Abstract B6: PEGPH20 depletion of pericellular hyaluronan sensitizes high hyaluronan-producing tumor cells in antibody-dependent cell-mediated cytotoxicity

Monoclonal antibodies (MAbs) have been developed successfully as part of the anticancer armamentarium. However, even in cases where the target of the antibody is stably and abundantly expressed, efficacy often falls below expectations. The tumor microenvironment (TME) has previously been shown to contribute to resistance to MAb therapy because matrix-associated high interstitial fluid pressure inhibits efficient drug penetration (1). Hyaluronan is an important component of the extracellular matrix (ECM) in about 25% of tumors, and 56% of breast cancers. This glycosaminoglycan is produced by malignant and stromal cells and acts to cross-link ECM proteins, contributing to the desmoplastic phenotype. Accumulation of HA has been associated with more aggressive malignancy (2).In this work, we show that >50% of p185HER2-overexpressing breast tumors have a high accumulation of HA (HA+3). We hypothesized that tumor cells which overexpress HA create a protective “turtle shell” when it cross-links other ECM proteins (e.g., versican), and this pericellular structure protects the HER2-overexpressing tumor cell from attack by antibody-mediated cellular cytotoxicity (ADCC). SKBR3 (overexpressing p185HER2) and MDA-MB-231 (expressing EGFR) breast tumor cells were transfected with hyaluronan synthase 2 (HAS2), creating SKBR3/HAS2 (HA+3) and MDA-MB-231/HAS2 (HA+2), enabling these cells to form substantial pericellular matrices. NK-cell/trastuzumab/cetuximab-mediated ADCC is diminished in both of the HAS2-transfected cell types, suggesting that HA accumulation may contribute to tumor cell resistance to ADCC. Similarly, when these cells are treated in vitro with PEGPH20, a pegylated PH20 hyaluronidase (3), sensitivity to trastuzumab/cetuximab-mediated ADCC of the HAS2-overexpressing tumor cells is increased by up to 2-fold. These experiments show that the pericellular matrix coat (“turtle shell”) of high HA-producing tumor cells can inhibit the access of human immune cells to target cells, and that this effect can be reversed by depleting HA using PEGPH20. In vivo studies have shown that only PEGPH20, and not native PH20, has the bioavailability to induce antitumor activity in vivo. Models are currently in development to test whether PEGPH20 will enhance the activity of therapeutic MAbs in vivo.