The vanishing role of whole brain radiation for limited brain metastases
Editor’s note: This commentary is in response to an article by Vyfhuis and colleagues available here.
Brain metastases significantly impact the lives of many patients with cancer.
Given that patients with brain metastases present with varying degrees of intracranial involvement, neurologic symptomatology, systemic disease burden and ability to tolerate various brain-directed interventions, the optimal management of brain metastases balances these factors for the individual patient. Most patients with brain metastases will receive external, brain-directed radiation therapy at some point during their disease course, largely due to the limited ability of many systemic agents to penetrate the blood-brain barrier.
Two distinct forms of external, brain-directed radiation exist: whole brain radiation therapy (WBRT), in which the entire brain is treated with radiation over 1 to 3 weeks, and stereotactic radiation, in which individual metastases, or surgical cavities in patients who have underwent neurosurgical resection of one or more brain metastases, are targeted over 1 to 5 days.
Results of randomized trials
The gold standard to identifying the superiority of one medical treatment over another is a randomized clinical trial. Randomized clinical trials are powered a priori to detect a specific difference in a primary endpoint between two populations, and their structure minimizes the impact of the measured and unmeasured confounding variables that plague retrospective studies and secondary analyses of clinical trial data.
Among patients with a limited number (one to four) of brain metastases, several randomized trials have evaluated the relative benefits of stereotactic radiation alone vs. stereotactic radiation in combination with WBRT. These trials showed consistent results — namely that the addition of WBRT to stereotactic radiation improves intracranial disease control but does not result in better OS. In addition, patient quality of life and neurocognitive function are worse with the addition of WBRT. The results of these trials definitively demonstrate that WBRT has little to no role in the management of patients with a limited number of brain metastases.
In their statement supporting consideration of WBRT among patients with a limited number of brain metastases, Vyfhuis and colleagues state that the addition of WBRT can be justified by the “improved local control, decreased intracranial progression and reduction of neurologic deaths.” We would argue that these endpoints are inconsequential relative to the lack of OS benefit and decreased quality of life observed when WBRT is employed.
In addition, Vyfhuis and colleagues reference several retrospective analyses and secondary analyses of existing randomized trials to justify their argument; such analyses are best characterized as hypothesis generating. The conclusions that can be drawn from them cannot compare to the definitive primary analyses of the six randomized trials referenced above.
Lack of OS benefit
There are two main reasons why WBRT does not improve OS among patients with a limited number of brain metastases.
First, patients who are treated with radiation for brain metastases are followed routinely for intracranial progression with MRI of the brain, with early salvage therapy employed if such progression is observed. In the context of regular surveillance MRIs of the brain, prophylactic treatment of microscopic intracranial disease is likely of minimal benefit. Surveillance MRIs of the brain permit gross disease to be treated stereotactically as initial therapy, reserving WBRT as salvage for significant distant intracranial progression thereafter.
Second, the competing risk for systemic death and the guarded life expectancy of many patients with brain metastases further decrease the benefit of measures to achieve distant intracranial control. The median survival of patients with newly diagnosed non-small cell lung cancer — the primary malignancy which most frequently metastasizes intracranially based on absolute incidence — and brain metastases is 4 to 6 months; the likelihood that a subclinical, micrometastatic deposit will grow to the point that it can result in neurologic symptomatology over this time period is doubtful. Therefore, the guarded life expectancy of many patients with brain metastases — primarily due to systemic disease progression — decreases the relevance of prophylactic treatment of the uninvolved brain.
The toxicity imparted by WBRT therapy also is an important consideration.
As noted above, numerous clinical trials have demonstrated the deleterious effect of WBRT on quality of life and neurocognitive function when employed among patients with a limited intracranial disease burden. The fact that a significant difference in such outcomes can be identified despite the multiple other factors that can impact quality of life and neurocognitive function among patients with advanced cancer (eg, steroids, antiepileptic drugs, other medications, pain, systemic disease burden, etc) is indicative of the real toxicity that WBRT can yield.
In their statement supporting consideration of WBRT for patients with a limited burden of intracranial disease, Vyfhuis and colleagues contend that a high percentage of patients managed with stereotactic radiation alone experience neurocognitive decline and that intracranial progression may account for such decline. They downplay the fact that a significantly higher percentage of patients who receive WBRT vs. stereotactic radiation alone experience neurocognitive decline and, regardless of the etiology, multiple randomized trials clearly demonstrated worse neurocognitive function among patients receiving WBRT. The absolute numbers of neurocognitive decline in a single arm are not relevant in determining which therapy best preserves cognitive function; rather it is the delta between the study arms of WBRT and stereotactic radiation that is clinically significant.
In addition, Vyfhuis and colleagues state the magnitude of neurocognitive deterioration deemed to constitute meaningful decline in the Alliance trial (> 1 standard deviation from baseline on at least one cognitive test at 3 months) was not clinically significant; it is important to note that the difference between stereotactic radiation and WBRT remained statistically significant, and similar differences were noted between study arms, regardless of the magnitude of the neurocognitive decline evaluated (eg, 3 standard deviation decline in cognitive function at 3 months: stereotactic radiation 24% vs. WBRT 52%; P < .01).
Another toxicity of relevance is radiation necrosis — or radiation injury to the brain abutting a treated metastasis. The risk for radiation necrosis with WBRT alone is negligible, but it is the primary long-term toxicity of concern among patients treated with stereotactic radiation. When WBRT is added to the management plan of patients treated stereotactically, the risk for radiation necrosis increases, and such toxicity may be one of the reasons that quality of life is worse when WBRT and stereotactic radiation are combined for the initial treatment of brain metastases.
In their statement supporting consideration of WBRT for patients with a limited number of brain metastases, Vyfhuis and colleagues promote the use of memantine and hippocampal-sparing techniques to minimize the neurocognitive impact of WBRT. However, it should be noted a significant percentage of patients receiving WBRT and memantine still experience neurocognitive decline. In addition, hippocampal-sparing WBRT is currently only supported by a single arm prospective trial and, therefore, there is no definitive evidence at this time that such a technique is beneficial for patients.
Finally, we want to emphasize that these commentaries pertain to patients with a limited number of brain metastases (ie, one to four brain metastases). We agree with Vyfhuis and colleagues that the trials evaluating patients with a limited number of brain metastases should not be directly extrapolated to other patient populations, such as those with a significantly greater number of brain metastases; patients with small cell lung cancer (who were excluded from the above trials); or patients with leptomeningeal disease, where WBRT remains more standard. It is important to recognize the risk-benefit ratio may shift in favor of WBRT for patients with more extensive intracranial disease.
When discussing WBRT with patients, the goal is not to instill fear or dread in patients, but to provide unbiased discussion of potential risks and benefits with an understanding that the majority of reported trials to date have focused on patients with limited brain metastases.
The results of future clinical trials are needed to provide further insight regarding appropriate management for patients with more extensive intracranial disease. These trials include:
- the Canadian Cancer Trials Group/Alliance CE.7, comparing WBRT with radiosurgery — with primary endpoints of OS and cognitive PFS — for patients with five to 15 brain metastases;
- a randomized study underway at Brigham and Women’s Hospital/Dana-Farber Cancer Institute comparing hippocampal-sparing WBRT with stereotactic radiation for patients with five to 20 brain metastases, with a primary endpoint of quality of life (NCT03075072); and
- an ongoing phase 3 trial being conducted at The University of Texas MD Anderson Cancer Center comparing WBRT with stereotactic radiation for patients with four to 15 brain metastases, with primary endpoints of neurocognitive function and intracranial control (NCT01592968).
Any argument supporting the role of WBRT for patients with a limited intracranial disease burden is in direct opposition of the results of six randomized studies and is founded upon lower tiers of clinical evidence. It is clear that WBRT does not prolong survival among patients with a limited intracranial disease burden, but does lead to decline in quality of life and neurocognitive function.
When brain-directed radiation is employed for patients with a limited intracranial disease burden, stereotactic radiation alone should be utilized. Thereafter, patients should be followed with frequent MRIs of the brain and whole brain radiation should only be employed if significant intracranial disease progression is observed.
Aoyama H, et al. JAMA. 2006;295:2483-2491.
Brown PD, et al. JAMA. 2016;doi:10.1001/jama.2016.9839.
Brown PD, et al. Lancet Oncol. 2017;doi:10.1016/S1470-2045(17)30441-2.
Brown PD, et al. Neuro Oncol. 2013;doi:10.1093/neuonc/not114.
Chang EL, et al. Lancet Oncol. 2009;doi:10.1016/S1470-2045(09)70263-3.
Cagney DN, et al. Neuro Oncol. 2017;doi:10.1093/neuonc/nox077.
Gondi V, et al. J Clin Oncol. 2014;doi:10.1200/JCO.2014.57.2909.
Kayama T, et al. J Clin Oncol. 2003;doi:10.1200/JCO.2016.34.15_suppl.2003.
Kocher M, et al. J Clin Oncol. 2011;doi:10.1200/JCO.2010.30.1655.
Minniti G, et al. Radiat Oncol. 2011;doi:10.1186/1748-717X-6-48.
Muldoon LL, et al. J Clin Oncol. 2007;25:2295-2305.
Mulvenna P, et al. Lancet. 2016;doi:10.1016/S0140-6736(16)30825-X.
Soffietti R, et al. J Clin Oncol. 2013;doi:10.1200/JCO.2011.41.0639.
Takahashi T, et al. Lancet Oncol. 2017;doi:10.1016/S1470-2045(17)30230-9.
For more information:
Ayal A. Aizer MD, MHS, is assistant professor of radiation oncology in the department of radiation oncology at Dana-Farber/Brigham and Women’s Cancer Center and Harvard Medical School. He can be reached at firstname.lastname@example.org.
Paul D. Brown, MD, is professor of radiation oncology in the department of radiation oncology at Mayo Clinic in Rochester, Minnesota. He can be reached at email@example.com.
Disclosures: Aizer reports research funding from Varian Medical Systems. Brown reports honorarium from UpToDate.