For the growing cohort of elderly patients with considerable co-morbidities, treatment concepts balancing tumor control against risk are needed in cases of failed “wait and scan” strategy. In this particular context, stereotactic radiosurgery may result in high tumor control with low toxicity rates.
The number of reported SRS series for elderly patients (> 65 years) is astonishingly low. Although a literature query of the NCBI database using the headings "radiosurgery elderly vestibular schwannoma" OR "radiosurgery elderly meningioma" FOR the last decade revealed more than 620 results, only the study by Hasegawa et al. [
10] reported single fraction radiosurgery of meningiomas in an elderly cohort (> 65 years). Two other series report SRS with different fractionation concepts [
9,
11]. For vestibular schwannoma, no study with special regard to elderly patients is available. Therefore, this present study is the first to report a detailed analysis of the treatment of SRS for this subpopulation.
Tumor control
Our study confirms the high rate of tumor control (> 90%) and effectiveness of SRS, as reported in many other series [
7,
8,
19,
20]. However, the influence of age on tumor control in benign brain tumors is currently not fully understood. Although several studies demonstrated no impact of age on tumor control, others suggest age as a risk factor for tumor recurrence [
5,
21]. Although a clear cut-off value for age in general not seem to exist, Starke et al. report about 65 years as a higher risk of local control failure [
21]. Since the anti-proliferative effect of radiosurgery for benign tumors depends not only on cytotoxic but also on delayed vascular effects [
22], one might speculate that these processes are less effective in older patients. If so, one might assume that patients with vascular diseases have a higher risk for recurrence. However, our results do not support this hypothesis. The important conclusion of our finding is that patients over 70 years of age do not have a higher risk for local control failure.
Patients with recurrence after previous treatment (surgery in most of the cases) had a higher risk for local control failure after SRS. This is in accordance with a study by Hasegawa et al. [
19], suggesting higher aggressiveness in recurrent tumors. Similar results for meningiomas [
23,
24], as reported e.g. by Kim et al. [
23] in more than 700 patients, explained the higher local recurrence rate after previous microsurgery by a surgery-related breakdown of the stroma capsule, rendering radiosurgery less efficient. Furthermore, after previous treatment scarring might arise, which hinders defining the exact target volume of SRS, especially in situations of dural insertion of the tumor. Therefore, further studies in a larger collective may help to elucidate the patterns of failure. However, a particular reason for treatment failure of radiosurgery after previous surgery remains to be identified.
Overall, the present local tumor control rates (93% at 5 years) are in the upper range compared to those observed in other series of elderly patients [
10]. In particular, these results were obtained in a distinctive collective with 50% treatments primarily due to proven tumor growth prior to SRS. During follow-up, approximately 1/3 of all tumors decreased measurably in diameter during the observation period. Thus, SRS provided control of tumor growth for the majority of patients in our series, but did not provide rapid tumor shrinkage. If the latter is necessary for symptom alleviation, surgical removal is mandatory [
25,
26].
On the other hand, microsurgery bears relevant risks for elderly patients, particularly in the presence of severe and/or multiple co-morbidities, even in a situation of a space-occupying, symptomatic lesion. Therefore, if the primary treatment goal is the mere control of tumor growth, surgery should be weighed against SRS and radiotherapy. In a large series reported by Sughrue et al. [
27], the 5-year recurrence rate after resection of WHO I meningioma (n = 373) for patients receiving a Simpson Grade I, II, III, or IV resection was 95, 85, 88, and 81%, respectively. The authors concluded that a Simpson Grade I resection is beneficial if it is easily obtained with a low risk. But the primary goal of meningioma surgery should be to remove as much tumor as possible, e.g. to reduce pressure. In cases where there is an increased risk of neurological or vascular injury, or CSF leak, the authors found it hard to justify performing more aggressive attempts of resection only to improve the rate of recurrence by a few percent, even if the recurrence rates match the rates of SRS.
Toxicity
When analyzing toxicity, we included any symptom occurring after treatment, without regarding any causal relationship, and classified it according to the CTCAE classification. Some of these symptoms may have been caused by the tumor itself or could have developed anyway. Thus, our results might overestimate the risk of toxicity to a certain amount. In addition, symptoms that occurred immediately after treatment were reversible in total.
In contrast to these low toxicity rates observed after SRS, surgical treatments are often associated with higher complication rates. In a current meta-analysis of Poon et al. [
28], a general complication rate of 20% was observed for surgical resection of meningioma in elderly people. Furthermore, it has also been shown that after resection of benign brain tumors, older patients have higher hospital mortality rates and longer hospital stays than younger patients [
29], and one-year mortality rates in these elderly patients may reach 15% [
30,
31].
The assessment of treatment-related imaging changes such as edema and radiation necrosis can help to objectify the toxicity of SRS. In a current review by Milano et al. [
32] the frequency of radiation-induced edema was reported to amount to between 2% up to 50%. We observed a favorable low crude rate of edema formation nearly similar to current observation studies [
33]. The causes for edema formation are discussed widely and clear relationships to potential risk factors such as tumor volume, radiation dose, previous treatment with radiation, location of the tumor, presence of edema before treatment or extent of tumor-brain were identified. Unger et al. [
34] considered a large tumor volume and single-fraction irradiation as main risk factors for edema formation after treatment. In the case of tumor volume, our results confirm the findings in literature [
33,
34].
In conclusion, our findings suggest that elderly patients with larger tumors may have an increased risk for edema development after SRS but most of these imaging changes remain symptomless. According to Chin et al. [
35], radionecrosis (RN) is the most important complication of SRS and it depends on tumor volume, 10-Gy volume [
36] and re-irradiation of the same tumor, and shows an onset time of about four months. Compared to the reported RN frequencies ranging from 2 to 25% [
35,
37], the incidence of RN in our study (about 0.5%) is extremely low. One reason might be the moderate size of tumor volumes treated in our cohort.
Finally, an important finding of our study is that patients' co-morbidities have no influence on the effectiveness and especially the toxicity of the treatment, whereas for surgery it is always an issue. A recent review of meningioma surgery in elderly patients [
38] often found that postoperative mortality is most commonly associated with co-morbidities. Eksi et al. [
39] also found in their meta-analysis that co-morbidities are a strong predictor of postsurgical neurologic complications. Especially in the group of elderly patients, these aspects should be considered, while taking into account the results of tumor control after SRS presented here. Thus, it is worth considering SRS as primary treatment of meningioma and schwannoma in the group of elderly patients even if they have severe co-morbidities.
Limitations of our study
Due to the retrospective nature, follow-up times are limited in our study. The reasons for this was lacking compliance, long travelling distances and changes in residence location preventing patients from attending follow-up at the referring hospital. Furthermore, the study is based on a heterogeneous cohort with potential bias induced by large divergences in premorbid factors, but perhaps this collective best reflects daily clinical practice.