Differential in vitro and in vivo antitumor effects mediated by anti-CD40 and anti-CD20 monoclonal antibodies against human B-cell lymphomas.

The antitumor effects of CD40 and CD20 monoclonal antibodies (mAbs) were compared on various human B-cell lymphomas by using both in vitro and in vivo assays. Anti-CD40 directly inhibited the proliferation of human B-cell lymphomas in vitro, whereas anti-CD20 exerted no inhibitory effects on the growth of any lymphoma tested. These lymphomas were then injected into immunodeficient mice to examine the antitumor efficacy of these unconjugated mAbs in vivo. This xenogeneic model was used in the evaluation of various potential therapeutic agents against human cancers in an in vivo setting. Surprisingly, in contrast to its negligible effects on lymphoma growth in vitro, anti-CD20 was more efficacious than anti-CD40 in promoting the survival of mice bearing some but not all lymphoma lines. To determine whether the antitumor effects of these mAbs were direct or indirect in vivo, we concurrently treated tumor-bearing mice with mAbs to the murine Fc receptor to block antibody-dependent cell-mediated cytotoxicity (ADCC). When these neutralizing antibodies against Fc receptors were administered at the same time as mAb treatment, the antitumor effects of anti-CD20 in vivo were completely abrogated, whereas anti-CD40 treatment, although also diminished, still provided significant antitumor effects. These results indicate that the in vivo antitumor activity of the murine anti-human CD20 mAb was primarily due to ADCC by murine effector cells, which may not translate into comparable effects in humans. By contrast, anti-CD40 may be of potential clinical use in the treatment of lymphomas in humans because of its additional direct anti-proliferative effects. The results also demonstrate a possible difficulty in accurately evaluating the potential clinical efficacy of murine antibodies against human tumors in a human/mouse model system. Murine monoclonal anti-human antibodies may produce greater effects in human/mouse xenogeneic models, in which they are more likely to elicit host effector systems than when used in vivo in humans.