When cancer cells are treated with chemotherapy, changes in gene transcription occur. Paul Polakis and colleagues investigated the possibility of exploiting this effect to target cancer cells. Using a cytotoxin-armed antibody, they targeted an antigen whose expression in cancer cells is induced by the topoisomerase-I inhibitor irinotecan.

The authors compared the expression of transcripts from human colorectal cancer xenografts grown in mice treated with either irinotecan or a saline control. Using expression profiling, they identified transcripts coding for cell-surface proteins that were significantly upregulated in the irinotecan-treated tumours relative to the controls. The LY6D gene (also known as E48), which codes for a small glycosylphosphatidylinositol-linked cell-surface protein, had the most consistent and strongest induction, and was therefore a potential antibody target. Crucially, irinotecan did not induce Ly6d in normal intestinal tissue from the treated mice. The only genes that were induced in the normal gut were those involved in a response to inflammation or infection.

So, could targeting the LY6D antigen improve on the tumour response to irinotecan? The authors obtained antibodies to LY6D from immunized mice. The naked antibody did not show any cytotoxic activity, so the LY6D monoclonal antibody was conjugated to the cytotoxin monomethyl auristatin E (MMAE). Tumour-bearing mice that were treated with irinotecan together with anti-interleukin-8 linked to the cytotoxin or with the vehicle alone as controls, showed attenuated tumour growth for 3 weeks before the tumours regrew. By contrast, treatment with irinotecan followed by the conjugated LY6D antibody led to complete responses in six of eight mice. Treatment with conjugated antibody alone did not have any anti-tumour activity.

LY6D has been used as a therapeutic target and diagnostic marker for head and neck cancer, and is expressed in stratified squamous and transitional cell epithelia. The effect of the LY6D antibody on these normal cells is not known, but Polakis and colleagues have shown that targeting a cell-surface antigen induced by chemotherapy is a worthwhile approach to pursue.