Journal of Cancer Research and Therapeutics

: 2014  |  Volume : 10  |  Issue : 1  |  Page : 29--37

Critical neurological structure sparing radiosurgery of vestibular schwannoma: Dosimetric comparison of different techniques and dose prescription methods

Shamurailatpam Dayananda Sharma1, Pranav Chadha1, Kaustav Talapatra1, Vaibhav Mahtre1, Abhaya P Kumar2, Anandh Balasubramaniam2,  
1 Department of Radiation Oncology, Kokilaben Dhirubhai Ambani Hospital and Medical Research Institute, Andheri, Mumbai, India
2 Department of Neurosurgery, Kokilaben Dhirubhai Ambani Hospital and Medical Research Institute, Andheri, Mumbai, India

Correspondence Address:
Shamurailatpam Dayananda Sharma
Department of Radiation Oncology, Room No AS/55024, Kokilaben Dhirubhai Ambani Hospital and Medical Research Institute, Four Bungalows, Andheri (W), Mumbai - 400 053


Aim: To investigate potential sparing of critical neurological structures (CNSs) during radiosurgery of vestibular schwannoma (VS) employing different techniques and dose prescription methods. Materials and Methods: Fused CT and MRI datasets of eight patients with unilateral VS representing a wide range of target volume (0.48 to 12.08 cc; mean = 3.56 cc), shape and proximity to CNSs such as cochlea, trigeminal nerve and brainstem were re-planned employing static conformal field (SCF), dynamic conformal arc (DCA) and intensity modulated radiosurgery (IMRS) techniques. For every patient, five plans were created for a fixed margin dose of 12 Gy prescribed at 80% in three plans (SCF_80%, DCA_80%, and IMRS_80%) and 50% in another two plans (SCF_50% and DCA_50%). All plans were compared using standard dosimetric indices. Results: Primary goal of every plan to cover ≥99% of target volume with 12 Gy was fulfilled for all patients with minimum significant dose to target (D 99 ) ≥11.99 Gy. Best conformity index (CI Paddick = 0.62 ± 0.12) was observed in SCF_80% and DCA_80% plans whereas; sharpest dose gradient index of 3.40 ± 0.40 was resulted from DCA_50%. All five plans resulted similar maximum dose to brainstem (11.04 ± 2.23 to 11.53 ± 1.10 Gy), cochlea (9.02 ± 1.79 to 10.15 ± 1.26 Gy) and trigeminal nerve (11.55 ± 1.38 to 12.19 ± 2.12 Gy). Among 80% prescription plans, IMRS_80% reduces mean and D 5 (P < 0.05) to all CNSs. Prescription of dose at 50% isodose sharpened the dose gradient and significantly (P < 0.05) reduced mean dose and D 5 to all CNSs at the cost of target conformity (P = 0.01). Mean dose to cochlea and trigeminal nerve were least at 4.53 ± 0.86 and 6.95 ± 2.02 Gy from SCF_50% and highest at 6.65 ± 0.70 and 8.40 ± 2.11 Gy from DCA_80% plans respectively. Conclusion: This dosimetric data provides a guideline for choosing optimum treatment option and scope of inter institutional dosimetric comparison for further improvement in radiosurgery of Vestibular Schwannoma (VS).

How to cite this article:
Sharma SD, Chadha P, Talapatra K, Mahtre V, Kumar AP, Balasubramaniam A. Critical neurological structure sparing radiosurgery of vestibular schwannoma: Dosimetric comparison of different techniques and dose prescription methods.J Can Res Ther 2014;10:29-37

How to cite this URL:
Sharma SD, Chadha P, Talapatra K, Mahtre V, Kumar AP, Balasubramaniam A. Critical neurological structure sparing radiosurgery of vestibular schwannoma: Dosimetric comparison of different techniques and dose prescription methods. J Can Res Ther [serial online] 2014 [cited 2021 Sep 25 ];10:29-37
Available from:

Full Text


Vestibular Schwannoma (VS) is relatively rare, slow-growing benign tumor arising from the Schwann cells lining the vestibular branch of the eighth (vestibulocochleae) cranial nerve (CN). The management options of VS include observation, microsurgery and radiotherapy. Stereotactic radiosurgery (SRS) is now established as a standard alternative to microsurgical excision with reported long-term progression-free survival better than 90% and lower risk of morbidity. [1],[2],[3],[4],[5],[6],[7],[8] This encouraging clinical result leads to a growing concern on the treatment related side effects and optimum management option must aim at the reduction in toxicity profile.

The primary side effects after radiosurgery/microsurgery are worsening of hearing, facial (CN VII) and trigeminal (CN V) nerve dysfunction. Worsening of hearing after radiosurgery was reported due to either radiation injury to the cochlea or the dose delivered to the auditory canal or both. [9],[10] Concurrently, the risk of facial and trigeminal neuropathy is associated with the irradiated length of CN and brainstem dose. [11],[12],[13],[14] A few studies have reported retrospective measurement of radiation doses to some of the critical neurological structures (CNSs) such as cochleae, CN and brainstem separately during gamma knife radiosurgery (GKS) of VS. [9],[10],[14],[15],[16] However, dosimetry data on these CNSs from a single treatment plan is rather limited to correlate with the toxicity profile. Radiosurgery treatment planning and delivery techniques of VS is still evolving and need further improvement in regards to reduction of dose to the CNSs without compromising target coverage. Although GKS was considered as the gold standard for SRS of many cranial tumor including VS, latest linear accelerator (Linac) based radiosurgery system such as Novalis Tx (Varian Medical System, USA and BrainLAB AG, Germany) has emerged as a promising alternative with additional advantages of image guidance and frameless option. Limited data available on the dosimetric comparison of various radiosurgery systems employing different techniques for VS demonstrate wide variations in the set planning goal which eventually influence the results. [17],[18],[19],[20] The recently introduced high definition multi-leaf collimator (HDMLC) in Novalis Tx have demonstrated dosimetric advantage over other MLC in the radiosurgery planning of intracranial tumours. [21] HDMLC together with the capability of modern treatment planning system (TPS) offer various options to choose for. Therefore an appropriate radiation planning and delivery technique should be designed, tested and adopted to reduce the dose to these CNSs while maintaining adequate target coverage. In this comprehensive dosimetric study of VS, the potential of static conformal field (SCF), dynamic conformal arc (DCA) and intensity modulated radiosurgery (IMRS) was investigated for sparing of cochlea, brainstem and CN without compromising the target coverage. As majority of the reported clinical outcome and their correlation with the doses to cochlea, CN and brainstem were from gamma knife series, we also investigated impact of the traditional gamma knife dose prescription methods on the dose to target and CNSs.

 Materials and Methods

Our institutional protocol for radiosurgery of VS includes acquisition of double-dose gadolinium contrast enhanced T1- and T2- MRI sequence of the patient one day prior to frame fixation. MRI sequences were acquired at 1mm slice thickness using a 3 Tesla MRI scanner (Magnetom Trio ® Siemens Medical Systems, Germany). Target was delineated on iPlan (BrainLAB, Germany) treatment planning system (TPS) according to tumor visualization on T1 contrast and T2- weighted MR image sequences. CNSs such as brainstem, cochlea, semicircular canal, trigeminal nerve, and vestibule were delineated on the same MRI sequences. Stereotactic planning CT scan of the patient with the BrainLab stereotactic invasive frame (BainLAB, Germany) in place was acquired at 1 mm slice thickness on Somatom Sensation CT scanner (Siemens Medical System, Germany). The CT and MRI images of the patient were fused automatically on iPlan TPS using image registration software based on mutual information. Treatment plan was created on iPlan TPS using SRS6MV X-rays and 7-11 static conformal fields (SCF), shaped using a recently available high definition multi-leaf collimator (HDMLC). The HDMLC is a tertiary collimator in Novalis Tx linear accelerator and has a projected leaf width of 0.25 cm and 0.5 cm at isocenter. The detailed physical and dosimetric characteristics of HDMLC especially for radiosurgery and IMRT were reported elsewhere. [22] In comparison to other mMLC/MLC, HDMLC has thinner leaf width, less effective penumbra and leakage. A margin dose of 12-13 Gy (mean 12.3 Gy) was prescribed at 80% isodose while maintaining the maximum dose to brainstem at 12 Gy and minimizing the dose to cochlea as low as possible. Treatment was delivered on a NovalisTx Linac at a high dose rate of 1000 MU/min under cone beam computed tomography (CBCT) image guidance.

The fused CT and MRI datasets of eight consecutive patients previously treated for unilateral VS were selected for this dosimetric study. The patient demographic and target and CNSs volumes characteristic are summarized in [Table 1]. The selected cases represent a wide range of target volume (0.48 cc to 12.08 cc; mean = 3.56 cc; median = 2.45 cc), shape and proximity to cochlea, trigeminal nerve and brainstem. All patients were retrospectively re-planned on iPlan TPS using eight SCF and five dynamic conformal arc (DCA) techniques. For each technique, two plans (e.g., SCF_80% and SCF_50%) were created with margin dose of 12 Gy at 80% and 50% isodose respectively. Treatment plan with the margin dose of 12 Gy at 50% isodose was created to closely simulate gamma knife surgery plan. This was realized by partially shielding the periphery of the target using HDMLC and prescribing a higher dose of 24 Gy at isocenter as compared to 15 Gy for 80% prescription method. For every patient IMRS plan was also created on iPlan TPS using the same non co-planar beam arrangement used in SCF plans. A margin dose of 12 Gy at 80% was prescribed. Thus, for every patient, five plans (SCF_80%, SCF_50%, DCA_80%, DCA_50% and IMRS_80%) were created for a fixed margin dose of 12 Gy. The primary goal of every plan was to cover at least 99% of the target by the 12 Gy isodose line while maintaining a maximum dose of 12 Gy to brainstem. The secondary endpoint was to minimize the dose to cochlea as low as possible. The plan was still considered to be acceptable with a minor deviation if 12 Gy volume of brainstem was less than or equal to 1 cc. [23] Though no threshold was set for cochlea and trigeminal nerve dose, attempt was made to reduce the dose to these CNSs to as low as possible.{Table 1}

Plan evaluation

The dosimetric outcome of SCF, DCA and IMRS techniques was compared qualitatively and quantitatively in terms of target volume coverage, conformity, dose homogeneity, dose gradient, dose to CNSs and irradiation of normal brain at various dose levels. Target coverage was estimated as the percentage of the target volume (V T ) covered by the prescription dose (V T, Pi ) of 12 Gy. Minimum dose to target was reported as the dose to 99% volume (D 99 ) of the target. The conformation of therapeutic dose volume to the target volume was estimated using the conformity index (CI) defined by Paddick [24] as CI Paddick = {V T, Pi × V T, Pi}/{ V T × V Pi}, where V Pi is the volume of brain tissue including the target enclosed by the prescription isodose. This conformity index takes into account of the location of the prescription isodose volume relative to the target volume. The dose homogeneity within the target was estimated as the ratio of the maximum dose to prescribe dose (MDPD). Dose gradient index (DGI) defined as the ratio of the volume of 6 Gy isodose to 12 Gy isodose was also compared from the different plans. [25] For brainstem, maximum dose, dose to 1 cc and volume receiving ≥12 Gy were estimated for each plan. The minimum, maximum and mean doses to cochlea, vestibule, semicircular canal and traigeminal nerves were reported. The maximum significant dose represented by maximum dose to 5% volume (D 5 ) of all these critical organs was also compared from the different plans. Volume of brain including the target receiving high (≥9 Gy), intermediate (≥6 Gy), and low (≥2.4 Gy) doses were also compared.

Statistical analysis

Statistical analysis was performed using commercially available SPSS version 17 (SPSS, Chicago, IL). Comparison of the dosimetric inferences from various plans was performed using nonparametric Wilcoxon signed rank test. Analysis of variance using two tailed was deemed to be statistically significant when the probability value was less than 0.05 (posthoc P < 0.05).


Target coverage, homogeneity and conformity

[Figure 1] (a-e) shows the dose distribution resulted from the five competing plans of a representative case. Five isodose lines (15 Gy, 12 Gy, 9 Gy, 6 Gy and 2.4 Gy) were shown to illustrate the differences present in target coverage and healthy tissue irradiation at high, intermediate and low doses amongst the techniques. Qualitative analysis of the dose distribution reveal no appreciable difference except spillage of low dose (≤2.4 Gy) to normal brain. Quantitative dosimetric parameters related to target coverage, minimum and mean dose, dose homogeneity and DGI derived from the cumulative dose-volume-histogram (DVH) of different plans were presented in [Table 2]. The value of each parameter in every plan represents the mean of eight patients. The primary goal of the dosimetric study to cover at least 99% of the target volume with 12 Gy isodose was fulfilled in all plans and for every patients with mean ± SD target coverage of 99.04% ± 0.05, 99.08% ± 0.07, 99.09% ± 0.11, 99.08% ± 0.14, and 99.05% ± 0.05 for SCF_80%, SCF_50%, DCA_80%, DCA_50% and IMRS_80% plans respectively. The variation of target coverage amongst the techniques was not statistically significant. The minimum significant dose to target, represented by D 99 was ≥11.99 Gy for all patients and plans with estimated mean ± SD values ranging between 12 ± 0.04 Gy to 12.14 ± 0.21 Gy. Although the variation in D 99 was very small amongst the five plans, the increase of D 99 in treatment plan with 50% prescription method was significant (P = 0.02) compared to corresponding plan with 80% prescription. Of the three plans with prescription dose of 12 Gy at 80% isodose, IMRS_80% plan showed significantly higher mean dose in all patients with overall mean ± SD of 14.49 ± 0.34 Gy as compare to 13.81 ± 0.13 Gy of SCF_80% plan (P = 0.01) and 13.79 ± 0.13 Gy of DCA_80% plan (P = 0.01) respectively. As expected from the planning design, for the same technique, the prescription of dose (12 Gy) at 50% isodose resulted significantly higher (P = 0.01) mean dose and MDPD compared to the corresponding plan with 80% prescription isodose including IMRS_80% plan. However, the mean value of the MDPD was comparable for SCF and DCA plans having the same prescription isodose. IMRS_80% showed slightly higher but significant (P = 0.03) mean ± SD MDPD at 1.32 ± 0.05 as compared to 1.27 ± 0.01 of SCF_80% and 1.26 ± 0.01 of DCA_80% respectively. The DGI was sharpest for DCA_50% plan (3.40 ± 0.40) followed in descending order by SCF_50% (3.68 ± 0.72), IMRS_80% (3.91±0.90), DCA_80% (4.47 ± 1.17) and SCF_80% (4.50 ± 1.23) respectively.{Figure 1}{Table 2}

[Table 3] represents CI Paddick calculated for every plan and patients. In general, conformation of target with prescription dose increases with increase in target volume for all plans. SCF and DCA plans with 80% prescription method demonstrate the best conformity index, both plans resulting the same mean CI Paddick of 0.62 (SD = 0.12 and 0.11). The corresponding plans with 50% prescription method significantly reduce CI Paddick (P = 0.01) with mean ± SD value of 0.57 ± 0.14 and 0.58 ± 0.14 respectively. CI Paddick (mean = 0.60, SD = 0.09) from IMRS_80% plan was slightly better than 50% but inferior to 80% prescription methods of other techniques. However the variation was not significant.{Table 3}

Dose to brainstem and brain

The maximum dose to brainstem from the five plan was within 12 Gy in six (75%) of the eight patients. In the two patients, volume of brainstem receiving ≥12 Gy from the five plans ranges from 0.01 to 0.05 cc with highest point dose of 14.55 Gy. [Table 4] summarizes the mean of maximum dose and dose to 1 cc of brainstem for the eight patients. The mean ± SD of maximum dose to brainstem from the five plans ranges from 11.04 ± 2.23 Gy to 11.53 ± 1.10 Gy and the variation amongst the plans was not statistically significant. The mean dose to 1 cc of brainstem from the five plans varies from 4.66 ± 2.37 Gy to 5.72 ± 2.96 Gy. Amongst the three planning techniques with 80% prescription, IMRS_80% delivered significantly less (P = 0.01) maximum dose and dose to 1 cc of brainstem as compared to SCF_80% and DCA_80% plans respectively. In general, prescription of dose at 50% isodose showed lesser maximum point dose (P > 0.05) and dose to 1 cc (P < 0.05) of brainstem as compared to corresponding plan with 80% prescription isodose. The reduction of dose to 1 cc of brainstem (mean; 13.86% for SCF and 13.11% for DCA) in 50% as compared to 80% prescription method was statistically significant (P = 0.02).{Table 4}

[Table 4] also shows the volume of brain (including the target) receiving high (≥9 Gy), intermediate (≥6 Gy) and low (≥2.4 Gy) doses of radiation. No significant difference in high and intermediate dose volume of brain was observed amongst the techniques except for DCA plan wherein prescription of dose at 50% isodose showed significant reduction (P = 0.03). In regard to low dose volume of brain, DCA plan showed significant reduction (P = 0.01) as compared to SCF and IMRT plans.

Dose to CNSs

[Table 5] represents the dose parameters such as minimum, mean, maximum and D 5 to different CNSs from the five plans. The values are presented as the mean of the eight patients. Amongst the 80% prescription methods, all the dose parameters were least in IMRS_80% and highest in DCA_80% plan for all CNSs. In comparison to SCF_80% and DCA_80% plans, IMRS_80% plan significantly (P < 0.05) reduces the values of dose parameters to all CNSs except maximum dose to cochlea and trigeminal nerve, which remains similar in all the five plans. The maximum point dose from the five plans ranges from 6.13 Gy to 12.82 Gy for cochlea and 9.05 Gy to 15.35 Gy for trigeminal nerve respectively. However, consideration of D 5 as clinically relevant maximum dose to the CNSs showed maximum dose ranging from 5.07 Gy to 10.63 Gy for cochlea and 7.17 Gy to 13.93 Gy in trigeminal nerve. The mean dose to cochlea ranges from 3.41 to 6.11 Gy for IMRS_80%, 4.87 to 7.1 Gy for SCF_80% and 5.53 to 7.52 Gy for DCA_80% respectively. The corresponding values for the trigeminal nerve were 4.83 to 10.31 Gy, 5.42 to 10.8 Gy and 6.01 to 10.88 Gy respectively. Semicircular and vestibular showed lesser maximum and mean dose compared to other CNSs in all respective plans. Prescription of dose at 50% isodose significantly reduces (P < 0.05) the dose parameters to all CNSs compare to corresponding plans with 80% prescription. The mean dose to cochlea from SCF_50% and DCA_50% plans ranges from 3.55 to 6.15 Gy and 4.38 to 6.23 Gy. The corresponding values for trigeminal nerve were 4.76 to 9.35 Gy and 5.34 to 10.9 Gy respectively. Overall, SCF_50% plan showed the least dose to all CNSs followed by IMRT_80%, DCA_50%, SCF_80% and DCA_80%.{Table 5}


Treatment planning of VS is challenging due to its shape, proximity to cochlea, brainstem and other CN. VS also erode the auditory canal and compress adjacent structures such as the auditory portion of the VIII th nerve and facial nerve. We have attempted a unique comprehensive dosimetric study of VS comparing SCF, DCA and IMRS at different dose prescription methods. The wide range of target volume and hence their proximity to CNSs selected in our study will help in testing the strength of the different techniques under different clinical presentations. In order to establish a fair dosimetric comparison of the different plans, we maintained a strict planning goal of covering at least 99% of target volume by 12 Gy while at the same time restricting the dose the CNSs. Amongst the 80% prescription methods, we found no major difference between SCF and DCA in regard to target conformity and dose to brainstem. We have previously reported similar finding for comparatively large intracranial target (mean volume = 54.85 cc, SD = 24.46). [26] However, Perks et al., [17] reported significant improvement of target conformity (mean CI = 1.65; SD = 0.25) in DCA as compare to SCF (mean CI = 1.78; SD = 0.24), at the expense of slightly higher brainstem dose. Similar to our finding, Perks et al., also reported improvement of CI for larger target. In another dosimetric study comparing DCA and rapidarc (RA) plans on three VS targets, RA plan showed superior CI (mean 1.6 vs 2.07) and similar maximum doses to the CNSs. [18] Although we have not included RA plan in this study as it is not supported by iPlan TPS, IMRS plan with non-coplanar beam arrangement showed similar CI (mean 1.7 ± 0.31) as compared to SCF and DCA plans and was comparable to the CI of RA plan reported by Legerwaard et al., [18] The mean ± SD CI Paddick (0.62 ± 0.11 for DCA and 0.60 ± 0.09 for IMRS) in our study was slightly inferior compared to the corresponding values (0.67 ± 0.05 and 0.66 ± 0.06) reported in a very recent study by Gevaert. [20] Another recent study on small VS target has also reported mean conformity index of 0.53 and 0.58 from Linac based radiosurgery and CyberKnife system. [19] The variation in the reported target conformity could be due to the variation in the modality, target size, beam geometry, number of arcs and planning constraints used in the various studies. While deciding the beam geometry in our study, due consideration was given not only to cover at least 99% of the target by prescription dose but also attempted to reduce the maximum dose to nearby CNSs.

We have also investigated the impact of lower isodose (50%) prescription method on the dosimetric outcome of target conformity, dose gradient and dose to CNSs. This was done to closely simulate a typical gamma knife plan, although the penumbra characteristic and planning approach was widely varied. In gamma knife plan, generally very small collimator (e.g., 4 mm) was used to take the distinct advantage of sharp penumbra and multiple isocenters were employed to tightly conform the target by the prescription dose. In case of Linac plan, prescription with 50% isodose line was realized by partially shielding the target using HDMLC, a concept similar to stereotactic body radiosurgery. In both cases a higher dose was prescribed at isocenter to maintain a tumour margin dose of 12 Gy. Perks et al., [17] reported improvement in the CI (P < 0.02), at the debatable cost of lower minimum target dose, from the gamma knife plan as compared to SCF and DCA plans. Gilvert et al., [20] reported highest dose conformity (0.76 ± 0.04), sharp dose gradient (2.55 ± 0.07) and highest dose heterogeneity at the cost of treatment time from the new gamma knife (LGK-PFX) radiosurgery system as compared to Novalis Tx and CyberKnife for five VS targets. Contrary to gamma knife plan, prescription of dose at 50% isodose in SCF and DCA plans showed decrease in CI Paddick (P = 0.01) of all target volume, worst for smallest target, compare to corresponding plans with 80% prescription method. This could be due to the combined effect of the finite width of HDMLC relative to small and irregular shape target which becomes more pronounced when target was shielded partially and also to our strict planning constraint for target coverage and minimum dose to target. During the planning spillage of prescription dose to normal brain was compromised to maintain a strict planning goal of 99% target coverage with prescription dose. This was done to establish an unbiased dosimetric comparison of the different plans. However, reduction of prescription isodose to 50% significantly improves the dose gradient as compared to corresponding plans with 80% prescription including IMRS. Perks and colleagues proclaimed that the high internal dose gradient (MDPD) in gamma knife is advantageous for tumour ablation; but could be interpreted as a disadvantage with respect to hearing preservation. [17] The similar statement can be extrapolated for Linac based plan with 50% prescription. Although the MDPD of GKS plan and SCF and DCA plan with 50% prescription was similar, the internal dose gradient could be quite heterogeneous in GKS plan as compared to Linac plan. However, there was no data to support the correlation between internal dose gradient and tumour ablation rate and hearing preservation.

While some of the previous studies reported the dose to few CNSs, [15],[17],[18],[19] other have not attempted. [20] As the VS target is surrounded by the CNSs, which have their own tight tolerance limit, choice of optimum treatment technique must consider the dose to the CNSs without compromising the target coverage. However differences do exist on the threshold dose and dose parameters to CNSs. Linskey et al., [15] first reported the doses receive by normal temporal bone structures during GKS of VS. In the retrospective planning study of 54 patients with VS treated with a mean dose of 14.2 Gy (range 12-16 Gy), they estimated a dose of 8.9 ± 4.2 Gy at modiolus, 5.5 ± 4.0 Gy at inferior basal turn and 5.7 ± 3.2 Gy at apical turn of the cochlea. These authors postulated that a high dose of radiation to inner ear structures can be a cause of hearing deterioration after GKS of VS. Peak and coworkers have also measured the cochleae radiation dose in 25 patients treated for VS. [16] The measured mean point dose ranged from 3.6 ± 1.2 to 8.1 ± 3.1 Gy. In contrast to the postulate by Linskey, they found that patients with hearing preservation after GKS received a nonsignificantly higher dose to the cochlea than patients with audiological worsening. However, the methodology for measurement of the dose to the cochlea was not mentioned. The first publication of the importance of the cochleae dose to hearing preservation after GKS for VS was by Massager et al., [9] In their retrospective study of 82 patients treated with a fixed margin dose of 12 Gy, they reported a mean cochleae dose of 4.33 Gy (range 1.30-10 Gy). Unlike other previous publication, they measured the mean cochlea dose averaged over the whole 3D volume of the cochlea and found that those with preserved hearing had a mean cochlea dose of 3.7 Gy versus 5.33 Gy in those who lost useful hearing. In another study comprising 69 patients treated for sporadic VS using GKS, mean maximal dose to cochlea was reported at 10.27 Gy (range 3.1 Gy-16.1 Gy). [10] The authors have claimed that significant relations exist between the maximal cochleae dose and the difference in the pure tone average before and after GKS. Although no threshold has been suggested, the authors emphasized the need for exact radiation planning to reduce the cochleae radiation dose if the hearing is to be preserved. We have reported minimum, mean, maximum doses and D 5 for all CNSs. Although none of the technique fulfills the cochleae threshold dose (3.7 Gy) for preserved hearing as suggested by Massager, [9] prescription of dose at 50% isodose and IMRS_80% plan maintained the mean dose below the upper threshold (5.33 Gy) for lost of useful hearing. While the mean maximal dose to cochlea in our study (range 9.02 Gy for SCF_50% to 10.15 Gy for DCA_80%) was less than that of gamma knife plan, [10] it was well above the threshold dose (6 Gy) suggested in a recent review article for hearing preservation. [23] However, mean cochlea dose and D 5 value in our study was less than previously reported values from two separate studies using similar Linac techniques. [18],[20] The higher cochleae dose in our study was a compromise of excellent target coverage (>99%) which was not address and/or kept roughly in between 93% to 99.8% in other studies, besides differences in the modalities. But, during the actual clinical treatment planning we have accepted target coverage range from 95% to 98% in five young patients to limit the maximum dose to cochlea within 6 Gy. Maximum and mean cochleae dose in actual clinical plan ranges from 3.01 to 9.48 Gy and 1.85 Gy to 5.75 Gy.

Different opinions appear in the literature in regard to cranial neuropathy. The Pittsburgh group demonstrated that the irradiated length of cranial nerve correspond to subsequent neuropathy. [11],[12] In contrast, brainstem dose was demonstrated to be the most significant predictor of trigeminal neuropathy in studies by Foote et al., [13] Recently, Hayhurst and colleagues reported maximum dose to brainstem, trigeminal nerve and dose gradient index besides tumour volume as a predictor of adverse radiation effect (ARE) following GKS for VS. [14] The authors claimed that in those with ARE, the mean maximum brainstem and trigeminal nerve dose was 13.82 Gy and 11.96 Gy compared with 11.16 Gy and 9.12 Gy in those with no ARE. They further recommended critical dose threshold for the brainstem of 13 Gy and trigeminal nerve of 9 Gy. This study has not reported the dose to the cochlea. The maximum dose to brainstem in our study was less than 12 Gy and comparable in all plan. The maximum dose increases with increase in target volume. Similar; finding was also reported by Perks et al., with mean maximum dose of 11.67 ± 2.93 Gy, 12.36 ± 2.01, 10.8 ± 5.21 Gy from SCF, DCA and GKS plans. [17] The authors have not studied the doses to cochlea and trigeminal nerve. Irrespective of the technique, the median maximum dose to trigeminal nerve (range 11.75 to 12.37) observed in our study was slightly higher (11.15 Gy) but well above the threshold (9 Gy) reported by Hayhurst et al., [14] although the clinical outcome of our ongoing prospective study on the CNSs sparing radiosurgery of VS was beyond the scope of this study, we have reported the doses to CNSs which was not completely address in the previous publications. We have evaluated the benefit of various Linac based radiosurgery technique in reducing the toxicity profile using published dose-effect relationship to different CNSs. The emergence of new dosimetric data like ours can provide a guideline and scope for inter and intra institutional comparison for further improvement of radiosurgery of VS.


We have reported a comprehensive CNSs sparing dosimetric study of VS comparing SCF, DCA and IMRS at different dose prescription methods. Although all techniques are largely comparable in regard to target coverage, minimum dose to target, maximum doses to the CNSs, differences was observed for target conformity, dose gradient index and mean dose to CNSs. Overall, among the three 80% isodose prescription methods, dosimetrically IMRS was slightly superior to SCF and DCA technique in regard to mean dose to target and CNSs and also dose gradient. The prescription of dose at 50% isodose further sharpens the dose gradient and reduces the mean dose and D 5 to all adjacent CNSs at the minimal cost of target conformity. The results of such comparative study can help in deciding an optimum treatment technique for radiosurgery of VS. The emergence of new dosimetric data like ours can provide a guideline and scope for inter and intra institutional comparison of dosimetric outcome from various techniques for further improvement of radiosurgery of VS. However, systematic prospective clinical trial is needed to clinically validate the dosimetric data.


1Pollock BE, Driscoll CL, Foote RL, Link MJ, Gorman DA, Bauch CD, et al. Patient outcomes after vestibular schwannoma management: A prospective comparison of microsurgical resection and stereotactic radiosurgery. Neurosurgery 2006;59:77-85.
2Kaylie DM, Horgan MJ, Delashaw JB, McMenomey SO. A meta-analysis comparing outcomes of microsurgery and gamma knife surgery. Laryngscope 2000;110:1850-6.
3Lunsford LD, Niranjan A, Flickinger JC, Maitz A, Kondziolka D. Radiosurgery of vestibular schwannomas: Summary of experience in 829 cases. J Neurosurg 2005;Suppl 102:S195-9.
4Hasegawa T, Fujitani S, Katsumata S, Shun K, Yoshimoto M, Koike J. Stereotactic radiosurgery for vestibular schwannomas: Analysis of 317 patients followed more than 5 years. Neurosurgery 2005;57:257-65.
5Friedman WA, Bradshaw P, Myers A, Bova FJ. Linear accelerator radiosurgery for vestibular schwannomas. J Neurosurg 2006;105:657-61.
6Combs SE, Thilmann C, Debus J, Schulz-Ertner D. Long-term outcome of stereotactic radiosurgery (SRS) in patients with acoustic neuromas. Int J Radiat Oncol Biol Phys 2006;64:1341-7.
7Rutten I, Baumert BG, Seidel L, Kotolenko S, Collignon J, Kaschten B, et al. Long-term follow-up reveals low toxicity of radiosurgery for vestibular schwannoma. Radiother Oncol 2007;82:83-9.
8Chopra R, Kondziolka D, Niranjan A, Lunsford LD, Flickinger JC. Long-term followup of acoustic schwannoma radiosurgery with marginal tumor doses of 12 to 13 Gy. Int J Radiat Oncol Biol Phys 2007;68:845-51.
9Massager N, Nissim O, Delbrouck C, Delpierre I, Devriendt D, Desmedt F, et al. Irradiation of cochleae structures during vestibular schwannoma radiosurgery and associated hearing outcome. J Neurosurg 2007;107:733-9.
10Timmer FC, Hanssens PE, van Haren AE, Mulder JJ, Cremers CW, Beynon AJ, et al. Gamma knife radiosurgery of vestibular schwannomas: Results of hearing preservation in relation to the cochleae radiation dose. Laryngoscope 2009;119:1076-8.
11Linskey ME, Flickinger JC, Lunsford LD. Cranial nerve length predicts the risk of delayed facial and trigeminal neuropathies after acoustic tumor stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1993;25:227-33.
12Flickinger JC, Kondziolka D, Lunsforda DL. Dose and diameter relationships for facial, trigeminal, and acoustic neuropathies following acoustic neuroma radiosurgery. Radiother Oncol 1996;41:215-9.
13Foote KD, Friedman WA, Buatti JM, Meeks SL, Bova FJ, Kubilis PS. Analysis of risk factors associated with radiosurgery of vestibular schwannoma. J Neurosurg 2001;95:440-9.
14Hayhurst C, Monsalves E, Bermstein M, Gentili F, Heydarian M, Tsao M, et al. Predicting nonauditory adverse radiation effects following radiosurgery for vestibular schwannoma: A volume and dosimetric analysis. Int J Radiat Oncol Biol Phys 2012;82:2041-6.
15Linskey 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;98:800-6.
16Peak SH, Chung HT, Jeong SS, Park CK, Kim CY, Kim JE, et al. Hearing preservation after Gamma Knife stereotactic radiosurgery of vestibular schwannoma. Cancer 2005;104:580-90.
17Perks JR, George ED, Hamri KE, Blackburn P, Plowman PN. Stereotactic radiosurgery XVI: Isodosimetric comparison of photon stereotactic radiosurgery techniques (Gamma knife Vs. micromultileaf collimator linear accelerator) for acoustic neuroma-and potential clinical importance. Int J Radiat Oncol Biol Phys 2003;57:1450-9.
18Lagerwaard FJ, Meijer OW, Hoorn EA, Verbakel WF, Slotman BJ, Senan S. Volumetric modulated arc radiotherapy for vestibular schwannomas. Int J Radiat Oncol Biol Phys 2009;74:610-5.
19Dutta D, Subramanian SB, Murli V, Sudahar H, Kurup PG, Potharaju M. Dosimetric comparison of Linac-based (BrainLAB) and robotic radiosurgery (CyberKnife) stereotactic system plans for acoustic schwannoma. J Neurooncol 2012;106:637-42.
20Gevaert T, Levivier M, Lacornerie T, Verellen D, Engels B, Reynaert N, et al. Dosimetric comparison of different treatment modalities for stereotactic radiosurgery of arteriovenous malformations and acoustic neuromas. Radiother Oncol 2012;106:192-7.
21Tanyi JA, Kato CM, Chen Y, Chen Z, Fuss M. Impact of the high-definition multileaf collimator on linear accelerator-based intracranial stereotactic radiosurgery. Br J Radiol 2011;84:629-38.
22Sharma DS, Dongre PM, Mhatre V, Heigrujam M. Physical and dosimetric characteristic of high-definition multileaf collimator (HDMLC) for SRS and IMRT. J Appl Clin Med Phys 2011;12:142-60.
23Milano MT, Usuki YK, Walter AK, Clark D, Schell MC. Stereotactic radiosurgery and hypofractionated stereotactic radiotherapy: Normal tissue dose constraints of the central nervous system. Cancer Treat Rev 2011;37:567-78.
24Paddick I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. J Neurosurg 2000;93 Suppl 3:S219-22.
25Paddick I, Lippitz B. A simple dose gradient measurement tool to complement the conformity index. J Neurosurg 2006;105:194-201.
26Sharma SD, Jalali R, Phurailatpam RD, Gupta T. Does intensity-modulated stereotactic radiosurgery achieve superior target conformity than stereotactic radiotherapy in different intracranial tumours? Clin Oncol (R Coll Radiol) 2009;21:408-16.