Keywordsbrain metastases, lung cancer, surgical resection, stereotactic radiosurgery, whole brain radiation therapy, prophylactic cranial radiation, toxicity
Treatment Options 228
Surgical Resection 228
Stereotactic Radiosurgery 228
Whole-Brain Radiation Therapy 229
Prophylactic Cranial Irradiation 233
Toxicity of Radiation to the Brain 237
Brain metastases from lung cancer represent a significant portion of all brain metastases. Between 10% and 25% of patients with lung cancer have brain metastases at diagnosis, and about 40–50% of all patients with lung cancer will develop brain metastasis during the course of the disease. Some evidence exists to suggest that improved control of locally advanced disease may be associated with increased incidence of brain metastases ( ). The prognosis for patients with brain metastasis is poor, with median survival times usually less than 1 year.
The manifestations of brain metastases vary and depend on the location and number of lesions and the amount of associated edema, hemorrhage, or both. Presenting symptoms can include headache, nausea and emesis, focal weakness, seizures, confusion, ataxia, visual disturbances or, occasionally, cranial nerve palsies. Magnetic resonance imaging (MRI) is currently the gold standard for identifying brain metastases and is more sensitive than computed tomography (CT) scanning for this purpose.
Initial management of brain metastases usually involves oral or intravenous corticosteroids ( ). Patients with seizures should be treated with antiseizure medications. However, prophylactic use of antiseizure medication is controversial because of the high risk of adverse effects. Subsequent management of brain metastases depends on the size, number, and location of the lesions as well as the presence of extracranial disease and the general condition of the patient. Whole brain radiation therapy (WBRT) can be used as primary therapy for brain metastases, as adjuvant treatment after surgery or stereotactic radiosurgery (SRS), or as salvage therapy after local treatment. SRS can produce local control rates that are comparable to those after surgery with minimal toxicity, and SRS can be used as primary therapy or salvage therapy. Its versatility makes SRS useful for multiple or deep-seated lesions and for patients with medical conditions that render them poor candidates for surgery. Surgery provides rapid relief of mass effects and may be the best choice for large single metastases.
This chapter focuses on the roles of various forms of therapy for treatment of brain metastases from lung cancer, including surgical resection, SRS, WBRT, and prophylactic cranial irradiation (PCI).
The surgical resection of brain metastases has traditionally been reserved for palliation of symptomatic lesions or for situations in which pathologic confirmation is necessary. However, some subsets of patients with favorable prognostic factors have undergone surgical resection with the intent of improving survival. A pivotal trial conducted in the 1980s by evaluated the use of surgery for 48 patients with a single brain metastasis who were also receiving WBRT. Specifically, patients with a suspected brain metastasis were randomly assigned to undergo either biopsy followed by WBRT or surgical resection followed by WBRT. WBRT was to be delivered to a dose of 36 Gy in 12 fractions. Resection was found to improve local recurrence rates (52% vs 20%) and quality of life (9 months vs 2 months of functional independence). Surgery was also found to extend overall survival time (9 months vs 3 months) and time to death from neurologic causes (14 months vs 6 months). Rates of operative mortality (4%) and morbidity (8%) in that study were deemed acceptable.
Another phase III trial conducted in the Netherlands involved patients with a single brain metastasis who were given WBRT (40 Gy in 2-Gy fractions twice a day) either alone or with surgical resection ( ). Surgical resection was again associated with extended median survival time (10 months vs 6 months), with particular benefit seen among those with stable extracranial disease. Indeed, local therapy is generally advocated for those with favorable performance status, a single brain metastasis, and stable extracranial disease. The need for adjuvant radiotherapy after surgical resection to maintain intracranial disease control is discussed further later in this chapter.
The neurosurgeon Lars Leksell formulated the principles of radiosurgery in 1951, about 17 years before the launch of the first Gamma knife prototype at the Karolinska Institute. SRS is commonly defined as a single high dose of radiation directed by stereotaxis conformally to the target, minimizing the dose to the normal surrounding tissue. The first Gamma knife unit in the USA was installed at the University of Pittsburgh Medical Center in 1987. Today, brain metastases represent the most common indication for SRS ( ). The goals of SRS can be achieved with several types of technologies, including the Gamma knife, linear accelerators, or the Cyberknife.
The Radiation Therapy Oncology Group (RTOG) has conducted several trials investigating SRS for brain metastases. The prospective trial RTOG 90-05 sought to determine the optimal dose for radiosurgery of brain metastases by evaluating the maximum tolerated radiosurgical dose in 168 patients who had previously undergone irradiation for primary or metastatic brain lesions ( ). The maximum tolerated doses depended on the size of the lesion, being 24 Gy for tumors≤20 mm in diameter, 18 Gy for those 21–30 mm, and 15 Gy for tumors 31–40 mm. However, the actual maximum tolerated dose for lesions≤20 mm was not met because the investigators were reluctant to exceed 24 Gy. Notably, the rate of radionecrosis at 2 years was 11%. Subsequent research suggested that treating tumors≤20 mm with 20 Gy rather than 24 Gy produced equivalent control with lower complication rates ( ).
RTOG 95-08 was a randomized controlled trial comparing WBRT alone versus WBRT with SRS for 333 patients with one to three brain metastases ( ). The WBRT dose was 37.5 Gy delivered in 15 fractions, and the SRS dose was delivered according to the guidelines established by RTOG 90-05. Although no difference was found in overall survival between the two groups, subgroup analyses revealed that SRS improved overall survival for patients with a single brain metastasis. The addition of SRS to WBRT also led to higher response rates and better local control rates at 1 year (82% vs 71%); receipt of SRS was the only factor predictive of local control in a Cox proportional hazards analysis. The risk of developing a local recurrence was 43% greater in the WBRT-alone group relative to the WBRT with SRS group. Moreover, patients receiving SRS were more likely to have had stable or improved Karnofsky performance status scores at 6 months after treatment.
To date, no randomized controlled trials have been undertaken to compare surgery with SRS for single brain metastases. Findings from retrospective analyses are mixed and fraught with selection bias. The indications for SRS remain controversial, with factors contributing to the choice of treatment including the number of metastases, performance status, status of extracranial disease, and tumor histology. Further clinical trials are necessary to define an SRS treatment algorithm for patients with brain metastases.
Whole-Brain Radiation Therapy
WBRT has been used to treat brain metastases since the early 20th century. The first report of its efficacy in a large series came from Memorial Hospital, where 38 patients with brain metastases were treated from 1949 to 1953 with fractionated WBRT. Palliation was successful for 63% of patients so treated, and the recommended dose was 30–40 Gy given over 3–5 weeks. A subsequent report, published in the early 1960s, of an additional 218 patients at Memorial Hospital given WBRT to a goal dose of 30 Gy in 3 weeks revealed a survival benefit among those who responded to this therapy, further establishing the role of WBRT in the treatment of intracranial metastatic disease. WBRT is still used for palliation of gross brain metastases and as prophylaxis against the development of new brain metastases. These topics are described further in the following sections.
WBRT as Primary Therapy for Gross Disease
Prospective trials led by the RTOG have been used to help ascertain the optimal dose and fractionation for WBRT. In general, WBRT to a dose of 20–40 Gy given over 1–4 weeks has resulted in median survival times of 4–6 months for patients with brain metastases ( Table 20.1 ) ( ). Although the treatment schedules varied somewhat among studies, they were comparable with regard to the incidence and the duration of improvement, time to progression, survival, and palliative index. In RTOG 7361, approximately 900 patients were randomly assigned to receive 20 Gy in five fractions over 1 week, 30 Gy in 10 fractions over 2 weeks, or 40 Gy in 15 fractions over 3 weeks. The median survival times were similar among the three treatment groups, at 15 weeks for the 20-Gy and 30-Gy groups and 18 weeks for the 40-Gy group. WBRT was shown to improve neurologic function for about half of the patients in this study, without much improvement in survival, depending on dose or fractionation. Another randomized comparison of an ultra-rapid high-dose radiation schedule (10 Gy in one fraction or 12 Gy in two fractions) revealed median survival times that were not much different than those in RTOG 7361; however, patients who received the ultra-rapid high-dose WBRT had more treatment-related neurologic toxicity. The rapidity of the response to therapy was also similar between the two treatment groups in this study, but the duration of improvement, time to worsening of neurologic status, and rate of complete disappearance of neurologic symptoms were generally less for patients receiving 10–12 Gy in one or two fractions than for those receiving more prolonged treatment such as 30 Gy in 10 fractions.
|Protocol||Period of Study||No. of Patients||Treatment Scheme||Median Survival Time|
|RTOG 6901||1971–1973||233||30 Gy in 10 Fx over 2 wks||21 wks|
|217||30 Gy in 15 Fx over 3 wks||18 wks|
|233||40 Gy in 15 Fx over 3 wks||18 wks|
|227||40 Gy in 20 Fx over 4 wks||16 wks|
|RTOG 7361||1973–1976||447||20 Gy in 5 Fx over 1 wk||15 wks|
|228||30 Gy in 10 Fx over 2 wks||15 wks|
|227||40 Gy in 15 Fx over 3 wks||18 wks|
|RTOG 6901 ultra-rapid||1971–1973||26||10 Gy in 1 Fx over 1 day||15 wks|
|RTOG 7361 ultra-rapid||1973–1976||33||12 Gy in 2 Fx over 2 days||13 wks|
|RTOG 7606 favorable pts||1976–1979||130||30 Gy in 10 Fx over 2 wks||18 wks|
|125||50 Gy in 20 Fx over 4 wks||17 wks|
|RTOG 8528||1986–1989||30||48 Gy in 1.6-Gy bid||4.8 months|
|Accelerated||53||54.5 Gy in 1.6-Gy bid||5.4 months|
|HFX||44||64 Gy in 1.6-Gy bid||7.2 months|
|36||70.4 Gy in 1.6-Gy bid||8.2 months|
|RTOG 9104||1991–1995||213||30 Gy in 10 Fx||4.5 months|
|216||54.4 Gy in 1.6-Gy fx bid||4.5 months|
|RTOG 7916||1979–1983||193||30 Gy in 10 Fx over 2 wks||4.5 months|
|Misonidazole||200||5 Gy in 6 Fx over 3 wks||4.1 months|
|196||30 Gy in 10 Fx+Miso||3.1 months|
|190||5 Gy in 6 Fx+Miso||3.9 months|
|RTOG 8905||1989–1993||36||37.5 Gy in 15 Fx over 3 wks||6.1 months|
|BrdU||34||37.5 Gy in 15 Fx+BrdU||4.3 months|
The RTOG 8528 trial investigated accelerated hyperfractionated WBRT, with patients receiving 1.6-Gy twice-daily fractionation to a total dose of 48 Gy, 54.4 Gy, 64 Gy, or 70.4 Gy. Although median survival times seemed to increase with increasing dose (4.8 months, 5.4 months, 7.2 months and 8.2 months), this apparent difference was not statistically significant. A follow-up randomized phase III RTOG study included 445 patients with Karnofsky performance status scores of≥60 who were given either hyperfractionated RT at 1.2-Gy twice-daily fractionation to a total tumor dose of 54.4 Gy or 30 Gy in 10 fractions ( ). The median survival time was 4.5 months in both treatment groups. Given the lack of superiority of the hyperfractionated regimens and the toxicity of ultra-rapid high-dose fractionations, modern clinical trials involving WBRT for gross disease typically use either 30 Gy in 10 fractions or 37.5 Gy in 15 fractions.
Various radiosensitizers have also been investigated in prospective randomized clinical trials. The RTOG 7916 trial evaluated one such radiosensitizer, misonidazole, in patients with brain metastases. This study involved 779 patients who received 30 Gy in 10 fractions over 2 weeks with or without 1 g/m 2 of misonidazole or 30 Gy in six fractions over 3 weeks with or without 2 g/m 2 misonidazole. The median survival times ranged from 3.1 months to 4.5 months and were not affected by the use of misonidazole. Other sensitizers tested have included the halogenated pyrimidine bromodeoxyuridine (BrdU) and gadolinium texaphyrins. To investigate the role of BrdU as a radiosensitizer for brain metastases, RTOG 89-05 enrolled 72 patients and treated them with WBRT to 37.5 Gy in 15 fractions over 3 weeks, with BrdU at 0.8 g/m 2 per day for 4 days given each week for 3 weeks. Five patients who received BrdU manifested significant grade 4 and 5 hematologic or skin toxicity, and no significant survival benefit was found from BrdU.
Another randomized controlled trial, RTOG 9801, evaluated the role of motaxefin gadolinium (MGd) as a radiosensitizer for patients with brain metastases treated with WBRT ( ). In that study, 401 patients received either WBRT alone (30 Gy in 10 fractions) or WBRT with MGd, injected intravenously at a dose of 5 mg/kg/d at 2–5 hours before each fraction. Receipt of MGd did not affect median survival times, at 5.2 months for MGd and WBRT and 4.9 months for WBRT alone. Regression after WBRT correlated with survival and preservation of neurocognitive function ( ); in that analysis, time to neurologic progression was longer among patients with metastatic lung cancer receiving MGd. In a subsequent phase III study, 554 patients with brain metastases from non-small cell lung cancer (NSCLC) were randomly assigned to receive the same treatments as those given in RTOG 9801 ( ). The median interval to neurologic progression seemed to be better in the MGd group (15.4 months vs 10.0 months), although this apparent difference was not statistically significant at p =0.11.
Thalidomide was investigated as a radiosensitizer in patients with brain metastases in RTOG 0118 ( ), in which patients were randomly assigned to receive WBRT alone (37.5 Gy in 15 fractions) or WBRT with thalidomide. Nearly half of the patients had to discontinue thalidomide because of side effects, and no survival benefit was noted. In the latest published study of radiosensitizers with WBRT, temozolomide or erlotinib was given concurrently with WBRT and SRS for patients with NSCLC and one to three brain metastases in RTOG 0320 ( ). In that phase III trial, the addition of radiosensitizer increased the rates of grade 3–5 toxicity (11% control vs 41% temozolomide vs 49% erlotinib, p <0.001) and seemed to reduce the median survival time (13.4 months control vs 6.3 months temozolomide vs 6.1 months erlotinib). Other radiosensitizers continue to be investigated for their potential efficacy in combination with WBRT.
WBRT as an Adjuvant after Local Therapy
WBRT After Surgery
WBRT has an established role in preventing local recurrence after surgical resection of brain metastases. In one prospective, randomized trial reported by , 95 patients with a single brain metastasis underwent complete resection followed by either observation or postoperative WBRT to a dose of 50.4 Gy. Receipt of WBRT led to a significant reduction in recurrence anywhere in the brain compared with the observation group (18% vs 70%; p <0.001). Rates of local recurrence in the tumor bed were reduced from 46% to 10%, and rates of regional recurrence within the brain and outside the tumor bed were reduced from 37% to 14%. WBRT also reduced the rate of deaths from neurologic causes from 44% to 14% but did not affect median survival.
Most of the retrospective studies evaluating the role of WBRT after surgical resection ( ) found no significant reduction in brain recurrence after postoperative WBRT except for a study by DeAngelis et al. ( p =0.03) ( ). Only two retrospective studies have shown an improvement in median survival time after postoperative WBRT ( p =0.02) ( ).
In general, more patients who receive WBRT at recurrence die from neurologic causes than do those given WBRT immediately after surgery, findings that support the immediate postoperative delivery of WBRT. Although the routine use of postoperative WBRT seems to prevent deaths from neurologic causes, this practice has become controversial because of growing concerns about the long-term toxicity of WBRT, including neurocognitive decline and dementia. Withholding postoperative WBRT carries the risk of increased neurologic morbidity from tumor recurrence within the brain; thus, if postoperative WBRT is to be given, it should be given promptly, usually within a few weeks of resection, depending on the aggressiveness of the surgical procedure and the patient’s rate of postoperative recovery.
WBRT after Stereotactic Radiosurgery
WBRT after SRS has also been evaluated in prospective clinical trials. In a study reported by , 132 patients with one to four brain metastases smaller than 3 cm in diameter were randomly assigned to undergo SRS or SRS followed by WBRT to a total dose of 30 Gy in 10 fractions. The SRS dose was reduced by 30% for the patients who were to be given WBRT. The addition of WBRT to SRS improved the 1-year local control rate (89% vs 73%) and decreased the incidence of new brain metastases at 1 year (42% vs 63.7% for SRS alone). Because no differences were found in median survival times, percentages with neurologic death, or neurocognitive outcomes, the authors concluded that SRS alone could be considered a reasonable strategy, with WBRT reserved for salvage therapy as needed.
In a similar study, enrolled 58 patients with one to three brain metastases and RTOG recursive partitioning analysis class 1 or 2 status to receive either SRS alone or SRS followed by WBRT to a dose of 30 Gy in 12 fractions. Notably, the primary endpoint in this study was neurocognitive function as defined by the Hopkins Verbal Learning Test-Revised (HVLT-R) of total recall at 4 months. The study was stopped early when it became clear that SRS followed by WBRT led to declines in learning and memory function. Indeed, the addition of WBRT led to a significant decline in HVLT-R 4-month recall (52% vs 24%); however, use of WBRT led to improved rates of local control at 1 year (100% vs 67%) and distant brain control at 1 year (73% vs 45%). Conversely, median overall survival time was longer in the SRS-only group (15 months vs 6 months SRS+WBRT).
The European Organisation for Research and Treatment of Cancer (EORTC) also undertook a randomized trial with 359 patients undergoing either SRS ( n =199) or surgery ( n =160) for one to three brain metastases who were then randomly assigned to receive WBRT (30 Gy in 10 fractions) or observation ( ). Those who went on to receive WBRT had reduced rates of relapse at initial sites at 2 years (surgery: 27% vs 59%; SRS: 19% vs 31%) and relapse at new sites (surgery: 23% vs 42%; SRS: 33% vs 48%). The rates of neurologic death were also lower among those who received WBRT (28% vs 44%). No differences were found in time to worsening performance status or overall survival.
The need for WBRT after SRS for the initial treatment of brain metastases continues to be debated. Although distant intracranial control is undoubtedly improved by the addition of WBRT, the lack of survival benefit and toxicity argue against the routine use of WBRT after SRS. Clinical trials are ongoing to evaluate survival, quality of life, and functional independence for such patients.
Prophylactic Cranial Irradiation
PCI has been noted to improve survival among children with acute lymphocytic leukemia by preventing the disease from involving the central nervous system (CNS). With the increasing use of more effective types of systemic chemotherapy, isolation of the malignancy in the CNS has become more common among patients with acute lymphocytic leukemia, a phenomenon that has also been noted in patients with small cell lung cancer (SCLC).
PCI has been proposed for patients with SCLC since the 1970s because of the high incidence of brain metastases from SCLC. Approximately half of patients with SCLC who do not receive PCI will develop clinically significant brain metastases. However, the incidence of microscopic brain metastases and extracranial CNS metastases had been underestimated given the relatively short survival duration for such patients, and PCI became more important when improvements in systemic treatment and earlier use of thoracic RT were shown to improve survival rates for patients with limited-stage SCLC ( ). Because PCI can be neurotoxic and because it had not improved overall survival times, it generally had not been considered for routine use despite its ability to reduce the number of brain metastases until neuropsychological testing of patients before and after PCI revealed no significant mental deterioration ( ).
Trials of PCI for SCLC date back to the late 1970s, when the Veterans Affairs Lung Study Group found that giving PCI to a dose of 20 Gy in five fractions did not reduce brain metastasis. At the time, this was thought to have resulted from the low total dose of RT (20 Gy), although the fraction size was considerably larger than the 2- to 3-Gy fractions in subsequent trials of PCI. Indeed, subsequent studies involving 24 to 30 Gy, delivered in 3-Gy fractions, have shown that PCI can reduce the incidence of brain metastases, although this effect is generally not associated with prolonged overall survival. Other groups have also found that PCI produced statistically significant reductions in brain metastasis but did not affect overall survival. Use of radiation doses in excess of 30 Gy seemed to reduce brain tumor recurrence in studies conducted in the late 1980s, although this may also have been related to the advent of more sensitive means of detecting brain metastases, improvements in systemic chemotherapy, and the use of thoracic RT ( ). reported a series of 111 patients randomly assigned to receive or not receive PCI to 40 Gy, which did not significantly reduce the incidence of brain metastasis (9% with PCI vs 13% without PCI). This finding was thought to be related to the late use of the PCI, which was begun 12 weeks after completion of the prolonged systemic treatment.
Twenty years after publication of the previous report, a meta-analysis by showed a survival benefit from PCI for patients with SCLC in complete remission. That analysis involved individualized data from 987 patients who had participated in seven clinical trials comparing PCI with no PCI from 1977 to 1994 ( Table 20.2 ) ( ). A minority of patients (12–17%) had extensive-stage disease. An absolute increase of 5.4% in survival rate at 3 years was noted for patients undergoing PCI (20.7% vs 15.3% no PCI), and the relative risk of death for the treatment group compared with the control group was 0.84 (95% confidence interval 0.73–0.97, p =0.01). PCI also led to a significant increase in the rate of brain metastasis-free survival at 3 years (22.3% vs 13.5% in control patients, p <0.001), and PCI reduced the cumulative incidence of brain metastasis at 3 years from 58.6% to 33.3% ( p <0.001). Analysis of four total radiation doses (8 Gy, 24–25 Gy, 30 Gy, and 36–40 Gy) revealed that the larger doses led to greater decreases in the risk of brain metastasis ( p =0.02), but the effect on survival was not proportional to the dose administered. This meta-analysis also confirmed the finding reported by that earlier use of PCI after induction chemotherapy reduces the risk of brain metastasis ( p =0.01 for<4 months vs 4–6 months vs>6 months after chemotherapy). No significant neuropsychological deterioration after PCI was noted, but most of these studies did not include systematic evaluations with neuropsychological testing before and after PCI.