RESUMEN
Treatment of breast cancer meningeal carcinomatosis (MC) relies on intrathecal chemotherapy. Thiotepa is one of the few drugs approved in this setting, although no large cohort has been reported. The aim of our retrospective study is to describe survival and prognostic factors of breast cancer patients treated by intrathecal thiotepa. A search in the electronic database of the Institut Curie was performed and retrieved the patients diagnosed with breast cancer MC from 2000 to 2012 and who received at least one intrathecal injection of thiotepa. The standard regimen was intrathecal thiotepa (10 mg) and methylprednisolone (40 mg), repeated every other week. Clinical data were retrieved from the computerized medical file of each patient. Sixty-six patients have been treated with intrathecal thiotepa either as first line or second line of treatment for breast cancer MC. The median overall survival was 4.5 months (range 0.1-50). There was no significant survival difference between patients treated as first or second line. In multivariate analysis, main adverse prognostic factors at diagnosis were performance status >2 (p = 0.001, RR = 3.4, 95 % CI 1.6-7.2) and history of more than 3 previous systemic chemotherapy lines (p = 0.002, RR = 2.90, 95 % CI 1.50-5.65). After start of the treatment, high primary tumor grade, elevated Cyfra 21-1 levels in the cerebrospinal fluid, and lack of clinical improvement were also independent adverse prognostic factors in multivariate analysis. This is the largest retrospective cohort of breast cancer MC treated by intrathecal thiotepa ever reported. The median overall survival was short but some patients clearly benefited from this treatment, even used as second line.