The risk of central nervous system metastases after trastuzumab therapy in patients with breast carcinoma

BACKGROUND. Trastuzumab, which is a large monoclonal antibody that is efficacious in the treatment of patients with HER-2/neu-overexpressing, metastatic breast carcinoma, does not penetrate the blood-brain barrier and, thus, may allow the brain to become a sanctuary site for micrometastases. Few studies have compared the risk of central nervous system (CNS) metastases in patients treated with or without trastuzumab. METHODS. The authors conducted a retrospective cohort study that compared 264 patients who did not receive trastuzumab therapy with 79 patients who received trastuzumab therapy. The study was powered to detect an effect size of 0.3, which was deemed clinically significant to change future management. RESULTS. CNS metastases developed in 48.1% of patients on trastuzumab-based therapy and in 46.6% of patients on nontrastuzumab-based therapy. The association between trastuzumab therapy and subsequent CNS metastases (either brain or leptomeningeal) was not significant, with a multivariate- adjusted odds ratio of 0.91 (95% confidence interval, 0.64-1.88; P = 0.79). Similarly, there was no evidence of an association between trastuzumab and brain metastases alone (P = 0.67) or leptomeningeal metastases alone (P = 0.14). The median overall survival after the diagnosis of all CNS metastases was 26.3 months for patients who did not receive trastuzumab and 24.9 months for patients who received trastuzumab (P = 0.7). A multivariate logistic regression model found that patient age at diagnosis (P < 0.05), positive lymph node status at presentation (P < 0.01), and liver metastases (P < 0.01) were significant predictors of CNS metastases. Lung metastases showed a borderline significant P value (0.056). CONCLUSIONS. Despite the impression of many oncologists, the results of this study did not support an association between trastuzumab therapy and an increased risk of CNS metastases. (C) 2004 American Cancer Society.