwannoma has demonstrated excellent local control rates, hearing deterioration is often reported after treatment. We therefore wished to assess the change in hearing loss after SRT and to determine which patient, tumor and treatment-related factors influence deterioration. METHODS: We retrospectively analyzed progression of hearing loss in patients with vestibular schwannoma who had received stereotactic radiosurgery (SRS) or fractionated stereotactic radiotherapy (FSRT) as a primary treatment between 2000 and 2014. SRS had been delivered as a single fraction of 12 Gy, and patients treated with FSRT had received 30 fractions of 1.8 Gy. To compare the effects of SRS and FSRT, we converted cochlear doses into EQD(2). Primary outcomes were loss of functional hearing, Gardner Robertson (GR) classes I and II, and loss of baseline hearing class. These events were used in Kaplan Meier plots and Cox regression. We also calculated the rate of change in Pure Tone Average (PTA) in dB per month elapsed after radiation—a measure we use in linear regression—to assess the associations between the rate of change in PTA and age, pre-treatment hearing level, tumor size, dose scheme, cochlear dose, and time elapsed after treatment (time-to-first-audiogram). RESULTS: The median follow-up was 36 months for 67 SRS patients and 63 months for 27 FSRT patients. Multivariate Cox regression and in linear regression both showed that the cochlear V90 was significantly associated with the progression of hearing loss. But although pre-treatment PTA correlated with rate of change in Cox regression, it did not correlate in linear regression. The time-to-first-audiogram was also significantly associated, indicating time dependency of the rate of change. None of the analysis showed a significant difference between dose schemes. CONCLUSIONS: We found no significant difference between SRS and FSRT. As the deterioration in hearing after radiotherapy for vestibular schwannoma was associated with the cochlea V90, restricting the V90 may reduce progression of hearing loss. The association between loss of functional hearing and baseline PTA seems to be biased by the use of a categorized variable for hearing loss.