Improving the Efficacy of Rituxan by Rational Protein Engineering.

Antibody-dependent cell-mediated cytotoxicity (ADCC) is a key effector function for the clinical efficacy of monoclonal antibodies (Cartron et al., 2002; Weng and Levy, 2003), and is mediated primarily through a set of closely related Fc receptors (FcRs) with both activating and inhibitory activities. Using computational design algorithms and high-throughput screening, we have engineered a series of IgG1 Fc variants with optimized FcR affinity and specificity. These variant Fc domains are readily combined with the variable domains of any antibody and consist of from 1–5 changes in the amino acid sequence. With rituximab as a model system, we have generated variants with greater affinity for the activating receptor FcgammaRIIIa and weaker affinity for the inhibitory receptor FcgammaRIIb. These variants show a marked increase (over 2 logs) in ADCC using purified natural killer cells or PMBCs from healthy human donors against WIL2-S cells, a CD20-bearing tumor cell line. Moreover, PBMC genotyping indicates that this effect occurs irrespective of whether the donor is a high responder (FcgammaRIIIa dimorphism 158V homozygous) or low responder (FcgammaRIIIa dimorphism 158F homozygous or heterozygous) to rituximab therapy. In vivo studies show that our variants deplete circulating B-cells more effectively than rituximab in cynomolgus monkeys with an improvement in half-maximal potency of over 10-fold as measured by flow cytometry. The macaques showed no unusual clinical signs or depletion of other cells types. These agents hold promise as more potent and more efficacious treatments for NHL. Because these Fc mutations display a comparable effector function enhancement in other antibodies such as anti-CD52, anti-EGFR1 and anti-Her2, our technology provides a broadly applicable means for improving the efficacy of the next generation of therapeutic antibodies.