Background: In vitro studies have demonstrated that exposure of tumor infiltrating lymphocytes (TIL) to human recombinant interleukin-2 (rIL-2) will generate activated T-lymphocytes with major histocompatibility complex (MHC)-restricted and non-MHC-restricted cytotoxicity toward a panel of tumor target cells. In melanoma and ovarian carcinoma, TIL display MHC-restricted and autologous tumor-specific cytotoxicity. Such tumor-reactive cytotoxic T-lymphocytes (CTL) represent important material for understanding cellular immunity in cancer and developing specific immunotherapeutic approaches in melanoma and ovarian cancer. In breast cancer, some TIL have been demonstrated to secrete cytokines upon interaction with autologous tumor cells, indicating that autologous tumor-reactive lymphocytes may also exist among TIL in breast cancer. Therefore, the authors conducted a study to investigate the cytotoxic profile of rIL-2-activated lymphocytes in breast cancer.
Methods: Lymphocytes were isolated from primary solid tumors (TIL) of breast carcinomas (10 patients) and from peritoneal effusions (effusion-associated mononuclear cells [EAMNC]) from 2 patients with newly diagnosed metastatic breast carcinoma. Tumor infiltrating lymphocytes or EAMNC were cultured with rIL-2 in long term cultures whereby their expansion index, phenotype, and cytotoxic potential were studied.
Results: Both TIL and EAMNC proliferated by greater than 300-fold (370-3650; mean, 1656) after 23-82 days in cultures containing mixtures of TIL or EAMNC, autologous tumor cells, and rIL-2. By fluorescence analysis, freshly isolated TIL and EAMNC were found to consist of 77.5 plus or minus 10.7% CD3+ T-cells, 33.2 plus or minus 8.9% CD4+, and 47.2 plus or minus 16.8% CD8+ cells. Their CD4 to CD8 ratio was 0.70. After expansion of lymphocytes with rIL-2 in the majority of patients (9 of 12), CD3+CD8+ T-lymphocytes were present in greater numbers than CD3+CD4+ T-lymphocytes. Recombinant interleukin-2-activated CD3+CD8+ cells exhibited preferential cytolytic activity against autologous tumor cells. The cytolytic activity of CD3+CD8+ cells was inhibited either by anti-T-cell receptor (TCR)-alpha/-beta and anti-CD3 monoclonal antibodies (MoAb) or after pretreatment of tumor target cells with MoAb against the class I MHC antigens. Recombinant interleukin-2-activated CD3+CD4+ cells demonstrated potent cytolytic activity against both autologous and allogeneic tumor cells. CD3+CD8+ T-cell clones isolated from representative TIL exhibited preferential autologous tumor-specific cytotoxicity whereas the cytolytic activity of CD3+CD4+ T-cell clones was mostly (12 of 14 clones) nonrestricted to the autologous tumor.
Conclusions: To the authors' knowledge, this is the first report to demonstrate that TIL from primary tumors of breast carcinomas and EAMNC from metastatic disease can be propagated in large numbers in vitro with rIL-2 while retaining autologous tumor specific and MHC-restricted CTL activity. These findings may be of importance to ongoing clinical trials using TIL or EAMNC in the immunotherapy of patients with advanced breast cancer.