Purpose: Administration of recombinant interleukin-2 (rIL-2) causes activation of natural killer (NK) cells, which express CD16, the Fc receptor gamma III (Fc gamma III), and can mediate antibody-dependent cellular cytotoxicity (ADCC). In an effort to generate a more directed antitumor response by rIL-2-activated NK cells, we have investigated combination therapy with rIL-2 plus the antitumor monoclonal antibody (mAb) 14. G2A and its molecular derivatives, which react strongly with the GD2 ganglioside expressed on melanoma, neuroblastoma, and certain other tumors.
Patients and methods: The initial trial of this therapeutic strategy involved 33 evaluable neuroblastoma patients who received rIL-2 by continuous intravenous infusion in combination with the murine 14.G2A mAb. A follow-up phase I study was conducted in 24 evaluable adult melanoma patients with a mouse/human chimeric mAb (ch14.18). The ch14.18 mAb has subsequently been investigated in combination with rIL-2 and granulocyte-macrophage colony-stimulating factor in neuroblastoma patients following autologous bone marrow transplantation. Based on preclinical data, the administration of rIL-2 plus mAbs may be more effective in patients with minimal residual disease. An alternative strategy to induce ADCC has also been investigated because of the finding that many NK cells fail to express Fc gamma III. The IL-2-ch14.18 fusion protein has been tested in preclinical studies for its ability to induce an effective antitumor response and is currently being evaluated for future clinical testing.
Results: Based on in vitro assays, the serum of patients treated with rIL-2 plus either 14.G2A or ch14.18 mAbs demonstrated sufficient levels of antibody and NK cell activity to mediate effective ADCC; however, many patients developed antibodies against the administered murine and chimeric mAbs, leading to allergic complications. Concurrent administration of rIL-2 and ch14.18 resulted in a lower incidence of anti-ch14.18 antibodies. The IL-2-ch14.18 fusion protein has demonstrated dramatic antitumor effects in murine models.
Discussion: These studies have demonstrated the feasibility of concurrent administration of rIL-2 with the mouse/human chimeric ch14.18 mAb. Although the early pilot studies were conducted in patients with bulky disseminated disease, this approach may be more effective in patients with minimal residual disease. Further refinement of the reagents and phase II and III trials will be necessary to evaluate fully the safety and efficacy of this immunotherapeutic approach.