10579 Background: Breast cancer patients (pts) with liver metastases have a poor prognosis. HAI has been frequently used for liver metastases from colorectal cancer, with better response and possibly survival than with intravenous administration. We evaluated the activity of HAI in pretreated breast cancer patients with liver metastases only. Methods: From December 2001 to October 2005 thirty patients previously treated with anthracyclines, taxanes and vinorelbine were enrolled. Median age 58 years (44–68) and median PS 1 (0–2). Eight patients (27%) had 5 or more liver nodes, seventeen (57%) had 5 to 10 metastases and five (16%) more than 10 nodes. Eighteen patients received 2 previous lines of chemotherapy (67 %), eleven 3 lines (30%) and one (3%) three lines and one cicle of high dose chemotherapy followed by ABMT. Seventeen pts received prior hormonotherapy. Two patients had been submitted to liver metastasectomy. Patients underwent percutaneous implantation of an arterial hepatic port-a-cath. HAI regimen administered on days 1–3 every 4 weeks, consisted of cisplatin 10 mg/sqm bid, mitomycin-C 1 mg/sqm bid, 5-fluorouracil 1000 mg/sqm 72 hours continuous infusion. Results: On an intent-to-treat analysis, four pts had CR (13%) with a duration of responses of 12, 18, 24 and 36 months. Fourteen pts (47%) had PR with a median duration of 7 months. Six patients had stable disease (20%) with a median duration of 3 months. Six patients (20%) had PD. Responses were more frequent in patients with PS 0 (10/13, 77%) and in patients treated with 2 previous lines of chemotherapy (13/18, 72%) than 3 or more (5/12, 42%). Median survival for the whole population was 11.8 months (5–44; 95% CI, 10.0 to 13.9). For pts with CR or PR and those who had SD, median survival was 16 months (10–44) and 10 months (8–18), respectively. No treatment-related death occurred. Grade 3–4 neutropenia in 7 % of pts. Grade 4 thrombocytopenia and grade 3 anemia 3 % of patients. Most frequent G3 non-haematologic toxicities were nausea and vomiting (3%) and alopecia (3%). We observed 2 dislodgements of catheter (7%), 3 cases (10%) of gastroduodenal ulcers, neither cases of thrombosis, nor sclerosing cholangitis. Conclusions: This intra-arterial regimen seems active in this patient population, with a very good safety profile. No significant financial relationships to disclose.
Should we continue to study high-dose chemotherapy in metastatic breast cancer patients? A critical review of the published data
