Introduction
Gamma Knife radiosurgery (GKRS) has evolved into a practical alternative treatment to open microsurgical resection of vestibular schwannoma (VS) [
1‐
30]. GKRS as a treatment modality for VS typically does not require inpatient hospitalization, however acute and chronic complications can occur [
31‐
33]. In particular, radiation toxicity of neuro-anatomic structures adjacent to the tumor may develop and manifest as impaired function of the facial nerve, hearing loss, or loss of equilibrium and balance. [
14,
16,
17,
23,
27,
30,
34‐
41]. Hydrocephalus, cerebral edema, and other cranial neuropathies have also been documented after GKRS, and in some reported cases required shunting as a treatment for hydrocephalus [
4,
23,
37,
42‐
49].
Despite the available data on facial nerve outcome in VS patients treated with GKRS, there is no consensus as to what reported clinical parameters relate to facial nerve function. Most reported studies to date have been small to modest in size, frequently from a single institution, and lacking the statistical power and freedom from potential practitioner bias to draw concrete conclusions. Our review of the literature revealed widely varying results with reported facial nerve preservation between 55 and 100% after GKRS for VS (Table
1). Due to these factors and the multitude of methods to assess facial nerve preservation in the reported literature, facial nerve preservation after GKRS has not yet been fully characterized.
Table 1
Data summary from papers listed by Pub Med ID and institution
17379451 | University of Pittsburgh | 216 | 215 | 56.5 | 13.0 | 1.300 | 98.30 | 68.4 | 100.0 |
16741754 | Ludwig Maximilians University | 123 | 121 | 59 | 13.0 | 1.600 | 96.70 | 98.4 | 100.0 |
16094154 | Komaki City Hospital | 317 | 291 | 54 | 13.2 | 5.600 | 92.00 | 93.6 | 96.4 |
15854240 | Haukeland University Hospital, Norway | 103 | 102 | 59.7 | 12.2 | | 89.20 | 70.8 | 94.8 |
15662791 | Inst of Neural Org, Japan | 18 | 9 | – | – | 15.200 | 93.33 | 72.0 | 100.0 |
15662787 | Taipei Veterans Gen Hosp and Natl Yang Ming University | 195 | 135 | 51 | 13.0 | 4.100 | 95.00 | 36.0 | 100.0 |
15354007 | Medical College of Wisconsin | 29 | 25 | – | 13.5 | | 96.55 | – | 100.0 |
15337560 | University of Pittsburgh | 313 | 313 | 56 | 13.0 | 1.100 | 98.60 | 24.0 | 100.0 |
14617712 | Royal Hallamshire Hospital, UK | 232 | 179 | 56 | 14.6 | 3.350 | 92.00 | 35.0 | 99.1 |
14609174 | Gunma Univ Sch of Med, Japan | 1 | 1 | 63 | 12.0 | 0.520 | 0.00 | 27.0 | 0.0 |
14571654 | Hospital Academique Erasme, Belgium | 48 | 42 | 54.8 | 12.3 | 1.440 | 97.92 | 12.0 | 97.9 |
14519213 | University of Pittsburgh | 157 | 124 | 60 | 16.7 | – | 96.90 | 109.2 | 95.0 |
12520350 | Addenbrooke’s Hospital, England | 5 | 5 | 29 | – | – | 0.00 | – | 80.0 |
12459364 | Baylor memorial Hermann Hospital | 72 | 58 | 61.6 | 14.5 | | 91.00 | 48.0 | 97.4 |
12379008 | Karl-Franzens University, Graz, Austria | 60 | 52 | 58 | 13.0 | 3.400 | 96.00 | 76.0 | 85.0 |
11483338 | Thomas Jefferson Univ Hosp, PA | 69 | 57 | 61 | 12.0 | 2.920 | 98.00 | 119.0 | 98.0 |
11143268 | University of Tokyo | 1 | 1 | 25 | 14.0 | 0.180 | 100.00 | 60.0 | 100.0 |
10821551 | Northwestern Hospital | 9 | 9 | 39 | 19.6 | | 74.00 | – | 55.6 |
10030254 | Mayo Clinic and Mayo Foundation [reduced protocol] | 40 | 33 | 65 | 16.0 | 3.700 | 97.44 | 27.6 | 92.0 |
10030254 | Mayo Clinic and Mayo Foundation [standard protocol] | 42 | 35 | 63 | – | 3.000 | 97.44 | 27.6 | 62.0 |
9833820 | Mayo Clinic/University of Pittsburgh | 76 | 35 | 58 | 15.0 | 2.800 | 94.00 | 43.0 | 83.0 |
9392535 | University of Tokyo | 46 | 46 | 54 | 16.8 | – | 96.00 | 39.0 | 80.0 |
8588625 | House Ear Clinic and House Ear Institute | 1 | 1 | 39 | – | – | 0.00 | 24.0 | 100.0 |
7826279 | University of Pittsburgh | 31 | 19 | 55 | – | 0.600 | 90.00 | 26.0 | 95.0 |
Totals and Avg | | 2,204 | 1,908 | 55.3 | 13.1 | 3.2 | 82.5 | 54.1 | 96.2 |
Several potential factors affecting facial nerve preservation after GKRS have been suggested, including the dose of radiation delivered, tumor volume, and patient age. In this study, we performed an extensive review of the English Language literature to objectively analyze and methodically evaluate facial nerve outcomes of patients with VS treated with GKRS. The primary aims were to provide an objective summary of the published literature on facial nerve preservation and to evaluate specific prognostic factors that may influence facial nerve preservation after GKRS for VS.
Methodology
Article selection
Articles were identified via Boolean PubMed searches using key words “Gamma knife,” “radiosurgery,” “acoustic neuroma,” “facial nerve,” “vestibular schwannoma,” and “facial nerve preservation,” alone and in combination. This query identified 23 papers describing over 2,204 patients from which all quantifiable and assessable data regarding patients treated with radiosurgery were analyzed. Articles published up to and including the year 2007 were included in this analysis. Inclusion criteria for articles were: (1) Facial nerve preservation rates were reported specifically for VS before and after GKRS, (2) Facial nerve outcome was reported using the House–Brackmann classification (HBC) for facial nerve function [
5,
50‐
54], (3) Tumor size was documented, and (4) GKRS was the only radiation modality used to treat the tumor. The data were then aggregated and analyzed based on radiosurgery dose delivered, size of the tumor, and patient age.
Data from individual and aggregated cases were extracted from each paper. Cases with pre-operative facial dysfunction (HBC 3 or higher) were excluded. All recent cases of open microsurgery and radiotherapy other than GKRS were also excluded. “Facial nerve preservation” was defined as having a grade I or II HBC at the last reported follow-up visit. Overall average for facial preservation, patient age, and radiation dose were weighted accordingly to their sample size, so that larger and smaller series had an appropriate impact on the overall data. Data were analyzed as a whole and stratified into three groups. (1) Radiosurgery marginal dose ≤13 versus >13 Gy, (2) Tumor size ≤1.5 versus >1.5 cm3, and (3) Age ≤60 versus >60 years old.
Statistical analysis
The raw data were tabulated using Microsoft Excel (Microsoft Corp., Seattle, WA). All results were analyzed using a Fisher’s exact test or a t-test when appropriate for statistical evaluation of the data. For these statistical investigations, tests for significance were two sided, with a (two tailed) P-value threshold of 0.05 considered statistically significant. Unless otherwise stated, all continuous values presented were mean ± standard deviation or standard error of measurement when appropriate.
Discussion
Facial nerve preservation continues to be a primary concern of patients undergoing Gamma Knife radiosurgery for vestibular schwannomas. Despite the currently available data there have been few efforts to combine this research into accurate estimates of facial nerve preservation with GKRS for VS. In this study we performed a comprehensive analysis of facial nerve functional preservation in a large aggregated population of patients who underwent GKRS for vestibular schwannomas.
Our methodical analysis revealed that patients treated with a marginal dose of less than 13 Gy were more likely to preserve facial nerve function after GKRS treatment than studies that delivered higher doses of radiation. Higher doses of radiation are associated with higher rates of cranial nerve toxicity [
67,
78‐
81]. One possible reason for this is the significant amount of fibrosis within and around the vestibular schwannoma, involving the adjacent cochlear and facial nerves. This finding has been noted in surgical salvage after failed irradiation [
82,
83]. Several recent studies have demonstrated that low dose radiosurgery has a favorable efficacy/toxicity ratio as compared to higher doses [
4,
23,
40,
44,
48,
57,
61,
84]. In our analysis patients treated with lower dose Gamma Knife radiosurgery (<13 Gy) had superior facial nerve preservation rates [<13 Gy = 98.5% vs. >13 Gy = 94.7%,
P < 0.0001 (Fig.
1)] with good tumor control rates of 96.7% at a reported average length of follow up duration of 54.1 months (Median 43.0 months).
In our objective analysis, patients with an average tumor volume of 1.5 cm
3 or less had a better facial nerve preservation rate compared to studies with tumors of larger volumes [<1.5 cm
3 99.5% vs. >1.5 cm
3 95.5%,
P < 0.0001 (Fig.
2)]. Smaller tumors had improved facial preservation rates and lower average radiation doses for smaller tumors (12.9 ± 0.8 Gy vs. 13.7 ± 1.3 Gy,
P < 0.0001). This data suggests that both smaller tumor size and lower radiosurgery dose are important risk factors for facial nerve preservation with Gamma knife radiosurgery treatment. Although it appears that radiation dose is an important associated factor with facial nerve preservation, our data does not permit the discrimination between size or radiation dose as the more significant parameter for facial nerve preservation as both smaller tumors and lower radiation doses both had improved outcomes. Our data does not clarify this ambiguity about whether size or radiation dose has a more significant impact on facial nerve preservation.
Older patients commonly have medically related comorbidities which can preclude them from open brain surgery. Our analysis indicates that older patients with age >60 years had inferior facial nerve preservation rates than younger patients [<60 years = 96.8% vs. >60 years = 89.4%,
P < 0.0001 (Fig.
3)]. Age may be an important prognostic factor for facial nerve preservation despite tumor size or radiation dose. Older patients had similar tumor sizes as younger patients (2.31 vs. 2.54 cm
3). Advanced age does appear to be a negative prognostic factor in facial nerve preservation outcomes in patients treated with GKRS for VS. Furthermore older patients (>57 years old), treated with high levels of radiation (>13 Gy) had significantly worse facial nerve outcomes than younger patient (<57 years old) treated with similarly high radiation doses of greater than 13 Gy (
P < 0.0010). Our data suggests that older age may be significantly associated with worse facial nerve preservation independent of radiation dose because older patients did worse with high radiation doses than their younger counterparts who also received high radiation doses (>13 Gy).
The various methods of data presentation reported in the papers for our systematic analysis precluded us from further investigation to stratify other statistically significant data points. Unfortunately actuarial time dependant data was not possible in our retrospective, systematic analysis as this is an inherent limitation in the methodology of our study. Similarly, multi-variable analysis and a logistic regression analysis are also problematic across multiple studies which adhere to differing formats of data presentation.
Prospective studies could further elucidate the actuarial nature of facial nerve preservation over time after GKRS and may also provide further insight into the exact relationship between the prognostic variables we investigated here and facial nerve preservation. Our systematic analysis is the first reported attempt to comprehensively evaluate the overall impact of GKRS for VS on facial nerve function as described in the published literature.
There are some inherent limitations with systematic reviews and analysis [
85]. One obvious limitation is that any aggregation of data is only as good as its composite studies. The quality of the data reported in the literature, the effect of failure to detect, or unwillingness to report complications, and other such omissions would inevitably change and skew the result reported in our aggregated analysis. Furthermore, small sample size reports that met our inclusion criteria were also included in our analysis. Although their contribution is small, we mitigated the effect of case reports and small samples by analyzing an aggregated database and by weighting the appropriate contribution of each paper by the number of patients with facial nerve intact before GKRS accordingly. Hence in our analysis, smaller sample sizes and case reports had a proportionate effect on our overall aggregated facial nerve preservation data. However, the large nature of our systematic review minimizes the biases and dilutes the inherent error of any individual study in our comprehensive report and also has the advantage of expansive results from multiple international centers.
In conclusion, we report the results from a large aggregated analysis of facial nerve outcomes in patients with vestibular schwannoma treated specifically with Gamma Knife radiosurgery. Utilizing this systematic data set from the available published literature, minimizes the effect of bias and dilutes the inherent error from individual institutions, increases the statistical power of our analysis, and aggregates expansive results to determine an accurate and overall facial nerve preservation for patients treated with Gamma Knife radiosurgery for vestibular schwannomas. This systematic analysis suggests that radiation dose is an important and critical prognostic factor for facial nerve outcomes in VS patients treated with GKRS. Our data also confirms that patients treated with 13 Gy or less of radiation, with tumors less than 1.5 cm3 in size, and younger patients have improved facial nerve outcomes.