Therapy Directed at Brain Metastases
The therapeutic modalities that can be used to treat metastases are surgery and radiation (Table 1). Modalities to enhance responses to radiation and to reduce therapy-induced neurocognitive deficits are currently being evaluated. Since brain metastases respond only occasionally to systemically delivered anticancer therapies (eg, intravenous chemotherapy), such therapy does not yet have an established role in the management of brain metastases—this issue will be discussed below.
Multiple Brain Metastases
Patients with multiple brain metastases (generally defined as more than four lesions) do not benefit from surgical resection, given the associated morbidity. The benefits of extensive surgery may also be limited as there may be several other, yet undiagnosed, microscopic metastases at other sites in the brain, which result in a very high rate of relapse in the brain. The main roles of surgery in this patient population are to relieve mass effect from a large dominant lesion and to obtain a tissue diagnosis if necessary (when the identity of the lesion is unknown).
The optimal therapy is to deliver whole-brain radiation therapy (WBRT), with the most commonly used schedule being 30 Gy in 10 daily fractions. Other fractionation schedules are frequently utilized in specific situations, such as 20 Gy in 5 fractions for patients with a poorer prognosis and 37.5 Gy in 15 fractions for those with a better prognosis. With such therapy, up to 50% will have an improvement in neurologic symptoms and 50% to 70% of patients have an objective response.[5,20,21] The role of stereotactic radiation (SRS) boost in this population is controversial, and the approach should not be used routinely.
• Safety of WBRT—The acute effects of WBRT are reversible and usually well tolerated. These include alopecia, fatigue, and scalp erythema. Some less common complications, primarily attributed to increased cerebral edema, are headache, nausea, emesis, and worsening of preexisting focal symptoms. Treatment of these symptoms with steroids usually results in symptomatic improvement. Of more concern, especially for patients with a better prognosis who may be expected to have longer survival, is the development of late delayed effects (generally appearing more than 6 months after WBRT) such as neurocognitive deficits.
Several early retrospective studies have emphasized the late neurotoxicity of radiotherapy. However, these retrospective studies have several methodologic flaws, the most important of which is the lack of detailed baseline neurocognitive testing for comparative purposes. Such testing would be imperative when trying to determine the etiology of any cognitive deficits posttreatment. It has long been clinically recognized that patients with brain metastases often have cognitive difficulties at baseline. The high prevalence and severity of these deficits was confirmed in a recent large prospective clinical trial of patients with brain metastases. This study noted significant cognitive impairment at baseline (ie, even before WBRT) in one or more neurocognitive domains in 90% of those enrolled.
In addition, prospective studies with baseline and serial neurocognitive testing have found tumor progression to be the dominant cause of cognitive decline in patients with brain metastases. Patients with stable or decreasing tumor volumes after therapy have stable or improving cognitive function.[23,24] Some studies have found that larger radiation fraction sizes appear to be associated with dementia and cognitive decline after WBRT. Therefore, dose-fractionation schedules should be determined according to the patient's prognosis, with more protracted schedules used in those with the possibility of long-term survival.
Although these prospective trials do not rule out detrimental neurocognitive effects of WBRT or even individual patient declines after WBRT, they do suggest that, on the whole, any detrimental effects on cognitive function seem to be balanced by the beneficial neurocognitive effects of improved tumor control in the brain. In summary, the available evidence supports the safety of properly administered WBRT.
Single Brain Metastasis
Approximately one-quarter of all patients with brain metastases have a single metastasis. The current standard of care is to perform surgical resection in patients with solitary accessible brain metastases, and to follow this with WBRT. Patients who undergo a complete surgical resection in conjunction with WBRT have better outcomes than those treated with WBRT after a diagnostic biopsy. The benefits of performing a gross surgical resection were demonstrated in two prospective randomized trials,[6,26] where patients who underwent a surgical resection had longer survival than those who did not. These studies did not formally measure quality-of-life parameters, but did detect a statistically (and clinically) meaningful prolongation of symptomatic neurologic stability. A third study done to evaluate the same issue did not demonstrate any benefit to a surgical resection, but this may have been due to a higher proportion of patients with poor prognostic factors. These findings can be interpreted to indicate that WBRT alone (ie, without surgical debulking) cannot adequately sterilize the large number of tumor cells present in a macroscopic metastasis.
The benefits of WBRT in improving the rate of local control beyond that achieved by surgical resection alone have been confirmed by a phase III study. While the overall survival was not improved in those receiving adjuvant WBRT after surgery, WBRT use did result in a reduced risk of local and distant brain relapses (10% and 14%, respectively), as compared to those who did not receive WBRT (46% and 37%). Thus, neither modality (WBRT or surgery) can adequately sterilize the large number of tumor cells in a macroscopic metastatic focus, and both need to be used together to achieve the best outcomes.
Stereotactic radiation appears to be effective as local therapy for brain metastases and can be considered a reasonable alternative to surgical resection in selected patients. Some of the issues related to the use of SRS in this patient population are discussed below.
Limited Number of Brain Metastases
The availability of radiosurgery or SRS has increased therapeutic options for patients with a limited number of metastases (usually defined as four or fewer). SRS allows for the delivery of a high dose of focal irradiation in a single fraction to the tumor from multiple geometric positions. The advantages of SRS include ease of use (single-day therapy), ability to treat patients who are not surgical candidates (due to inaccessible tumor location or comorbid medical conditions), and reduced morbidity. Brain metastases are ideal targets for SRS as they are small, spherical, and have distinct margins. SRS minimizes the amount of radiation delivered to nontarget areas of the brain. Methods of delivering SRS (Gamma Knife, linear accelerator-based, charged particles, and CyberKnife) are becoming increasingly available. The widespread availability (and use) of these modalities has resulted in a debate about the role of SRS vis-à-vis surgery to treat brain metastases.
An evidence-based approach is extremely important in this area, as the routine use of SRS has the potential to significantly increase health-care costs. The main controversies in this area and the relevant data are discussed below.
• Is SRS as efficacious as surgery in achieving local control? Only one small prospective trial (reported in abstract form) has addressed this question, randomizing 64 patients with a single metastasis (maximum diameter: 3 cm) to either neurosurgical resection followed by whole-brain radiotherapy or radiosurgery alone. All patients enrolled had brain metastases and were eligible for treatment using either modality. No significant difference was shown in any of the main endpoints—overall length of survival, mortality due to neurologic complications, or local tumor control—although there was a trend in favor of SRS with regard to local control (82% vs 97%; P = .06). Overall toxicity of therapy was similar, but the SRS group had a greater improvement in quality of life when evaluated 6 weeks posttherapy. Patients treated with SRS had a higher rate of distant brain failure (as might be expected since they did not receive WBRT) and required salvage therapy more frequently. Due to the small number of patients in this study, these results need to be interpreted with caution.
Other comparative studies that compared outcomes with these modalities have been retrospective in nature.[30,31] Such analyses are inherently flawed due to a number of selection biases, given that SRS and neurosurgery are applicable to different and somewhat nonoverlapping patient populations. With these caveats in mind, the literature on this topic suggests that local control rates are equivalent between those treated with SRS and surgery. The available data also suggest that SRS is more convenient, effective for patients with few (usually defined as less than four), small lesions (usually defined as < 3 cm) and for those with tumors in surgically inaccessible locations. SRS is also the preferred modality for patients who are not surgical candidates for medical reasons. On the other hand, surgery is clearly the optimal modality for lesions causing a mass effect. In summary, SRS and surgical resection should be seen as complementary, but different, modalities to be utilized in the treatment of selected patients with brain metastases.
• In patients treated with WBRT, does the use of SRS as a 'boost' improve outcomes? The rates of recurrence after WBRT at the site of the initial metastases remain high and can be > 50% with prolonged follow-up. This suggests that WBRT cannot adequately sterilize the macroscopic tumor mass. Escalation of radiation dose by delivering a boost using SRS is a logical method to improve long-term local control.
Attempts have been made to evaluate the utility of using SRS in this manner in randomized clinical trials.[32,33] These studies demonstrate that SRS boost does not improve the primary outcomes of overall survival or mortality attributable to neurologic complications of brain metastases. Nevertheless, SRS boost does appear to have some benefits, primarily resulting in improvements in local control rates, time to failure in the brain, symptoms, and in some subsets of patients, survival. Based on these results and data from other retrospective studies, a boost using SRS may be considered as a reasonable option for selected patients who present with a limited number of brain metastases.
• If SRS is used for local therapy, can WBRT be safely omitted? The issue of whether WBRT can be omitted in patients treated with SRS is currently a matter of much debate and ongoing clinical trials.[28,34] Clearly, omitting WBRT avoids the associated toxicities. Moreover, retrospective and prospective data suggest that omitting WBRT does not result in shorter survival,[11,35] which suggests that in many patients, the status of systemic disease (rather than that of the brain metastases) is the predominant determinant of a patient's prognosis.
On the other hand, arguments in favor of routinely administering WBRT include the premise that in a vast majority of patients, several occult lesions may be present at the time of initial diagnosis, and their further growth would be prevented by using WBRT. In addition, the high risk of brain recurrences (when WBRT is omitted) mandates a rigorous monitoring program with frequent use of imaging studies, and a frequent need for salvage therapy. Recurrences may also result in complications precluding any other therapy (by causing a significant deterioration of performance status). Finally, proponents of the procedure point out that adverse events due to WBRT are usually minor (as noted above), and that no prospective trial to date has been designed or powered to address the impact of WBRT on survival.
Only one reported randomized trial has addressed the issue of whether SRS can be used as sole therapy for those with a limited number of metastases. In this study, 132 patients with four or fewer brain metastases were randomized to receive either SRS alone or a combination of SRS and WBRT. The neurologic function, toxicity rates, and overall survival were similar in both groups. As may be expected, patients in whom WBRT was omitted were found to have a higher rate of failure in the brain (76% vs 48%) and required salvage therapy more often for recurrent brain metastases. Significant limitations of this study include a lack of rigorous neurocognitive testing and a lack of statistical power to determine an effect of the therapy on overall survival. These results can be interpreted to suggest that WBRT can be omitted in patients being treated with SRS, if close follow-up can be ensured, with the understanding that the risks of local and distant brain failure are higher.[28,36]
An ongoing intergroup phase III clinical trial (N0574) led by the North Central Cancer Treatment Group (NCCTG) and available through the National Cancer Institute Clinical Trial Support Unit (CTSU) is attempting to address this question. In this study, patients with three or fewer brain metastases will be randomly assigned to receive SRS or a combination of SRS and WBRT. This trial is adequately powered to detect a possible survival benefit with the use of WBRT. Secondary endpoints in this trial include an evaluation of changes in several measures of neurocognition and quality of life, to assess the impact of therapy on these factors.