The Long-term Outcome of CyberKnife-Based Stereotactic Radiotherapy for Intra/ Extracranial Non-Vestibular Schwannomas: A Single-Center Experience

Background The long-term outcomes of CyberKnife-based hypofractionated stereotactic radiotherapy (SRT) for intra/ extracranial non-vestibular schwannomas (nVS) need to be accumulated. Method Patients who received SRT by CyberKnife for nVS from 2010 to 2019 were retrospectively reviewed. Results A total of 45 patients with nVS were identied. The mean age was 53 (± 18) years old, and 23 patients (51%) were female. Twenty-nine patients (64%) had previous procedures. As for the tumor extension, 22 (49%) nVS were classied as primary intracranial, ve (11%) were classied as intra/ extracranial (dumbbell shape), and 18 (40%) were classied as primary extracranial. The median prescribed dose, covering 95% of the planning target volume, was 21 (IQR 21 – 25) Gy, and the median target volume was 7 (IQR 3.6-13.1) cm 3 . The local control rate of nVS for patients without neurobromatosis type 2 (NF2) was 100%. Old age (OR 0.92, p-value 0.03) and previous surgery (OR 0.02, p-value 0.02) were signicant risk factors for no symptomatic improvement. The progression-free survival was 74 (±33) months clinically and 69 (IQR 36 – 94) months radiologically. During follow-up, two cases (4%) with NF2 resulted in treatment failure, 13 cases (41%) resulted in transient tumor expansion (TTE), 10 (22%) suffered from transient adverse radiation effect (ARE), and two (4%) resulted in permanent ARE. Hypofractionated SRT for head, neck, and spine nVS was an effective treatment regardless of tumor extension relative to the cranium. Although the risk of permanent ARE was low, some patients experienced transient clinical worsening due to TTE. rate, y: year, mos: month, cSRT: conventional SRT, fSRT: fractionated SRT, SRT,


Introduction
Despite the progressive advancement of surgical armamentarium, schwannomas in the head, neck, and spine may be di cult to treat because of proximity to nerves and arteries. [1,2] They are World Health Organization grade I benign tumors and commonly arise outside the central nervous system. Most intracranial schwannomas arise from the vestibular nerve, and schwannomas from other cranial nerves are not frequently encountered. Intracranial and spinal schwannomas are treated by either or combinations of microsurgical resection, stereotactic radiosurgery (SRS), and stereotactic radiotherapy (SRT). [3] To date, compared to the data on SRS outcome on vestibular schwannomas, we have less data on non-vestibular schwannomas (nVS). And since most of the radiation therapy is performed by gamma knife [4,5], we have even fewer data of SRT performed by CyberKnife, especially on nVS.
CyberKnife can treat lesions outside as well as inside the cranium, which is a key difference from gamma knife. Some schwannomas around the skull base foramina or canals, such as jugular foramen and hypoglossal schwannomas, present with lesions occupying both inside and outside the cranium (dumbbell shape) or primarily extracranial regions. These tumors are theoretically more suitable for CyberKnife than gamma knife.
Our facility is a referral center for CyberKnife, and we have treated quite a few nVS occupying extracranial spaces with or without intracranial space. Considering the rarity of SRT outcome on nVS, we thought our data might add further evidence to the current literature.

Patient selection
Consecutive patients who received SRT for nVS in the head, neck, and spine between 2010 and 2019 were included. Of note, since the authors' hospital is a referral center for CyberKnife therapy, most of the included patients were referred from different hospitals and heterogeneous in baseline characteristics. Various data were recorded, including age, sex, symptoms at the time of SRT, history of previous operation or radiation, the status of neuro bromatosis type 2 (NF2), the form of treatment (upfront, adjuvant, upfront for relapsed cases, and adjuvant for relapsed cases), parameters for SRT, radiological response, changes in symptoms, adverse radiation effect (ARE), and length of follow-up, among others.
The diagnosis of schwannoma was made radiologically by MRI for non-operated patients and further con rmed pathologically for operated patients. There were no strict criteria for surgical resection, as the patients were referred from various hospitals. However, surgery tends to have been performed for younger patients. Postoperative SRT was recommended for those with residual tumors. The tumor locations were classi ed into the following three types relative to the cranium: primary intracranial, intra/ extracranial (dumbbell shape), and primary extracranial. This is the result of modifying the classi cation described by Martin et al[6] SRT at our facility Target volumes were delineated on thin-slice CT with or without gadolinium-enhanced MRI. A 6D-skull tracking algorithm was used. In planning treatment, we selected the prescription dose and fractionation according to the size and site of the lesions based on the senior authors' (K.S. and R.N.) experience. Most of the tumors were treated 21Gy over 3 fractions or 25Gy over 5 fractions.

Patient evaluation and tumor response
Based on the MRI, we classi ed the response to SRT into two categories of under control and treatment failure, like a previous study [1] using perpendicular diameters as reference. The senior authors (K.S. and R.N.) classi ed tumor response into the two categories, with con rmation by the primary author (S.H.).
Adverse radiation effect (ARE) was evaluated by Common Terminology Criteria for Adverse Events (CTCAE), if listed. [7] Transient tumor expansion (TTE), which is known to be a common phenomenon after SRS to vestibular schwannoma [8], and concurrent clinical changes were recorded and counted as ARE if clinically symptomatic.

Statistical analysis
From the acquired data, univariate and multivariate analyses were performed to identify risk factors for symptom improvement. A multicollinearity test was done to identify factors to exclude in multivariate analysis. Statistical analysis was done using SPSS version 25.0 (IBM Inc., Armonk, NY, USA). Values were listed as means for parametric data, and median for nonparametric data. The Shapiro-Wilk test was used to differentiate parametric from nonparametric data. Binary logistic regression was used to evaluate risk factors for symptom improvement. A p-value of 0.05 or less was considered statistically signi cant.
Our institutional review board did not require informed consent for study participation because this study relied on information obtained as part of routine clinical practice.

Baseline characteristics
A total of 45 schwannomas in 45 patients were identi ed (Table 1). Females constituted 51% of the patients. Common presenting signs and symptoms at the time of SRT were diplopia, facial dysesthesia, trigeminal neuralgia, dysphagia, dysarthria, hoarseness, and radiculopathy depending on the tumor location and previous surgical complications. Of the ve (11%) patients who had no symptoms, three patients received SRT for enlarging tumor size after wait and scan, one patient received the treatment for relapse, and the other one received for residual tumor after surgery. Most common tumor were trigeminal schwannomas (27%). A previous operation had been performed in 29 cases (64%). No cases had a previous history of radiation treatment. Two cases (4%) were neuro bromatosis type 2. The most common form of SRT was upfront treatment for newly-diagnosed tumors (18 cases, 40%), followed by upfront treatment for relapsed tumors (16 cases, 36%) and adjuvant treatment (9 cases, 20%). The median time between the previous surgery and SRT was 2 months and that for the relapsed tumor was about 35 months (Table 1). The parameters of SRT are summarized in Table 2. Most therapies were done in multiple sessions, and the median target tumor volume was 7 (3.6-13.1) cm 3 . Relative to the cranium, 51% of tumors had extracranial tumor portions (Table 3). We illustrate two cases as examples ( Fig.).   Table).

Discussion
We showed that SRT effectively controlled nVS in the head, neck, and spine, regardless of the spatial relation to the cranium. As for non-NF2 patients, the local control rate was 100% in our cohort (Table 4). Although the severity and rate of permanent ARE were mild (CTCAE grade 1) and low (4%), clinical worsening due to TTE should be kept in mind, as 25% resulted in transient symptomatic deterioration, which was controlled with or without medications. The comparison of the treatment result of our study population to the past studies is summarized in Table 5. [9-14, 3, 15, 16, 2, 17, 5, 4, 1] Except for spinal schwannomas, we still have few long-term data on nVS treated by SRT.  (Table 3). We showed that CyberKnife could treat the extracranial tumor segment as effectively as the intracranial segment (Table 4). Two past studies on jugular foramen schwannomas reported that dumbbell-type tumors and large tumor volume as risk factors for tumor control failure. [16,5] The cohorts of these studies were treated by gamma knife SRS. In our cohort, all tumors were treated by CyberKnife SRT, and all tumors regardless of portions of intra and extracranial space were controlled. This difference is attributed to the difference in way of radiation delivery between gamma knife and CyberKnife. Gamma knife is more suitable for intracranial targets whereas CyberKnife is suitable for extracranial targets as well as intracranial targets. Since both the dumbbell-shaped schwannomas and primary extracranial schwannomas were effectively treated (Tables 3 and 4), these tumors may be more suited for CyberKnife.
Having said that, we still need to accumulate evidence as we have few studies on treatment outcomes of these tumors performed by CyberKnife ( Table 5).
As for clinical symptoms, our data showed older age and previous surgical history were statistically signi cant risk factors against symptom improvement (Supplementary Table). Based on this, surgical resection, if chosen, should be pursued under the top priority of not damaging the surrounding nerves, especially in older patients. However, since our center is a referral center and whether to proceed with upfront SRT or adjuvant SRT was not decided based on the strict criteria, our data was at risk for selection bias.

Limitations
The limitations were the retrospective nature of the study and that this study was carried out in a single institution. Since our institution was a referral center for SRT, incomplete clinical and radiological information was available. These may have affected the exact rate of ARE and TTE and our data was at risk of selection bias.

Conclusions
Hypofractionated SRT is a valid modality for head/ neck/ spine non-vestibular schwannomas over the long term and CyberKnife-based SRT effectively treats schwannomas of dumbbell-shaped as well as located in the extracranial space. However, some patients experience symptoms from transient tumor expansion.

Declarations
Funding: Not applicable.
Con icts of interest/Competing interests: Not applicable.
Availability of data and material: Data transparency was con rmed.
Code availability: Not applicable

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. SupplementaryTable.docx