Risk factors for brain metastases in patients with a metastatic relapse of breast cancer

Abstract

837 Background: Brain metastases are a more and more frequent, difficult clinical problem in advanced breast cancer (BC) patients (pts). We searched for risk factors of brain metastasis (BM) among pts diagnosed with a metastatic relapse of BC. Methods: A total of 557 premenopausal BC pts receiving systematic adjuvant chemotherapy were accrued at Institut Gustave-Roussy in a randomized trial of ovarian suppression [Arriagada et al, ASCO 03]. After a median follow-up of 9.5 years, 215 developed a first metastasis at a site other than brain alone. Of these 215 metastatic patients, 31 developed BM. The risk of BM in these 215 patients was analyzed as a function of patient, tumor and treatment characteristics by Cox models. Risk factors identified in the multivariate analysis were re-evaluated in a confirmatory series of 199 patients with metastatic relapse registered at Institut Gustave-Roussy in two randomized adjuvant chemotherapy trials (691 postmenopausal node-positive pts [Arriagada et al, Ann Oncol 2002] and 247 premenopausal node-negative pts [Arriagada et al, submitted]). Of these 199 pts, 19 developed BM. Results: In the first series, risk factors for BM in univariate analyses were lung metastasis as first site (yes/no): 2-year incidence rates of BM 24% vs 7% (p<0.001), negative hormone receptor (HR- vs HR+): 2-year BM rates 34% vs 5% (p<0.001) and short interval between BC and metastatic relapse (≤ vs >24 months): 18% vs 6% (p<0.008). In the multivariate analysis, only lung metastases remained prognostic (RR 5.3, 95% confidence interval (95%CI): 2.6–11, p<0.001). This was confirmed in the second series where the 2-year rates of BM were respectively 44% [95% CI: 26%–64%] and 3% [95%CI: 1–7%] in pts with or without lung metastases (RR 10 [95%CI: 3.8–27], p< 0.0001). Conclusions: This study shows that lung metastasis is a strong risk factor for the occurrence of brain metastases among BC patients. The occurrence of brain and lung metastases could be due to a common, albeit unknown, biological determinant. We suggest that prophylaxis of brain metastases may be evaluated in BC patients with lung metastases. No significant financial relationships to disclose.