Gemcitabine and cisplatin combination (Gem-Cis) is a commonly used regimen in metastatic breast cancer (MBC), with proven activity in phase II trials. It is mostly used as a salvage regimen for progressive disease refractory to anthracyclines and taxanes, and when liver dysfunction secondary to liver metastasis precludes these drugs. Retrospective review of medical charts was conducted for patients treated with Gem-Cis for MBC in a single institution in Brazil between January 2004 and July 2007. The purpose of this study was to evaluate the outcomes and toxicity of Gem-Cis in a broad indication, including patients with deteriorated performance status (PS) and liver dysfunction, which were excluded from clinical trials. Fifty-six patients were included. Median age was 52 years, 46.4% were hormone-receptor negative, 57.2% received 3 or more prior chemotherapy lines, and 34 had liver metastasis. The median overall survival (OS) was 7.6 months, the median progression-free survival was 3.3 months, and the response rate was 21.2%. In variable analysis, PS was significantly associated with OS, even after adjusting to other factors. Toxicities included grades 3 or 4 anemia in 19.3%, neutropenia in 21.1%, and thrombocytopenia in 12.3%. Gem-Cis was a relatively active combination in this population that typically carries a poor prognosis. The subgroup of patients with favorable PS experienced longer survival, even when liver metastasis and hepatic dysfunction were a concern. Toxicity was manageable and it was not correlated with PS or liver dysfunction.

e17568 Background: Liver metastasis (mets) from breast cancer is typically associated with poor prognosis. GC is a good option, since patients (pts) have often failed anthracycline and taxane therapy and liver dysfunction may preclude these regimens. Methods: Retrospective study designed to evaluate the clinical outcomes and predictive factors in pts with metastatic breast cancer treated with GC, with special interest for liver mets. Results: From 2004 to 2007, 56 pts were treated with GC. Median age was 52.1 years, 33 pts had PS 0–1, 26 were hormone receptor (HR) negative, 32 had been treated with 3 or more chemotherapy (CT) regimens and 34 had liver mets. The median overall survival (OS) was 7,1 mo (95% CI; 4.3–9.7), the progression free survival was 3.3 mo (95% CI; 2.2–5.5) and the clinical response rate was 25%. OS was 4.0 mo (95% CI; 2.3–8.9) for pts with liver mets and 9.7 mo (95% CI; 6.9–12.9) for pts without liver mets (p = 0.03). No factor showed correlation with OS in pts with liver mets, including age < 45 years (median OS [95% CI]: 2.8 mo [1.8–11.5] versus 4.5 mo [2.5–8.9]; p = 0.74), PS 0–1 (median OS [95% CI]: 7.3 mo [2.5–11.5] versus 2.8 mo [0.8–9.7]; p = 0.55), HR positivity (median OS [95% CI]: 7.3 mo [2.0–11.5] versus 5.9 mo [1.9–9.7]; p = 0.87), bilirubin level ≥ 5 times the superior limit of the normality (median OS [95% CI]: 5.9 mo [0.5–16.7] versus 4.5 mo [2.5–9.7]; p = 0.45), progression free interval after the previous CT ≤ 1 mo (median OS [95% CI]: 4.0 mo [2.0–9.7] versus 7.1 mo [2.5–14.7]; p = 0.93) and previous treatment with ≥ 3 CT regimens (median OS [95% CI]: 7.3 mo [1.0–11.0] versus 4.0 mo [2.0–9.7]; p = 0.93). Conclusions: These data is in accordance with the literature concerning the dismal prognosis in liver mets from breast cancer and the clinical activity of GC. We could not define predictive factors in this cohort, which was probably due to the relatively small number of pts. No significant financial relationships to disclose.