In an effort to broaden the applicability of adoptive cellular immunotherapy toward nonmelanoma cancers, we have designed chimeric antibody/T-cell receptor genes composed of the variable domains from mAbs joined to T-cell receptor-signaling chains. We have demonstrated that T cells retrovirally transduced with these genes can recognize antibodydefined antigens and that this recognition leads to T-cell activation, specific lysis, and cytokine release.

In this study, we have examined the in vivo activity of murine T cells transduced with a chimeric receptor gene (MOv-γ) derived from the mAb MOv18, which binds to a folate-binding protein overexpressed on most human ovarian adenocarcinomas. Nude mice that were given i.p. implants of human ovarian cancer (IGROV) cells were treated 3 days later with i.p. murine tumor-infiltrating lymphocytes (TIL) derived from an unrelated tumor. Mice treated with MOv-γ-transduced TIL (MOv-TIL) had significantly increased survival compared to mice treated with saline only, nontransduced TIL, or TIL transduced with a control anti-trinitrophenyl chimeric receptor gene (TNP-TIL). In another model, C57BL/6 mice were given i.v. injections of a syngeneic methylcholanthrene-induced sarcoma transduced with the folate-binding protein (FBP) gene. Three days later, mice were treated i.v. with various transduced murine TIL (derived from an unrelated tumor), followed by low-dose systemic interleukin 2. Eleven days after tumor injection, mice were sacrificed, and lung metastases were counted. In multiple experiments, mice receiving MOv-TIL had significantly fewer lung metastases than did mice treated with interleukin 2 alone, nontransduced TIL, or TNP-TIL.

These studies indicate that T cells can be gene modified to react in vivo against tumor antigens, defined by mAbs. This approach is potentially applicable to a number of neoplastic and infectious diseases and may allow adoptive immunotherapy against types of cancer not previously amenable to cellular immunotherapy.