Whole-Brain Radiotherapy (WBRT) for Brain Metastases: Does the Interval Between Imaging and Treatment Matter?

Discussion

Many patients with brain metastases have a very poor OS (1, 2, 8, 9). In order to improve the prognoses of these patients, personalized treatment approaches have become more popular during recent years also for this group. Personalization of treatment is generally based on decisions made by multidisciplinary tumor boards often including radiation oncologists, medical oncologists, neurosurgeons, radiologists and pathologists. In cases of a single lesion, it can be difficult for the radiologist to distinguish between metastasis and a primary brain tumor, and additional procedures may be helpful. Furthermore, histology is important if no previous malignancy is known. These procedures may cause a delay regarding the beginning of the treatment. If a patient is selected to receive radiotherapy, this could be WBRT, SRS/FSRT or a combination of both modalities. However, SRS/FSRT is generally limited to very few (up to 3-5) brain metastases (2, 10). The majority of patients presenting with brain metastases have more than 3-5 lesions and receive WBRT alone. Regarding treatment, the question arises whether a delay in the start of WBRT has a negative impact on the patient's prognosis. The present study aimed to answer this question in a cohort of 573 patients treated with longer-course WBRT alone. According to its results, the interval between first diagnosis of brain metastases on MRI or CT and the start of WBRT had no significant impact on OS up to 12 months following WBRT.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Comparison of the three time intervals between diagnosis of brain metastasis on MRI or CT and start of WBRT ((0-7 vs. 8-14 vs. 15-21 days) with respect to OS (univariate analysis).

In contrast to the interval between diagnosis of brain metastases and WBRT, several other factors were significantly associated with OS on multivariate analysis, including age, ECOG-PS, metastases outside the brain, and completion of the WBRT as planned. In addition, control of the primary tumor showed a trend to increase OS. Most of these factors have been previously identified by other studies performed to create predictive tools for estimating OS in patients irradiated for brain metastases. The first and most common of such tools, the recursive partitioning analysis (RPA) classification, was published in 1997 and based on the data from 1,200 patients of three randomized trials who had received WBRT (11). This tool included three prognostic groups (RPA classes I, II and III) with median OS times of 7.1 months, 4.2 months and 2.3 months. The three classes were designed based on four independent prognostic factors for OS, including Karnofsky performance score, age, metastases outside the brain, and control of the primary tumor. Another very commonly used tool is the graded prognostic assessment (GPA) score published in 2008 (12). This prognostic instrument was created based on the data of patients from five randomized trials who received WBRT, which was supplemented by an SRS boost in one arm of a trial. The GPA score included four prognostic groups with median OS times of 2.6, 3.8, 6.9 and 11.0 months, respectively, and was based on Karnofsky performance score, age, metastases outside the brain and number of brain metastases (12). In the present study, the number of brain metastases did not achieve significance on univariate analysis but showed a trend for an association with OS (p=0.056). An independent predictor of OS that was not previously described for WBRT of brain metastases is completion of the WBRT as planned. In the present study, this factor was highly significant (p<0.001). The reason for non-completion of WBRT mostly is a decline in the patient's performance status due to progression of the malignant disease or to other complications such as pneumonia or pulmonary embolism leading to an early death. Only 4% of these patients survived 3 months or longer following WBRT.

The comparably poor OS in our study may reflect the current trend to avoid WBRT whenever reasonably possible, particularly in patients with few brain metastases and a favorable OS prognosis, and to use SRS/FSRT alone instead. Two randomized trials have demonstrated that the addition of WBRT to SRS/FSRT leads to a significantly more pronounced decline in neurocognitive function than SRS/FSRT alone (13, 14). In the trial of Chang et al., the rates of neuro-cognitive deficits (compared to baseline) at 4 months following irradiation were 96% with and 24% without WBRT (p<0.001) (13). In the Alliance trial, the corresponding rates at 3 months were 92% and 64%, respectively (p<0.001) (14). Another trial from Japan suggested that using SRS/FSRT alone without WBRT was reasonable for up to 10 metastases (15, 16). The authors found no significant difference between patients with 2-4 lesions and those with 5-10 lesions with respect to median OS (10.8 vs. 10.8 months), grade 3-4 adverse events (2% vs. 3%) and neuro-cognitive decline after 4 years (14% vs. 11%). However, despite the data that led to a decrease in the proportion of patients with brain metastases receiving WBRT alone, many patients, particularly patients with multiple brain metastases and/or a poor performance status, will continue to be candidates for WBRT alone. When WBRT is used, modern approaches such as hippocampal sparing and concurrent treatment with memantine should be considered to reduce the risk of treatment-related neuro-cognitive deficits (17, 18).

In the present study, the time interval between diagnosis of brain metastasis using MRI or CT and start of WBRT had no impact on OS. Thus, if difficult treatment decisions require multidisciplinary coordination, or waiting for a definitive histology is important prior to the start of treatment, WBRT may be postponed for some days. However, since brain metastases are often associated with significant clinical symptoms, WBRT should be started as soon as reasonably possible. Corticosteroids can be used to help with symptoms as well. In contrast to the interval between diagnosis of brain metastasis and WBRT, OS was significantly associated with the pre-treatment factors age, performance status and metastases outside the brain. These factors have been previously identified as independent predictors of OS in patients irradiated for brain metastases, which demonstrates consistency of the data of the present study with the literature and supports validity of the results.