Rades et al.1 state their case for a palliative whole-brain radiotherapy treatment (WBRT) course of 20 grays (Gy) given in 5 fractions, halving the treatment length of the more common 30 Gy in 10 fractions WBRT regimen. The data they present are not prospective, randomized, or novel in their suggestion that local control and survival are equivalent with either of these 2 regimens.2 Nevertheless, the authors' arguments are presented with common sense, and the conclusions of their article are difficult to refute. The report is timely in that the palliative care community has been reaching out for a discussion of the benefits and drawbacks of shorter treatment courses for the 20% to 40% of patients with solid tumors who will develop brain metastases at some point during their illness.3

Approximately 75% of hospice and palliative care professionals believe that radiation oncologists are too reluctant to offer short-course therapy, thereby placing the burdens of both time and transportation on the morbidly ill and their caregivers. What is more, nearly half of those same professionals believe that WBRT may be omitted in favor of supportive care for patients with brain metastases who fall into the poorer prognostic categories.4 For their part, radiation oncologists have sponsored woefully few hypofractionation studies to evaluate the wide spectrum of end-of-life symptoms observed in cancer patients. The majority of the hypofractionation trials that have been reported to date lack assessment with a validated quality-of-life instrument. The exception to this relative disinterest in palliative care research resides in the setting of painful bone metastasis, in which multiple prospective, randomized trials overwhelmingly suggest the equivalency of single-fraction treatment to multiple fraction therapy.5 In spite of this evidence, the specialty has been slow to change practice patterns.

Why are radiation oncologists reluctant to shorten their treatment courses for palliative care patients? In an attempt to help patients, is it natural to believe that doing more equates doing better? Are radiation oncologists reluctant to change from the dosing schemes that they learned in training and which have been handed down from the pioneers of the field? Or, as some have suggested, are economic considerations a motivating factor?6 Although these concerns may influence some practitioners, discussions with my fellow radiation oncologists would suggest to me that their apprehension results from an increased risk of long-term side effects due to treatment with large daily fractions.

Although the acute side effects of RT are influenced by total dose and length of the treatment course, the late effects are much more dependent on the daily fraction size. Central nervous system tissue in particular appears to suffer more pronounced late effects with hypofractionated RT. In patients with brain metastasis, a larger volume of brain receiving >4 Gy during stereotactic radiosurgery correlates with worsened long-term cognitive functioning.7 Previous studies have shown a greater cognitive decline in patients with limited stage small cell lung cancer who receive larger fraction prophylactic cranial irradiation. Studies have also suggested that cognitive difficulties in the memory domain of patients with low-grade glioma correlate with a fraction size of >2 Gy, although overall cognitive decline has been found to more often represent disease recurrence rather than treatment-related side effects.8, 9 It is interesting to note that other reports of patients who received WBRT for metastasis also suggest that tumor progression correlates more closely with diminished neurocognitive functioning than does the delivery of WBRT.10, 11 A landmark study of patients randomized to undergo surgical resection with or without WBRT revealed that foregoing the RT increased death due to neurologic causes, exposing patients to tumor-induced loss of mental and physical abilities.12 Regine et al. confirmed the concern that these intracranial recurrences produce symptomatic neurologic deficits.13

Neurocognitive testing functions as both a sensitive measure of brain functioning as well as a prognostic variable that predicts survival.14 Analysis of the influence of RT fraction size on neurocognitive functioning is confounded by factors in addition to tumor progression. These factors include patient age, concurrent medical illness, neurologic comorbidity, medication use, and previous treatment with surgery or chemotherapy.11, 15 As an example of the multifactorial nature of the causes of neurologic dysfunction, 1 study revealed that 97% of patients with limited stage small cell lung cancer were found to experience neurocognitive decline before the delivery of prophylactic cranial irradiation.16 The most frequent impairments noted after combined treatment have been described as decreased verbal memory, frontal lobe dysfunction, and motor uncoordination.

This central nervous system damage is mediated on the cellular level by both apoptosis and secondary damage that leads to white matter necrosis or vasculopathy.17 Although the extent of white matter disease may in some circumstances correlate with impairments in memory, attention, and language, patients often experience cognitive damage without exhibiting overt pathologic changes.15, 17 Studies of rats treated with RT to the brain suggest that hippocampal injury may play a significant role in the formation of long-term side effects after RT.17 RT may cause critical damage that leads to the loss of neural precursor cells specifically located in the subgranular zone of the hippocampal dentate gyrus.18 One dosimetric analysis suggested that conformal RT may be calculated to spare the hippocampus in patients receiving RT for multiple brain metastasis, with little risk of diminished tumor coverage.19 This sparing of the hippocampal dentate gyrus may offer a new means by which to decrease the risk of the long-term neurocognitive side effects of RT.

Nevertheless, even with the potential risks of conventional WBRT, one could effectively argue that the long-term side effects of RT are largely irrelevant for the patient group described by Rades et al..1 Common sense dictates that cognitive decline is more notable in patients treated with prophylactic cranial RT for small cell lung cancer or in the postoperative treatment of young, healthy patients with low-grade glioma. Simply stated, those patients may live to experience the effects of the choice of fraction size, whereas the majority of patients with brain metastasis will not. The Radiation Therapy Oncology Group (RTOG) has reported on its analysis of multiple randomized trials totaling 1200 patients with brain metastasis from a range of primary disease sites. The RTOG defined recursive partitioning analysis (RPA) prognostic groups as follows: RPA class 1 comprises patients with a Karnofsky performance score (KPS) ≥70 and age <65 years, and with controlled primary tumor and no extracranial metastases. RPA class 3 is comprised of patients with a KPS <70, and RPA class 2 encompasses all other patients. Only a small percentage of patients with brain metastases fall into the RTOG RPA class 1 category, although they do have a life expectancy of 7.2 months. Therefore, the large majority of patients fall into RPA classes 2 and 3, with life expectancies of only 4.2 months and 2.3 months, respectively.20 These survival averages are less than the 6-month or greater time frame associated with the irreversible and progressive late effects of WBRT.21

Therefore, the time has come for the radiation oncology and palliative care communities to unite and collaborate in the formation of prospective randomized trials to measure the appropriate interventions for patients with common end-of-life cancer symptoms. Focused questions should be asked for groups of patients with brain metastases by separating them into each of the 3 RTOG RPA classes. Class 3 patients should be randomized to supportive care alone versus supportive care plus hypofractionated RT. Class 2 patients might best undergo randomization between RT given to 20 Gy in 5 fractions versus that given to 30 Gy in 10 fractions. RPA class 1 patients should be considered for the evaluation of more time-consuming and costly treatment options including protracted RT or, for patients with a limited number of metastases, surgical excision, stereotactic radiosurgery, or gamma knife treatment. The trials need to include validated quality-of-life scales, and the costs of treatment should be tabulated. Although the ability to organize prospective randomized trials to compare fractionation schema is limited by the poor health of the patient population, the potential for gains in patient care are too great to ignore.


  1. Top of page