Jonathan A. Borden, M.D., Advanced Neurosurgery, Inc. 10496 Montgomery Road Cincinnati Ohio 45242 firstname.lastname@example.org
When visual loss is associated with a mass lesion, the cause may result from either mass effect or another interference with the function of the optic apparatus. Such interference may be the result of abnormalities in the vasculature to the apparatus as a result of the tumor, or chemical factors secreted by the tumor.
When visual loss is associated purely with mass effect, and the tumor is in an accessible location, surgery is usually the best option. In many cases, particularly with optic nerve sheath meningiomas, attempts at complete resection of the lesion result in vascular damage to the optic nerves and/or chiasm. Stereotactic radiation has emerged as an effective treatment. With frequency we treat a tumor with an initial debulking, aimed at relieving mass effect, followed by stereotactic radiosurgery. With such a combined approach complex lesions are treated with low morbidity.
Traditional neurosurgical approaches
When orbital tumors are situated lateral to the optic nerve, the lateral orbitotomy may be performed with relative safety (Maroon and Kennerdell, 1984, Acciarri et al. 1995). Such an approach is undertaken for cavernous hemangiomas (Harris and Jakobiec, 1979). If the tumor is situated posterior in the cone, involvement of branches of the opthalmic nerve risks diplopia. If the tumor is large, a consequent reduction in the volume of orbital fat may cause postoperative enopathalmos unless care is taken in the orbital reconstruction. Tumors superior to the optic nerve can be approached via an orbitofrontal craniotomy. Tumors below the optic nerve can be approached via a transethmoid or transmaxillary approach in selected cases. Surgical excision of optic nerve sheath meningiomas has traditionally involved en-bloc resection (Eggers et al. 1976) but we do not advocate this given the more modern radiosurgical alternatives. When vision is present, subtotal resection has traditionally been followed by some form of radiation (Kennerdell et al, 1988)
Tumors of the Optic canal, Superior orbital fissure and sphenoid wing
Tumors such as optic nerve sheath, and or cavernous sinus meningiomas may involve the optic canal or superior orbital fissure respectively. An orbitofrontal craniotomy allows approach to such lesions (Al-Mefty, 1987, Zabramski et al. 1998) however it is difficult to achieve complete excision unless the sheath is resected and this has a significant risk of visual loss. We have generally moved toward stereotactic radiosurgery or fractionated radiation in such cases, which will be discussed below.
In the situation when a meningioma or fibrous dysplasia results in hyperostosis of the optic canal, craniotomy allows bony decompression using a high-speed diamond drill.
Meningiomas involving the sphenoid wing may actually invade the bone, greatly thickening it and causing exopthalmos. In such cases the involved bone is widely resected with a high-speed drill via a pterional or frontotemporal craniotomy (Maroon et al. 1994).
Tuberculum sella meningiomas
Meningiomas involving the tuberculum sella classically cause visual loss through involvement of an optic nerve and the anterior chiasm. Such tumors are approached from the most affected side via a pterional craniotomy. When an arachnoid plane persists between the tumor and the optic nerve and/or chiasm, the tumor can usually be safely resected. An attempt is made to aggressively resect the dural attachment and underlying hyperostotic bone with a high speed diamond drill. When the tumor is adherent to the optic apparatus, particularly in a circumferential fashion, resection can lead to a delayed visual loss presumed due to vascular compromise. When tumor fragments are left attached to the optic apparatus, stereotactic radiation is a useful adjunct.
Macroadenomas which present with visual loss are usually large and are best treated with transphenoidal resection. In the past we had used a transnasal transseptal approach to the sphenoid sinus. More recently we have adopted an endoscopic transnasal transphenoidal approach to the sella (Heilman et al. 2000, Jho and Carrau, 1996). Such tumors are often soft or liquefied and the chiasm can be easily decompressed with low morbidity. Residual or recurrent tumor is often treated with Gamma Knife radiosurgery.
Craniopharyngiomas are often treated by a combined approach. When located in an infrachiasmatic position, they are approached via a pterional or orbitofrontal craniotomy. When located in the third ventricle, particularly in the supraoptic recess, a translamina terminalis approach via a frontal craniotomy is preferred. This approach may allow removal of tumor from the optic apparatus however it is difficult to remove tumor from the more superior part of the third ventricle. A transcallosal approach allows better visualization of the upper part of the third ventricle, but from this approach it is difficult to adequately see the supraoptic recess (Yasargil et al. 1990). Aggressive resection of tumor wall from the optic apparatus and/or third ventricle risks neurologic deficit. Our approach is to remove what is easily separated from the neurologic structures, and decompress the optic apparatus. Any residual is often easily dealt with by stereotactic radiation.
Stereotactic Radiosurgical Approaches
The description of the principles and device for stereotactic neurosurgery is credited to Robert Henry Clarke and Victor Horsley in 1906. The Swedish neurosurgeon Lars Leksell developed the arc centered stereotactic system in 1949, and applied this technique to the delivery of focused radiation several years later to which he coined the term stereotactic radiosurgery. He, along with the physicist Borje Larsson introduced the Gamma Knife in 1968. The Gamma Knife employs 201 Cobalt 60 gamma emitters focused through a set of collimators onto a small point in space. (Leksell, 1968)
The advent of the CT scanner in the early 1970s and the MRI in the 1980s has revolutionized the practice of stereotactic radiosurgery allowing direct three dimensional target imaging, treatment planning and treatment. In the past, surgeons have been loath to use radiosurgery for tumors of the optic apparatus due to concern of visual loss. With recent improvements in MRI resolution and spatial accuracy, we now safely treat lesions involving the optic nerves and chiasm, and are able to reverse visual loss in selected cases. Before treating such cases it is essential to verify the spatial accuracy of the MRI because magnetic field inhomogeneities can result in significant image warp and other errors.
Gamma Knife Radiosurgery
Tumors of the Globe
Choroidal/uveal melanomas that are too large or otherwise not appropriate for plaque therapy may be treated with Gamma Knife radiosurgery (Rand et al. 1987, Logani et al. 1993, Langmann et al. 2000) . Our practice is to paralyze the affected eye with a Marcaine block prior to MRI and during treatment. We generally prescribe 20 Gy to the 50 % isodose margin though smaller tumors may be safely treated with 25 or even 30 Gy. We have not seen a significant rate of disease progression at 20 Gy within the first 2 years but await longer term results. Our morbidity has been limited to several instances of self limited eyelid erythema and/or ulceration in large anteriorly placed tumors. We take care to limit the dose to the lacrimal gland when possible.
Benign tumors of the orbit
Gamma Knife radiosurgery has been recently used to treat optic nerve sheath meningiomas (Klink et al. 1998), cavernous hemangiomas (Thompson et al. 2000) and neurofibromas of the orbit . We have seen several instances of surprising and dramatic recovery of visual loss due to presumed meningiomas. Such recovery of vision has been seen without change in size of the tumor leading us to suspect that visual loss may result from factors other than direct compression. It is not known how radiosurgery achieves this benefit. Such tumors have been treated with 12 – 13 Gy in a single fraction. Great care is taken to three dimensionally reconstruct the tumor and optic nerve and to produce a highly conformal plan. Short segments of the optic nerve appear able to tolerate 12 Gy at least in certain cases. In all such cases, the patients presented with poor vision in the affected eye and were willing to sacrifice remaining vision in that eye in order to maximize chances of tumor control.
In contrast, when we attempt to preserve useful or normal vision, we limit optic nerve dose to 10 Gy and have had no incidence of visual loss at this dose (Leber et al. 1998).
When treating cavernous hemangiomas presenting with near complete and long-standing visual loss we have prescribed 18 to 20 Gy. In such cases there has been a relatively early and dramatic reduction in tumor size, but vision has not returned and often any remnant of light perception is lost. Based on this early and dramatic response to radiosurgery at 18 Gy, we are confident that we can achieve control of cavernous hemangiomas at lower radiation doses, and have begun dose de-escalation particularly when useful vision remains.
Meningiomas involving the superior orbital fissure and cavernous sinus
When such meningiomas present with visual loss we prefer surgical decompression prior to Gamma Knife radiosurgery. Attempts to cure such tumors with aggressive skull base surgery results in further neurologic deficits. Using a combined craniotomy/radiosurgical approach we are able to effectively treat the most complex tumors with minimal morbidity. We prefer to prescribe 13 Gy to the 50 % isodose margin for meningiomas. For large meningiomas that continue to demonstrate optic apparatus compression despite craniotomy, our practice is to limit dose to the optic nerves and chiasm to 10 Gy (Morita et al,. 1999) which may result in a very small portion of the tumor that receives an inadequate treatment. After the large bulk of the tumor decompresses, and the optic compression resolves, and if this portion of tumor regrows, it can be safely retreated with a latter Gamma Knife procedure.
In order to minimize dose to the optic nerves and chiasm, we employ multiple small isocenters, and aggressively "block" any radiation beams that may pass through the chiasm. Such treatments are often complex and lengthy but are required for the most difficult lesions, and ones that have failed prior fractionated radiotherapy.
When pituitary macroadenomas present with visual loss our preference is to decompress the optic chiasm via endoscopic transnasal transphenoidal resection. Following surgery we often wait several months for the residual tumor to collapse away from the chiasm. When the tumor capsule is two millimeters away from the chiasm it is generally safe in most all cases to treat with Gamma Knife radiosurgery. In selected cases when there is no distance between tumor and chiasm, particularly when the amount of tumor adjacent to the chiasm is a small percentage of the total tumor, we limit to 10 Gy, dose to this small portion of tumor. Gamma Knife radiosurgery of tumor extending into the cavernous sinus is generally safe with very low risk of injury to cranial nerves. Diplopia due to cranial nerve involvement in the cavernous sinus is more likely to improve following radiosurgery than to worsen.
For hormone secreting tumors we generally prescribe 20 Gy to the 50 % isodose margin. Non-hormone secreting tumors require a lower dose, often 14 Gy, in order to arrest growth. We have performed radiosurgery in cases that have progressed despite craniotomy and prior fractionated radiation. In such cases we limit dose to the chiasm even further depending on the time between prior fractionated radiation and radiosurgery. In such cases we have limited chiasm dose to a maximum of 4 to 9 Gy depending on the details of the case.
Craniopharyngiomas appear exquisitely sensitive to radiosurgery and even though they are often intimately involved in the optic apparatus a marginal dose of 9 Gy appears effective. Care must be taken to ensure the optic apparatus is at the margin of the tumor and detailed MRI imaging is essential in such situations. We generally combine a three dimensional post gadolinium enhanced T1 weighted sequence (e.g. MPRAGE) with a three dimensional T2 weighted sequence such as CISS. Even with single fraction treatment of 9 Gy, surprisingly rapid tumor shrinkage is seen, often within a number of months. Treatment of lesions of the optic apparatus with Gamma Knife is a relatively new indication (Chung et al. 2000). We await long term follow-up data on tumor control.
Given the apparent excellent response and low morbidity of low dose radiosurgery overly aggressive surgery for craniopharyngiomas is becoming a less frequent indication.
Studies of chordomas suggest that marginal radiosurgical dose directly affects long term tumor control (Austin et al., 1993). Consequently craniotomy is recommended to remove tumor from near the optic structures and away from the brainstem. Using single fraction Gamma Knife radiosurgery we prescribe a 20 Gy marginal dose (Kondziolka et al., 1991). When tumor remains close to the optic apparatus, dose escalation with a fractionated regime of intensity modulated radiotherapy (IMRT) is performed.
Fractionated stereotactic radiotherapy
In cases when a small residual tumor is directly attached to the optic chiasm, we treat with three dimensional conformal fractionated intensity modulated radiotherapy (IMRT) using mask immobilization (Tsai et al., 1999). IMRT is a relatively new radiation technique in which the photon beam emanating from a linear accelerator is modulated and shaped by a multileafed collimator as the accelerator travels in an arc around the patient’s head. Such techniques require sophisticated computer planning and control of the metallic leaves during the treatment. The result is an improved conformality of radiation dose to the desired target.
It is well known that the optic chiasm can tolerate a much higher total dose of radiation when given in small doses over a six-week period. However, the effect on the tumor is also lessened. The benefits of fractionation depend on a selective response of the tumor in contrast to the surrounding normal tissues. For malignant tumors, particularly when portions of the tumor is hypoxic, this response is well described and fractionation does appear to be beneficial. For well vascularized, slowly growing, benign tumors, this selective response is less clear. This issue remains one of controversy. Unfortunately there is no definitive long-term study that compares both tumor control and complication rate between Gamma Knife radiosurgery and IMRT.
Other characteristics of the different radiation delivery systems also come into play and the decision to treat with one or another method involves a number of technical factors. We have developed a model that allows comparison of certain selected variables between different stereotactic radiation systems yielding an objective ‘quality factor’ for radiosurgical dose plans (Borden et al. 2000). Based on this study and other factors, we no longer employ single fraction stereotactic radiosurgery with our linear accelerator system, treating either with single fraction Gamma Knife radiosurgery, or fractionated IMRT as the individual case warrants. Generally IMRT is superior for very large tumors, ones too large to be safely treated with Gamma Knife radiosurgery, and ones not amenable to surgical debulking.
Optic nerve astrocytomas
When tumor invades the optic nerve, fractionated radiation (IMRT) is the best treatment. Astrocytomas demonstrated improved control with increased radiation dose. The tolerance of the optic chiasm and surrounding hypothalamus are the limiting factors. Surgical biopsy to establish pathology is recommended. Due to the infiltrative nature of the astrocytoma, radical surgical resection is not curative and hence is not warranted. Such tumors often invade both the hypothalamus and optic chiasm. Chemotherapy is occasionally useful as an adjunct. We have had some recent success with the oral agent Temozolomide (O’Reilly et al. 1993), which has relatively low side effects, and is often prescribed on a long term basis when effective. Our treatment program emphasizes maximal quality of life. Since no current treatments have been demonstrated curative, the risks of aggressive treatments must be counterbalances against the risks of debilitating neurologic deficits.
In selected cases, particularly for the case of a prolactin secreting pituitary adenoma, and in which the visual loss has been mild or very gradual, medication such as bromocriptine can have a surprisingly rapid and significant effect. Such cases are the exception rather than the rule.
We have not had success with chemotheraputic agents for meningiomas such as hydroxyurea or RU-486.
The role of chemotheraputic agents for the treatment of optic nerve astrocytomas is discussed above.
Treatment of tumors involving the optic nerves and chiasm can be successful and spare vision. We have had the best success using a combined approach involving skull base approaches, Gamma Knife radiosurgery and IMRT and the case warrants. Gamma Knife radiosurgery is a promising emerging treatment modality for such tumors.
Acciarri N, Giulioni M, Padovani R, Gaist G, Pozzati E, Acciarri R.
Orbital cavernous angiomas: surgical experience on a series of 13 cases.
J Neurosurg Sci. 1995 Dec;39(4):203-9.
Supraorbital-pterional approach to skull base lesions.
Neurosurgery. 1987 Oct;21(4):474-7
Austin JP, Urie MM, Cardenosa G, Munzenrider JE. Probable causes of recurrence in patients with chordoma and chondrosarcoma of the base of skull and cervical spine.
Int J Radiat Oncol Biol Phys. 1993 Feb 15;25(3):439-44.
Chung WY, Pan DH, Shiau CY, Guo WY, Wang LW. Gamma knife radiosurgery for craniopharyngiomas.
J Neurosurg. 2000 Dec;93 Suppl 3:47-56.
Eggers H, Jakobiec FA, Jones IS.
Tumors of the optic nerve.
Doc Ophthalmol. 1976 Apr 28;41(1):43-128.
Harris GJ, Jakobiec FA.
Cavernous hemangioma of the orbit.
J Neurosurg. 1979 Aug;51(2):219-28
Heilman CB, Shucart WA, Rebeiz EE, Gopal H.
Endoscopic pituitary surgery.
Clin Neurosurg. 2000;46:507-14
Jho HD, Carrau RL.
Endoscopy assisted transsphenoidal surgery for pituitary adenoma. Technical note.
Acta Neurochir (Wien). 1996;138(12):1416-25.
Kennerdell JS, Maroon JC, Malton M, Warren FA.
The management of optic nerve sheath meningiomas.
Am J Ophthalmol. 1988 Oct 15;106(4):450-7.
Klink DF, Miller NR, Williams J. Preservation of residual vision 2 years after stereotactic radiosurgery for a presumed optic nerve sheath meningioma.
J Neuroophthalmol. 1998 Jun;18(2):117-20
Kondziolka D, Lunsford LD, Flickinger JC. The role of radiosurgery in the management of chordoma and chondrosarcoma of the cranial base.
Neurosurgery. 1991 Jul;29(1):38-45; discussion 45-6
Langmann G, Pendl G, Klaus-Mulln er, Papaefthymiou G, Guss H. Gamma knife radiosurgery for uveal melanomas: an 8-year experience.
J Neurosurg. 2000 Dec;93 Suppl 3:184-8
Leber KA, Bergloff J, Pendl G. Dose-response tolerance of the visual pathways and cranial nerves of the cavernous sinus to stereotactic radiosurgery.
J Neurosurg. 1998 Jan;88(1):43-50.
Leksell L. Cerebral radiosurgery. I. Gammathalanotomy in two cases of intractable pain.
Acta Chir Scand. 1968;134(8):585-95
Logani S, Helenowski TK, Thakrar H, Pothiawala B. Gamma Knife radiosurgery in the treatment of ocular melanoma. Stereotact Funct Neurosurg. 1993;61 Suppl 1:38-44
Maroon JC, Kennerdell JS.
Surgical approaches to the orbit. Indications and techniques.
J Neurosurg. 1984 Jun;60(6):1226-35
Maroon JC, Kennerdell JS, Vidovich DV, Abla A, Sternau L.
Recurrent spheno-orbital meningioma.
J Neurosurg. 1994 Feb;80(2):202-8.
Morita A, Coffey RJ, Foote RL, Schiff D, Gorman D. Risk of injury to cranial nerves after gamma knife radiosurgery for skull base meningiomas: experience in 88 patients.
J Neurosurg. 1999 Jan;90(1):42-9
O'Reilly SM, Newlands ES, Glaser MG, Brampton M, Rice-Edwards JM, Illingworth RD, Richards PG, Kennard C, Colquhoun IR, Lewis P, et al. Temozolomide: a new oral cytotoxic chemotherapeutic agent with promising activity against primary brain tumours.
Eur J Cancer. 1993;29A(7):940-2
Rand RW, Khonsary A, Brown WJ, Winter J, Snow HD. Leksell stereotactic radiosurgery in the treatment of eye melanoma.
Neurol Res. 1987 Jun;9(2):142-6.
Thompson TP, Lunsford LD, Flickinger JC. Radiosurgery for hemangiomas of the cavernous sinus and orbit: technical case report.
Neurosurgery. 2000 Sep;47(3):778-83
Tsai JS, Engler MJ, Ling MN, Wu JK, Kramer B, Dipetrillo T, Wazer DE. A non-invasive immobilization system and related quality assurance for dynamic intensity modulated radiation therapy of intracranial and head and neck disease.
Int J Radiat Oncol Biol Phys. 1999 Jan 15;43(2):455-67.
Yasargil MG, Curcic M, Kis M, Siegenthaler G, Teddy PJ, Roth P. Total removal of craniopharyngiomas. Approaches and long-term results in 144 patients.
J Neurosurg. 1990 Jul;73(1):3-11
Zabramski JM, Kiris T, Sankhla SK, Cabiol J, Spetzler RF.
Orbitozygomatic craniotomy. Technical note.
J Neurosurg. 1998 Aug;89(2):336-41