Hypofractionated high-energy proton-beam irradiation is an alternative treatment for WHO grade I meningiomas

Meningiomas are benign tumours, accounting for about 14–37% of primary intracranial tumours [8, 17, 29]. Total surgical resection of the meningioma and its associated dural base is generally the treatment of choice, since it results in long-term disease-free survival in the majority of patients [4, 39, 43]. However, not all skull base meningiomas can be totally resected without an unacceptable risk of postoperative neurological deficits, particularly cranial nerve palsies [6, 16, 27]. Further treatment is needed for patients with postoperative residual tumour masses, recurrent tumours, multiple tumours, or primarily inextirpable tumours [24, 35].

There is compiling evidence indicating that radiotherapy has a valuable role in patients suffering from both benign and atypical meningiomas. This is in spite of a lack of well-performed randomised studies [19]. Both shrinkage of existing tumour burden and prevention of regrowth are possible gains [26, 31]. In retrospective studies conventionally fractionated radiotherapy has been reported to have an effect. Common schedules utilised have generally been 1.8– to 2.0-Gy fractions 5 days/week to a total dose of 50–60 Gy [2, 5, 14, 40]. Stereotactic radiosurgery using higher fraction doses have also been reported to play a role in the treatment of meningiomas. Most authors have utilised single fractions in the range of 15–25 Gy given either with the Leksell Gamma Knife® or with Linear Accelerators (LINAC) equipped with stereotactic instrumentation [7, 9, 10, 21, 34, 38]. In general, reported times of follow-up have been rather short in these studies and in series of meningiomas treated with proton radiation [15, 45, 47].

The dose distribution of a high-energy proton beam facilitates irradiation with a high dose to the target and comparatively low dose to the surrounding normal tissue due to the proton beam’s innate ability of conforming to the target by losing a large amount of energy at the end of the beam path (the Bragg peak) [11, 32]. A typical dose plan with modified proton beams given from two portals results in full dose coverage of a benign meningioma. No secondary photons or electrons are delivered downstream. The integral dose is lower compared to most photon techniques. A complicating factor with proton therapy seems to be the varying value of relative biological effectiveness (RBE). The total dose, number of fractions and type of tissue may affect the value, as pointed out by Bleddyn Jones [19]. RBE is not known for brain tissue, the optical apparatus and brainstem. However a factor of 1.1 has been accepted for treatment protocols at most proton therapy centres [32].

In the present study, we evaluated proton beam irradiation for the treatment of benign meningiomas in relation to clinical benefit and possible side effects. The aim was to find out if stereotactic irradiation with protons inhibits regrowth of residual tumour or progress of inextirpable meningiomas without causing undesirable effects on the brain.

The overall mortality rate during the period of the study (1994–2012) was 13.5% (23/170 patients). The cause of death was identified in 17 cases with cancer being the most common. In three cases death could be related to the treated meningioma resulting in a disease-specific mortality of 1.7%. In one case the tumour increased in size 3 years after irradiation but no further treatment was initiated due to the patient’s age (81 at the time of progression). The patient died 2 years later. In another case the tumour recurred 3 years after initial treatment with resection plus proton radiotherapy and the biopsy after reoperation showed transformation to WHO grade II tumour. The patient underwent additional radiotherapy with photons and died 5 years later from further tumour progression. Finally, the third case was a patient with neurofibromatosis type 2, who died due to refractory status epilepticus after showing tumour growth shortly post irradiation. However, this patient had multiple other tumours as well and the irradiated meningioma was located at the clivus, thus making it improbable that the meningioma was the actual cause of status epilepticus and death.

The aim of this study was to evaluate proton irradiation technique in the treatment of benign meningiomas focusing on tumour growth control rates and possible complications and to compare the results with other more frequently used methods of irradiation. For that purpose 170 patients with benign meningiomas who were treated consecutively with protons were enrolled and followed-up for at least 5 years after irradiation (unless the tumour recurred or the patient died before). One patient was completely lost from follow-up. Nineteen of those patients were first included in a pilot study that was published in 1999 [13]. The patients were treated with a total dose of 24 Gy given in four 6-Gy fractions and were followed up for three years. No patient showed signs of tumour progression and in one patient slight deterioration of the clinical status was noticed 6 months after treatment but the situation improved at 12 months.

The 5- and 10-year PFS rate in the current study were 93% and 85% respectively which is satisfactory, especially with regard to the large tumour volumes (mean target volume 13 cm³), and comparable with the results from other studies [15, 26, 41, 42]. No deaths could be directly attributable to the radiation treatment. Complications were seen in 9.4% of the patients but the majority were either asymptomatic or treated pharmacologically with good effect (i.e. hypopituitarism).

More serious complications were also noticed. One patient became blind after irradiation. In that case the tumour was located on the optic nerve and despite the fact that smaller fraction doses were used (2-Gy fractions up to 46 Gy) the blindness in the ipsilateral eye was inevitable. Four patients developed milder visual deficits (two cases of deterioration of visual acuity and two of visual fields). In the cases with visual field deficits the tumour was located on the optic nerve and the patients had slight deficits already before the irradiation which worsened afterwards. In one case with deterioration of visual acuity, the treated meningioma was located in the medial sphenoid wing and the problem was identified to be the direction of one of the two fields that included the optic nerve and artery. In the other case the tumour was located in the medial sphenoid wing and the treatment was given in 4 × 6-Gy fractions to a total dose of 24 Gy including parts of the optic pathways. Visus deterioration might have been avoided with lower fraction doses. Finally, one patient with cavernous sinus meningioma developed radiation necrosis in the ipsilateral temporal lobe that required resection due to mass effect. Retrospective analysis of the treatment protocol for this patient revealed that the meningioma was treated with 4 × 6-Gy fractions to a total of 24 Gy through two portals. For two proton-beam portals the entrance dose lies in the range of 30-50% of the target dose, i.e. approximately 2 Gy × 4 or 2.5 Gy × 4. In our opinion, these dose levels are well below the risk for temporal lobe necrosis. The beam delivery technique used for these patients was passive scattering, which means that the conformality was not as good as can be achieved today with intensity-modulated proton beams (spot scanning). The larger dose margin around the target given with passive scattering could explain the development of temporal lobe necrosis.

The large number of patients included and the long follow-up period are the main strengths of this study. Regarding weaknesses, the study comes from a single centre and there is some inhomogeneity in the fraction doses as two patients were treated with 2-Gy fractions which is conventional fractionation and 14 with doses between 3.5–4.5 Gy (though the latter doses can be considered hypofractionated). In addition, 46 patients were lacking histological confirmation of their tumours and were treated for presumed grade I meningiomas. Finally, the pathological classification of the tumours was made based on the WHO guidelines from 2003. The revised guidelines that came out in 2007 changed the classification criteria in a way that some tumours which were classified as grade I based on the previous system now became grade II; that could possibly affect our material since some grade I tumours could actually be grade II, had they been classified with the 2007 system [22].

Radiotherapy/radiosurgery primary or secondary after surgery is nowadays a well-established treatment method for residual, recurrent or primarily unresectable meningiomas. Several modalities are available for that purpose; for example, external photon beam radiation therapy with conventional fractionation or intensity-modulated radiation therapy (IMRT), volumetric arc therapy (VMAT), stereotactic radiosurgery with photons (Gamma Knife-LINAC) either in a single fraction or with hypo-fractionation, high-energy protons or photon/proton combinations. Besides surgery and radiotherapy, there are currently no efficient medical options in the treatment of meningiomas. There are examples from palliative treatments, e.g. somatostatine analogues, however with negligible effect. Chemotherapy has not been proven to give any satisfactory responses [3, 23].

Minniti et al. [26] reviewed studies about treatment of skull base meningiomas with conventionally fractionated radiotherapy reporting 5 years PFS rates of 76–95% (90% on average). Complications reported in these series were radiation necrosis (rare), from optic pathways (0–3%), other cranial nerves deficits (1–3%) and hypopituitarism (5%). IMRT series were also reviewed in the same article. The largest series by Milker-Zabel et al. [25] included 94 patients, some with atypical or malignant meningiomas, treated with doses of 57.6 Gy with a median follow-up of 4.4 years. The reported tumour control rate was 93.6% at 5 years with neurological worsening occurring in 4% of the patients. Other smaller series reported tumour-control rates of 93–97% but with fewer patients and shorter follow-up [36, 44].

Stereotactic radiosurgery with Gamma Knife (GKS) or Linear accelerators (LINACs) has been widely used in the treatment of meningiomas with several studies published in the literature with varying numbers of patients and follow-up duration. The GKS experience shows that this technique is more suitable for smaller tumours (less than 10 cm³) with well-defined margins and sufficient distance from critical structures [1]. In a recently published large multicentre study Santacroce et al. [37] reported on 4,565 patients with 5,300 benign meningiomas treated with GKS. The mean tumour volume was 4.8 cm³, median dose to tumour margin was 14 Gy and the median follow-up 63 months. The 5-year PFS rate was 95.2% and the permanent morbidity rate was estimated at 6.5%. Starke et al. [42] reported a 5-year PFS rate of 96% in a series of 255 patients treated with GKS for skull base meningiomas where the mean tumour volume prior to treatment was 5 cm³ and the mean follow-up 6.5 years. In another study from the same group, 75 patients with large skull base meningiomas, defined as volume >8 cm³ (diameter >2.5 cm), were treated with GKS with a mean follow-up of 6.5 years [41]. The 5-year PFS rate in this study declined to 88% (compared to 96% in the previous study) with deterioration of neurological function in 17% of the patients. Pollock et al. [33] treated 416 patients harbouring benign meningiomas (252 image-defined and 164 histologically confirmed) with single fraction GKS. The mean tumour volume was 7.3 cm³, the mean dose 16 Gy and the median follow-up 60 months. The reported 5-year PFS was 96% and permanent radiation-related complications were seen in 11% of the patients. Kimball et al. [20] reported on 55 patients treated with LINAC radiosurgery for cavernous sinus meningiomas (median volume 5.9 cm³, median follow-up 50 months) where the actuarial 5-year local tumour control rate was 100% with one patient developing new cranial nerve (CN) deficit and one patient showing worsening of previously existing CN deficit.

Protons have been used in the treatment of meningiomas to a lesser extent compared to GKS and in varying settings (with conventional fractionation, hypofractionation, single-session SRS or combined with photons). One of the earliest published studies by Wenkel et al. [46] reported on 46 patients with benign meningiomas treated with combined photon and proton beam therapy at the Massachusetts General Hospital in Boston, USA. Recurrence-free rate at 5 and 10 years were 100% and 88% respectively but severe complications were also seen with one death case from brain stem necrosis and eight patients developing severe long-term toxicity of radiotherapy. Noel et al. [30] reported a 4-year local control rate of 87.5 ± 12% on 17 patients (5 with atypical meningiomas, 12 with benign but recurrent or rapidly progressive) treated with highly conformal photon and proton beam. Weber et al. [45] utilised proton radiotherapy with conventional fractionation for the treatment of intracranial meningiomas in 16 patients (median gross tumour volume 17.5 cm³, median dose 56 Gy[RBE-corrected] at fractions of 1.8–2 Gy[RBE-corrected] and median follow-up 34.1 months) and reported a 3-year PFS of 91.7%, with three patients developing radiation-related complications. Hypofractionated protocols have also been used, with Vernimmen et al. [47] reporting on 23 patients, 5 of whom were treated with stereotactic radiotherapy (≥16 fractions) and 18 with hypofractionated radiotherapy. The mean target volume in the latter group was 15.6 cm³ and the mean clinical and radiological follow-up periods were 40 and 31 months respectively. Sixteen out of 18 (88.8%) patients in that group showed good clinical status and radiological tumour control at the end of the follow-up period, 2 out of 18 suffered permanent neurological deficits. Finally, in a more recent study, Halasz et al. [15] reported the first series of proton stereotactic radiosurgery for the treatment of meningiomas. Fifty patients were included in the study with median tumour volume of 2.1 cm³, median prescribed dose 13 Gy[RBE-corrected] and median follow-up 32 months. The 3-year actuarial control rate was 94% and the rate of potential permanent adverse effects of radiosurgery was 5.9%.

The development of radiosurgery and stereotactic radiotherapy, including proton-beam therapy, offers a more precise treatment compared to older conformal 2D and 3D techniques with photons. With improved fixation techniques, more sophisticated analysis with CT and MRI for target definition and treatment with protons, a reduction is to be expected in late reactions in surrounding normal tissue both in frequency and intensity.

The results of this study suggest that protons are an equivalent alternative to other more frequently used methods of radiation in the treatment of meningiomas with high rates of long-term tumour growth control. The physical and dosimetric features of proton beams (no dose delivered “downstream” of the target and high conformality even to irregularly shaped targets) make them particularly interesting for patients with relatively large tumours, since the mean target volume in our material (almost 13 cm³) was clearly larger than in many other similar series in the literature. The somewhat problematic incidence of radiation-related complications may reflect some difficulties in defining the target based on CT images alone (which was exclusively used in the first patients of the series, MRI images as support became available after year 2000) but also the larger target volumes in our material.