Brain metastases are the most common type of brain tumors, with the total number diagnosed annually outnumbering all other intracranial tumors combined (22, 30). With the increasing survival of patients with systemic (extracranial) disease, the incidence of the most common cancers (lung, breast, melanoma, renal and colon) is thought to be rising. Autopsy data show that the frequency of brain metastases in patients dying from cancer varies from 20 to 50%, and may be higher if dural, leptomeningeal, or spinal metastases are taken into account. As the incidence of brain metastases rises due to improved cancer therapy for systemic disease (18), it is imperative that improved intracranial therapy be developed as well.
The most common source of brain metastases in males is lung cancer and in females is breast cancer (29), but with the increasing frequency of lung cancer in females, it is expected that for females this too will be the primary cause of metastatic brain tumors (22).
How Do Tumors Metastasize?
The mechanisms by which primary tumors produce brain metastases is thought to be hematogenous spread from primary or secondary sites in the lung. Since the brain has no lymphatic system, all tumors metastasizing to the brain do so by spreading through the bloodstream. Arterial blood passes through the lungs before entering the brain, and collects tumor cells filtered out in capillaries, which subsequently embolize to the brain. This is correlated with sites of localization: the cerebrum is involved in 80 to 85% of all brain metastases, the cerebellum in 10 to 15% and the brainstem in 3 to 5% (6, 10, 29). The overall distribution corresponds roughly to the relative size of blood flow regions in the brain.
Different types of primary tumors have different relative frequencies of single versus multiple metastases. Melanoma has the highest tendency to produce multiple lesions, followed by lung and breast cancers (19, 22). Though many studies have indicated that 37 to 50% of patients present with a single metastasis (6, 29), recent studies have shown that patients with one lesion detected by CT may demonstrate multiple lesions detected by MRI (4, 28). These findings clearly agree with our data in which the majority of patients presented with multiple lesions upon contrast dye with MRI. (1).
Metastatic brain tumors present with the usual signs and symptoms of any expanding intracranial mass lesion. These include increased intracranial pressure and focal neurological deficits with focal irritations. Such symptoms include headaches, focal weakness, mental status changes, seizures, ataxia [inability to coordinate voluntary muscular movements] and sensory and visual changes. Though most of these symptoms are of gradual onset, acute episodes may occur due to hemorrhages into a metastasis (15). When such an event occurs, either choroid carcinoma or melanoma must be considered, because these have the greatest tendency to hemorrhage (15). Because of the greater incidence of bronchogenic metastasis, these lesions represent the most common source of a hemorrhagic lesion (15, 21).
Whole Brain Radiation Therapy (WBRT)
Brain metastases carry an ominous prognosis regardless of primary status or treatment given. The median survival of untreated patients, or those treated with corticosteroids alone to reduce brain edema, is about one month (32). Whole brain radiation therapy (WBRT) is the most widely used method of treating brain metastasis, despite the fact that patients treated this way have an expected survival of only three to four months. Death from recurrent or persistent tumors occurs in about 50% of the patients (12, 29).
The radiosensitivity of the tumor itself is not taken into account when these patients are being treated. Most tumors that metastasize to the brain, such as non-small cell lung, renal, colon, and melanoma are radioresistant [resistant to radiation therapy]. Worse yet, many treating facilities continue to use prophylactic cranial radiation despite the fact that only one study has ever demonstrated a statistically significant increase in life span (20). (Prophylactic radiation therapy is treatment given before lesions have appeared within the brain.)
Significant neurotoxicity has been reported with the use of WBRT. Acute effects include hair loss (alopecia), nausea, vomiting, lethargy, otitis media and severe cerebral edema. Though some of these effects can be transient, dermatitis, alopecia, and otitis media can persist for months after irradiation (23). Chronic effects are even more serious, and these include atrophy, leukoencephalopathy, radiation necrosis, neurological deterioration and dementia (5). Reports of development of severe radiation induced dementia have varied between 11% in one-year survivors (23, 24, 27) to 50% in those surviving two years (7, 23). The time involved in this therapeutic intervention frequently is over two weeks, in itself a burden to many patients (5, 8).
Surgery and WBRT
Surgical removal of solitary and occasionally multiple lesions has been reported to enhance survival (2, 6, 7, 10, 16, 26), with several reports indicating improvement of neurological function. Recently, the concept of multiple craniotomies for multiple lesions has been promoted (2), though only in those patients with "accessible locations" and "good clinical condition." The risks of postoperative morbidity in "eloquent" areas must also be considered when contemplating surgical intervention. The complications of the surgery itself include hemorrhages and wound infection.
Pseudomeningoceles form in 8 to 9% of patients, and an estimated 10% of patients develop clinically evident thromboembolic complications such as deep vein thrombosis or pulmonary embolisms (3, 9). Recent reports have also indicated an operative mortality of approximately 3%. Though adjunct WBRT has been prescribed in the past , and Patchell et al (16) showed that a subset of patients with favorable prognosis and a single brain metastasis that had surgery followed by adjunct WBRT had a median survival of 10 months, other subsequent randomized trials failed to show a benefit to surgical resection (14).
Radiosurgery is a technique which allows the delivery of a single high dose of radiation in a highly accurate manner (24, 25). The Gamma Knife (a dedicated neuro-surgical instrument) allows numerous beams of radiation to converge on a target site, resulting in a high dose of radiation delivered to the target site with a sharp dose gradient at the target edge. A recent report by Somaza et al (25) revealed that even in patients with radioresistant tumors (such as melanoma), local tumor control was achieved in 97% of patients and neurological improvement occurred in 53% of affected patients.
Median survival with radiosurgery alone improved from two to three months to nine months in patients with single or multiple metastatic melanoma lesions to the brain (25). Despite such results, radiosurgery has not been considered a primary therapy. In the recent past most treatment centers treat only unresectable tumors or recurrent tumors with this modality (17, 31, 32).
The issue of multiple metastases has become important, as has the issue of lesion size. From our perspective, neither number of lesions nor the size of the lesions has been shown scientifically to be a limiting factor in single session Gamma Knife treatment. Multiple metastases may be more of an issue in terms of the equipment itself not allowing multiple lesions to be treated in a single sitting. At The Miami Neuroscience Center, at Doctors Hospital in Coral Gables, Florida, we have treated 460 patients (261 females and 199 males) with a mean of four lesions per treatment. The patients had the following types of cancers: 111 males and 111 females had lung cancer, 32 males and 16 females had melanoma, 7 males and 20 females had colon cancer, and 8 males and 16 females had renal cancer.
When we looked specifically at the outcome of metastatic breast carcinoma (1), we found the following results: 68 women were treated, ranging in age from 25 to 83 years, and the median age was 52. Thirty-eight patients had previously received conventional modalities, including WBRT. A total of 110 treatments were given to the 68 women with an average of eight tumor sites per patient. Twenty-seven (40%) of 68 survived one year, seven (10%) survived two years, and two (3%) survived more than three years. Twenty-six patients with one to three lesions were treated, 18 with four to seven lesions, and 24 with more than eight lesions. Their overall local control rate was 94%, with 39 (91%) of the 43 patients expiring, dying of causes unrelated to their brain metastases. There was no significant difference in survival and local control based on the number of lesions treated. Survival was clearly found to be independent of the number of lesions treated.
Similarly, when we looked at our renal cell carcinoma patients, we found similar results. Twenty-two patients were treated: 8 females and 14 males. The range of lesions was between 1 and 21, with a median of 3.4 per patient. Twelve of 22 (55%) had WBRT. Age ranged from 38 to 80, with a median age of 60. The median survival was 8.7 months (3 to 55 months), with local control in 20 of 22 patients (91%). Eight patients (36%) required re-treatment for new lesions. Survival at one year was 24% in patients older than 60, but 54% in those younger than 60. Once again, the number of sites or prior WBRT did not have statistically significant effects on survival.
In our study, Gamma Knife radiosurgery shifted the question of survival to that of systemic control. Previous whole brain radiation therapy results have yielded no survival advantage to the treatment. The overall complication rate with one-session Gamma Knife has been 1.2%, in which patients having biopsy proven radiation necrosis required treatment with stereotactic aspiration and corticosteroids. This is a very low rate of complications.
In conclusion, we believe that one-session Gamma Knife radiosurgery for brain metastases is a superior mode of treatment for either single or multiple metastases. Survival rates match or exceed those previously reported for surgery with whole brain radiation or whole brain radiation alone. Radiosurgery yields added advantages: outpatient treatment, lower morbidity, greater flexibility in terms of local and number of tumors treated, and the ability to treat the patient over multiple periods of time for the development of new lesions.
We have not found that WBRT leads to a survival benefit nor that it prevents later onset of remote metastases in other brain locations. In our opinion, radiosurgery alone is the primary mode of therapy for brain metastases, unless the patient presents with neurological deficits resulting from mass effect, thus requiring surgical intervention. Radiosurgery clearly provides a very high rate of local control and preservation of neurological function with minimum effort and morbidity to the patient.