Gamma Knife Radiosurgery for Large Vestibular Schwannoma More Than 10 cm ³ : A Single-Center Indian Study

Abstract

Introduction  Gamma Knife radiosurgery (GKRS) is an effective treatment for benign vestibular schwannomas (VSs). The established cutoffs have recently been challenged, as recent literature expanded the horizon of GKRS to larger tumors. Even though microsurgery remains the primary option for large VS, GKRS can provide reasonable tumor control and is more likely to avoid cranial neuropathies associated with open surgery.

Methods  We analyzed patients with VS with volume exceeding 10 cm ³ who underwent GKRS at our center from January 2006 to December 2016. Clinicoradiological and radiosurgical data were collected from medical records for statistical analysis. Follow-up was performed every 6 months with a clinical assessment along with magnetic resonance imaging (MRI) of the brain and audiometric evaluation in patients with serviceable hearing.

Results  The study included 34 patients (18 males and 16 females) with an average age of 45.5 years. The mean tumor volume was 10.9 cm ³ (standard deviation [SD], ± 0.83), with a median tumor dose of 12 Gy (interquartile range, 11.5–12) and a mean follow-up of 34.7 months (SD, ± 23.8). Tumor response was graded as regression in 50%, stable in 44.1%, and increase or GKRS failure in 2 cases (5.8%). Treatment failure was noted in five cases (14.7%), requiring microsurgical excision and a ventriculoperitoneal shunt post-GKRS. The tumor control rate for the cohort is 85.3%, with a facial preservation rate of 96% (24/25) and hearing loss in all (5/5), while three patients developed new-onset hypoesthesia. We noted that gait ataxia and involvement of cranial nerve V or VII at initial presentation were associated with GKRS failure in univariate analysis.

Conclusion  Microsurgery should remain the first-choice treatment option for large VSs. GKRS is a viable alternative with good tumor control and improved or stabilized cranial neuropathies with a low complication rate.

Introduction

Vestibular schwannoma (VS) is a common benign tumor seen in the neurosurgical clinic. Being a benign pathology, preservation of facial innervation and preservation of serviceable hearing are prime objectives in any form of treatment strategy. The effectiveness of Gamma Knife radiosurgery (GKRS) in medium- and small-sized VS is well documented. ^{1 2 3 4 5}

GKRS has been usually preferred for small lesions less than 10 cm ³ , as the control rates depend on the tumor's size. Microsurgery is preferred when tumors are large and cause mass effect. In a subset of VS greater than 3 cm in size or tumor volume exceeding 10 cm ³ , which is a usual standard cutoff for GKRS, there are recent attempts worldwide for using radiosurgery as a primary or secondary treatment after initial microsurgery. For large VS, microsurgery remains the mainstay of treatment but, with advancements in radiosurgical techniques and dose planning software, good tumor control rate and prevention of cranial neuropathy are possible with GKRS.

Our institute is a premier tertiary care neurosurgical center in southern India and is well equipped for both microsurgery and radiosurgery for large VS. We analyzed our institutional data on the use of GKRS for primary or postmicrosurgical residual large VS and defined tumor control rates and predictors of the same along with factors impacting the preservation of cranial nerve functions.

Materials and Methods

At our institute, all larger VS are generally considered for microsurgery in view of their mass effect. The decision is individualized and varies on a patient-to-patient basis. The decision to consider microsurgery or GKRS was taken by the treating neurosurgeon considering various factors, namely, clinical condition, mass effect, tumor size, patient's wishes, etc. Most patients with tumor size less than 3 cm were considered for GKRS. Patients who have a high anesthesia-related risk of surgery such as cardiac comorbidities were also given the option of GKRS over microsurgery. In this study, all VS patients treated with GKRS, with tumor volume exceeding 10 cm ³ at the time of GKRS treatment, from January 2006 to December 2016 at the National Institute of Mental Health and Neurosciences (NIMHANS), Bengaluru, India, were included. The treatment options were discussed with the patient and caregivers.

Clinical Data

Complete clinicoradiological data including demography, tumor size, and pre-GKRS neurologic dysfunction were collected from medical records. Facial nerve dysfunction was graded based on the House–Brackmann (HB) grading scale, while hearing impairment was assessed using pure tone audiometry and classified by Gardner–Robertson (GR) grading for hearing. Audiometry was performed at each visit unless it became nonserviceable at any point during the study period. Radiosurgical data that included tumor treatment volume, maximum tumor dose, and the marginal dose were collected.

Radiosurgical Technique

GKRS was performed using Leksell 4C (Elekta) at our institute. Patients who were diagnosed with a case of primary or residual large VS based on pre-GKRS magnetic resonance imaging (MRI), without significant mass effect on brainstem or obstructive hydrocephalus and willing for radiosurgical treatment, underwent Leksell stereotactic frame fixation under local anesthesia. After frame fixation, the patients underwent MRI brain, and 1-mm thin-cut, continuous-slice, high-resolution postgadolinium T1W sequences were acquired. The images in the axial plane were used to measure the volume in each slice. Marking of tumor contour and brainstem was done by a senior neurosurgeon in the department. The number of shots, isodose line, maxiumum dose, and the marginal dose were computed using the latest software by an expert medical physicist and final plan was decided by consensus among neurosurgeon, radiation oncologist, and medical physicist. Posttreatment patients were observed overnight in a ward with a full oral diet and analgesics.

Follow-up

Clinical and radiological follow-up was performed at every 6-month interval. At each follow-up visit, all patients underwent clinical assessment for cranial nerve deficits and pure tone audiometry if the hearing was serviceable at a previous visit. MRI brain was performed to assess for changes in the size and to grade the response. Tumor size was measured on T1W contrast images in all the three planes. Treatment response was graded into three subgroups based on the variations in the size and/or volume of the tumor at follow-up imaging:

1. Tumor regression: >10% decrease in the volume.

2. Stable response: volume change <10%.

3. Tumor progression or failure of GKRS: >10% increase in the volume. ^{2 6 7 8 9 10}

Treatment failure is defined as any case that required microsurgical decompression of the tumor or ventriculoperitoneal shunt post-GKRS due to either increase in the tumor size or worsening of the clinical symptoms. Tumor control is achieved when there is no need for further microsurgery, radiosurgery, or any other form of treatment after primary GKRS.

Statistical Analysis

For analysis of factors associated with treatment failure, univariate Cox proportional hazards regression analysis was used for continuous variables, while Kaplan–Meier curves with logrank tests were used for categorical variables. For all tests, p -values < 0.05 were considered statistically significant. All analyses were performed using SPSS 21.0 statistical software (SPSS, Inc.).

Results

Clinicoradiosurgical Parameters

In the study cohort, 34 patients with a mean tumor size of 2.9 cm (standard deviation [SD], ± 0.28; range, 2.5–3.9 cm) and mean tumor volume of 10.9 mL (SD, ± 0.83; range, 10–13.25 mL) were considered for analysis. The mean follow-up duration was 34.7 months (SD, ± 23.8). Out of the 34 patients, 30 had more than 24-month follow-up period. Four patients had microsurgery, and were followed up till 6 to 20 months. Since the end point was reached further, these four patients were not followed for this study. The median dosage to the tumor was 12 Gy (interquartile range, 11.5–12), with a median isodose line of 50% (range, 48–53%). The mean dose to brainstem was 0.95 Gy (SD, ± 0.42; range, 0.2–2 Gy) ( Table 1 ).

Table 1. Clinical and radiosurgical data.

Characteristics	n  = 34 (range)
Age	45.5 (15–72)
Gender, male/female	18/16
Laterality, right/left	15/19
Primary/residue	26/8
Mean follow-up (mo)	34.7 (6–120)
Pre-GKRS
Hearing, preserved/lost	5/29
Seventh nerve, intact/involved	25/9
Fifth nerve, intact/involved	24/10
Gait, normal/ataxia	20/14
Hemiparesis, no/yes	32/2
Mean tumor volume (cm 3 )	10.9 (10.0–13.2)
< 11	20 (58.8%)
≥11	14 (41.2%)
Mean tumor size (cm)	2.9 (2.5–3.9)
< 3	19 (55.9%)
≥3	15 (44.1%)
Median marginal dose (Gy)	12 (11–12)

Abbreviation: GKRS, Gamma Knife radiosurgery.

Tumor Characteristics and Response

All 34 patients underwent an MRI brain with gadolinium contrast at 6-month follow-up and then at every year. Tumor regression at follow-up imaging was noted in 17 (50%) patients ( Figs. 1 and 2 ). Fifteen patients (44.11%) had a stable tumor, while another two patients (5.8%) had an increase in the tumor size, i.e., GKRS failure. The mean follow-up was 28.8 months (24–60 months) in stable tumors, while the mean follow-up was 44.1 months (24–120 months) in tumors with regression. In tumors that showed growth, the mean follow-up was 13 months (one at 6 months and one at 20 months). Both patients had significant tumor growth of more than 4 mm with the development of new-onset gait ataxia and recurrence of headache with raised intracranial pressure. One of these patients developed expansion of cystic component with mass effect and distortion of the brainstem ( Fig. 3 ). They underwent early microsurgery for the tumor following development of new symptoms. Two more patients underwent surgery at different intervals of follow-up, although they had a stable response. One of those patients had new-onset intractable trigeminal neuralgia and the other patient had worsening ataxia. Both patients were given a choice for close clinical observation with symptomatic management, but they opted for microsurgery. All four cases underwent microsurgery within 20 months after GKRS (range, 6–20) ( Table 2 ). Histopathological examination was suggestive of benign schwannoma. One patient (2.9%) required ventriculoperitoneal shunting post-GKRS due to the development of new-onset hydrocephalus with raised intracranial pressure, although the tumor size was stable.

Fig. 1.

Contrast-enhanced T1W MRI of a 46-year-old man with 10.1-cm ³ left VS who had tumor regression. ( a ) Planning MRI of GKRS showing heterogeneously enhancing lesion with a margin dose of 12 Gy. ( b ) 6-month follow-up MRI showing a decrease in the central enhancement. ( c ) 12-month follow-up MRI showing a decrease in the size with central necrosis. ( d–f ) 24-, 36-, and 60-month follow-up images showing progressive shrinking in the size of the lesion.

Fig. 2.

Contrast-enhanced T1W MRI of a 38-year-old man with 11.3-cm ³ left VS who had tumor regression. ( a ) Planning MRI of GKRS with heterogeneously enhancing lesion with a margin dose of 12 Gy. ( b ) 24-month follow-up MRI showing central necrosis. ( c, d ) 60- and 72-month follow-up showing a progressive decrease in size with decreased mass effect on the brainstem.

Fig. 3.

Contrast-enhanced T1W MRI of a 45-year-old woman with 10.5-cm ³ right VS who had treatment failure. ( a ) Planning MRI of GKRS showing homogeneously enhancing lesion with a margin dose of 12 Gy. ( b ) 12-month follow-up MRI showing central necrosis with cystic degeneration. ( c ) 18-month follow-up showing tumor enlargement and expansion of the cyst with an increase in the mass effect and raised intracranial pressure for which she underwent microsurgery 20 months later.

Table 2. Clinicoradiosurgical details of cases with treatment failure.

No.	Age/gender	Prior resection	Tumor volume (cm 3 )	Tumor size (cm)	Marginal dose (Gy)	Time of microsurgery/VP shunt after GKRS (mo)	Tumor response	Reason for surgery
1	45/F	Yes	10.5	3.9	12	20	Increase	Expansion of tumor with cyst enlargement. Development of severe headache with new-onset ataxia
2	48/F	No	10.1	2.8	12	6	Increase	Pseudoresponse with central necrosis. Increased mass effect and raised ICP
3	45/F	No	11.3	3.2	11	8	Stable	New-onset severe ataxia
4	28/F	No	10.8	3	11	9	Stable	New-onset intractable trigeminal neuralgia
5	57/F	No	10.5	2.7	12	14	Stable	Raised ICP. Underwent VP shunt

Abbreviations: F, female; GKRS, Gamma Knife radiosurgery; ICP, intracranial pressure; VP, ventriculoperitoneal.

Hence, treatment failure was noted in 5 cases (14.7%), while 29 cases (85.3%) showed tumor control and needed no further intervention. Four patients had undergone ventriculoperitoneal shunting for hydrocephalus at primary diagnosis of VS for raised intracranial pressure, before GKRS treatment.

Kaplan–Meier analysis of overall tumor control rate during the follow-up after GKRS is shown in Fig. 4 .

Fig. 4.

Kaplan–Meier analysis of overall tumor control rate during follow-up after GKRS.

Factors Influencing Outcome

The cohort was divided into two subgroups based on the pre-GKRS tumor volume as (1) less than 11 mL or (2) 11 mL or more, and compared for tumor control. The two subgroups contained 20 (58.8%) and 14 (41.2%) patients, respectively. The larger initial tumor volume statistically proved to be insignificant ( p  = 0.475) when compared with the other group, indirectly confirming good tumor control regardless of the initial tumor volume. To our surprise, we noticed that the two patients who had an increase in tumor size at follow-up and labeled as a failure of GKRS had tumor size less than 11 mL at presentation. Among the other two patients who underwent surgery despite stable response, only one of them had tumor volume exceeding 11 mL, while the other had tumor volume less than 11 mL. Kaplan–Meier analysis of tumor control rate for these two subgroups (<11 and ≥11 cm ³ ) during follow-up after GKRS is shown in Fig. 5

Fig. 5.

Kaplan–Meier analysis of tumor control rate for <11-cm ³ subgroup and ≥11-cm ³ subgroup during follow-up afterward.

Clinical Outcomes after GKRS

• Hearing preservation : In our series of 34 cases, 29 patients (85.3%) presented with hearing impairment as the commonest presenting complaint. Five patients who had preserved hearing pre-GKRS lost the serviceable hearing post-GKRS at 6-month follow-up on pure tone audiometry. The mean cochlear dose for the five patients was 4.94 Gy. We noticed complete hearing loss in 100% cases in our series.

• Seventh nerve : Twenty-five patients (73.5%) had intact facial animation at presentation. Three patients had HB grade 4 facial palsy, while five had HB grade 3 facial palsy, and another one had HB grade 2 facial palsy at presentation. At various points of follow-up of nine patients with facial weakness, two patients were noted to have worsening of HB grade (from grade 3 to 4), while two other patients had improvement in facial weakness (from grade 3 to grade 2 and grade 4 to grade 3). One patient developed new-onset facial weakness (grade 2), while 24 out of 25 patients (96%) had preserved facial animation at the end of the mean follow-up period.

• Fifth nerve : At presentation, 10 patients (29.4%) had trigeminal symptoms in the form of paresthesia or decreased facial sensation. Post-GKRS, six patients (60%) had improvement in trigeminal symptoms, while three patients developed new-onset hypoesthesia at follow-up. Overall, 27 patients (79.4%) had intact facial sensation at the end of mean follow-up period.

• Gait : Fourteen patients (41.2%) presented with gait ataxia. Nine patients (64.3%) improved at follow-up with a normal gait and only five remained the same with no worsening of ataxia compared with pretreatment based on a subjective opinion of the patient.

• Motor weakness : Two patients (5.9%) had a mild hemiparesis at presentation. One patient (50%) had improved motor power at 6-month follow-up, while the other patient remained the same even at the end of the mean follow-up of 3 years.

Risk Factors for Treatment Failure after GKRS

We analyzed various clinicoradiological and treatment parameters as possible risk factors for treatment failure of GKRS. Factors such as age, gender, side of the lesion, primary versus residual tumor, largest dimension, the volume of the tumor, and presenting symptoms were analyzed. In univariate analysis, we noted that the age ( p  = 0.861), side of the lesion (0.278), primary versus residual tumor ( p  = 0.788), largest dimension ( p  = 0.108), and the volume of the tumor ( p  = 0.499) were not associated with treatment failure. Interestingly, we noted that gender ( p  = 0.011), gait ataxia ( p  = 0.003), and involvement of cranial nerve (CN) V/VII along with CN VIII ( p  = 0.002) were associated with treatment failure. It was noted that all the five cases of treatment failure were noted in patients who initially had ataxia or involvement of CN V or VII in addition to CN VIII involvement, while no failure occurred in those with normal gait initially and no involvement of CN V or VII. Multivariate analysis could not be performed in view of small numbers.

Discussion

VS is a benign, slow-growing nerve sheath tumor. The tumor's natural history suggests that the growth rate is variable in sporadic VS and is around 2.9 ± 1.2 mm per year. ^{11 12} The predominant presenting complaint is hearing loss, seen in 90% of the cases. ¹³ The various options in the management of VS depend not only on the size of the lesion but also on the preoperative symptomatology, comorbidities, and patient's preference. Management options include watchful observation with serial imaging, ^{14 15} microsurgery, ¹⁶ radiosurgery, ¹⁷ and fractionated radiotherapy. ¹⁸

The application of GKRS in small- and medium-sized VS is well established, with tumor control rate as high as 90 to 100%, ^{1 17 19 20 21 22} hearing preservation rate ranging from 51 to 80%, ^{1 18 20 21 22 23} and facial nerve preservation rate of up to 98.2%. ^{24 25} In our series, in small- and medium-sized VS at our institute, we achieved tumor control rate of 96.1% and hearing preservation rate of 79.2% following GKRS at the end of the mean follow-up period of 30 months. ⁴

With the advent of radiosurgery and due to difficulties in achieving gross total resection in larger tumors, another option has emerged and is currently preferred over gross total resection, which is maximal safe resection followed by close clinicoradiological observation or radiosurgery for the residual tumor. ²⁶ Fractionated radiotherapy is another option for large VS with results comparable to stereotactic radiosurgery. ²⁷ Since the turn of the century, few case series, albeit of small sample size, have been published utilizing GKRS as a primary modality of treatment in large tumors and rarely after subtotal decompression with large residual tumors or following regrowth in the residual tumor. ^{9 28 29 30 31 32} It is mostly considered for patients with significant comorbidities, which make them medically unfit for a microsurgical procedure. Presently, the literature lacks a randomized trial comparing microsurgery and GKRS for large VS. However, the available literature shows favorable results for GKRS in large tumors and it may be more beneficial than microsurgery on certain postintervention clinical outcomes. ^{6 9 31 33}

Tumor Control

Radiosurgery aims to control the growth potential of the concerned tumor. GKRS has achieved 100% tumor control in some of the case series for small and medium VS, and is considered an optimal treatment option. In large VS, the tumor control rate varies from 54.2 ⁶ to 100%. ³⁴ ( Table 3 ) Our present series achieved a radiological control rate of 94.1%. Tumor control, in which patients do not undergo further microsurgery, ventriculoperitoneal shunt, or radiosurgery, in our series is 85.3%. Although most of the series mentioned in Table 3 did not evaluate the factors predicting GKRS failure, certain studies statistically analyzed pre-GKRS clinicoradiological parameters for the same. Larger tumor volume (>15 mL), size, its mass effect in the form of brainstem compression, and effacement of fourth ventricle are certain known causes for GKRS failure. ⁶ One of the major limitations in utilizing them as a yardstick for formulating treatment protocols is the lack of uniformity in the definition of large VS. We noted that gait ataxia and involvement of CN V or VII were associated with treatment failure in univariate analysis; however, these observations need to be confirmed in a larger cohort.

Table 3. Review of the literature for large vestibular schwannoma treated with GKRS.

Author and year	No. of cases	Definition of large VS	Median tumor volume	Tumor control (%)	Hearing preservation (%)	Facial preservation (%)	Mean follow-up	Factors affecting failure
Litvack et al, 2003 34	9	Size > 3 cm	NR	100	33.3 (1/3)	NR	31.7 mo	NR
Rowe et al, 2003 46	35	Size > 3 cm	NR	85.7	NR	NR	35 mo	Dose to tumor margin < 14.6 Gy
Inoue et al, 2004 47	5	Size > 4 cm	21 cm 3	100	100 (1/1)	100 (5/5)	10.6 y	NR
Inoue, 2005 48	20	Size > 3 cm	15.2 cm 3	93.4	80 (4/5)	100 (20/20)	>6 y	NR
Hasegawa et al, 2005 6	24	Vol > 15 cm 3	NR	54.2	NR	NR	93 mo a	Tumor volume > 15 cm 3 , brainstem compression, effacement of the fourth ventricle
Wang et al, 2009 49	50	Size > 3 cm	14.8 cm 3 b	94	100 (14/14)	96 (2/50)	86 mo	Size ≤ 4 cm, age > 60, no brainstem compression, nonenhancing tumor or small cyst on contrast MRI
Chung et al, 2010 28	21	Size > 3 cm	17.3 cm 3	90.5	100 (8/8)	0 (0/12)	5.5 y	T2 signal ratio between tumor and brainstem, duration of tumor swelling
Yang et al, 2011 31	65	Size > 3 cm	9 cm 3	87	82 (18/22)	98 (49/50)	36 mo a	Prior resection, Koos grade 3
van de Langenberg et al, 2011 26	33	Vol > 6 cm 3	8.8 cm 3	88	58 (7/12)	100 (33/33)	30 mo a	None
Sun and Liu, 2012 10	28	Vol > 10 cm 3	14.3 cm 3 b	78.6	NR	NR	74 mo	Vol ≥ 10 cm 3 , symptomatic brainstem compression, homogenous tumor enhancement
Milligan et al, 2012 9	22	Size > 2.5 cm	9.4 cm 3	82	30 (3/10)	85.7 (18/21)	66 mo a	None
Williams et al, 2013 30	24	Size > 3 cm	9.5 cm 3	75	NR	66.7 (16/24)	82.5 mo a	None
Zeiler et al, 2013 32	28	Size > 3 cm	9.6 cm 3	92	100 (7/7)	100 (12/12)	34.5 mo	NR
Iorio-Morin et al, 2016 50	68	Vol > 4 cm 3	7.4 cm 3	94.1	49 (20/41)	100 (62/62)	47 mo	None
Bailo et al, 2016 33	59	Size > 25 mm	5.6 cm 3	98.3	31.3 (5/16)	94.8 (55/58)	79.4 mo	NR
Huang et al 2018 29	35	Size > 3 cm, vol > 10 cm 3	14.8 cm 3	85.7	33.3 (1/3)	100 (28/28)	48 mo a	Tumor volume > 15 cm 3
Mezey et al, 2020 40	103	Vol > 10 cm 3	13.6 cm 3	78.6	95.5 (63/66)	100 (24/24)	6.2 y	Tumor volume > 20 cm 3
Present series	34	Vol > 10 cm 3	10.7 cm 3	85.3	0 (0/5)	96 (24/25)	34.7 mo	Female gender, gait ataxia, involvement of CN V or VII

Abbreviations: MRI, magnetic resonance imaging; NR, not recorded; vol, volume.

^aMedian follow-up.

^bMean.

Functional Outcome

Microsurgical resection for large VS can jeopardize facial nerve function significantly even in experts' hands, and preserving facial function (HB grade 1 or 2) postoperatively is possible in only 27 to 58% of cases. ^{26 35 36} There are reports of facial nerve preservation in tumors greater than 2.6 cm following surgery, ranging from 56 to 91% and 86 to 100% in gross total resection and planned subtotal resection, respectively. ^{37 38 39} Similarly, facial animation is preserved in 67 to 100% of cases following GKRS in large tumors. ^{29 30 32 40} We achieved a facial preservation rate of 96% (24/25) in our series, which is comparable to outcomes in small- or medium-sized VS where GKRS is an established modality. This preservation rate is also far superior to primary microsurgery and is comparable with subtotal resection followed by GKRS. ^{26 41}

The factors affecting hearing outcomes in small- and medium-sized VS are age, mean cochlear radiation dose, and pre-GKRS GR grade, which includes both pure tone threshold and speech discrimination score. ⁴ In large VS, by virtue of the tumor's size, hearing is lost in the majority even before patients present for treatment and hence preservation of hearing becomes impossible. In larger VSs, functional hearing preservation following microsurgery is 13.1%. ^{42 43 44} The rate of hearing preservation in large VS post-GKRS varies from 13 to 82%. ^{9 26 29 32} However, on a careful review of the individual case series, it is obvious that the absolute number of patients who preserve their hearing post-GKRS is very less compared with the number of patients considered for evaluation. This again points to the loss of hearing at presentation due to the tumor's mass effect before any treatment. In our series, only five cases had a serviceable hearing at presentation, but we could not preserve hearing in any of these cases. This may be due to a higher mean cochlear dose of 4.9 Gy delivered. All five patients lost their hearing in their first follow-up at 6 months.

The fifth nerve involvement in large VS is common and affects the sensory component of the nerve more frequently than the motor segment. Radiosurgery proved to be curative in 60% (6/10) of cases with atypical trigeminal neuralgic pain. New-onset facial hypoesthesia was noted in only three cases (12.5%) following GKRS. The majority of the series concentrates on hearing preservation and facial preservation while neglecting other cranial neuropathies that may benefit or worsen with . Improvement in facial sensory symptoms is seen in 36 to 66% of cases, while the symptoms remain stable in most other cases; rarely, any worsening is noted. ^{29 30 32}

Gait ataxia is another neurological dysfunction that gets benefited with GKRS. In our series, 64.3% improved at 6-month follow-up and no deterioration was noted. Chung et al ²⁸ and Huang et al ²⁹ noted 23% (5/23) and 11% (4/35) worsening of gait, respectively. Ventriculoperitoneal shunting was done pre-GKRS in four cases (11.7%) for raised intracranial pressure. One patient (2.9%) underwent VP shunt despite no increase in tumor swelling post-GKRS. An additional form of treatment as microsurgical resection was required in four cases. The treatment failure rate varies from 0 to 17.1% ( Table 4 ). Radiologically, only two tumors had a significant increase in the tumor size, with one of them having the mass effect on brainstem due to enlargement of the cystic component and the other patient having pseudoprogression with central necrosis ( Fig. 6 ). Transient tumor expansion is a possible cause for morbidity 6 to 18 months following GKRS in large tumors. ⁴⁵

Table 4. Treatment failure requiring additional treatment.

Author and year	No. of patients	Surgery/Shunt post-GKRS (%)
Litvack et al, 2003 34	9	0 (0)
Rowe et al, 2003 46	35	6 (17.1)
Inoue et al, 2004 47	5	0 (0)
Inoue, 2005 48	20	1 (5)
Wang et al, 2009 49	50	1 (2)
Chung et al, 2010 28	21	2 (9.5)
Yang et al, 2011 31	65	7 (10.8)
van de Langenberg et al, 2011 26	33	5 (15)
Milligan et al, 2012 9	22	2 (9)
Williams et al, 2013 30	24	3 (12.5)
Zeiler et al, 2013 32	28	1 (4)
Huang et al, 2018 29	35	5 (14.3)
Mezey et al, 2020 40	103	14 (13.6)
Present series	34	5 (14.7)

Fig. 6.

Contrast-enhanced T1W MRI of a 48-year-old woman with 10.1-cm ³ right VS who had treatment failure. ( a ) Planning MRI of GKRS showing homogeneously enhancing lesion with a margin dose of 12 Gy. ( b ) 6-month follow-up MRI showing pseudoresponse with an increase in the size of the lesion with central nonenhancing necrotic degeneration. Increase in the mass effect on brainstem with development of raised intracranial pressure is observed for which she underwent microsurgery 6 months after GKRS.

This is a retrospective review of a small number of cases ( n  = 34) with a mean follow-up of only 34.7 months; it has an inherent selection bias. Although clinical outcomes are comparable to any microsurgery series or even better on some parameters such as facial preservation, we cannot extrapolate these findings to all cases with large VSs. Also, we noted that gait ataxia and involvement of CN V or VII were associated with GKRS failure in univariate analysis; however, the observations need to be interpreted with caution, in view of small sample size and relatively shorter follow-up. Hence, a larger cohort study may help validate these findings, as randomizing cases to either microsurgery or GKRS may pose ethical challenges.

Conclusion

We recommend GKRS for large VSs with tumor volume exceeding 10 cm ³ in carefully selected cases. We could achieve good tumor control and improvement or stabilization of cranial neuropathies with an extremely low complication rate.