Radiation Associated Meningiomas:The Case for Repeat Radiotherapy

Abstract

Purpose

To evaluate the circumstances relating to Radiation Associated Meningiomas (RAM) and whether repeat radiotherapy can be used.

Materials and Methods

In this Ethics Approved Study, databases of the Long Term Follow-Up Clinic at Sydney Children’s Hospital and the Prince of Wales Cancer Centre were audited for patients who received radiotherapy to the head as a child, and subsequently developed an intracranial meningioma. Features noted were, initial diagnosis, extent of prior treatment, dose, latency to diagnosis, subsequent treatment, and outcome.

Results

There were 18 patients with an equal gender mix of males/females (9), the majority of patients being treated with whole brain radiotherapy for leukaemia prophylaxis. 6 patients had radiotherapy, 4 had follow-up out to 17 years with no further adverse event, and control of the meningioma.

Conclusion

Surgery may still be the treatment of choice, however where appropriate radiotherapy can be considered, with good prospect of benefit to the patient.

THE CASE FOR REPEATED RADIOTHERAPY

Radiotherapy has an established role in the management of many paediatric tumours, benign and malignant. Whilst many are central nervous system (CNS) tumours where radiotherapy may be a central component of management, treatment is also directed to the brain/meninges for other conditions, such as acute leukaemia to address the risk or definite evidence of meningeal involvement. The unfortunate circumstance is that radiotherapy is still an essential component of many paediatric protocols for prolonged survival. However, there can be a price paid by the long term survivors of this treatment. Cognitive deficits,[1,2] musculo-skeletal abnormalities,[3] cardiovascular events[4, 5] and endocrine deficiencies[6] have all been described. However for many survivors it is the development of a 2nd tumour which carries the most emotive concern.[7] Whilst malignancies have been reported, the tumour occurring with the greatest frequency is that of intra-cranial meningiomas.[8] Friedman et al. in an analysis of the Childhood Cancer Services Study (CCSS) in which follow-up information was available in 14,359 survivors, noted that of the 2,703 tumours diagnosed in 1,402 patients, 3.1% of these were meningiomas.[7] Paulino recently reported on a literature review of 66 studies where a total of 143 patients with radiation associated meningiomas (RAM) were documented.[9] This review describes a median time for development of a meningioma of 19.67 years, with the primary medial condition being treated by either small or wide field radiotherapy, with no dose dependence. Treatment details are not always recorded in individual series, however the importance is that none of the recorded patients in that review were treated by further radiotherapy for their RAM. This review evaluates patients seen, and treated in a single centre where repeat radiotherapy, in the form of stereotactic intensity modulated radiotherapy (SIMRT) and stereotactic radiosurgery (SRS), was used in selected patients to achieve local control.

MATERIALS AND METHODS

For this Ethics approved study two complimentary databases were audited. The eligibility criteria were: treatment by radiotherapy of a paediatric patient with a malignant tumour (includes one child with a craniopharyngioma), and subsequent development of an intracranial meningioma. These databases were the long term follow-up registers of the Paediatric Oncology Department at Sydney Children’s Hospital (SCH), and the Meningioma Database at the Prince of Wales Hospital (POWH) Cancer Centre. Whilst the Radiation Oncology Department at POW Hospital provides radiotherapy services to the SCH (and its predecessor POW Children’s Hospital) a number of adults were referred with the more recent diagnosis of meningioma; the past history of prior radiotherapy as a child and treatment elsewhere was noted, and separately recorded in the database. In this review information retrieved included: patient (age at initial diagnosis, sex), disease (initial tumour), partial / whole brain radiotherapy, dose, time interval to RAM diagnosis, and subsequent treatment factors with emphasis upon the use of SIMRT and SRS to address a typically non-resectable RAM. The latency period is the time interval between the initial delivery of radiotherapy and the subsequent diagnosis of the RAM. There was no planned screening program for these patients, thus the diagnosis of an RAM was based more typically on new symptoms. Since a proportion of patients were treated initially outside of POW Cancer Centre, it is not possible to define a denominator in terms of developing an RAM versus total number of patients treated. This is a descriptive study of a relatively small number of patients with no statistical analysis supplied.

SIMRT is applied using a relocatable head device (GTC, Radionics, Boston) with X-Knife 2-4 version software for planning with all treatment using a 4mm leaf width mini multi-leaf collimator (MMLC) on a Siemens Linac. A rigid quality assurance program exists for the delivery of all stereotactic radiotherapy. Dose and fractionation was individualised for each patient. Where SRS was used this involved application of a BRW (Radionics, Boston) head ring using the same software as above with however cones being used to define the volume to be treated. The differentiating feature between using SRS and SIMRT is the volume of the lesion to be treated.

RESULTS

Table 1 indicates all the relevant details regarding the patient population of 18 patients. All patients are aged (at the time of this review) less than 42 years. Of note is that 16 of 18 patients received cranial radiotherapy (doses ranging from 18-47Gy.) Thus only 2 patients had partial brain radiotherapy although patient 10 had initial convention radiotherapy then an SRS procedure for progressive disease. His craniopharyngioma remains locally controlled 16 years after his SRS (initial radiotherapy being in 1988). The median interval to diagnosis of an RAM is 27.5 years for males and 19.34 years for females. There are no reported atypical or malignant meningiomas in this series although histology was not available for all patients. There were 6 patients who had repeat radiotherapy (2 with SRS, and 4 with SIMRT). Patient 9 required 2 episodes of SIMRT to a prepontine meningioma to achieve stable disease, in the context of progression following his initial SIMRT delivered 4 years previously.

With the constraints of this study having relatively short follow-up following an intervention (including observation), all patients remain free of features of clinically progressive meningioma. Those patients having surgery have no further problems related to the resected meningioma. Of the 6 patients having a form of radiotherapy, all patients remain locally controlled, with one patient requiring repeat treatment to achieve this. There is no added toxicity to any of this group, given the focused nature of treatment. However, morbidity data is not available for the surgical patients, merely a ‘tumour outcome’.

A case history demonstrates the clinical journey for one of the SIMRT patients. Patient 8 presented in February 2000 with severe right facial pain and worsening vision in the right eye, imaging indicated a right sphenoid ridge meningioma extending into the right optic canal. Resection of 80% of the meningioma was performed with improvement in pain and vision. Early in 2001 there was worsening right facial pain, again quite severe in nature. Repeat MR imaging indicated progression of the meningioma: involving the right sphenoid ridge, cavernous sinus, pterygo maxillary fissure, posterior aspect of right maxilla, right nasal cavity, floor of the right middle cranial fossa, and limited extra cranial extension through the right foramina of ovale and rotundum. The only neurological deficit was the known impaired vision in the right eye. This was not resectable disease and too large for SRS with a volume of 20.5 cubic centimetres. This was treated with SIMRT (Figure 1), within 6 months all pain had cleared, and 8 years later MR imaging remains stable.

DISCUSSION

Radiological abnormalities are a relatively common event in children many years after cranial radiotherapy for leukaemia. Fasci reported on 56 patients with a median follow-up of 11 years post treatment, with cavernous haemangiomas the most common tumour, and an asymptomatic meningioma noted in 1 patient.[10]

Banerjee reported on a population of adult survivors of childhood leukaemia who had all received cranial irradiation.[11] Of note is that 49 of 60 survivors had follow-up MRI scans performed. Meningiomas were noted in 11 of the 49 (22%), with a mean latency period of 25 years. The incidence appeared to be increasing the longer the follow-up, a feature noted by others. In a review of 14,359 5 year survivors by Friedman, 1,402 patients developed 2703 neoplasm, of which 3% were meningiomas.[7] It is unknown how many patients had follow-up imaging to gain a better understanding of the time incidence.

Table 1.

Initial Diagnosis, Latency Period and Meningioma Details by Meningioma Diagnosis

Patient	Sex	D.O.B.	Diagnosis	Time	Extent		Dose (Gy)	Fx	Meningoma Diagnosed	Years	Single/	Treatment	Outcome
					WB	PB					Mutliple
12	M	01/03/1976	NHL	Jun-79	Y	-	24	12	10/02/1990	10.66	S	Surg	Stable
3	F	22/06/1972	ALL	Jun-74	Y	-	30	18	12/06/1994	19.99	M	Surg	Stable
14	M	07/02/1969	E	Jun-85	Y	-	36	20	20/06/1996	11.01	M	Surg	Stable
13	F	07/05/1982	ALL	Jun-86	Y	-	24	16	15/02/2000	13.67	S	SIMRT	Stable
5	M	29/04/1980	ALL	Feb-87	Y	-	18	10	17/03/2003	16.08	S	Surgery	Stable
11	F	23/06/1980	AE	April 93	-	Y	60	34	20/11/2003	10.6	S	SRS	Stable
15	M	09/10/1980	Cranio	Jun-88	-	Y	54	+30	01/04/2004	15.79	S	SIMRT + Observation	Stable
6	M	07/09/1976	ALL	Dec-78	Y	-	24	14	15/06/2006	27.5	S	Surgery	Stable
1	F	01/03/1971	ALL	Feb-87	Y	-	25	13	20/06/2006	19.34	S	Surgery	Stable
16	F	21/06/1985	E	Aug-91	Y	-	47	26	15/08/2007	16	S	Surg + SIMRT	Stable
8	M	18/12/1965	Histo	Aug-71	Y	-	20	8	12/01/2008	36.41	S	Surgery	Stable
2	F	16/02/1972	H	Jul-87	Y	-	30	15	10/12/2008	21.41	S	Surgery	Stable
10	M	03/07/1966	AML	Nov-78	Y	-	36	20	12/01/2010	31.16	S	Observation	Stable
4	M	20/11/1974	ALL	Jul-77	Y	-	30	15	20/09/2010	33.18	S	Surgery	Stable
9	F	29/04/1980	ALL	Jul-82	Y	-	18	12	18/10/2010	28.26	M	Surgery	Stable
17	M	21/04/1974	ALL	Jun-78	Y	-	24	12	15/11/2010	32.42	M	SIMRT	On Treatment
7	F	09/10/1985	Med	Oct-91	Y	-	36	20	20/11/2010	19.1	S	Nil	Stable
18	F	26/04/1979	ALL	My 83	Y	-	18	10	20/11/2010	28.4	S	SIMRT	On Treatment
Median years										19.67

D.O.B:	date of birth	SURG:	surgery
ALL:	Acute Lymphatic Leukaemia	+:	received initial conventional radiotherapy, then SRS of 18Gy single
HISTOLOGY:	Histiocytosis		fraction for recurrence in May 1994, craniopharyngioma, stable since then
AE:	Anaplastic Ependymoma	WB:	whole brain
E:	Ependymoma	PB:	part brain
CRANIO:	Craniopharyngioma	SIMRT:	Stereotactic Intensity Modulated Radiotherapy
AML:	Acute Myeloblastic Leukaemia	S:	Single
MED:	Medulloblastoma	M:	Multiple
TIME:	Time of initial radiotherapy

Galloway in perhaps the most descriptive study, endeavoured to review 370 patients aged < 19 years, treated to a site in the cranio spinal axis from January 1963 – December 2006.[12] 81% of living patients were contacted, with a total of 16 patients reporting a second tumour, potentially treatment related, a median of 18, 9 years following radiotherapy. Amongst these 16 patients, 10 had meningiomas. Repeat radiotherapy was used in 3 patients, and 5 had gross total resection, all 8 patients have controlled disease. Sub-total resection occurred in 2 patients, 1 of whom died due to progression of the meningioma.

In Paulino’s review of 66 published series for a total of 143 patients, it was noteworthy that there were only 16% of patients who had a diagnosis of acute leukaemia, although of the total number of diagnoses 53% of patients had whole brain treatment.[9] Of concern was the fact that there were 17 patients (12%) who had benign tumours, with 14 having radiotherapy for tinea capitis. The median latency period for the diagnosis of a RAM was 19.67 years. In this series there are equal numbers of males and females (9), a situation at variance with the normal circumstance in which there is a female preponderance for meningiomas. All patients were relatively young to be recorded as having meningiomas (oldest patient being 41 years at diagnosis). Whole brain radiotherapy doses ranging from 18-47Gy were delivered in 16 of 18 patients. The 2 patients with part brain treatment received 54+Gy to the lesion being addressed. Only 2 patients had multiple meningiomas. For the 2 patients with part brain treatment the respective meningioma developed in the lower dose volume. There was only 1 patient who didn’t receive chemotherapy, the patient with a craniopharyngioma. This may be of relevance since Paulino commented that those patients with leukaemia diagnosed had a shorter latency period.[9] This cannot be assessed for our population.

In terms of treatment Paulino commented that surgical information was available for 126 of the 143 patients.[9] A gross total resection was performed in 113 patients, and a sub-total resection in 14 patients, with no patient reported as having radiotherapy. There was insufficient follow-up data in the reported series to document the effectiveness of surgery. In this population 6 patients had their meningiomas treated by radiotherapy. Patients 17 and 18 have recently commenced SIMRT and thus there is not complete follow-up information available. Patient 17 has an optic nerve sheath meningioma, with recent review of this Centre’s results indicating a 96% local control.[13] For 4 patients with a follow-up range of 1–17.3 years since stereotactic treatment, all patients have stable disease, a pattern comparable to other published series,[14, 15] including this Centre’s, for non RAM managed by a form of radiotherapy.[16]

Banerjee endeavoured to define predictive features. [11] There appeared to be no impact by the type of chemotherapy nor the duration. Nor did the age of the child at treatment have an effect. The dose delivered also seemed to have no predicted effect. It should be noted that this review by Banerjee only addressed a defined population, those children treated for acute leukaemia,[11] the denominator still remains unknown.

It is feasible and appropriate that surgical resection of a meningioma should occur, as happened for some of the patients in this series. However the fact that the meningioma may be induced by radiotherapy does not automatically exclude the use of further radiotherapy to achieve a stable growth pattern, a feature regarded as the aim of management in non RAM meningiomas. Obviously the patient numbers in this series are small, as is each of the individual 66 series in Paulino’s review. [9] The 18 reported patients in this study would make this the largest series.

There are uncertainties in this report. Would additional further radiotherapy increase, or change the risk of these patients developing additional meningiomas (or any other tumour including cavernous haemangiomas), and does this additional radiotherapy have to be delivered stereotactically?

For the former, only time will tell. There are no additional meningiomas in this series with follow-up now out to 17 years. In terms of the latter, it is a more precise treatment with the ability to define dose critical structures, and lower the dose to adjacent normal tissues. Convenience may also allow further treatment to be delivered as SRS. However there is more of the brain that receives this lower dose. Again only time will tell, with the lowest dose delivered being 18Gy, it is completely unknown whether lower doses are capable of inducing meningiomas, other than to note that there are 17 patients in Paulino’s review who had doses less than 10Gy to small areas of the scalp.[9]

CONCLUSION

Fortunately RAMs are infrequent, it is not possible from this series or Paulino’s review to define a risk of this event occurring. Of importance is that each patient’s meningioma should be addressed on its own merits, be it single or multiple. Observation can be the preferred approach, at least initially, depending upon the location and the patient’s symptoms. Surgery should be performed as appropriate. However, it is entirely indicated, given the above caveats, to consider utilising repeat radiotherapy, delivered stereotactically, for an RAM with high likelihood of long term local control.