A phase 2 trial of whole-brain radiotherapy combined with intravenous chemotherapy in patients with brain metastases from breast cancer




A study was conducted to determine the efficacy, tolerability, and safety of concurrent cisplatin and vinorelbine chemotherapy and radiotherapy in patients with previously untreated brain metastases from breast cancer.


Twenty-five patients with untreated brain metastases from breast cancer were treated with cisplatin (at a dose of 20 mg/m2/day, Days 1-5) and vinorelbine (6-mg/m2 bolus on Day 1 and 6 mg/m2/day continuous infusion on Days 1-5) chemotherapy combined with concurrent 30-gray fractionated external-beam radiotherapy. Chemotherapy was given at 3-week intervals for a total of 4 cycles. Primary endpoint was the rate of radiologic response of brain metastases.


Complete response in the brain was observed in 3 patients, and partial response was noted in 16 patients, yielding a 76% response rate in the brain. The overall systemic response rate was 44%. Progression-free and overall survival were 3.7 months and 6.5 months, respectively. Overall toxicity was acceptable; nonhematologic grade 3-4 events were noted in 5 (20%) patients, and there were no toxic deaths.


Concurrent chemoradiation with cisplatin and vinorelbine for brain metastases from breast cancer appears to be active and well tolerated. Cancer 2008. © 2008 American Cancer Society.

Approximately 10% to 30% of patients with metastatic breast cancer will develop central nervous system (CNS) involvement (ie, leptomeningeal involvement or brain metastases) in the course of their disease.1 CNS involvement is associated with debilitating neurologic symptoms, such as focal weakness, headache, and seizure. Furthermore, brain metastases are associated with poor prognosis; the median survival of these patients is usually ≤6 months, and has not improved in the last 15 to 20 years.2–4

Whole-brain radiotherapy (WBRT) remains the cornerstone of treatment in the management of patients with brain metastases, resulting in symptomatic improvement in the majority of patients. However, response to WBRT is usually short-lasting, and impact on survival is usually limited. Aggressive local therapies such as radiosurgery or neurosurgery combined with WBRT have shown promising results, but only apply to patients with few (1-3) brain metastases.5, 6 The role of systemic chemotherapy in the management of patients with brain metastases remains unclear, largely because of issues regarding diffusion of cytotoxic drugs through the blood-brain barrier. However, some investigators have reported response rates in the 30% to 50% range, using in most cases cisplatin-based regimens.7–9

Concomitant chemotherapy and radiotherapy have been shown to improve the rate of local control over radiotherapy alone in diseases such as cancers of the lung, head and neck, or rectum and glioblastoma.10–13 The addition of chemotherapy to standard radiation could improve the activity of radiation on brain disease through radiosensitizing mechanisms and allow the concomitant treatment of other sites of disease.

We conducted a prospective phase 2 trial to assess the activity and toxicity of cisplatin and vinorelbine (Navelbine; Pierre Fabre Médicament, Boulogne, France) combination chemotherapy with concurrent external-beam radiotherapy in patients with previously untreated brain metastases from breast cancer. Cisplatin was chosen 1) because cisplatin-based combinations were shown to be active in the treatment of brain metastases14; 2) because autopsy series showed therapeutic concentration of cisplatin in brain metastases, but not in the adjacent normal brain tissue15, 16; and 3) because of its known role as a radiosensitizer. Vinorelbine was chosen because this drug had clinical activity in anthracycline- and taxane-pretreated breast cancer,17 and we believed that cisplatin alone would not be sufficient to treat systemic disease. Furthermore, data from studies in patients with brain metastases from lung cancer demonstrated that concomitant chemoradiation with cisplatin and vinorelbine was feasible.18 The primary endpoint was radiologic response rate of brain metastases, and secondary endpoints included overall and progression-free survival (PFS), safety, and tolerability. Preliminary results of this phase 2 trial were presented as a poster at the 27th San Antonio Breast Cancer Symposium, December 8 to 11, 2004.19


Study Population

Patients with histologically proven metastatic breast cancer were eligible if they met the following criteria: female; aged between 18 and 70 years; World Health Organization (WHO) performance status of 0, 1, or 2; life expectancy of >3 months; absolute neutrophil count >1.5 × 109/L; platelet count >100 × 109/L; and serum creatinin <106 μmol/L (11.8 mg/L). Prior chemotherapy, whether administered in the adjuvant setting or for metastatic disease, was allowed. Hormonal treatment started before study entry could be continued during the study. Patients were required to provide written informed consent.

Patients bearing a single metastasis that could be treated by either surgery or radiosurgery were not eligible. Other exclusion criteria included the following: progression during previous cisplatin-containing and/or vinorelbine-containing chemotherapy; pregnancy or lactation; progressive brain metastases inside a previously irradiated field; other serious illness or medical or psychiatric condition; and history of other neoplasm (except for curatively treated basocellular carcinoma of the skin or cervical carcinoma in situ).

The study was approved by the institutional ethics committee of the Leon Berard Center in Lyon, and all patients provided written informed consent.

Study Design and Statistics

This was a multicentric, open-label, phase 2 trial designed to assess the efficacy of WBRT combined with cisplatin and vinorelbine (the CIVIC regimen17). The primary endpoint was the response rate of brain metastases. Other endpoints included assessment of toxicity, PFS, overall survival (OS), and systemic response rate.

Statistical design was done according to the Gehan method.20 Briefly, 14 patients were to be accrued at first; if no response was observed then the trial was to be terminated, because the probability of further observing a response rate of at least 20% would be <5%. If at least 1 response was observed among the first 14 patients, then more patients were to be included.

Data were analyzed on an intent-to-treat basis. PFS was calculated from the day of study entry to disease progression or death, whichever came first. PFS in the brain was calculated from the date of study entry to the date of brain progression (progression at other sites and death were considered as censored data), and OS was calculated from the date of study entry to death from any cause. All curves were displayed using the Kaplan-Meier method.


Patients were to receive a standard 30-gray (Gy) course of WBRT over 2 weeks (3.0 Gy/fraction/day × 10 days), using high-energy photons. Chemotherapy was to be initiated within 7 days after the start of radiation. The CIVIC regimen has been previously described17 and comprises vinorelbine in a 6-mg bolus on Day 1, followed by vinorelbine at a dose of 6 mg/m2/day by continuous infusion on Days 1 to 5 and cisplatin at a dose of 20 mg/m2/day on Days 1 to 5. Cycles were repeated every 3 weeks for 4 cycles. In case of febrile neutropenia, grade 4 neutropenia without fever lasting >6 days, grade 3 or 4 thrombocytopenia, or grade 2 nonhematologic toxicity (except alopecia), the dose for the following cycles was to be reduced by 20%. Patients experiencing febrile neutropenia or grade >2 nonhematologic toxicity after a dose-reduced cycle were withdrawn from the study. During radiotherapy, patients received corticosteroids at a dose of 1 to 2 mg/kg/day and 20% mannitol if signs of increased intracranial pressure appeared.


Prestudy evaluation included a history, physical examination, hematology, serum biochemistry tests, and serum CA 15-3. Brain metastases were assessed using magnetic resonance imaging (MRI) or a computed tomography (CT) scan if MRI was not possible. Other targets were evaluated using standard techniques including CT scan, ultrasonography, and bone scintigraphy when indicated. Physical examination and complete blood cell count were repeated before each cycle. Complete evaluation was repeated, using the same imaging techniques used for prestudy evaluation, after Cycles 2 and 4. Response of brain metastases (also referred to as “brain response”) was assessed using criteria defined by Macdonald et al.21 Extracerebral response (also referred to as “systemic response”) was assessed according to WHO criteria22 and included the assessment of nonmeasurable disease. When multiple brain lesions were present, only 3 were serially measured for response assessment. Patients experiencing overt clinical progression or in whom progression was documented were withdrawn from the study and were followed until death; further treatments were left to the referring physician's discretion. All toxicities were graded according to the National Cancer Institute Common Toxicity Criteria v2.



Between March 1998 and January 2000, 14 patients were included in the first step of this trial. Seven partial and 3 complete responses in the brain and 5 partial responses and 1 complete systemic response were observed. Nine more patients were therefore included in the second step, for a total of 25 patients accrued between March 1998 and March 2001. Their main characteristics upon study entry are summarized in Table 1. Briefly, the median age of the patients was 47 years (range, 31 years-70 years), 17 patients (74%) had Scarff-Bloom-Richardson grade 3 tumors, 13 (52%) patients had previously received adjuvant chemotherapy, and 14 (56%) patients had previously received chemotherapy for metastatic disease. Twenty-one patients (84%) had previously received an anthracycline-containing regimen, 10 had previously received taxane, and 9 had received both. Overall, only 2 patients were chemotherapy naive before entering the study. All patients had brain metastases; the median number of brain metastases was 2 (range, 1 to >10 brain metastases), 12 (48%) patients had ≥3 brain metastases, and the median number of metastatic sites (including the brain) was 3 (range, 1-4 metastatic sites). Although it was not an exclusion criterion, none of the patients included in this study had previous surgery or radiosurgery for brain metastases. Four patients who had a single brain metastasis were included after they were deemed inoperable by a neurosurgeon.

Table 1. Patient Characteristics
  • WHO PS indicates World Health Organization performance status; ER, estrogen receptor; +, positive; PR, progesterone receptor; −, negative.

  • *

    Data missing for 2 patients.

  • Data missing for 10 patients.

  • Data missing for 4 patients.

Median age, y (range)47.4 (31.2-69.9) 
Histologic type*  
 Invasive ductal carcinoma2087
Tumor grade* 0
 HER-2 overexpression747
Hormone receptor status  
Brain metastases  
Other sites of disease  
 Other (eg, skin, soft tissue)416
 Median no. of metastatic sites (range)3 (1-4) 
Prior chemotherapy  
 Median no. of previous regimens (range)1 (0-3) 

Early Treatment Discontinuation and Toxicity

The median time between the initiation of WBRT and the initiation of chemotherapy was 2 days (range, −2 to +7) days; only 2 patients began chemotherapy >4 days after the beginning of WBRT (Day 6 and Day 7, precisely). Eleven (44%) patients discontinued treatment early either because of disease progression in 7 (28%) patients or toxicity in 4 (16%) patients. Early progression was both systemic and cerebral in 4 patients and was systemic only in 3 patients. Two patients experienced grade 3 toxicity that could be related to WBRT (nausea in 1 patient and depression in 1 patient), and there were no grade 4 adverse events during concomitant chemoradiation. Overall, the majority of toxicities from WBRT were mild (Table 2). However, 23 of 25 patients received prophylaxis for intracranial hypertension with either steroids, mannitol, or both. The most common grade 3-4 toxicities are displayed in Table 2. Grade 3-4 hematologic toxicities were noted in 20 (80%) patients, and was the most common toxicity; 5 (20%) patients experienced a grade 3 nonhematologic toxicity (asthenia in 3 patients, depression in 1 patient, and vomiting in 1 patient).

Table 2. Main Toxicities and Reasons for Early Study Discontinuation
  1. WBRT indicates whole-brain radiotherapy.

Main Grade 3 Toxicitiesn%
Hematologic toxicity >grade 2  
Grade 3 asthenia (no grade 4 reported)624
Neurologic (WBRT)14
Nausea/vomiting (WBRT)14
Other Toxicities of WBRTGraden%
Erythema of the forehead114
Reason for Early Study Discontinuationn%
Disease progression728

Response Rate, PFS, and OS

Response rates are displayed in Table 3. Response in the brain could not be evaluated in 1 patient because of the use of different imaging techniques between baseline and other assessments. A second patient could not be evaluated because she died before the first assessment. Complete disappearance of brain metastases was observed in 3 (12%) patients, and a partial response was noted in 16 (64%) patients, yielding a 76% response rate for brain metastases. Two further patients had stable disease for 2.4 months and 3.4 months, respectively. Eight and 17 patients were assessed with CT scan and MRI, respectively. Brain response rates were not found to be significantly different between patients assessed with MRI and CT: 13 of 15 (87%) patients (2 patients assessed with MRI at baseline could not be evaluated) and 6 of 8 (75%) patients, respectively (P = .49). The median PFS in the brain, calculated from the date of study entry to the date of brain progression, was 5.2 months (range, 0.5 months-53.3 months) (Fig. 1). The brain was the first site of disease progression in 16 patients. For responding patients, the median duration of response was 8.5 months.

Figure 1.

Progression-free survival (PFS) in the brain is shown.

Table 3. Response Rates per Site
Site/Responsen%ORR (CR+PR), %Median PFS, mo (Range)
  1. ORR indicates overall response rate; CR, complete response; PR, partial response; PFS, progression-free survival; SD, stable disease; PD, progressive disease.

Brain  765.2 (0.5-53.3)
 Could not be assessed28  
Systemic (brain+other sites)  443.7 (0.2-46.5)

Systemic response (ie, cerebral and extracerebral disease) was complete in 1 patient (4%) and partial in 10 (40%), for an overall systemic response rate of 44%. Five patients had stable disease for a median of 4.2 months (range, 2.8 months-6.8 months). Systemic PFS, calculated from the date of study entry to the date of disease progression (brain or other organ) or death, whichever came first, was 3.7 months (range, 0.2 months-46.5 months) (Fig. 2). Four (16%) patients were progression free at 12 months. In 3 patients a partial response of brain metastases was observed, whereas extracerebral disease progressed.

Figure 2.

Overall (systemic and brain) progression-free survival (PFS) and overall survival (OS) are shown.

At the time of disease progression 18 (72%) patients received salvage therapy; some of them received several lines of chemotherapy, either for cerebral or for systemic disease progression or both.

The median OS was 6.5 months (range 0.5 months-62.1 months) (Fig. 2). Seven (28%) patients were alive at 12 months. In an exploratory analysis, patients were stratified by recursive partitioning analysis (RPA) class, based on the work by Gaspar et al from the Radiation Therapy Oncology Group.23 Patients from classes I and II were pooled because of the small number of patients in class I (1 patient). As shown in Figure 3, patients classified as RPA class III had a significantly poorer outcome, with a median OS of 4.2 months, compared with 8.4 months for patients classified as RPA class I or II (P = .0026).

Figure 3.

Survival is shown according to the Radiation Therapy Oncology Group recursive partitioning analysis (RPA) class. Classes I and II were pooled.


The management of patients with brain metastases is challenging, and treatment options have remained virtually unchanged for more than 2 decades. Concurrent chemoradiation is currently being investigated with the aim of improving the control rate of both cerebral and systemic disease. Molecularly targeted agents are also currently under investigation for the treatment of patients with brain metastases.24, 25 Another direction of current research is focusing on patients' selection, to offer more aggressive therapy for patients who are likely to benefit from the added treatment intensity. The RPA classification is currently recognized as a major prognostic factor in patients with brain metastases and is based on simple clinical factors: Karnofsky performance status, age, and the control of extracerebral disease.

The goal of the current trial was to assess the efficacy and toxicity of radiotherapy and concurrent chemotherapy with cisplatin and vinorelbine for patients with brain metastases from breast cancer. This combination shows a high rate of objective response in the brain (76%) combined with an interesting systemic response rate (44%) in the setting of taxane- and/or anthracycline-pretreated patients. By using this regimen, the median brain and systemic PFS were 5.2 months and 3.7 months, respectively, with a median OS of 6.5 months. Overall grade 3-4 nonhematologic toxicities were infrequent, which combined with the systemic response rate, is consistent with our previously reported experience with the CIVIC regimen.17

Other studies of concurrent chemoradiation for brain metastases of solid tumors have reported response rates of 50% and up to 96%.26, 27 However, the median PFS and OS times remain disappointingly low in most studies. In their study of concurrent topotecan and WBRT, Gruschow et al reported a median survival of 5 months,26 which is comparable to what was previously reported for WBRT alone.4 Antonadou et al reported a significant improvement in response rate using temozolomide and concomitant radiation as compared with radiation alone in a randomized phase 2 trial. This improvement in response rate did not translate into a significant advantage in median OS, although there appeared to be a trend toward improved survival in the chemoradiation arm.27 The PFS and OS in both of these studies is comparable to those reported in our study.

Several studies have assessed the role of chemotherapy for the treatment of brain metastases. Rivera et al reported that combination chemotherapy using temozolomide and capecitabine was efficient and well tolerated in treating brain metastases from breast cancer, including brain metastases recurring after radiotherapy.28 Christodoulou et al reported on the use of cisplatin-temozolomide combination chemotherapy for patients with brain metastases from solid tumors. In their study, the response rate was 30%, and 16% of patients had stable disease.8 However, in both of these studies PFS and OS were short.8, 28 These studies demonstrate that, as was previously shown in nonsmall cell lung cancer, the response rate and disease control rate of brain metastases with chemotherapy are comparable to those of systemic disease18; the addition of WBRT to chemotherapy could improve the response rate at this site and therefore reduces the prevalence of neurologic symptoms, thereby improving quality of life.

In conclusion, the results of the current study show that chemotherapy with cisplatin and vinorelbine combined with concurrent WBRT is feasible and induces a high response rate in brain metastases from breast cancer while providing an interesting response rate of systemic disease, and has an acceptable toxicity profile. However, despite this high rate of cerebral control, benefit in OS appears to be marginal compared with radiotherapy alone. This combination could therefore have interest for patients with both cerebral and extracerebral disease progression.