Simon Lo and colleagues are to be congratulated for their eloquent summary of the role of stereotactic radiosurgery (SRS) for patients with one to three parenchymal brain metastases Citation[1]. I would, however, like to make several comments regarding the conclusions of Lo and colleagues.
The inclusion of whole-brain radiotherapy (WBRT) following the treatment of a single brain metastasis has been prospectively evaluated in only a single randomized study (level 1 evidence) notwithstanding several retrospective studies Citation[2]. If conclusions regarding treatment of patients with parenchymal brain metastases are to be based on high-level evidence, then the Patchell study [2] (resection of a single metastasis followed or not with WBRT) suggests WBRT should follow surgery (or SRS) for the treatment of single brain metastasis for the following reasons:
| • | Recurrence in brain metastases is two- to threefold greater in patients treated with surgery or SRS only, and recurrences occur primarily at brain sites distant to surgery or the site of SRS | ||||
| • | Recurrence in brain metastases results in neurologic impairment in 70% of patients and affects quality of life | ||||
| • | Recurrence in brain metastases results in a two- to fourfold increase in the need for salvage therapy (usually WBRT in patients previously treated with SRS or surgery only) for single parenchymal brain metastasis | ||||
| • | Notwithstanding no difference in survival (inclusion of adjuvant WBRT or not to either primary surgery or SRS), an increased incidence of neurologic death is observed (threefold increase in the Patchell study) | ||||
The second issue warranting comment regards the delayed late effects of WBRT on neurocognition. The often-quoted DeAngelis paper reported an 11% incidence of dementia in 1-year survivors of parenchymal brain metastases treated with WBRT Citation[3]. However, and as Lo and colleagues comment upon, idiosyncratic WBRT schedules (accelerated hypofractionation) were used and rarely would be administered today. Also reported in this paper was a cohort of patients treated with standard WBRT schedules in which no delayed late-radiation injury (defined as dementia) was observed in 1-year survivors. This suggests that usual and customary WBRT schedules have a low likelihood of resulting in dementia in patients surviving to 1 year with parenchymal brain metastases. Furthermore, the Radiation Therapy Oncology Group (RTOG) has conducted a variety of WBRT trials (varying both dose and schedule) and have not demonstrated any improvement with one schedule relative to another. Despite this data, radiation oncologists (a philosophy espoused by Lo) continue to offer longer treatment schedules (e.g., 20 fractions to a total dose of 40 Gy) to patients believed to have longer survival, arguing that this will likely diminish delayed late radiation side effects and, in particular, effects on neurocognition. The second relevant paper in this regard is that of Meyers, who reported on prospective neuropsychologic assessments performed prospectively in patients with multiple brain metastases enrolled in the WBRT and concurrent motexafin gadolinium trial Citation[4]. As Lo and colleagues comment upon, and as shown in the study by Meyers, patients with multiple brain metastases overwhelmingly (>90%) manifest abnormal neuropsychologic evaluations prior to initial treatment. Additionally, when such patients are treated and respond to treatment (in this instance motexafin gadolinium and WBRT), neuropsychologic pretreatment deficiencies improve (unlike the cohort of patients treated and not responding). Furthermore, this improvement in neurocognition is maintained at 6-month follow-up, suggesting little in the way of late detrimental effects of WBRT. Similar findings have also been reported in patients with low-grade gliomas, indicating that the delayed late-radiation injury to the brain may be overstated and less relevant with respect to outcome, as often suggested.
How best to manage radioresistant tumor histologies such as renal cell cancer, melanoma and sarcoma remain problematic. Again, a number of retrospective studies (see Table 3 of Citation[5]) purportedly show no benefit by addition of WBRT to SRS with respect to overall survival, local control rate (brain metastases treated with SRS) or distant brain failure. Not evaluated, however, is a comparison between WBRT alone versus SRS with or without WBRT. The premise for adjuvant SRS in these tumor histologies is based upon the concept of radioresistant tumors. Adjuvant WBRT followed at the time of relapse treatment with SRS may be an equiefficacious treatment. As suggested by Lo (although in a different study design), a randomized trial is needed to provide evidence for these hypotheses. The RTOG 9508 study, which included approximately 5% of patients with melanoma did not demonstrate benefit when comparing WBRT with WBRT plus SRS in patients with one to three parenchymal brain metastases Citation[5]. The Eastern Cooperative Oncology Group performed a small trial (n = 36) of radioresistant brain metastases (melanoma, sarcoma and renal cell) treated with SRS and no WBRT Citation[6]. All patients had a single parenchymal brain metastasis and only 19% of patients had active systemic disease (i.e., the majority of patients were RTOG recursive partitioning analysis [RPA] class 1). Local control was 60% (as compared with 85% in patients with radiosensitive tumor histologies), distant brain failure was 39% (twice that observed in the RTOG 9508 trial when WBRT was administered), and brain metastasis as a cause of death was 38% (twice that observed in the RTOG 9508 trial). These studies question the often-quoted position that radioresistant tumor histologies warrant differing treatments and suggest the need for prospective randomized trials.
The RTOG 9508 trial has been interpreted by many as an endorsement to treat patients with one to three parenchymal brain metastasis with SRS Citation[5]. Lost in the interpretation of this seminal study (the only prospectively randomized trial of SRS) is the fact that all patients received WBRT, and SRS when added to WBRT was beneficial only to patients with single metastasis and with an excellent performance status. If the study is critically evaluated with respect to protocol prespecified end points, even this conclusion is tenuous as the primary hypothesis was a 50% improvement in survival in patients treated with SRS and WBRT compared with WBRT only, an objective not achieved (median survival of 6.5 months in patients receiving WBRT plus SRS compared to 4.9 months in patients treated with WBRT only). Furthermore, multiple subset analyses were undertaken (although not as part of the initial study design), suggesting a benefit in patients less than 50 years of age, with non-small cell lung cancer tumor histology, no active systemic cancer and RTOG RPA class 1. Notwithstanding modest improvements in patients with two or three parenchymal brain metastases (with respect to Karnofsky performance status and steroid use) treated with both WBRT and SRS, the significance level was marginal and does not provide sufficient evidence to justify catholic use of SRS in patients with multiple brain metastases.
In the opinion of this reviewer, treatment with SRS in patients with brain metastases is best reserved for patients with single metastasis and an excellent performance status. Consequently, the majority of patients are best managed with WBRT (in standard schedules of ten fractions and a total dose of 30 Gy), reserving SRS for patients with brain metastases recurrence and in whom there is a reasonable expectation for survival. There is a clear need for further prospective randomized trials in patients with one to three parenchymal brain metastases. An example of such a trial is that of the American College of Surgeons Oncology Group mentioned by Lo (which has been resurrected and is now directed by the North Central Cancer Treatment Group). The study proposes to determine the difference between SRS versus SRS plus WBRT in patients with one to three parenchymal brain metastases. The recently reported Japanese study (Japanese Radiation Oncology Study Group 99-1) that evaluated SRS versus WBRT plus SRS in 148 patients with one to four parenchymal brain metastases demonstrated no difference in median survival (7.6 vs. 7.9 months), however, local control rate (85 vs. 75%), distant brain failure rate (18 vs. 52%) and brain metastasis as a cause of death (15 vs. 16%) favored WBRT plus SRS Citation[7]. Similar trials will hopefully further define the role of SRS in the adjuvant setting of patients with one to three parenchymal brain metastases.
References
- Lo SS, Chang EL, Suh J. Stereotactic radiosurgery with and without whole-brain radiotherapy for newly diagnosed brain metastases. Expert Rev. Neurotherapeutics 5(4), 487–495 (2005).
- Patchell RA, Tibbs PA, Regine WF et al. Postoperative radiotherapy in the treatment of single metastases to the brain: a randomized trial. JAMA 280(17), 1485–1489 (1998).
- DeAngelis LM, Mandell LR, Thaler HT et al. The role of postoperative radiotherapy after resection of single brain metastases. Neurosurg. 24, 798–805 (1989).
- Meyers CA, Smith JA, Bezjak A et al. Neurocognitive function and progression in patients with brain metastases treated with whole-brain radiation and motexafin gadolinium: results of a randomized Phase III trial. J Clin. Oncol. 22(1), 157–165 (2004).
- Andrews DW, Scott CB, Sperduto PW et al. Whole-brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: Phase III results of the RTOG 9508 randomised trial. Lancet. 363(9422), 1665–1672 (2004).
- Mañon R, Oneill A, Mehta M et al. Phase II trial of radiosurgery (RS) for 1 to 3 newly diagnosed brain metastases from renal cell, melanoma, and sarcoma: An Eastern Cooperative Oncology Group Study (E6397). J. Clin. Oncol. ASCO Annual Meeting Proceedings (Post Meeting Edition). 22, 14S, (Abstract).
- Aoyama H. ASCO Annual Meeting Proceedings (Post Meeting Edition). J. Clin. Oncol. 22, 18S(Abstract) (2004).