Chapter 46 Radiosurgery of Central Nervous System Tumors
• Randomized clinical trials confirm the survival benefit of stereotactic radiosurgery for patients with single brain metastases and the local control benefit for patients with multiple (two to four) tumors.
• Four matched cohort studies (class II evidence) that compare Gamma Knife radiosurgery to resection for patients with acoustic neuromas less than 2.5 cm in diameter show improved outcomes after radiosurgery with lower morbidity rate. Indications for microsurgery at institutions with radiosurgery expertise include disabling symptomatic brainstem compression, severe headache, hydrocephalus, trigeminal neuralgia, and patient choice.
• Long-term data are available over the past 10 years from numerous institutions on the value of stereotactic radiosurgery for different benign intracranial tumors.
Stereotactic radiosurgery (SRS) has become one of the most important concepts in the management of patients with central nervous system tumors. The neurosurgeon uses SRS to destroy a tumor using precise, image-guided ionizing irradiation in a single procedure.1 Over the past two decades, radiosurgery has supplanted fractionated radiation therapy in the care of many tumors, and has forced an evolution of thinking regarding the roles of resection or observation. At a biological level, radiosurgery can halt cell division, cause vascular occlusion, induce apoptosis or necrosis, and affect blood-brain barrier integrity.2–6
Radiosurgical devices include the cobalt-60 Gamma Knife, modified linear accelerator systems, and charged particle generators.7–10 Although initially designed for functional neurosurgery, the use of brain tumor radiosurgery took off in the era of higher-resolution parenchymal brain imaging with magnetic resonance imaging (MRI). Now robotic devices such as the Perfexion or model 4C gamma units or the Cyberknife facilitate dose planning and delivery.11,12 For use in the brain, accuracy and reliability are crucial. At present no pharmacological agents are used to modify the target response to radiosurgery.13,14 A listing of indications is shown in Table 46.1.
Diagnosis | Number of Procedures |
---|---|
Vestibular schwannoma | 1439 |
Trigeminal schwannoma | 45 |
Other schwannomas | 65 |
Meningioma | 1377 |
Pituitary tumor | 300 |
Craniopharyngioma | 76 |
Hemangioblastoma | 50 |
Hemangiopericytoma | 41 |
Glomus tumor | 24 |
Pineocytoma | 16 |
Malignant pineal tumor | 13 |
Chordoma | 31 |
Chondrosarcoma | 25 |
Choroid plexus papilloma | 12 |
Hemangioma | 7 |
Glioblastoma multiforme | 349 |
Anaplastic astrocytoma | 134 |
Fibrillary astrocytoma | 43 |
Mixed glioma | 78 |
Pilocytic astrocytoma | 89 |
Ependymoma | 74 |
Medulloblastoma | 24 |
CNS lymphoma | 12 |
Hypothalamic hamartoma | 6 |
Brain metastasis | 3264 |
Invasive skull base tumor | 32 |
Other tumors | 66 |
Total | 7692 |
CNS, central nervous system.
∗ Data from clinical series at the University of Pittsburgh: total procedures = 10,158; total brain tumors = 7692.
Not all radiosurgery systems are the same and they differ in both hardware and software. One should seek a system that allows the surgeon to efficiently create highly conformal and selective volumetric dose plans for irregular lesion volumes, because tumors are rarely “spherical.”15 The steep falloff of radiation into the surrounding structures (selectivity) helps to achieve safety. The location of many tumors either within or adjacent to critical brain or nerve makes the requirement for conformal and selective radiosurgery paramount.
Radiosurgery or radiotherapy?
Fractionated stereotactic irradiation has been used by some centers to treat benign tumors,16,17 and of course, large-field conventional or whole-brain radiotherapy has been used for decades. Any advantage for the use of fractionated radiotherapy becomes important when large volumes of sensitive surrounding normal tissue need to be included in the treatment volume, such as with the standard 2-cm margin around a malignant glioma. If the volume of normal tissue irradiated inside or outside the target volume is small, then fractionated radiation may be of no additional value. Certain tumors appear to be more resistant to traditional radiotherapy and include meningiomas, schwannomas, and some metastatic tumors such as melanoma, sarcoma, and renal cell carcinoma. The radiobiological response can be addressed with the α/β ratio. A lower number refers to later responding tissue.18–20
Vestibular Schwannomas
The goals of vestibular schwannoma radiosurgery are to prevent tumor growth, preserve cochlear and other cranial nerve function, maintain general function and quality of life, and avoid the risks associated with open surgical resection. Available long-term data have established radiosurgery as an important alternative to resection. Radiosurgery was first offered to patients who were elderly or medically infirm, but with experience, was found ideal for patients of all ages.21–24 We have found consistent results across age groups.21,25
To date we have managed 1397 patients with vestibular schwannomas using Gamma Knife radiosurgery (Fig. 46.1). The mean patient age in our series was 57 years (range, 12 to 95). Eight percent had neurofibromatosis (93 patients). Symptoms before radiosurgery included hearing loss (92%), balance symptoms or ataxia (51%), tinnitus (43%), or other neurological deficit (19.5%). Thirty-four percent of our patients had useful hearing, Gardner-Robertson grade I (speech discrimination score ≥70%; pure tone average ≤30 dB) or grade II (speech discrimination score ≥50%; pure tone average ≤50 dB). Since 1992, the average dose prescribed to the tumor margin was 13 Gy. The 50% isodose line was used in 90% of patients.
Our long-term study documented a 98% clinical tumor control rate (no requirement for surgical intervention) at 5 to 10 years.21,23 Since the institution of MRI-guided dose planning, there has been a significant reduction in morbidity.26,27 Currently, the risk for any grade delayed facial nerve dysfunction is below 1%.26,27 Patients with useful hearing before radiosurgery continue to report an approximate 75% overall rate for maintenance of useful hearing, depending on tumor size, with even better results for intracanalicular tumors.27–29 In comparison studies, radiosurgery has been shown to be a cost-effective alternative to microsurgery for these patients.
For smaller tumors, it is likely that more patients now receive radiosurgery as primary care. No randomized clinical trials have been conducted, but there are now four matched cohort studies (class II evidence). These studies evaluated patients with similar sized tumors, and evaluated clinical, imaging, and quality of life outcomes. All showed better results after radiosurgery for most clinical measures, similar results for the symptoms of tinnitus and imbalance, and similar freedom from tumor progression rates.30–33 It is important to tell patients that there can sometimes be a transient expansion of the tumor capsule after radiosurgery and that it usually can be observed without further treatment.34,35 The value of radiosurgery in neurofibromatosis type 2 has been studied.36 Based on these data, we believe that there are few remaining indications for surgical resection in a patient with a small to moderate size tumor. These indications include disabling symptomatic brainstem compression, severe headache, hydrocephalus, trigeminal neuralgia, and patient choice. Radiosurgery has also been performed and evaluated for patients with other cranial nerve schwannomas (Fig. 46.2).
Meningiomas
Radiosurgery has transformed the management of intracranial meningiomas, particularly for tumors of the skull base.37 As for most indications, radiosurgery was first considered for residual or recurrent tumors after prior resection.38 The steep radiation falloff can be directly conformed to the well-defined tumor margin. The role of aggressive skull base surgery has waned as reports showing excellent clinical outcomes for small basal tumors have been published.39,40 Larger tumors with mass effect benefit from subtotal resection followed by radiosurgery if complete resection is not feasible. However, because of their more aggressive nature, atypical or anaplastic/malignant meningiomas (WHO [World Health Organization] grade II or III) are best managed with complete resection followed by fractionated radiotherapy because of their tendency to extend beyond the borders seen on imaging.41 Radiosurgery also is of value in the setting of incomplete resection.
Our 22-year experience includes 1302 intracranial meningiomas. Recently, we published data from 972 patients with 1045 meningiomas (Fig. 46.3).42 Half of the patients had undergone a prior resection and the average age was 57 years. Tumor locations included middle fossa (351 patients), posterior fossa (307), convexity (126), anterior fossa (88), parasagittal region (113), or other sites (115). Follow-up over the past 5, 7, 10, and 12 years was obtained in 327, 190, 90, and 41 patients, respectively.
The control rate for patients who had radiosurgery for known WHO grade I (benign) meningiomas (after prior resection) was 93%. Primary radiosurgery patients (no prior histological confirmation; n = 482), had a tumor control rate of 97%. Adjuvant radiosurgery for patients with WHO grades II and III tumors had tumor control rates of 50% and 17%, respectively. Delayed resection after radiosurgery was performed in 51 patients (5%) at a mean period of 3 years. Additional radiosurgery was performed in 41 patients, usually for new tumors. At 10 years or more, adjuvant grade I tumors were controlled in 91% (n = 53), and primary tumors in 95% (n = 22). There was no case of a subsequent radiation-induced tumor. The morbidity rate was 7.7%. Most centers restrict the dose received by the optic nerve or chiasm to 8 to 9 Gy or less, which should keep the risk for delayed radiation-related optic neuropathy very low.
We believe that SRS has changed meningioma management significantly. Rather than performing a subtotal resection and “following the patient,” we now advocate postoperative radiosurgery to reduce the risk of delayed progression.43–45