Single‐session image‐guided robotic radiosurgery and quality of life for glomus jugulare tumors

Limited data are available on the efficacy and impact on the quality of life (Qol) of single‐session image‐guided robotic radiosurgery (RRS) for glomus jugulare tumors (GJTs). This study investigates the role of RRS in the management of GJTs and reviews the RRS literature.


| INTRODUCTION
Paragangliomas, also known as chemodectomas, are highly vascularized, rare, slow-growing tumors deriving from extra-adrenal parasympathetic or sympathetic paraganglia. 1 Depending on their location, size, and hormone activity, they can cause a broad spectrum of symptoms ranging from pulsatile tinnitus, headache, hearing loss, vertigo, and lower cranial nerve palsies in glomus jugulare tumors (GJTs) to tachycardia and labile blood pressure in catecholaminesecreting paragangliomas. [2][3][4] Even though histologically considered as benign tumors, paragangliomas can locally infiltrate surrounding tissue such as bones or vessels. Moreover, tumors can metastasize, leading to a significantly decreased overall survival rate. 5 However, only approximately 3% of the found tumors are considered to be malignant, bearing the risk of distant metastases. 5,6 Until the development of radiotherapy, the primary treatment option for paragangliomas was the surgical resection of the tumor. Due to its high vascularization and its localization near the skull base and large vessels, surgical procedures may cause severe complications including high-grade cranial nerve dysfunctions, significant blood loss, and strokes. 7,8 Despite technical advancements, the microsurgical tumor extirpation yields substantial morbidity for patients, also when preoperative embolization procedures are performed to reduce the intraoperative bleeding risk. 8,9 This work will be presented at the 2020 meeting of the Radiosurgery Society in Washington, DC, USA. Today, possible treatment options range from surgery alone, tumor embolization with and without surgical resection to various radiation techniques including fractionated radiotherapy, proton therapy as well as radiosurgery. [10][11][12][13][14][15] There is still no consensus on the optimal management in regard to local tumor control, treatment-related morbidity, and quality of life (Qol). 16 However, since its introduction, primary and adjuvant radiotherapy played a pivotal role in the management of glomus jugulare patients, showing similar or even better treatment outcomes than surgical approaches. 8,17 Several studies have investigated the role of radiosurgery in the treatment of GJTs either using single-session or fractionated treatments. 18 Especially the use of Gamma Knife (GK)-and linear accelerator-based radiosurgery were analyzed by different researchers over the past decades and showed good results. 17,[19][20][21][22] Due to the inherent limitations of the stereotactic frame-based radiosurgery, it may not be performed for lower located GJTs. Image-guided robotic radiosurgery (RRS) has no such spatial limitation. Moreover, only sparse data are available on the role of RRS for the management of GJTs.
Furthermore, almost no data are available on posttreatment Qol changes in this tumor entity.
The aim of this retrospective, monocentric study is to report and evaluate the efficacy and safety of RRS and its impact on the posttreatment Qol at our institution. Finally, we compare our results with the existing RRS literature.

| Patients
Fifty-three patients with GJTs were consecutively treated between July 2005 and November 2018 and enrolled in this retrospective, monocentric study. Medical history, previous treatments, clinical symptoms as well as treatment and follow-up data were collected in a dedicated database for radiosurgery. Local tumor response, clinical symptoms, and adverse events were evaluated clinically and by MRI assessment every 6 months for the first year after treatment and then every 12 to 24 months for the following years. This study received the approval of the institutional review board. Informed and written consent was obtained from all patients prior to data assessment, evaluation, and analysis.

| Treatment procedure and outcome
Prior to RRS, patients underwent a planning CT scan and MRI of the head, both with 1-mm slice thickness and contrast agent. The CT was subsequently overlaid with secondary MRI, including gadolinium-enhanced T1 and T2 sequences, as well as vessel-focused time of flight series when available. Inverse treatment planning was performed with various versions of the MultiPlan and Precision software (MultiPlan, Precision, Accuray Inc, Sunnyvale, California). All treatments were delivered in a single-session outpatient setting using a CyberKnife RRS system (Accuray Inc). The CyberKnife utilizes a lightweight 6 MeV linear accelerator mounted on a sixaxis robotic arm, a stereoscopic kV imaging system, and a robotic treatment table. In all patients, the 6D-Skull tracking software was used to track the position of the patient's head, whereas custom-fitted thermoplastic face masks were employed for light, noninvasive fixation. 23 Radiographic assessment of the treatment outcome was defined as follows. Complete remission (CR) is the disappearance of the whole tumor, partial response (PR) is at least a 30% decrease in tumor volume, minor response is a decrease of tumor volume up to 30%, stable disease is the unchanged tumor volume, and progressive disease (PD) is an increase of the overall tumor volume of at least 20% or tumor growth of at least 5 mm. Local control (LC) was defined as no radiographic evidence of PD.

| Quality of life
Health-related Qol was assessed using the 12-item health survey questionnaire (SF12v2, 1992, 2000 Health Assessment Lab, Medical Outcomes Trust and QualityMetric Inc). The SF12v2 assesses various health concepts (HC) of Qol including physical functioning (PF), role physical (RP), bodily pain (BP), general health, vitality (VT), social functioning (SF), role emotional, and mental health (MH).
Patients were asked to complete the questionnaire before treatment delivery and during follow-up visits. All survey results were evaluated using the SF Health Outcomes Scoring Software (Qualimetric Inc, Lincoln, Rhode Island). Only patients who answered all questions were included in the analysis. The received answers were transformed into a standardized continuous scale ranging from 0 to 100, with 50 being the mean. Higher SF12v2 scores indicate a better Qol and better overall function. The data were tested for normality by graphic appearance, skewness, kurtosis, and the Shapiro-Wilk test. Baseline data and values at first and last follow-up were compared using paired two-tailed t tests or Wilcoxon signedrank tests in STATA 15.1 (StataCorp, College Station, Texas). Statistical significance was set at a P value ≤.05.

| Literature review
We conducted a PubMed-based literature research by using various keyword combinations including CyberKnife, glomus jugulare tumor, paraganglioma, chemodectoma, robotic radiosurgery, radiosurgery, radiotherapy, and stereotactic to search the National Library of Medicine database. Only articles that reported the use of single-session or multisession (fractionated) RRS for GJTs or paragangliomas were included in this study, even if they included the treatment of other tumor entities or the use of other radiation techniques. Articles published after February 1, 2019, were not considered.

| Treatment results
All 53 patients obtained a clinical as well as radiographic follow-up which ranged from 4 to 160.8 months with a median of 38 months. The 5-year actuarial LC was 100%, the crude LC was 98% at last follow-up. Fifty-two tumors either shrunk or remained unchanged in size (Table 2). After 70.8 months, one patient developed a local recurrence and was subsequently treated with proton radiotherapy. As of today, there is no new evidence of a local recurrence in this patient. Another patient developed lymph node metastases 4 months after treatment delivery while his primary treatment site remained controlled. During the last follow-up, 18 patients had entirely recovered and 17 had experienced symptom improvement while all three patients without pretreatment deficits remained symptom-free (Table 2). In 13 patients, the symptoms remained unchanged, whereas one patient experienced a transient worsening before improvement.
Only one patient reported a symptom worsening, consisting of a newly occurred pulsatile tinnitus during the first follow-up at 6 months.

| Complications
After treatment delivery, only one acute complication occurred. One patient developed an edema around the tumor leading to mild to moderate pain irradiating to the ear, neck, and mandibula. The complication was treated with dexamethasone in an outpatient setting and did resolve shortly after (<72 hours). No radiation necrosis, seizures, acute bleedings, or other complications have been observed in the other 52 patients. Throughout follow-up, no radiation-induced malignancies occurred.

| Quality of life
Before treatment delivery, 47 of 53 patients completed the SF12v2 questionnaire and answered all questions (return rate 88.6%). During first follow-up, which took place around 6 months after treatment (median: 6.1 months), 35 patients had filled out the questionnaire again (66.0% return rate). Analysis of the data showed no significant decline in any of the eight Qol HC (Table 3; Figure 1). The category BP significantly improved compared with the baseline (P = .04), whereas MH showed a trend toward posttreatment improvement without reaching significance (P = .08). During the last patient follow-up, after a median of 38 months, 28

| DISCUSSION
Herein, we report our long-term experience with RRS and its safety, efficacy as well as the impact on Qol of treated patients. To date, this study comprises the largest and most homogeneous series of patients treated with RRS and it is the first to provide standardized Qol measurements before and after treatment delivery. In agreement with previous RRS and radiosurgical studies, our LC rate was close to 100% (Table 4). 8,17 We are the first to report a local recurrence of a tumor as other studies consistently reached LC rates of 100% with sometimes even longer median follow-up periods. Given the sparse data available, any prognostic factors are unclear so far. The recurrence after 5.9 years in one of our patients highlights the fact that GJTs can relapse even after a long follow-up, as discussed in various other studies. 9,17,18,34,35 In addition, GJTs are usually slowly growing tumors with most of them having a tumor doubling time of more than 10 years and an average growth rate of less than 1 mm per year. 36 Subsequently, long follow-up periods are needed to detect relapses and to confirm the high LC rates initially reported in the current literature and herein. Still, there are no consensus guidelines on how and when to treat GJTs. Some colleagues have argued that only fast-growing, large, catecholamine-secreting or symptomatic tumors should be treated. 8,17,37 In young patients with hereditary tumors, chances of malignant GJTs are higher and, thus, should be treated in a timely manner as well. 38,39 Lieberson and colleagues have proposed a treatment guideline including four primary options-surgery ± adjuvant radiation (conventional radiotherapy or radiosurgery), radiation alone or watchful waiting. 17 Nevertheless, the decision making is not linked to the actual tumor size as the LC rates do not seem to be correlated with tumor size. This is also in agreement with the experience at our institution-larger tumors can be treated with good success. However, it is unclear to what extent fractionation plays a role in longterm tumor control as fractionated radiotherapy studies showed high tumor control rates as well. 8,17 Based on our experience, RRS can be used in most of the cases and surgery may be reserved for patients with rapid neurological worsening and peripheral tumors. The primary use of RRS is still feasible regardless of the histological confirmation as the imaging findings and caused symptoms are usually sufficient to diagnose GJTs. 40,41 For catecholamine-secreting tumors, it remains unclear which treatment is safe and most beneficial. We treated two hormone-secreting tumors, which caused syncope, tachycardia, and labile blood pressures. Despite short-term LC, cardiovascular symptoms remained unchanged and patients were lost to follow-up, emphasizing the need for interdisciplinary treatments including surgery. With regard to the proposed guidelines, more studies, ideally of prospective nature, are needed for verification.
In addition, future treatment guidelines should implement patient-centered outcomes including Qol analyses. Only sparse Qol data after radiotherapy and radiosurgery are available for paragangliomas and, more specifically, GJTs. It has been described that head and neck paragangliomas have a considerable impact on the Qol of affected patients. 42,43 Galland-Girodet and colleagues reported that Qol significantly differs in a group of 30 head and neck paraganglioma patients according to the treatment modalities (surgery/embolization ± radiotherapy). 44 Patients undergoing radiotherapy alone had better values for speech, hearing, trismus, and total score. 44 Results were obtained retrospectively by mailing the EORTC-QLQ-C30 and EORTC-QLQ-H&N 35 questionnaire at least 12 months after treatment delivery. 44 Recently, more data have been published by Patel and colleagues, investigating changes in Qol stratified by primary or secondary GK-based radiosurgery after a median of 97 months posttreatment delivery. 45 Swallowing function was better in a group of 26 glomus jugulare patients undergoing primary radiosurgery. Overall and disease-specific measures did not significantly differ. 45 To the best of our knowledge, we report the first standardized Qol data for GJT patients with pretreatment and follow-up comparisons using the SF12v2 questionnaire. Results show no decrease in any of the Qol concepts either during first or last follow-up with positive trends for RP, MH, and VT, whereas BP significantly improved in our series. These results confirm one of the essential features of RRS besides its high local tumor control, namely limiting negative impacts on the Qol from treatment delivery, especially in comparison with surgical treatment options. 8,17 This is underlined by the observed symptom control and deficit improvements seen in twothird of our patients with pretreatment deficits. Various other studies using radiosurgery or radiotherapy reported similar experiences, with rates of stable or improved pretreatment deficits ranging from 42% to 85%. 17,21,32,46

| Limitations
As inherent to retrospective clinical studies, reporting as well as selection biases cannot be ruled out.
In our case, we included every patient undergoing RRS since the availability of the technique at our institution in 2005. However, as the prevalence and incidence of GJTs is very low, we were only able to include 53 patients. Still, this is the most extensive series using RRS to date ( Table 4) and one of the largest radiosurgical studies in the literature. This indicates the problem that even with our series, our patients might not be representative of the whole glomus jugulare population. Furthermore, our series include 33 primarily treated patients and histological confirmation of the tumor was not conducted.
Even though the radiological appearance including tumor location, contrast uptake, and clinical symptoms were characteristic for GJTs, we cannot completely exclude other possible diagnoses. Still, modern imaging techniques like CT, MRI, and angiography show high rates of sensitivity and specificity in the diagnosis of GJTs. 40,41,47 In addition, our follow-up duration is still not long enough to reliably detect recurrences after 5 years or more. As previously described, relapses can occur even 40 years after treatment delivery. 35 Finally, not all patients have fully completed the SF12v2 questionnaires before treatment and during follow-up. Hence, there could be selection and reporting biases in our Qol analyses.

| CONCLUSION
RRS is a safe, reliable, and efficient tool for the primary and secondary treatment of GJTs even for larger tumors. RRS achieves high rates of LC and leads to improved or stable pretreatment deficits in most patients. This is also reflected by the Qol analysis showing no significant decrease in any of the Qol concepts during first or last follow-up with positive trends for RP, MH, and VT while BP significantly improved. RRS may be considered as a primary treatment option for most GJTs.