Abstract
Antibody-drug conjugates (ADCs) are gaining increasing traction in the treatment of oncological diseases; however, many clinical failures have also been observed. One key factor limiting ADC effectiveness is the heterogeneous expression of their target antigen. While the vast majority of ADCs in clinical development target antigens on cancer cells (cancer-targeting), they can also target antigens expressed on non-cancerous stromal cells in the tumor microenvironment (stroma-targeting). It remains unclear if ADCs targeting stromal cells can outperform cancer-targeting ADCs. Here, we present three related mathematical models to evaluate: (1) cancer-targeting ADCs with homogeneous target antigen expression, (2) cancer-targeting ADCs with heterogeneous target antigen expression, and (3) stroma-targeting ADCs. Our simulations suggest that cancer-targeting ADCs can achieve high efficacy when their target antigen is homogenously expressed. However, in cases of heterogeneous antigen expression, cancer-targeting ADCs may lead to an initial reduction in tumor size, followed by regrowth due to the elimination of antigen-positive cells and expansion of antigen-negative cells. This limitation could potentially be overcome by stroma-targeting ADCs, as antigen-positive stromal cells may continue to be recruited into the tumor by the oncogenic factors produced by the remaining cancer cells. Furthermore, we demonstrate that ADCs with more permeable payloads and less stable linkers may offer improved efficacy in the context of heterogeneous target expression.