This is a multicenter, randomized, prospective, Phase III study in which patients with advanced breast carcinoma previously treated with anthracyclines and taxanes will be randomly assigned to receive one of two treatment options: vinorelbine (Arm A) or gemcitabine plus vinorelbine (Arm B).

The investigators assume that progression-free survival mean time for patients treated with vinorelbine will be 3 months, and for patients treated with gemcitabine plus vinorelbine will be 5 months. That implies a reduction in risk ratio of 40% (Hazard ratio = 1,67). Assuming a bilateral alpha error of 0.05 and beta error of 10%, and the number of events needed if 60% of patients have progressed after 1 year, the number of patients needed per treatment arm is 114. Considering a 10% post-randomization drop-out, the final number of patients is 252 (126 per arm).

Inclusion Criteria:

• Histological or cytological diagnoses of breast cancer, with metastases.
• Metastatic lesions should not be curable with surgery or radiotherapy.
• Women of age > 18.
• To have received a previous treatment with anthracyclines and taxanes.
• A maximum of 2 previous chemotherapy treatment lines for metastatic disease.
• Previous radiotherapy is allowed, whenever the radiated area is not the only disease location.
• At least 4 weeks since the last previous antineoplastic treatment; patient must have recovered from all previous toxicities.
• Performance status < 2 in World Health Organization (WHO) scale.
• Clinically measurable, non measurable or really non measurable disease, as per Response Evaluation Criteria in Solid Tumors (RECIST) criteria.
• Life expectancy of at least 12 weeks.
• Patients able to comply and to receive an adequate follow-up.
• Adequate bone marrow function: neutrophils ≥ 2 x 10^9/L; platelets ≥ 100 x 10^9/L; hemoglobin ≥ 100 g/L.
• Calcium within normal limits.
• Premenopausal women must adopt an adequate contraceptive method during the study and up to 3 months after treatment finalization.

Exclusion Criteria:

• Active infection or serious concomitant disease (investigator's criteria).
• Clinical evidence of metastases in the central nervous system (CNS).
• Blastic bone lesions as only disease.
• Previous neurological toxicity grade 3-4 National Cancer Institute (NCI) Common Toxicity Criteria (CTC) v.2.0.
• Previous treatment with gemcitabine and/or vinorelbine.
• More than 2 previous chemotherapy treatment lines for metastatic disease.
• Abnormal liver function (bilirubin > 2.0-fold upper normal limit (UNL); alanine transaminase (ALT) and aspartate transaminase (AST) >2.5-fold UNL). In patients with hepatic metastasis, a value of ALT and AST of up to 5-fold UNL is permitted.
• Unpaired renal function (creatinine > 2.0 mg/dL).
• Pregnancy or lactating.
• Treatment with any investigational agent in the previous 4 weeks.
• Second malignancy (except for cervix carcinoma in situ or skin carcinoma - no melanoma- with an adequate treatment). Previous malignancies are allowed if disease-free survival is superior to 5 years, except for renal carcinoma or melanoma.
• Males.