Welcome to the American Association of Neurological Surgeons continuing medical education presentation on metastatic brain tumors. During this presentation, we will cover the epidemiology and pathogenesis of these tumors, as well as initial management and treatment options for patients with metastatic brain tumors. Bear in mind that these treatment options are often employed in a multimodal approach over the course of a given patient's management. Current estimates place the number of new brain metastases in the U.S. around 300,000 per year, accounting for more than half of all brain tumors. Although many cases may be unrecognized prior to death, up to 40% of patients who will die of cancer will have brain metastases on autopsy. Fifteen to 20% of the time, a brain tumor is the first sign of systemic cancer. These tumors are termed precocious metastases. In 9% of cases, the brain tumor is the only indication of cancer at the time of diagnosis. The incidence peaks in the 50s to 70s, along with cancer peaks, and the development of a brain metastasis seems to be independent of gender, with perhaps a slight male predisposition in melanoma. The primary source of brain metastases has been studied several times with reproducible results which are exemplified by Nussbaum's study of 1996. Lung cancers account for the primary source in nearly half of all cases, followed by breast cancer at 15-20%, melanoma at 10%, and then renal and GI sources. When considering the likelihood of a given primary source of cancer to invade the CNS, lung cancer again tops the list with a 1-year cumulative incidence of approximately 15%. Breast cancer, although a common source for brain metastases, is less likely to invade the CNS than melanoma and renal cell carcinoma. Since the vast majority of metastases occur through hematogenous spread, the distribution of brain metastases in the brain is proportioned to the blood flow in a given area of the brain. 80% are located in the cerebral hemispheres, 15% in the cerebellum, and another 1-3% in the brainstem. The distribution of the middle cerebral artery is the most common location, and the tumors are often located at the gray-white junction. This diagram depicts the proportion of metastases by anatomical brain lobe. Again, the proportions follow blood flow and delivery. The pathogenesis behind the development of a brain metastasis involves a seed that escapes from the primary tumor and survives travel in the bloodstream to the CNS. Once in the CNS, the seed must escape the vasculature and develop an expanding tumor clone, thought to be in the Virchow-Robin space. From here, cells may escape and see the CSF. It is estimated that a cancerous tumor 1 cm in size can shed millions of tumor cells into the circulation per day. These cells may enter the brain through the carotid or vertebrobasilar circulation or through the valveless veins of the venous plexus. There are microenvironmental changes in the development of brain metastases, perhaps most notably the development of neoplastic blood vessels, but neoplastic angiogenesis is defective with an imperfect blood-brain barrier which allows radiographic contrast to leak through as well as contributes to the development of vasogenic edema, a hallmark of brain metastases. Unfortunately, the leaky blood-brain barrier does not seem to allow the entry of chemotherapy to any useful degree. There are many molecular mechanisms involved or potentially involved in brain metastasis development. E-cadherin-catenin complex is important for maintenance of cytoarchitecture. Tumors have disordered cytoarchitecture and cells can break away easily and enter the vascular system. Decreased expression of the E-cadherin-catenin complex has been correlated with the development of brain metastasis and poor prognosis in cancer patients. Tumor clones that have invaded the vasculature arrest and then extravasate from vessels in the brain. Survival of these clones depends on angiogenesis mediated through a number of growth factors and signals. In 1997, the Radiation Therapy Oncology Group published a recursive partitioning analysis in an effort to stratify the risk of patients with brain metastases. The factors that were identified as important were Karnofsky performance score with a cutoff of 70, patient age with a cutoff of 65, and the presence or absence of active systemic disease. Just as a reminder, KPS of 70 is the cutoff for functional independence, with patients under a KPS of 70 requiring assistance for self-care. What the RTOG RPA found was that in the best groups of patients, that is patients under 65 who were functionally independent with no active systemic disease, the median survival was 7.1 months. For functionally independent patients over the age of 65 and or with uncontrolled systemic disease, the median survival was 4.2 months. And for patients who had lost their functional independence, median survival was 2.3 months. This was validated and adopted quickly. However, there was a concern that the historical data pulled from patients enrolled in RTOG trials were outdated and no longer reflected current advances in the treatment of brain metastases. In 2008, the RTOG reported its Graded Prognostic Assessment, GPA score, developed from a database of 1,960 patients accrued to four RTOG trials. In this system, the factors impacting survival were age, Karnofsky performance score, number of CNS lesions, and the presence or absence of extracranial metastases. Each patient was assigned a score of 0, .5, or 1 for each factor, and the sum determined the GPA score. In the very best of circumstances, that is, in a patient under the age of 50 with a KPS of 90 to 100, a single brain lesion, and no extracranial metastases, the median survival was 11 months from diagnosis. An analysis of more than 4,000 additional patients led to the development of disease-specific GPA scales for specific subtypes of cancer, such as lung cancer, breast cancer, melanoma, and others. These were published by the RTOG in 2011. Along with corresponding median survival times based on GPA scores, these scales were able to better stratify patient risk based on individual disease circumstances. Clinical management of these patients consists of the usual necessary supportive care dictated by the clinical situation. Corticosteroids are a mainstay of early care for patients with metastatic brain tumors and are very effective at providing symptomatic relief from tumor-associated vasogenic edema. The steroid dose should be tapered when possible to minimize the development of side effects. Anticonvulsants are used only if patients have suffered a seizure. Phenytoin and carbamazepine are effective and inexpensive. Levotiracetam is frequently used as well because, unlike the other two, it does not induce the cytochrome P450 pathway. Induction of this pathway can have implications on systemic chemotherapy metabolism in cancer patients. In addition, patients with metastatic brain tumors should undergo a staging workup, particularly if one has not been done recently or if the brain tumor is the initial presenting sign of cancer. Treatment options for patients with brain metastases include observation, radiotherapy, radiosurgery, chemotherapy, and surgery. These options are often used in combination or in series in the management of these patients. Whole brain radiation therapy is a longstanding mainstay in the treatment of brain metastases. It has been used for many years and has been demonstrated to provide a benefit in CNS disease control and survival compared to corticosteroids alone. Fractionation schemes vary, but typically patients receive 20 to 40 gray delivered in daily fractions over 1 to 4 weeks, resulting in a median survival time of 4 to 6 months overall. The optimal dose fractionation schedule has been studied by the RTOG using linear accelerator and COBALT-60 equipment. This involved two studies reported together in which five different fractionation schemes were used. Median survival times were 15 and 18 weeks, with no difference detected between fractionation schedules. Neurological function was noted to improve in 50% of patients in the short term. Based on these data, most whole brain radiation therapy is delivered at 30 gray in 10 fractions. In 1998, PATCHEL reported a well-designed trial of postoperative radiotherapy in the treatment of patients with single brain metastases. In PATCHEL's study, the addition of whole brain radiation therapy after surgery was shown to be beneficial by reducing the rate of CNS disease progression from 70 to 18%, and specifically reducing the rate of recurrence at the surgical site from 46 to 10%. Overall survival was not impacted by the addition of whole brain radiation therapy, but this was explained by considerable crossover between the treatment groups. Overall, lung cancer patients develop brain metastases 15% of the time. For small cell lung cancer in particular, the rate of CNS metastasis is higher, approximately 10% at diagnosis and 25% overall. The Prophylactic Cranial Irradiation Overview Collaboration Group reported a meta-analysis of 987 patients with small cell lung cancer in complete remission. Patients treated with Prophylactic Cranial Irradiation had a 3-year survival of 20.7% compared to 15.3% for the untreated group. Whole brain radiation therapy also significantly decreased CNS disease recurrence and the cumulative incidence of brain metastases in patients with small cell lung cancer. Based largely on these findings, Prophylactic Cranial Irradiation has become standard of care for patients with small cell lung cancer. Certain histological subtypes of brain metastases are particularly sensitive to ionizing radiation. Histologies typically considered particularly sensitive to radiation include small cell lung cancer, germ cell tumors, lymphoma, leukemia, and multiple myeloma. Other histologic subtypes of tumors are considered relatively insensitive to radiation, specifically fractionated radiation. Melanoma and renal cell carcinoma are considered relatively radio-resistant and respond less reliably to radiation. Notably, these tumors seem to respond as well to single-fraction radiation, or radiosurgery, as histologic subtypes considered to be radiosensitive. Radiation in general induces histologic changes in normal brain tissue. In the acute period, sharply demarcated areas of coagulation necrosis are seen, with endothelial injury and fibrinoid vascular changes. In the subacute period, these areas of necrosis are invaded with macrophages, reactive gliosis develops, and proliferative vasculopathy begins. In the more chronic period, the necrotic areas are replaced by scar and hyaline infiltrates, and vascular reaction progresses with lumen obliteration. The clinical side effects of whole-brain radiation therapy also show an evolution over time. Acute effects include fatigue, hair loss, and scalp erythema and hyperpigmentation. Three to 10 weeks after therapy, patients frequently experience fatigue, anorexia, and irritability in what is referred to as somnolence syndrome. In the longer term, a number of patients develop a significant cognitive decline, which may be correlated with imaging findings. The risk of cognitive impairment seems to be higher with greater daily fractions. This slide demonstrates imaging changes in a patient treated with whole-brain radiation therapy. The images in A are pretreatment T2-weighted images, and the images in B are 46 months after treatment, demonstrating radiation-related leukoencephalopathy. It's also noteworthy that this patient is alive nearly four years after treatment with whole-brain radiation therapy. Neurocognitive function is correlated to quality of life in patients with brain metastases. In 2008, Leigh showed that whole-brain radiation therapy prolongs the time to neurocognitive function decline. A previous study classified significant neurocognitive decline as a three-point drop in the Mini Mental Status Examination. Compared to patients undergoing stereotactic radiosurgery alone, patients treated with stereotactic radiosurgery and whole-brain radiation therapy had less cognitive decline at one year, 25 versus 41 percent, but a dramatically higher rate of decline at three years, 85 percent versus 48 percent. The one-year results indicated that whole-brain radiation therapy likely controlled CNS disease progression, and progression of CNS disease is known to negatively impact neurocognitive function. In 2009, Chang studied stereotactic radiosurgery with or without whole-brain radiation therapy in patients with one to three brain metastases. The study was halted early due to interim findings. Neurocognitive functional decline was seen at four months in only 22 percent of patients treated with radiosurgery alone, compared to 49 percent of patients treated with whole-brain radiation therapy and stereotactic radiosurgery in combination. It seems relatively clear that whole-brain radiation therapy contributes to early neurocognitive decline manifested mainly in memory. Longer-term decline in patients surviving longer than 12 months is poorly quantified. This issue can be clouded by disease progression and the effects of systemic treatment. Again, control of overall CNS disease is critical for preservation of neurocognitive function. Despite growing recognition of its limitations and neurocognitive functional risks, whole-brain radiation therapy remains a mainstay in the treatment of brain metastases. In an effort to mitigate the cognitive impact of radiation therapy, the RTOG is currently undertaking a trial of whole-brain radiation therapy using hippocampal sparing techniques. Surgery also continues to play a central role in the management of patients with brain metastases. Clinical studies have confirmed the benefits of surgery with whole-brain radiation therapy when compared to whole-brain radiation therapy alone. Surgery provides the ability to address the mass effect caused by a tumor and to provide tissue for pathological diagnosis. A patchel study of 1990 demonstrated that surgery plus whole-brain radiation therapy improved local control, functional independence, and survival when compared to whole-brain radiation therapy alone. Clinical consideration in patients with metastatic brain tumors are similar to patients with other brain masses, but one must consider the extent of the patient's systemic disease as a part of this general health evaluation, in addition to their RPA class and or GPA score as predictors of survival. Furthermore, the size, number, and location of metastatic lesions are a consideration for surgical planning. Mass effect is caused by tumor bulk and peritumoral edema. Based on size alone, some tumors should be removed to address the tumor mass. There have been no studies analyzing the relationship between tumor size and surgery in terms of outcomes like survival or local recurrence. Local recurrence rates are high, even with MRI-proven complete resection, as reported in Patchell's 1998 study, with a 46% local recurrence rate with surgery alone. For patients with multiple metastatic tumors, the role of surgery is somewhat less clear. A 1993 study evaluated 56 patients with two or three brain metastases. Three groups were compared, one group in which one or more lesions was left unresected, a second group in which all lesions were resected, and a third group was the match control group of patients with a single, completely resected metastatic tumor. There were no differences in surgical mortality or morbidity. Patients in group B, the group with multiple metastases, all of which were fully resected, had a significantly longer survival time than patients with some unresected lesions. The group B survival time was also similar to patients with a single, fully resected metastatic tumor. This data suggests that the resection of multiple metastases is as effective as resection of a single metastasis, as long as all metastatic tumors are resected. In 2003, Pollock reviewed 52 patients who underwent radiosurgery or surgery for tumor progression after whole brain radiation therapy. Multivariate analysis showed that survival correlated only with RPA class and that the choice of a treatment option and number of metastases are less important than the general health and performance score of the patient and overall disease control. The location of the tumor impacts anticipated surgical risk, as with other pathologies. These considerations are well known to this audience and beyond the scope of this presentation. Clearly, there are risks associated with resection of brain metastases. New postoperative neurological deficits following surgery for glioblastoma have been correlated with decreased survival. The degree of neurological impairment in cancer patients has been correlated with outcome as well, and neurological impairment has a negative impact on quality of life, which has again been correlated with outcome and survival in cancer patients. However, surgery for metastatic brain tumors may also result in improved neurological function by alleviating tumor mass effect. A 2002 study found that 59% of patients had improved neurological function following surgery, whereas only 9% were worse. Stereotactic radiosurgery, a technique commonly employed in the management of metastatic brain tumors, is a term first coined by Lars Lexell in 1951 as the closed skull destruction of an intracranial target using ionizing radiation. Stereotactic radiosurgery is the delivery of a single fraction of radiation using image guidance allowing for a high target dose and a low dose to the surrounding tissues. This has a much different radiobiological effect than conventional fractionated radiotherapy. There are several radiosurgery systems currently employed. Charged particle beam or proton beam irradiation is effective, but there are financial and logistical constraints to these systems. Linear accelerator based systems are currently used in many centers. These must be specifically calibrated for stereotactic radiosurgery. There is a moving radiation source in these systems with a potential for inaccuracy. The Gamma Knife system is commonly employed as well using cobalt-60 photons in a highly accurate rigid and frame-based system. There are some advantages to the use of stereotactic radiosurgery in the management of metastatic tumors. Treatments can be delivered in a single session. There is a limited target volume with a limited dose to the non-target volume, so there is little to no negative impact on neurocognitive function. Metastatic tumors present a well-defined target, unlike malignant glial tumors. They often enhance, are fairly discreet, and sharply demarcated on imaging studies. Radiosurgery systems are able to treat multiple lesions in a single session. Patients can avoid the risks of open surgery. And radiosurgery appears to be effective for radio-resistant tumors such as melanoma, renal cell carcinoma, and others. There are some limitations to the use of radiosurgery in metastatic tumors as well. In general, treatment is limited to tumors 3 cm in size or less. Only tumors which can be radiographically identified can be treated. Tumors follow up with MRI as warranted to continue to survey for micrometastases below the resolution of MRI at the time of treatment. Subsequent lesions can be treated when they arise. Finally, because of the high focal dose of radiation, there is a small chance of the development of focal radiation necrosis, which must be investigated. These images demonstrate a case of radiation necrosis induced by radiosurgery. The image on the left shows a single lung metastasis in a patient treated with the Gamma Knife system using a dose of 20 gray at the 50% isodose curve. Three months later, a larger enhancing lesion with significant perilesional edema was identified. Out of concern for treatment failure, this was resected with open surgical resection. The pathological diagnosis suggested radiation-induced reactive changes and not tumor progression. The image on the right shows the same patient one year following surgery with no additional treatment. There has been no local recurrence. Typical single-fraction radiosurgery doses employed for management of metastatic tumors are listed below. Doses typically decrease as tumor size increases. A number of studies comparing stereotactic radiosurgery with and without whole brain radiation therapy have shown no significant improvement in median survival with the addition of radiotherapy. It doesn't appear that whole brain radiation therapy adds substantially to stereotactic radiosurgery as long as close follow-up is maintained. However, similar to the benefit of adding surgery to whole brain radiation therapy, the addition of stereotactic radiosurgery to whole brain radiation therapy affords a substantial improvement in survival, as documented in a number of studies. It's important to remember that most patients with brain metastasis eventually die due to systemic disease progression and not CNS disease progression. In Patchell's 1990 study, 70% of patients succumbed to systemic disease progression. We recently reviewed our own local data and found that only 4% of patients with lung cancer died of CNS disease progression. Heoplastic meningitis occurs in 4-15% of all patients with solid tumors, most commonly in breast cancer. Recognition of this disease entity is increasing, and diagnosis is becoming more frequent. Prognosis remains generally poor, and although there are no randomized studies comparing intrathecal and systemic chemotherapy, a growing body of reports suggest that intrathecal chemotherapy may be effective in the management of leptomeningeal disease. In addition to neurological examination and routine imaging studies, MRI spectroscopy showing high CSF lactate levels is strongly suggestive of neoplastic meningitis in cancer patients. Standard intrathecal chemotherapy currently includes methotrexate and liposomal cytarabine. Lumbar puncture has afforded marginal distribution, and for adequate distribution and treatment, most patients require placement of a ventricular access device. Contradictory evidence in the support of the role of intrathecal chemotherapy is probably based in part on poor administration and distribution provided through lumbar access rather than intraventricular access. The American Association of Neurological Surgeons and Congress of Neurological Surgeons jointly undertook a systematic evaluation of the evidence regarding metastatic brain tumors. This was a multidisciplinary, methodologically rigorous, evidence-linked clinical practice parameter guideline development project, and included representatives from surgical neuro-oncology, radiation oncology, medical oncology, neurosurgery, and other associated disciplines. The results were published in the Journal of Neuro-Oncology in 2010 and represent the most thorough review of this evidence to date. Level 1 recommendations reflected a high degree of clinical certainty, level 2 recommendations reflected clinical certainty, and level 3 recommendations were uncertain but suggestive. When the group's strict methodology was applied to the available literature, the number of strong recommendations were relatively few but noteworthy. In newly diagnosed patients with a single-brain metastasis, only two level 1 recommendations were supported regarding whole-brain radiation therapy. First, that whole-brain radiation therapy alone is inferior to surgical resection plus whole-brain radiation therapy, and second, that there is no benefit to altered dose or fractionation schedules when compared to the standard schedule. With regard to the role of surgery in patients with a single newly diagnosed brain metastasis, there were data to support a level 1 recommendation that the addition of whole-brain radiation therapy to surgery alone improves local and remote brain control, and there was support for a level 2 recommendation that resection plus whole-brain radiation therapy is equivalent to stereotactic radiosurgery plus whole-brain radiation therapy. With respect to the role of stereotactic radiosurgery, the group found underpowered class 1 evidence along with the preponderance of class 2 evidence to suggest that stereotactic radiosurgery alone may be adequate as long as adequate surveillance for distant failures is available. This brings up one of the shortcomings of the guidelines approach, namely that evidence of adequate evidentiary importance often lags behind clinical practice experience. There were data to support several high-level recommendations regarding the inferiority of whole-brain radiation therapy alone when compared to whole-brain radiation therapy and stereotactic radiosurgery. A level 1 recommendation was made in patients with one brain metastasis to support stereotactic radiosurgery and whole-brain radiation therapy, providing longer survival than whole-brain radiation alone. In patients with 1-4 metastases, a level 1 recommendation could be made that radiosurgery plus whole-brain radiation was better for local control and functional outcome. In a level 2 recommendation reflecting clinical certainty in patients with 2-3 brain metastases demonstrated that radiosurgery plus whole-brain radiation may lead to improved survival when compared to whole-brain radiation alone. With respect to chemotherapy, the group found that current available evidence did not support the use of chemotherapy in patients for whole-brain radiation therapy. However, they recognized that most of the data reported were for non-small cell lung cancer and breast cancer patients, but the selection for histologies more likely to benefit may eventually demonstrate additional benefit with chemotherapy. In summary, metastatic tumors are very common, accounting for about half of all brain tumors. They are typically hematogenously spread and distributed through the brain. Surgery alone leads to high local failure rates, which may be mitigated by the addition of whole-brain radiation therapy. There is often a significant neurocognitive functional decline 12-18 months following the administration of whole-brain radiation therapy. There are evolving data suggesting that stereotactic radiosurgery alone may be adequate in treating brain metastases in patients who can be followed closely with serial imaging studies. As surgical indications, the role of radiosurgery, and recognition of leptomeningeal disease grow, neurosurgeons are increasingly involved in the management of these patients, so it is incumbent upon them to understand the implications of the various treatment options. For more information, visit www.FEMA.gov

metastatic brain tumors

brain metastases

treatment options

whole brain radiation therapy

stereotactic radiosurgery

early diagnosis

management of brain metastases