Clinical Oncology 25 (2013) 356e361 Contents lists available at SciVerse ScienceDirect
Clinical Oncology journal homepage: www.clinicaloncologyonline.net
Original Article
Radiotherapy for Optic Nerve Sheath Meningioma: A Case for Earlier Intervention? G. Adams *, D.E. Roos *y, J.L. Crompton zx * Department
of Radiation Oncology, Royal Adelaide Hospital, Adelaide, South Australia, Australia Department of Medicine, University of Adelaide, Adelaide, South Australia, Australia z Neuro-Ophthalmology Service, Royal Adelaide Hospital, Adelaide, South Australia, Australia x Institute of Ophthalmology and Visual Science, University of Adelaide, Adelaide, South Australia, Australia y
Received 29 September 2012; received in revised form 3 December 2012; accepted 24 January 2013
Abstract Aims: To assess tumour control, visual outcomes and toxicity after radiotherapy for all patients with optic nerve sheath meningiomas (ONSM) treated by a single radiation oncologist at a single institution over a 15 year period. To explore potential predictors of outcomes. Materials and methods: All patients underwent ophthalmological and radiological assessments before radiotherapy. These were repeated at regular intervals after treatment. A retrospective analysis of clinical, dosimetric and radiological data was carried out. Patients with useful vision before radiotherapy were divided into two groups e those with maintained or improved vision and those with a deterioration in vision. The groups were compared using the ManneWhitney U-test with regard to eight potential predictors of outcome. Results: Seventeen patients with 18 ONSM were treated with fractionated radiotherapy (46.8e55.8 Gy in 26e31 fractions). No evaluable tumours grew after treatment: control rate 100% (95% confidence interval 82e100%). Using the most common definition of visual function described in the literature, vision was maintained or improved in 89% (95% confidence interval 67e97%) of cases. In those with useful vision before treatment (13 evaluable eyes), visual acuity was maintained or improved in eight (62%, 95% confidence interval 36e82%). There was a suggestion that the time from the onset of symptoms to radiotherapy may influence outcome. Those with stable or better visual acuity after radiotherapy had been observed for a shorter time compared with those who had worse visual acuity (median of 18 months versus 62 months). Acute and late toxicity from radiotherapy was manageable. Conclusion: Radiotherapy is an extremely effective modality in arresting the growth of ONSM. A longer time from symptom onset to the start of radiotherapy may predict for poorer outcomes. Crown Copyright Ó 2013 Published by Elsevier Ltd on behalf of The Royal College of Radiologists. All rights reserved. Key words: Meningioma; ONSM; optic nerve; optic nerve sheath; radiotherapy
Introduction Optic nerve sheath meningiomas (ONSM) are rare and account for 2% of all orbital tumours [1]. Like other meningiomas, they are more common in women and are typically slow-growing benign tumours. The usual presentation is with progressive visual loss or proptosis (Figure 1). Historically it was felt that complete visual loss was inevitable and that the only worthwhile intervention was to
Author for correspondence: G. Adams, NT Radiation Oncology, Alan Walker Cancer Care Centre, Rocklands Drive, Tiwi, NT 0810, Australia. Tel.: þ61-8-8944-8220; Fax þ61-8-8944-8222. E-mail address:
[email protected] (G. Adams).
surgically decompress the nerve in those with useful but deteriorating vision, with the aim of delaying complete visual loss [2]. However, over the last few decades, evidence has emerged that with radiotherapy, not only can the growth of the tumour be stopped or reversed, but also vision can be maintained or improved in most cases [3]. Because of the rarity of ONSM, evidence has come from case series with limited numbers [4e13] that have either combined the experience from several groups [4,5] or have presented single institutional data [6e13]. Various techniques have been used, including conventional threedimensional conformal radiotherapy (3DCRT) [4,10,12], fractionated stereotactic radiotherapy (FSRT) [4e9,11,13], stereotactic radiosurgery [11] or proton therapy [6]. It is not possible from these small studies to determine if one
0936-6555/$36.00 Crown Copyright Ó 2013 Published by Elsevier Ltd on behalf of The Royal College of Radiologists. All rights reserved. http://dx.doi.org/10.1016/j.clon.2013.02.004
G. Adams et al. / Clinical Oncology 25 (2013) 356e361
357
Fig 1. T1-weighted post-contrast magnetic resonance imaging of a neglected 35 27 26 mm left optic nerve sheath tumour causing blindness in a patient initially observed, then lost to follow-up before seeking treatment for cosmetic reasons. (a) Axial image showing severe left proptosis and ‘ghost’ of the atrophic optic nerve within the enhancing tumour posteriorly; (b) coronal image through the anterior aspect of the tumour showing the compressed left optic nerve (central dark spot), and the posterior aspect of the normal right globe (dark).
modality is superior to others. However, given that radiation tolerance of the optic nerve and chiasm may be improved by using conventionally fractionated (1.8e2 Gy per fraction) radiotherapy, and the limited toxicity data on hypofractionated radiotherapy on these structures [14], it is preferable to use conventionally fractionated radiotherapy when attempting to maintain vision. The timing of when radiotherapy should be administered is less certain. Due to concern regarding late effects on the optic apparatus, a period of observation is often advocated while there is no clinical or radiological deterioration [3,4,15]. Yet, the question arises as to whether prolonged observation may compromise preservation of vision as the tumour continues to grow-albeit slowly. The aim of this study was to assess tumour control, visual outcomes and treatment-related toxicity for all patients with ONSM treated with radiotherapy at the Royal Adelaide Hospital, South Australia, and to compare our experience with previously published series. In addition, we wanted to explore the influence on these outcomes of factors such as the prescribed dose of radiotherapy, the dose to critical structures, the tumour size or the timing of radiotherapy.
Materials and Methods All patients who received radiotherapy for ONSM between 1996 and 2011 were identified using the departmental electronic database. The diagnosis was based on the presence of typical clinical features and characteristic magnetic resonance imaging (MRI) findings [3,15]. Where there was sufficient doubt about the diagnosis, a biopsy was carried out. Relevant information was obtained from hospital records, radiotherapy treatment plans and serial imaging (baseline and follow-up MRI scans). For patients referred from elsewhere, their ophthalmologist was contacted directly for further information when required.
Ophthalmological Signs/Symptoms All patients had a complete assessment at first presentation to an ophthalmologist. This included best corrected visual acuity (Snellen chart), colour vision, visual fields, eye movements, optic discs, proptosis, relative afferent pupillary defect and pain. At the time of referral for radiotherapy, any changes in these findings from baseline were recorded. After completing treatment, patients were followed long term by their referring ophthalmologist with regular repeat assessments and (usually) yearly imaging. Cataract formation as a cause of declining visual acuity was identified and corrected where appropriate. Radiotherapy In all but one patient, radiotherapy was delivered using conventional conformal techniques using 6 MV photons with immobilisation in a thermoplastic shell. On the planning computed tomography scan, the gross tumour volume was defined as the visible (enhancing) meningioma and was marked with the aid of the diagnostic MRI (fusion images when available). A margin of 5e10 mm was added to create the planning target volume. Organs at risk, namely retina, optic nerve, chiasm and brain stem, were delineated. An optimal conformal plan was obtained with regard to the prescribed dose and tolerances of organs at risk. A typical beam arrangement involved a three- to five-field plan sometimes including a superioreinferior oblique field exiting via the oral cavity to help spare dose to anterior orbital structures. Image guidance protocols followed departmental protocols at the time. One patient was treated using FSRT, using three noncoplanar 6 MV photon arcs with circular collimators to a total dose of 52.2 Gy in 29 fractions prescribed to the 90% isodose curve (final fraction off retina). The gross
358
G. Adams et al. / Clinical Oncology 25 (2013) 356e361
tumour volume was defined the same way as in the 3DCRT cases, but without a planning target volume expansion. Radiology Standard diagnostic MRI scans (typical slice thickness 2e3 mm) were obtained before and after treatment. Attempts were made to ensure that when measuring tumour size in individual patients over time similar sequences were used (typically T1-weighted fat suppressed with gadolinium). In addition to the radiology report, copies of all individual patient imaging were reviewed at the same time by a single author. The maximum dimensions of each tumour along the course of the optic nerve and orthogonally in the supero-inferior and medio-lateral directions were measured. Outcome Assessment The primary outcome assessments were changes in visual acuity and tumour size after radiotherapy as well as treatment toxicity. In order to identify possible prognostic factors for visual outcomes after radiotherapy, a strict definition for deterioration in acuity was used. This was any decline in visual acuity in those patients with useful vision (best corrected visual acuity greater than 6/60) before radiotherapy. However, so that a comparison with other published series could be made, we also used other common criteria quoted in the literature, namely either a one or two line loss on the Snellen chart considering all treated eyes. A reduction in tumour size was defined as a 2 mm or greater sustained regression in one dimension without growth of 2 mm or greater in any other dimension. Growth of the tumour was defined as a sustained increase of 2 mm or greater in any dimension regardless of changes in other dimensions. Given the retrospective nature of the study, toxicity was assessed by the descriptive reports in the case notes. Insufficient information was available to assess pituitary dysfunction related to radiotherapy. An analysis of eight potential predictors of visual outcome was made. These factors were the time from the onset of symptoms to radiotherapy, the size of the tumour at the start of treatment, visual acuity at the start of treatment, age at the start of treatment, radiotherapy dose, the dose to the retina, the dose to the chiasm and the year of treatment. Statistics In order to assess the response to radiotherapy, findings at the last assessment before treatment were compared with the most recent post-therapy assessment. To test whether potential prognostic factors had a significant effect on visual acuity, tumours were divided into two groups, namely improved or stable versus worse visual acuity according to our strict definition of deterioration.
Comparisons between the two groups were carried out using the ManneWhitney U test [16]. Because eight prognostic factors were tested, for a single prognostic factor to be declared significant at the 5% level, its P value would have to be less than 0.0064. However, with the limited number of subjects, it was not considered appropriate to carry out a multivariate analysis.
Results Seventeen patients and 18 ONSMs were treated (one had bilateral tumours treated simultaneously). For 16 tumours the diagnosis was based on clinical and radiological features. For two there was sufficient doubt about the diagnosis to warrant biopsy. All patients were treated by a single radiation oncologist. Outcomes for eight of these patients were included in a previously published multi-institutional review of 34 ONSM patients, but their outcomes were not reported separately [4]. Table 1 summarises the baseline patient and tumour characteristics, with reported toxicity summarised in Table 2. Prescribed doses were 46.8e55.8 Gy in 26e31 fractions (all 1.8 Gy per fraction). One patient refused follow-up imaging, but was available for clinical assessments. The median follow-up from the start of radiotherapy to the latest assessment was 6.2 years (range 0.9e13.3 years) for MRI imaging and 5.3 years (range 0.6e13.4 years) for clinical assessment. No evaluable tumours enlarged, 10 (56%) reduced in size and seven (39%) remained stable. Of note, the patient who refused imaging has been followed-up for more than 5 years with improved visual acuity compared with base line. Five of the 18 eyes did not have useful vision prior to radiotherapy (defined as visual acuity of 6/60 or worse). Of the remaining 13 eyes (all with acuity of 6/18 or better before treatment), visual acuity improved in five, continued unchanged in three and deteriorated in five. Vision was therefore maintained or improved in 62% of the eyes with useful vision before treatment (95% confidence interval 32e86%). For comparison with other series, improved or stable vision rates were 78% (95% confidence interval 55e91%) and 89% (95% confidence interval 67e97%) for one and two line changes on the Snellen chart, respectively. Eight possible prognostic factors for a change in visual acuity after radiotherapy were tested in the subgroup with useful vision before treatment. Their median follow-up for visual assessment was 3.7 years (range 1.0e13.4 years) from the commencement of therapy. The results are shown in Table 3. In this small cohort with multiple statistical tests there were no statistically significant differences between the groups. However, there was a suggestion that one factor, the time from the onset of symptoms to the start of radiotherapy was different. The median time from the onset of symptoms was shorter for those with maintained or improved visual acuity compared with those where visual acuity deteriorated (18 versus 62 months, respectively, P ¼ 0.033).
G. Adams et al. / Clinical Oncology 25 (2013) 356e361
359
Table 1 Baseline characteristics of 17 patients (18 tumours) Patient’s gender Eye with tumour Age at the start of radiotherapy (years) Onset of symptoms to radiotherapy (months) Maximum tumour dimension (mm)
Male:female Right:left:bilateral Median (range) Median (range) Median (range)
2:15 8:8:1 47 (12e72) 43 (1e149) 25 (15e38)
Signs/symptoms at the time of radiotherapy Visual acuity
Pain Relative afferent pupillary defect Visual fields
Proptosis Ocular movement
Other signs
6/6 or better 6/7.5 e 6/18 No useful vision (6/60 or worse) No Yes Present Not recorded Normal Impaired Not recorded No Yes Normal Impaired Not recorded Optic disc oedema Optic disc atrophy
Of those with functional vision: Colour perception
Normal Impaired Not recorded
Discussion The results presented here represent the experience of a single institution (and a single radiation oncologist) in treating ONSMs over a 15 year period. Although technology has evolved, including a move recently to intensity-modulated radiotherapy, the vast majority of patients were treated using a 3DCRT technique with head shell and image guidance typical for brain or head and neck radiotherapy in most departments. As with other groups, our results confirm that fractionated radiotherapy will probably arrest tumour growth. Only two of 18 (11%) tumours were deemed to require biopsy by the referring ophthalmologists before radiotherapy. No attempt at surgical removal was made for any of this cohort. Therefore, the diagnosis of ONSM was
Number
% of 18
2 11 5 12 6 13 5 6 11 1 9 9 13 4 1 5 2
11% 61% 28% 67% 33% 72% 28% 33% 61% 6% 50% 50% 72% 22% 6% 28% 11%
Number
% of 13
3 7 3
23% 54% 23%
presumptive in 89% of tumours. Although this could be interpreted as a potential weakness of the data, the diagnosis of ONSM can be made by the presence of classic clinical and radiological findings, and so a biopsy is seldom required in practice [3,15]. Other series have reported variable rates of tissue diagnosis. Some had low rates (less than 20%) [5,12,13]; others intermediate (20e40%) [4,6,8], whereas a few had rates of around 50% [7,9]. This is partly explained by the fact that, in some series, there was a mixture of patients treated with either primary radiotherapy or radiation after surgery (either due to only partial debulking or relapse after attempted complete excision) [7e9,13]. Although this may represent different philosophies in management, it may also reflect different clinical or radiological features of tumours in the different series, which may have some influence on the response to
Table 2 Side effects of radiotherapy in 17 patients
Acute effects (11 patients)
Late effects (8 patients)
Hair loss Headache Otitis externa Xerostomia Conjunctivitis Xerophthalmia Cataract Optic disc atrophy
Number
% of 17
8 2 1 1 1 5 4 2
47% 12% 6% 6% 6% 29% 24% 12%
360
G. Adams et al. / Clinical Oncology 25 (2013) 356e361
Table 3 Possible prognostic factors for visual acuity after radiotherapy in 13 eyes with functional vision before radiotherapy (median and range of each variable) Prognostic factor
Improved or stable (n ¼ 8)
Worse (n ¼ 5)
P value*
Onset of symptoms to radiotherapy (months) Prescribed dose of radiotherapy (Gy) Dose to chiasm (Gy) Dose to ipsilateral retina (Gy) Maximum tumour dimension (mm) Visual acuity before radiotherapy (6/) Age at start of radiotherapy Year started radiotherapy
18 50.4 50.1 50.0 23 10.5 49.5 2006
62 50.4 51.2 50.5 26 9 44 2001
0.033 0.97 0.55 0.86 0.35 0.37 0.91 0.27
*
(1e89) (50.4e55.8) (5.0e52.0) (20.8e52.8) (17e38) (5e18) (27e61) (2002e2011)
(36e149) (46.8e55.8) (46.8e54.0) (45.0e52.0) (20e35) (5e18) (22e61) (1996e2011)
Exact P value from two-sided Mann-Whitney U test comparing the two groups.
treatment. For the two patients biopsied in our series, one had non-functioning vision before treatment and the other had stable vision on follow-up (using all three criteria). Therefore, when assessing functional results after radiotherapy, previous surgical intervention is unlikely to be a confounder in our series. We limited our primary analysis of visual function outcomes, and possible predictors, to those with useful vision before radiotherapy. We feel that this is appropriate, as, for eyes with established non-useful vision, the aim of treatment is not to restore sight but to control other symptoms such as pain and proptosis. Indeed, if the sight has already been lost, it is legitimate to relax or remove dose constraints to organs at risk such as the ipsilateral retina and nerve. Inclusion of such patients in any analysis of visual outcomes (especially the effect of dose to these structures) would seem flawed. This exclusion has been carried out in few other studies [5,8,13]. In addition, there are no clear criteria as to what constitutes a significant deterioration in vision. Most studies used a specified objective deterioration in vision. In some, this was a one line change on the Snellen chart [6], whereas in others a two line change was used [4,5,7,10]. Others have classified patients according to their own subjective reports or the non-quantified reports of their ophthalmologists [11,13]. In order to identify possible factors that may contribute to changing visual function, we chose to use a strict definition of any deterioration from baseline. Although we acknowledge that in some cases, the degree of deterioration may not significantly affect function in patients with very good eyesight, equally employing the other criteria may ignore an important deterioration in a patient with poorer but functional vision before treatment. Also, as patients were followed-up long term by their own ophthalmologist with repeat assessments over time, any deterioration tended to be noted on multiple examinations. Our results give maintenance of function rates of 62% using our criteria, but 78 and 89% using the one and two line Snellen chart criteria, respectively. Given the small numbers in all reports and these methodological differences in reporting visual outcomes, it is difficult to confidently compare results. However, a 100% tumour growth control rate in all evaluable cases, and maintaining or improving
vision in 89% using the most commonly applied measure of visual acuity outcomes, are consistent with previous reports (tumour control rates of 87e100% and stable or improved vision rates of 67e98%) [4e13]. Unfortunately, due to the retrospective nature of this study, no other findings (including pituitary function assessment) were recorded consistently enough throughout the follow-up period to allow for a detailed analysis. However, clinically significant toxicity (as reported in Table 2) was manageable. There is a suggestion from our results that the timing of radiotherapy may influence visual outcomes, as those whose vision was maintained or improved tended to be treated earlier than those whose vision deteriorated (median 18 versus 62 months, respectively). The timing of radiotherapy is controversial, with recommendations generally favouring delaying treatment if visual acuity is good and stable due to concerns about toxicity and the belief that a delay until signs of deterioration are apparent has no long-term detrimental effect. However, specific evidence for this is scant. Saeed et al. [4] found no association between the timing of radiotherapy and outcomes, whereas Abouaf et al. [10] did show a trend towards better acuity in those treated earlier. Unfortunately, no other studies have specifically addressed this issue. None of the other prognostic factors analysed showed any suggestion of influencing outcome. However, as with all other reports on radiotherapy for ONSM, the small numbers in our study mean that any analysis had low statistical power. Although our results could be interpreted as suggesting that radiotherapy should be delivered earlier rather than later, this observation should be regarded as hypothesis generating and should be validated in larger cohorts. Even if the association between timing and outcome is confirmed, it is not necessarily causal. The time from the onset of symptoms to the delivery of radiotherapy may be a surrogate for the biology of the individual tumour. Patients with a slower growing tumour (which may respond less well to radiation) may have more insidious symptoms and so present later for treatment. As well as the timing of radiotherapy, other controversies surround optimum modality and prescribed dose. Most reports have used FSRT for most of their patients [5e9,11,13]. We, together with others [10,12], have reported cohorts
G. Adams et al. / Clinical Oncology 25 (2013) 356e361
where most patients were treated with 3DCRT techniques. The multi-institutional study by Saeed et al. [4] included both modalities and found similar outcomes with either. FSRT offers advantages in terms of accuracy of delivery, smaller margins and sparing of critical structures. However, this technology is not universally available and requires specific training in order to be used safely. There are no obvious differences in control rates or toxicity between our study and those that used FSRT. However, due to the small numbers and methodological differences between studies, no definitive statement can be made on this question. In the literature, the prescribed doses have ranged widely from 45 to 64 Gy with no clear evidence of a dose response [4e13]. However, most tumours (including 78% of ours) were treated with doses between 50.4 and 54 Gy using conventional fractionation. Given the overall excellent control rates and the low risk to the optic apparatus at these doses [14], 50.4e54 Gy at 1.8 Gy per fraction would seem to be appropriate on the basis of current evidence.
Conclusion This study confirms that radiotherapy is an appropriate treatment modality for ONSMs. At our centre, it has provided control of tumour growth in all evaluable tumours and preservation of visual acuity in 62% of eyes with useful vision (89% using a less stringent criterion employed in other series). Our non-stereotactic techniques seemed to be suitable for this purpose. The study suggests that a longer time from symptom onset to radiotherapy may have a detrimental effect on the long-term preservation of vision. Therefore, early intervention with radiotherapy may be appropriate even for those patients with good visual acuity and stable radiological findings. Prospective evaluation in larger cohorts of patients is necessary for determining the optimal timing for radiotherapy delivery.
Acknowledgement The statistical analysis was carried out by J.G. Smith, PhD, AStat.
References [1] Shields JA, Shields CL, Scartozzi R. Survey of 1264 patients with orbital tumours and simulating lesions. Ophthalmology 2004;111:997e1008.
361
[2] Wright JE, Call NB, Liaricos S. Primary optic nerve sheath meningioma. Br J Ophthalmol 1980;64:553e558. [3] Carrasco JR, Penne RB. Optic nerve sheath meningiomas and advanced treatment options. Curr Opin Ophthalmol 2004; 15:406e410. [4] Saeed P, Blank L, Selva D, et al. Primary radiotherapy in progressive optic nerve sheath meningiomas: a long-term follow-up study. Br J Ophthalmol 2010;94:565e568. [5] Baumert BG, Villa S, Studer G, et al. Early improvement in vision after fractionated stereotactic radiotherapy for primary optic nerve sheath meningioma. Radiother Oncol 2004; 72:169e174. [6] Arvold ND, Lessell S, Bussiere M, et al. Visual outcomes and tumour control after conformal radiotherapy for patients with optic nerve sheath meningioma. Int J Radiat Oncol Biol Phys 2009;74:1166e1172. [7] Paulsen F, Doerr S, Wilhem H, Becker G, Bamberg M, Claben J. Fractionated stereotactic radiotherapy in patients with optic nerve sheath meningioma. Int J Radiat Oncol Biol Phys 2012;82:773e778. [8] Andrews DW, Faroozman R, Yang BP, et al. Fractionated stereotactic radiotherapy for treatment of optic nerve sheath meningiomas: preliminary observations of 33 optic nerves in 30 patients with historical comparison to observation with or without prior surgery. Neurosurgery 2002;51:890e904. [9] Milker-Zabel S, Huber P, Schlegel W, Dubus J, Zabel-du Bois A. Fractionated stereotactic radiation therapy in the management of primary optic nerve sheath meningiomas. J Neurooncol 2009;94:419e424. [10] Abouaf L, Girard N, Lefort T, et al. Standard-fractionated radiotherapy for optic nerve sheath meningioma: visual outcome is predicted by mean eye dose. Int J Radiat Oncol Biol Phys 2012;82:1268e1277. [11] Smee RI, Schneider M, Williams JR. Optic nerve sheath meningiomas e non-surgical treatment. Clin Oncol 2009;21: 8e13. [12] Tsao MN, Hoyt WF, Horton J, et al. Improved visual outcome with definitive radiation therapy for optic nerve sheath meningioma. Int J Radiat Oncol Biol Phys 1999;45(Suppl.): 324e325. [13] Solda F, Wharram B, Gunapala R, Brada M. Fractionated stereotactic conformal radiotherapy for optic nerve sheath meningiomas. Clin Oncol 2012;24:e106ee112. [14] Mayo C, Martel MK, Marks LB, Flickinger J, Nam J, Kirkpatrick J. Radiation doseevolume effects of optic nerves and chiasm. Int J Radiat Oncol Biol Phys 2010;76(Suppl.): S28eS35. [15] Kim JW, Rizzo JF, Lessell S. Controversies in the management of optic nerve sheath meningiomas. Int Ophthalmol Clin 2005;45:15e25. [16] VSN International. GenStat for Windows, 14th ed. Hemel Hempstead: VSN International, 2011.