Abstract #4124

Background: Vaccination with tumor cells engineered to secrete granulocyte-macrophage colony stimulating factor (GM-CSF) generates potent, specific, and long-lasting anti-tumor immunity in multiple tumor models. At DF/HCC, over 100 patients with melanoma, ovarian cancer, and lung cancer have been vaccinated. These studies have led to the identification of targets of immune-mediated tumor destruction. Here, we present analysis on the safety, feasibility, and biologic activity of an autologous GM-CSF-secreting breast cancer vaccine.
 Methods:27 patients with metastatic breast cancer have undergone tumor procurement for vaccine from malignant pleural effusions (23), ascites (1), tumor nodules (2), lymph node (1), and 12 patients have received vaccine. Cells were transduced with a replication defective adenoviral vector encoding GM-CSF, irradiated, and GM-CSF secretion was measured by ELISA. Patients were required to have received at least one prior chemotherapy for metastatic disease and have an ECOG performance status 0-1. Vaccine was delivered s.c. and i.d. weekly for three weeks, then every other week. Immune monitoring included skin biopsies of vaccine sites, measurement of leukocyte populations, and proteomic-based assessment of antibody responses.
 Results: Sufficient viable tumor cell yields for vaccination were obtained from all patients, with cellular yields ranging from 1.1-679 x 106 cells and viabilities ranging from 71-100%. Dose levels were based on cellular yield, ranging from 105 -107 cells/dose. The average yield of GM-CSF was 838 ng/106 cells/24 hrs (range: 24.4-5696), higher than the average yields for lung cancer and melanoma. Vaccinated patients were 34-69 years, and received 3-23 vaccines. 9 patients had progressive disease after 3-10 vaccines (1-4 months). One patient had stable disease after 23 vaccines (11 months), resumed vaccination at 16 months for progression and remains on study. One other patient remains on study, and one patient did not have measurable disease. Toxicities related to treatment include grade I/II local injection-site reactions, grade I/II inflammation at tumor sites as well as grade I/II fatigue, fever, nausea, and edema. Skin biopsies of vaccination sites revealed mild-moderate infiltration of lymphocytes, granulocytes, and macrophages.
 Conclusion: Breast cancer cells can be harvested in sufficient number for autologous vaccine production from solid tumor as well as from malignant effusions. Autologous vaccination can induce coordinated immune responses with limited toxicity. The proteomic-based identification of antigen-specific immune responses following vaccination will be presented.

Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 4124.

Thirty-first San Antonio Breast Cancer Symposium Dec 10-14, 2008; San Antonio, TX