Background
Meningiomas account for approximately one third of all primary brain tumors and tumors of the central nervous system [
1]. Most are benign, slow growing lesions originating from the arachnoidal cap cells, with the skull base being the most frequent localization [
2]. Besides benign histology, a smaller number of meningiomas can be of atypical or anaplastic histology, characterized by aggressive growth patterns and a high rate of recurrence [
3]. Many analyses focus on meningiomas of the skull base because of its intricate anatomy and the close proximity to vascular structures, cranial nerves and the brainstem; consequently treatment in those cases is challenging and treatment options are controversially discussed.
Surgical resection has for long been the treatment of choice but in the last decades advances in radiotherapy (RT) such as radiosurgery, fractionated stereotactic radiotherapy (FSRT), or intensity-modulated radiotherapy (IMRT) have made radiotherapy an important treatment alternative [
4,
5]. Due to the complex anatomy of the skull base, tumor adherence to bony structures and the close proximity to sensitive organs at risk (OAR), total resection is often not possible as it may cause to substantial morbidity. Consequently, as neurosurgical resection is subtotal in many cases, it cannot achieve high long-term local control and overall survival rates [
6]. Additional radiotherapy can improve chances for long-term tumor control [
7].
Meningiomas are often discovered incidentally or present with only mild symptoms and indolent growth patterns. In those cases, there is no urgent need for therapeutic intervention. Active surveillance can be a reasonable approach, focusing on precise high-resolution contrast-enhanced magnetic resonance imaging (MRI) and clinical examinations in regular intervals (e.g. every six to 12 months, provided an asymptomatic/stable clinical situation). However, if the tumor enlarges significantly during the course of neuroimaging or clinical symptoms develop or worsen, treatment becomes mandatory [
6].
Particle beams, such as protons and heavier ion beams like carbon ions offer high precision when it comes to dose application to the tumor volume so that OAR can be very effectively spared [
8,
9].
With its unique physical characteristics, including the inverted dose profile, high local dose deposition within the Bragg Peak and a steep falloff outside the treatment volume, particle radiation therapy leads to a greater dose conformity than photon RT [
10]. Compared to protons, carbon ions additionally offer the advantage of higher biological efficiency with a relative biological effectiveness (RBE) ranging between 3 and 5, potentially leading to higher local control rates [
11].
To date, particle therapy at Heidelberg Ion Therapy Center (HIT) has been integrated into the clinical environment at our institution for close to a decade and is constantly being validated for the treatment of skull base meningiomas. In this present study, we analyse our results for skull base meningiomas in 110 patients treated with particle therapy – protons as well as carbon ions – with special focus on treatment outcome and toxicity.
Discussion
The present manuscript evaluates the efficacy and toxicity profile of particle therapy for the treatment of 110 consecutive patients over a a period of 5 years, treated at a single institution. Histology was predominantly benign (WHO grade I) and mainly proton therapy was used, although a combination of photon IMRT and a carbon ion boost was used for a total of six patients with higher-grade histology. None of the treated meningiomas had previously been irradiated. Excellent overall local control at the cost of very light toxicity was achieved with 100% PFS after three and 96.6% PFS after 5 years and histology appearing to significantly influence PFS.
The treatment of skull base meningiomas is a complex clinical situation that requires careful interdisciplinary evaluation. Due to the intricate anatomy of the skull base and the distinct subset of symptoms and toxicities caused by tumors there located, it has been discussed that skull base meningiomas should be regarded as a separate entity regarding outcome and treatment-associated toxicity [
4].
Over the years, radiation therapy – and particularly high-precision techniques such as FSRT or IMRT – has evolved to become a central pillar in the multimodal treatment of meningiomas. Several groups have shown high efficacy with minimal toxicity [
4,
6,
15]: One of the largest collectives of skull base meningiomas treated with photon IMRT or FSRT and with a median follow-up of 107 months has been described at our institution, showing a local control rate of 95% at 5 years and 88% at 10 years [
4]. Histology (WHO grade I vs. grades II and III) proved to be an important prognostic factor, significantly impacting PFS. These data have been confirmed by several similar studies performed at other institutions: Kaul et al. have described PFS to be 93.8% after 5 years for 318 patients with benign meningeomas treated with FSRT [
16]. In a separate series focusing exclusively on skull base meningiomas PFS was similar for low-risk histology and 41.8% after 5 years for high-risk histologies [
17]. Minniti et al. found a PFS rate of 96% at 3 years and 93% at 5 years in a series of 52 patients with large skull base meningiomas treated with FSRT [
18]. Kessel et al. comprehensively reviewed recent literature on the subject and published another large series of 260 patients treated with FSRT or IMRT and including 16% high-risk histologies. They found a PFS rate after 5 years of 87.1% and 54.9% for low-risk and high-risk histologies respectively. Futhermore, patient-reported outcome showed very mild toxicity with no more than 3.0% of patients experiencing worsened or new symptoms ≥3 during RT and the first 6 months thereafter [
15].Our results have shown that proton therapy can achieve similarly excellent local control, though continuous long-term follow-up is warranted. Reported data on toxicity and symptom response to treatment is very similar to the results achieved in the current analysis with only mild acute toxicity, the majority of patients showing either stable or improved symptoms during long-term follow-up.
One of the main rationales for the use of particle therapy lies in its higher dose conformity, potentially allowing for better OAR sparing and the reduction of side effects [
19‐
21]. The energy-deposition of accelerated photons occurs continuously over a comparably wide range of penetration depths through tissue [
22]. The improved dose distribution of particle therapy is achieved by exploiting the physical characteristics of particle irradiation where the maximum dose deposition occurs within the sharply defined Bragg peak [
9]. By varying the particle energy, the position of the Bragg peak can be altered. Particle therapy has been shown to be superior to photon-based techiques in terms of sparing OAR and in terms of target dose homogeneity/conformity with carbon ions showing slightly superior dose distributions compared to protons [
23,
24]. Arvold et al. observed a significant dose reduction to neurocognitive, visual and auditory organs achieved by proton irradiation as compared to photon RT. Furthermore, they found protons to reduce the risk of developing a radiologically-induced or associated secondary malignancy by half [
1]. Other publications showed a sizeable improvement of pre-exstisting clinical symptoms in up to 47% of patients treated with proton radiotherapy for meningiomas, to which our results compare favorably [
10,
25]. We could observe a clear tendency towards improvement that was most prominent in patients suffering from visual impairment, mostly diplopia. 34.5% of all patients showed stabilization or improvement regarding eye-related symptoms and 41.8% regarding headaches, corresponding to 77.8% of the patients reporting pre-therapeutical eye-related symptoms and 94.3% of the patients reporting headaches respectively.
In recent years the body of literature on treating meningiomas with proton therapy has steadily grown and to date there are several publications describing adequate-sized collectives with a median follow-up of 32 to 84 months: Vlachogiannis et al. recently published a retrospective analysis on 170 patients with grade I meningiomas, 155 of which were located at the skull base, who received hypofractionated proton therapy over a period of 13 years. Median follow-up was 84 months and authors reported PFS rates of 93% and 85% at five and 10 years respectively. Main differences in comparison to the current work were the use of passive scattering and the hypofractionated dose regime of 3–8 fractions at 5 or 6 Gy(RBE) per fraction, translating approximately to an EQD2 (equivalent dose in 2-Gy-fractions) of 43 Gy.
Halasz et al. were the first to describe a radiosurgical approach for proton therapy in meningiomas [
10]. They analysed a group of 50 patients treated with proton stereotactic radiosurgery at a dose of 13 Gy prescribed to the 90% isodose, achieving a three-year acturial tumor control rate of 94% and toxicity rates similar to those described above. The regarded collective included only small tumor volumes and low-risk histologies. The data suggest that a hypofractionated or even radiosurgical approach, as has been extensively evaluated for photon therapy, might be a feasible and well-tolerated approach for proton therapy as well and achieve satisfactory results [
26].
A recent retrospective study by Murray et al. described the outcome of 96 meningioma patients treated with pencil beam scanning proton therapy at Paul Scherrer Institute in Switzerland over a 10 year period [
27]. 63.5% were low-risk and and 36.5% high risk meningiomas. The authors reported an estimated 5-year local control (5y-LC) rate of 95.7% for the low-risk group and 68% for the high-risk group, showing consistency with the previously discussed literature and the results of our current work. Five-year grade ≥ 3 toxicity-free survival was 89.1%. The authors reported on identifying several prognostic factors for local failure beside histology (
p < 0,001), One such factor was the timing of particle therapy (initial vs. for recurrence or progressive disease) with patients treated initially showing favorable outcome; furthermore tumors of the skull base showed favorable outcome vs. non-skull base (
p = 0,14), as did female patients vs. males (
p = 0,32). However, none of those factors was tested in multivariate analysis, thus their predictive value should be interpreted with care.
DiBiase and colleagues revealed the size of the GTV to be a significant prognostic factor, since in their described collective of 162 patients treated with Gamma Knife SRS, patients with smaller tumor volumes had longer survival rates with a 5-year overall survival of 100% compared to 59.7% for larger lesions [
28].
A small prospective randomized series by Sanford et al. has tested the effect of dose escalation using a combination of photon and proton therapy for the treatment of low-risk meningiomas with a median follow-up of 17,1 years [
29]. While overall local control of 98% at 10 years and 90% at 15 years was excellent, no significant benefit could be shown for the use of 63 Gy(RBE) over 55,8 Gy(RBE). However, dose escalation might be beneficial for the treatment of high-risk meningioma patients who show less favorable outcomes with established dose regimens. 5y-LC rate for high-risk meningiomas was 75% in our analysis and though patient number was small, results are comparable to figures reported in recent literature of 50–81% for IMRT or proton therapy, depending on WHO grade [
27,
30,
31].
Adeberg et al. could within the high-risk collective identify the WHO grade as prognostic factor for PFS with higher grade yielding inferior PFS (
p = 0,017) [
30]. Notably, the results of McDonald et al. support the rationale of dose escalation for high-risk meningiomas, achieving 5y-LC rates of 87.5% for a radiation dose > 60 Gy(RBE) compared to 50% for ≤60 Gy(RBE) of proton RT (
p = 0,038) [
31]. Regarding dose escalation in a highly radiosensitive region such as the skull-base, the use of heavier ions such as carbon ions with their potentially superior dose distribution and biological advantages attributed to the increased relative biological effiectiveness (RBE) could prove beneficial and may lead to higher local tumor control rates [
9,
32]. In a small prospective phase I/II trial conducted at our institution in 2010 on the administration of a carbon ion boost after photon radiotherapy for 10 patients with high-risk meningiomas, we achieved promising results with 5- and 7-year local control rates of 86% and 72% [
33]. Median cumulative dose in this analysis was 68 Gy(RBE) and the series included two previously irradiated tumors. Building upon those results, we have initiated the MARCIE trial, a prospective phase II trial evaluating PFS, OS and toxicity for the postoperative bimodal irradiation of atypical meningiomas Simpson grade 4 or 5 [
3]. The trial is currently recruiting and the dose regimen of 50 Gy photon RT combined with 6 × 3 Gy carbon ion boost that we applied to the high-risk patients in this analysis is analogous to the concept employed in the MARCIE trial. Though patient number was small for high-risk meningiomas in our current analysis, results are in agreement with previously published data for this dose regimen [
33].
Certainly a potential benefit of particle therapy over photon radiation techniques has to be verified clinically and prospective trials are warranted. Several treatment planning studies showed superiority for protons, particularly for larger target volumes: For example, Phillips et al. on reviewing different radiosurgical methods found that particle RT results in supervior dose distributions than photon-based linear accelarator (linac) methods for target volumes > 25 ccm, though for smaller volumes results are comparable while linac methods might offer higher flexibility [
34,
35]. Smith et al. supported these findings, comparing linac-based photon RT to Gamma Knife SRS and proton RT and calculating normal tissue complication probability indices (NTCP) based on the dose conformity of the resulting treatment plans and using a logistic model based on the tolerance data by Rubin et al. and Emami et al. [
36]. While photon SRS techniques proved superior for small spherical targets, protons had the lowest NTCP for large (> 15 ccm) and peripheral target volumes (13,5 for protons vs. 17,0–33,5 for linac) [
37].
To date, our analysis represents the largest group of patients with skull base meningiomas treated with particle therapy, including both protons and carbon ions, in a single institution. Limitations of this analysis include the relatively short follow-up period, its retrospective character and the small number of both high-risk histologies and patients treated with carbon ions, limiting the possibility to perform meaningful subgroup analyses. The median follow-up in this series at 46,8 months – though substantial – is still relatively short when compared to other studies available, especially in the field of precision photon RT. In the light of the benign nature of low-grade meningiomas and predicted long-term tumor-control and overall survival continuous follow-up is warranted. Regarding the different physical and biological characteristics of particle therapy, potential long-term effects are of special interest. Currently particle therapy patients at our institution are included in a close-knit and rigorous follow-up regimen and potential late side effects are documented in a prospective database with dedicated institutional funding for long-term evaluation [
12].
To conclusively demonstrate the clinical benefits of particle therapy there is currently a lack of a prospective comparison to advanced photons. Prospective clinical trials have since been initiated at several institutions to further establish the role of particle therapy for the treatment of certain subgroups of intracranial menigiomas.