Ten-year follow-up after Gamma Knife radiosurgery of meningioma and review of the literature

Objectives With regard to the generally slow growth of meningioma, it is essential to analyse clinical treatment results in a long-term perspective. The purpose of the present analysis is to provide clinical data after Gamma Knife radiosurgery of meningioma in a 10-year perspective together with a review of the current literature. Methods The current study is a retrospective analysis of 86 consecutive Swedish patients with meningiomas treated using Gamma Knife radiosurgery at the Karolinska Hospital Stockholm between March 1991 and May 2001. A total of 130 tumours were treated in 115 treatment sessions. The median radiological follow-up was 10 years (1.8–16.5 years), and the median clinical follow-up was 9.4 years (2.1–17.4 years). Results After a median follow-up period of 10 years, local tumour control was achieved in 87.8% of meningiomas (108/123 tumours). The median latency between initial treatment and local (in-field) recurrence (n = 15) was 5.8 years (1.9–11.5). Recurrences adjacent but outside the initial radiation field occurred in 15.1% of patients (13/86) at a median of 7.5 years (1.3–15.7). New meningiomas were seen in 10.5% after a median of 5.4 years (0.9–10.8). In 72% of patients, no further treatment was required, 17.4% (15/86) underwent a second Gamma Knife treatment, 4.7% (4/86) required later open surgery and 5.8% (5/86) required both secondary treatments. Eighty-six percent of patients were neurologically unchanged or improved. A significantly lower rate of local (in-field) recurrences was seen in meningiomas treated with a prescription dose of > 13.4 Gy (7.1% vs. 24%, p = 0.02). Conclusions The current retrospective analysis provides a 10-year follow-up and comprises one of the longest available follow-up studies of radiosurgically treated meningiomas. The current series documents a persistent high local tumour control after Gamma Knife treatment, while providing an estimation of a necessary minimum dose for long-term tumour control in meningiomas. The study confirms the validity of previous short-term data in a long-term perspective.


Introduction
The resection of intracranial meningiomas is a classical indication for open microsurgery. The infiltrative nature of meningioma can result in structural peri-operative damage of involved vessels, sinuses and cranial nerves resulting in increased potential morbidity in cases of aggressive tumour resection. Aggressive surgical approaches have increasingly been replaced by a disease management with lower invasiveness and lower peri-operative morbidity as current priorities strictly emphasize the patient's quality of life and avoidance of postoperative defects. 'Noli nocere' is the ancient term that ideally describes neurosurgery in the beginning of the twenty-first century.
The complete surgical removal of meningiomas in functional anatomical locations can be associated with a This article is part of the Topical Collection on Tumor -Meningioma significant risk of mortality and significant postoperative neurological deficits [6,37,43,55], and thus, the close involvement of eloquent or sensitive structures can make complete tumour resections virtually impossible [37]. Critical anatomical regions are the skull base, particularly the cavernous sinus and the petroclival region, but even complete resections of parafalcine meningioma can be complicated when the sagittal sinus is infiltrated. Meningiomas have often slow but highly variable growth rates with reported median doubling times ranging between 415 days and 8 years [22,24]. Although the clinical impact of postoperative remnants is occasionally questioned, long-term studies demonstrate that incomplete meningioma resections carry a significant risk for clinically relevant tumour recurrences [24,39].
Stereotactic radiosurgery has modified the therapeutic spectrum for meningioma and has gained an important role by reducing the risk for tumour recurrences in remnant meningiomas without significantly increasing the management risk. Stereotactic radiosurgery is more effective in smaller tumour volumes. In a successful combined management of meningioma, surgery is applied to reduce the tumour volume, while radiosurgery provides the tumour control in incompletely resected tumours. The option of a later complementary radiosurgical treatment has helped to significantly reduce the need for surgical radicality and hence the risk of potential perioperative complications [1,3,42].
The current retrospective consecutive cohort study reports the 10-year follow-up of meningioma patients treated with Gamma Knife radiosurgery according to current clinical and technical standards. The emphasis lies on the clinical and radiological long-term outcome of this radiosurgical approach that mostly had been combined with previous open tumour resections.

Patients and methods
All records of Swedish patients with meningiomas undergoing Gamma Knife radiosurgery between March 1991 and May 2001 at the Department of Neurosurgery at the Karolinska Hospital Stockholm, Sweden, were thoroughly reviewed. These patients were included in the long-term follow-up. Patients with anaplastic or atypical meningiomas were excluded from the current study. Hence, when histology was available, all radiosurgically treated meningiomas in the current study had been classified according to WHO grade I.

Radiosurgical treatment
The treatment was carried out using a 201 source Cobalt-60 Leksell Gamma Knife Model B (Elekta AB, Stockholm, Sweden). In all cases, a stereotactic frame was applied under local anaesthesia followed by a gadolinium-enhanced stereotactic MRI scan. The tumour outline was delineated on the T1weighted scans, which were imported into the planning software (Leksell Gamma Plan). The tumour margins including critical anatomical structures were outlined, and the dose plan was created with isodoses, prescription doses and maximum doses being determined by the responsible neurosurgeon in accordance with a radiosurgically trained medical physicist. The treatment protocol required that the prescription dose generally comprised more than 95% of the identified tumour volume. The contrast-enhancing dura adjacent to the meningioma, the so-called dural tail, was not included in the radiosurgical treatment field within the prescription isodose.

Follow-up
Clinical follow-up information was gathered by retrospective review of detailed patient records. In addition, patients were contacted by letter or occasionally by phone and asked to provide structured follow-up information based upon a questionnaire.
MRI data and clinical data were reviewed retrospectively as part of the clinical routine. In general, patients were followed with annual MRI in the first 5 years after radiosurgery and with bi-annual MRI thereafter, in very few exceptions with CT scans. These images as well as the radiological reports were used for the assessment of local tumour control after radiosurgery.

Definition of types of recurrences
The 'radiation field' was defined as the tumour volume contained within the prescription dose. Tumour (in-field) recurrence was defined as a progression of tumour volume within the initial radiation field, i.e. within the prescription dose. Out-of-field recurrence was defined as tumour progression immediately adjacent to the radiation field and hence outside the initial prescription isodose. The occurrence of a new meningioma was defined as a distant tumour unrelated to the radiation field.

Statistical analysis
Kaplan-Meier plots were used to estimate actuarial growth control rates. Additional comparisons applied the Fisher's exact test as appropriate.

Follow-up
The median radiological follow-up period after initial GKRS was 10 years (1.8-16.5 years) (including deceased patients). In 3 meningiomas, the follow-up period was considered too short after secondary treatment, and in 4 meningiomas, radiological follow-up was unavailable. Hence, conclusive radiological follow-up was available in 123 meningiomas (94.6%). Ten percent of the patients had a radiological or clinical follow-up of less than 5.7 years and 5.6 years, respectively. The patients were followed clinically for a median of 9.4 years (2.1-17.4 years) with clinical follow-up being available in 95.3% (82/86).

Neurological status
A total of 87.8% of patients (72/82) confirmed an unchanged or improved clinical status after a median clinical follow-up of 9.4 years (2.1-17.4 years) with complete clinical follow-up information available in 95.3% (n = 82). A clinical deterioration associated with the meningioma occurred eventually in 12.2% (10/82). In 7 out of 10 patients with later clinical deterioration, the symptoms were related to a tumour recurrence (5 local (in-field) recurrences and 2 out-of-field recurrences) while 3.7% of patients with available follow-up (3/82) developed side effects associated to the Gamma Knife treatment ( Table 2). One of these patients presented with confusion as a result of an adverse radiation effect after a re-treatment of a recurrent clinoid process meningioma.

Mortality
A total of 18 patients deceased during the long clinical followup. There was no short-term treatment-related mortality (e.g. > 1 year following GKRS) and no radiation-related mortality. In 8 patients, death was confirmed to be unrelated to the meningioma. In 4 patients, mortality was attributed to late meningioma recurrences: 1. 9 years after the first of five resections and 5 years after the last of three Gamma Knife treatment with signs of tumour progression of a petroclival meningioma. 2. At the age of 91, 15 years after the first operation and 8 years after radiosurgery with an extensive meningioma progression. 3. At the age of 73 with tumour progression 6 years after the first resection and 3.3 years after radiosurgery (MIB index 20%). 4. At the age of 77, a patient died as a result of a new and untreated tumour progression that had developed quickly with a volume of 19 cm 3 , which had then been considered too large to be treated with radiosurgery, 4 years after the first operation and 2 years after two Gamma Knife treatments for four and five highly cellular but non-atypical meningiomas.
In further three patients where the actual cause of death could not be established, a potential relation to the treated meningioma could not completely be ruled out: 1. One patient with unknown cause of death died 15 months after radiosurgery of a large (8 cm 3 ) meningioma of the foramen magnum at the age of 78. 2. One patient had been retreated for a recurrence, had developed an ARE and died 2 years and 4 months after the second and 9 years after the first radiosurgical treatment with no information concerning the actual cause of death at the age of 89. 3. Another patient with a large meningioma that had been treated incompletely with a low dose (prescription dose 11 Gy) experienced a later recurrence and a pituitary insufficiency 7 years after Gamma Knife treatment and died due to unknown reasons.
In further 3 patients, the cause of death was unknown. Out-of-field recurrences after Gamma Knife radiosurgery of a meningioma initially originating from the clivus. Example for a longterm response and local tumour control within the radiosurgically treated target. A recurrence outside the initial radiation field developed from the tumour's 'dural tail', which is generally not included in the radiosurgical treatment field. The patient was retreated with Gamma Knife resulting in tumour regression even in the recurring/progressive parts but developed a further 'out-of-field recurrence' within the right cavernous sinus 54 months after the initial treatment

Discussion
The present study provides the outcome data of patients with benign meningiomas treated with stereotactic Gamma Knife radiosurgery with a median radiological follow-up of 10 years (1.8-16.5 years) and a median clinical follow-up of 9.4 years (2.1-17.4 years) and provides evidence for the efficacy of radiosurgery in a long-term perspective. This is among the longest follow-up studies available in the literature of radiosurgically treated meningioma [2, 4, 5, 7, 9-11, 13, 14, 16-18, 20 [49].
Generally, local control rates are slightly lower in series with longer observation periods [27] and are very similar to the tumour control of 87.8% at 10 years found in the present study. Cohen-Inbar reported virtually identical local tumour control in 88.1% in a series with a median follow-up of 102 months [4], while the North American Gamma Knife Consortium published actuarial progression-free survival rates of 84% at 10 years after Gamma Knife treatment of petroclival meningiomas in a multicentre study of 254 patients [57].
The limitations of the current study lie in the limited total number of patients and in the fact that many of the meningiomas had been treated in the early phase of the development of radiosurgery with treatment regimens and dose planning systems that were significantly less sophisticated than currently available Gamma Knife techniques .
With an MR follow-up of 10 years, however, the current retrospective analysis comprises one of the longest available follow-up investigations in a larger series after stereotactic radiosurgery of meningiomas. It documents a persistent high local tumour control after Gamma Knife treatment, which is only slightly lower than in published observations with shorter follow-up.  In-field recurrences and out-of-field recurrences Meningioma recurrences within the radiation field should be differentiated from recurrences outside the initial radiation field. In-field recurrences ultimately represent intended treatment parameters, which potentially can be optimized, whereas recurrences outside the initial radiation field reflect the tumour biology and potentially progression of undetected tumour residuals. It can be argued that out-of-field recurrences can potentially be avoided through more sensitive and specific imaging and hence inclusion of tumour tissue adjacent to the outlined main treatment target. For example, imaging with Ga-DOTATOC might hold promise for sensitive targeting [33]. In contrast, remote meningiomas that appear many years after treatment must be considered as de novo tumours and unpredictable at initial radiosurgery. While the treatment cannot prevent these late recurrences, the patient's later management has to take this potential risk into account. The general policy behind the current study was based on the availability of MR follow-up, and the resulting clinical judgement that an eventual re-treatment for a documented tumour development outside the initial radiation field would carry a lower risk for side effects that the prophylactic inclusion of larger areas of potential and generally ill-defined 'dural tails'.
The current study demonstrates meningioma recurrences within the treated volume in 12.2% and out-of-fieldrecurrences in 15.1% of patients. This relatively high risk for development of meningioma outside the initial treatment area is generally not reflected by the current literature. Some late phenomena may be underestimated in the present literature since so far only 5 out of 39 published Gamma Knife series with a total of 1364 out of 12,431 published patients reported a follow-up of more than 94 months after radiosurgical treatment [4,26,31,34,61]. The present long-term series documents new remote meningiomas in 12.8% after a median of 5.4 years and a median time span of 7.5 years between the initial radiosurgical treatment and recurrences outside the initial radiation field. This latency is reflected by another recent long-term series by Kondziolka and colleagues who reported tumour growth adjacent to the treated volume at a median of 62 months after radiosurgery [30].
The current study demonstrates a latency of 5.8 years between GKRS and in-field recurrence, which is almost identical to the 59.8-month mean latency to local recurrence as documented in a long-term series by Skeie et al. [56] in which 67% of recurrences progressed within the first 2.5 years [56]. The present results show that in-field recurrences and out-offield-recurrences are rare and appear late, nevertheless having an important impact on clinical routines.
A long-term management is paramount for patients with meningioma where multiple interventions are often unavoidable. In the present study, 76.7% of patients had undergone an open tumour resection before radiosurgery and 17.4% underwent a second Gamma Knife treatment, 4.7% required later open surgery and 5.8% required both secondary surgery and a second Gamma Knife treatment. Since multiple interventions can be necessary, each intervention should be performed at the lowest invasive level. Long-term follow-up is absolutely essential in meningioma as the disease must be considered to be a chronic condition in many patients.

Tumour size reduction
Volume changes after stereotactic radiosurgery of meningioma are commonly moderate. The necessity for a reduction of tumour volume is not a central issue in the radiosurgical management of meningioma, since radiosurgery should generally be avoided in large meningiomas or when symptoms result from the tumour's mass effect. In larger meningiomas, a surgical resection remains the treatment method of choice. In cases where the tumour volume does not cause symptoms, a potential further volume reduction would be insignificant for the clinical outcome as long as further tumour progression is prevented. In these cases, the meningioma should be treated with stereotactic radiosurgery.
Size reductions were quantified in a study from the Mayo Clinic by Morita and colleagues with a typical distribution: 8% of the 88 meningiomas decreased in volume by more than 25%, 60% decreased by less than 25% and 29.5% remained unchanged [41]. Similarly, a report by Hayashi and colleagues documents a more than 50% volume reduction in 23%, and lower volume reduction in 59% and stabilization of the tumour in further 17% [17]. Shrinkage rates were significantly correlated to the amount of radiation energy delivered per tumour volume [17]. Others reported that tumours decreased in 46% and were unchanged in 44% of cases [9]. A report from Pittsburgh demonstrated a median 40% meningioma volume regression in 67% of patients and tumour stabilization in 26% [14]. Meningiomas that regressed demonstrated an 18% decrease in the first 3 months post-SRS with regression stabilizing after approximately 6 months, but a transient enlargement was observed in 9% of tumours that ultimately regressed [14]. A summary of reported tumour volume changes is provided in Table 3.

Side effects
In the present series, a total of 87.8% of patients (72/82) were neurologically unchanged or improved. The data revealed a clinical long-term management risk of 12.2%, which was higher than reported in comparable series with short-term follow-up, but the majority of side effects (7/10) were unrelated to the radiosurgical treatment but appeared late and were associated with tumour recurrences (5 in-field recurrences and 2 out-of-field recurrences). Only 3.7% of patients (3/82) with available follow-up developed side effects that were directly related to the Gamma Knife treatment (Table 1). Hence, the risk for treatment-related side effects was low. Similar to the present study, Starke and colleagues reported that tumour progression was present in 64% of patients with new or worsening neurological decline [58].
After linear accelerator-based radiosurgery of benign meningioma, the 5-year actuarial rate for the development of post-radiosurgical symptoms was 26.0%, which appears to be relatively high [32]. After Gamma Knife treatments, Kondziolka reported in a long-term study that 94% asymptomatic patients remained asymptomatic [30]. Other studies reported transient radiosurgical sequelae in 3.5% and permanent side effects in 1% [44], but in general, the reported risk for clinical side effects after Gamma Knife treatment ranges between 4 and 8% [10,28,48,49,53,57,61].

Adverse radiation effects
In serial structures such as the optic nerve, side effects are predominately related to dose thresholds, while a parenchymal radiation-induced tissue irritation is related to the applied dose and the volume of the irradiated brain tissue. The consequence of a radiation-induced tissue irritation is the occurrence of adverse radiation effects (AREs) that rarely occurs in radiosurgical treatment of meningiomas, as the applied radiation doses are relatively low.
AREs are seen on T1-weighted MRI images as secondarily increasing oedema and also as ring-shaped peripheral contrast enhancement. The radiation-induced oedema in meningiomas appears to occur late with the highest risk at about 11 months after SRS [18], is transient in most cases and generally regressing 18 months after SRS [51]. The clinical manifestation depends on the anatomical location of the secondary oedema. In the present study, only two patients presented with clinically symptomatic ARE, one of these patients after a second stereotactic radiosurgical treatment. Hence, the present study does not contribute data for the description of ARE in meningioma.
One early long-term study comprising patients who had been treated in the 1990s and published in 2001 showed transient radiation-induced oedema in 10.3% after Gamma Knife treatment at a frequency that must be considered as unusually high according to present standards [27], but 8/9 patients who had developed oedema had been treated at higher prescription doses [27]. Lee and colleagues who had summarized the radiosurgical experience in Pittsburgh after treatment of 964 patients with meningioma noted that the incidence of adverse radiation effect ranged from 5.7 to 16%, but that side effects were gradually reduced with better imaging and lower dosing [35]. In small-sized meningiomas, peri-lesional oedema occurred in 6.1% [36]. The large study from the Hospital Na Homolce in Prague published by Kollova and colleagues had demonstrated peri-lesional oedema after radiosurgery in 15.4% and temporary and permanent morbidity rates of 10.2% and 5.7%, respectively [28]. Virtually the same rate for the risk of oedema (15%) was recently reported in large series by Jang (15%) [23] or by Hoe (15.3%) [18].

Tumour volumes
As larger tumours are associated with a higher risk for radiation-induced oedema, the tumour volume is generally seen as the most complicating factor in stereotactic radiosurgery. A recurring question concerns the largest possible volume that can be treated with stereotactic radiosurgery. This volume may differ depending on the tumour location. Hoe and colleagues found increased management risks above relatively low tumour volumes of 4.2 cc [18]. Petroclival meningiomas with volumes of 8 cc and larger showed a significantly increased risk for tumour progression [9], and similarly in an earlier series, the outcome after Gamma Knife treatment was significantly worse in parasagittal meningiomas larger than 7.5 cc [29]. In cavernous sinus meningioma, the complication rate was considerably higher (21% vs. 3%) in meningiomas larger than 9.4 cc [47].

Pre-treatment oedema
Pre-treatment oedema and hemispheric tumour location have been associated with an increased risk for peri-tumoural oedema after radiosurgery [18,50]. In some extreme cases, preradiosurgical oedema in convexity, parasagittal or falcine meningiomas was even associated with the occurrence of severe persistent [51] secondary oedema after Gamma Knife treatment. In a series by Hasegawa, 4 out of 6 patients with preradiosurgical oedema from convexity, parasagittal or falcine meningiomas developed severe panhemispheric oedema after GKRS [15]. Based on these studies, pre-existing oedema should be considered a relative radiosurgical contra-indication as the oedema can increase significantly and can persist after radiosurgery.

Dose threshold for tumour control
The necessary dose threshold for successful tumour control in meningiomas remains to be defined, but the current series allowed an estimation of a necessary minimum dose (of 13.4 Gy) for tumour control in meningiomas while validating and confirming the short-term data from previous series in a long-term perspective. The European multicentre Gamma Knife study documented effective control in 92.5% of 3768 evaluated meningiomas that had been treated at a median prescription dose of 14 Gy [49]. The present long-term study demonstrated that patients who had been treated at prescription doses less than 13.4 Gy showed a significantly higher risk for a local recurrence of the meningioma (24.2%). At higher prescription doses, the risk for recurrence was only 7.1%. Similarly, tumour margin dose below 13 Gy significantly increased the likelihood of tumour progression in 763 patients with sellar or parasellar meningiomas treated with GKRS [54]. Kollova observed that a significantly higher incidence of tumour volume increase occurred in meningiomas treated with a margin dose lower than 12 Gy [28]. Skeie reported that in cavernous sinus meningiomas, lower prescription doses of 11.5 Gy were significantly associated with further tumour growth [56] and the team from Charlottesville noticed an increasing risk for tumour progression with decreasing dose to tumour margin [59]. On the other hand, Iwai et al. published the long-term outcome after Gamma Knife treatment of 108 patients and proposed to use lower prescription doses ranging from 8 to 12 Gy (median 12 Gy) [20]. With these lower doses, the actuarial progression-free survival rate was 93% at 5 years and 83% at 10 years, which is not different from other series applying higher doses [20].
With regard to the results from the present long-term study and the data cited above, it is safe to claim that meningiomas should be treated at prescription doses above 13-14 Gy for an improved chance to achieve long-term tumour control.

Gender differences
In the present long-term study, the risk for tumour recurrences was 28% for men and 12.1% for women (7/58). Due to the relatively low numbers, however, the difference was not significant. It is interesting, however, that men who had been treated at low prescription doses (< 13.4 Gy) carried a 50% risk for local recurrence of the meningioma. The present numbers are too small to draw conclusions based on this study alone, but serve as supporting evidence with regard to similar results that had been demonstrated by others: Kollova and colleagues had reported a lower local tumour control in male patients with meningioma [28], and the European multicentre meningioma study found that significantly higher tumour control in female patients [49]. Similarly, multivariate predictors of favourable outcome included female gender in the multicentre study of benign petroclival meningioma from the North American Gamma Knife Consortium [57], and male sex was a significant risk factors for tumour progression [25] in petroclival meningiomas and cerebello-pontine angle meningiomas [7,9].

Cavernous sinus and skull base meningiomas
In a large series of 255 patients with skull base meningiomas treated with Gamma Knife, the actuarial progression-free survival at 5 and 10 years was 96% and 79%, respectively [59], while Igaki documented actuarial local tumour control rates of 86.9% and 78.9% at 5 and 10 years, respectively [19]. Higher local control was seen in smaller tumours (≤ 4 cc) and in meningiomas treated with prescription doses above 14 Gy [19]. New cranial neuropathies occurred or worsened in 8.6% and decline in cognition or memory or cerebellar deficits, etc. in 2%, with petrous or clival location being predictive factors for side effects versus parasellar, petroclival and cerebello-pontine angle location [59]. In cavernous sinus, meningiomas local tumour control rates were 99% at 5 years [47] and 90.4% at a mean follow-up of 82.0 months with a resulting 10-year actuarial tumour growth control rate of 83.8% [56]. The complication rate of 6% included optic neuropathy, worsened diplopia or pituitary dysfunction while 21.0% of patients experienced improvement of symptoms [56].

Sellar and parasellar meningiomas
Due to their involvement of neurovascular and endocrine structures, complete resection of parasellar and sellar meningiomas can be associated with significant morbidity and incomplete resections are common. A multicentre study of ten centres of the North American Gamma Knife Consortium of patients identified 763 patients with benign sellar and parasellar meningiomas with median follow-up of 66.7 months. At the last follow-up, tumour volumes remained stable or decreased in 90.2% of patients with 88% actuarial progression-free survival rate at 8 years with new or worsening cranial nerve deficits occurring in 9.6% and additional 4.2% of patients experiencing other forms of symptom progression [54].

Posterior fossa meningioma and cerebello-pontine angle meningiomas
The large multicentre study of 675 patients treated with Gamma Knife radiosurgery for posterior fossa meningiomas documented tumour control in 91.2% at a mean follow-up of 60.1 months with resulting actuarial 10-year tumour control of 81% and a total of 27.4% of patients showing improvement in clinical outcome [53]. Trigeminal dysfunction was the most frequent new or deteriorated cranial nerve symptom followed by dysfunction of Nn. III/IV/VI [53]. Clival tumour locations, petrous or CPA locations rather than petroclival, tentorial and foramen magnum locations were predictive of neurological deterioration [53]. A virtually identical outcome was reported from the multicentre study of cerebello-pontine angle meningiomas with actuarial rates of progression-free survival of 93% and 77% at 5 and 10 years, respectively, with permanent neurological deterioration occurring in 8.5% with most common worsening neurological deficits being dizziness, imbalance, hearing loss or permanent trigeminal nerve dysfunction in 4% (transient in 54.5%) [7].

Petroclival meningiomas
In a large multicentre study of 254 patients with benign petroclival meningioma, 140 patients had been treated with upfront radiosurgery while 114 patients were treated with Gamma Knife following surgery. Kaplan-Meier actuarial progression-free survival rates at 10 years were 84%, and at last clinical follow-up, only 6.4% of patients had experienced progression of symptoms [57]. Similarly, Flannery, Kondziolka and colleagues had reported 10-year progression-free survival rates of 86% in their series with petroclival meningiomas with meningiomas larger than 8 cc having a significantly increased risk for tumour progression [9].

Parasagittal and parafalcine meningiomas
Neurological function did not deteriorate, and no additional therapy was required in patients after Gamma Knife treatment of parasagittal meningiomas smaller than 7.5 cc [29]. The experience at the University of Virginia with Gamma Knife treatment of parasagittal and parafalcine meningiomas (WHO grade I) was published in two different cohorts: 65 patients with 90 meningiomas with median treatment volume of 3.7 cc (range 0.7-33.1 cc) treated between 1991 and 2006 [8] and 61 patients with 77 meningiomas and a median volume of 5.6 cc were treated between 1991 and 2013 [51]. The actuarial tumour control rate was 70% at 5 years [8]. In both cohorts, new or worsened peri-tumoural oedema occurred in 40% [8,51] (40.4% [8]) with 8.2% being symptomatic in the earlier cohort [8]. Post-radiosurgery seizures were seen in 14.3% including 6% patients who had not experienced any seizures prior to radiosurgery [8]. The median interval between GKRS and oedema peak ranged between 6 and 24 months with a median at 18 months [51]. Tumour volume above 10 cc and prior existing peri-tumoural oedema were factors for new or worsening oedema [51]. Hasegawa reported a similar rate of symptomatic radiation-induced oedema in 7% of convexity, parasagittal or falcine meningiomas undergoing Gamma Knife radiosurgery with actuarial 5-and 10-year local tumour control rates of 87% and 71%, respectively [15]. Radiosurgery of parasagittal and parafalcine meningiomas may be associated with a higher risk for the development of radiation-induced oedema, and hence, the volume of radiosurgically treated tumours in this region should be consequently restricted. The necessity of prior resection has to be scrutinized with radiosurgical options remaining for smaller or unresectable tumours, for postoperative tumour remnants and meningiomas with sinus invasion.

Planned surgical/radiosurgical cooperation
As meningiomas are frequently large at diagnosis and the risk profile of radiosurgery is significantly increased in large tumours, prior surgical resection and volume reduction generally create the prerequisites for a successful long-term outcome. Since, however, incomplete resections carry a significant risk for recurrence, a completing postoperative radiosurgical treatment is often necessary. From a surgical point of view, the necessity for a postoperative treatment is occasionally questioned since benign meningiomas develop at variable growth rates with often low doubling times [22,24] and significant growth of postoperative tumour remnants may not always be noticed in a shorter follow-up. As a result, the surgical approach to tumour remnants that are deemed unresectable is often a 'wait-and-see' attitude with a potential additional surgical removal only if the re-growing tumour has reached a clinically relevant mass effect.
The risk for long-term growth, however, is significant: tumour growth occurred in 76% of petroclival meningiomas after a median follow-up of 85 months with associated functional deterioration in 63% of cases when tumours were growing [60]. Similarly, Mathiesen and colleagues had shown that after deliberate non-radical surgery (Simpson grade IV), tumour recurrences appeared in 72% [40], while a combined treatment of immediate Gamma Knife radiosurgery after a tailored microsurgical resection provided a low recurrence rate of 10% [40]. The documented long-term tumour control achieved by stereotactic radiosurgery has contributed to a policy change in open resective neurosurgery as tumour components close to functional brain or cranial nerves can generally be left in place during surgery and can later be treated with radiosurgery, thus reducing the need for an aggressive resection [1,3,42].
Since the tumour volume is a significantly complicating factor for stereotactic radiosurgery, it would not be beneficial to monitor a meningioma without treatment until the tumour progresses significantly. Hence, from a radiosurgical point of view, any radiosurgical treatment should be carried out while tumours are still as small as possible.

Conclusions
The current retrospective analysis provides a 10-year follow-up and comprises one of the longest available follow-up studies of radiosurgically treated meningiomas. The current series documents a persistent high local tumour control and a persistently low risk for side effects after Gamma Knife treatment, while providing an estimation of a necessary minimum dose for long-term tumour control in meningiomas. The study confirms the validity of previous short-term data in a long-term perspective. In critical locations, stereotactic radiosurgery can replace a complicated surgical resection. In a planned surgical/radiosurgical cooperation, the need for an aggressive tumour resection is reduced as stereotactic radiosurgery provides a documented long-term control of tumour remnants.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. As this manuscript is based on a retrospective study, for this type of study, formal consent is not required.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.