Development of models of both acquired and de novo resistance to Herceptin

676 Despite the proven clinical benefits of Herceptin therapy, either as a single agent or in combination, not all patients that overexpress HER2 respond to Herceptin. In fact, less that 35% of patients with HER2 overexpressing breast cancer respond to Herceptin as a single agent and a number of those who achieve an initial response acquire resistance within one to two years. The aim of this study was to develop and characterize cell line models of both acquired and de novo resistance to Herceptin.
 Acquired resistance was established by culturing BT-474 and SKBR3 cells in high dose Herceptin (105 μg/ml) for 18 months and characterized for response to Herceptin and the dual EGFR-HER2 small molecule inhibitor, lapatinib by cell count and colony formation assay. De novo resistance was determined by measuring the response to Herceptin and lapatinib in a panel of 18 HER2 positive breast cancer cell lines. Protein levels of HER2, and other proteins associated with Herceptin resistance were quantified by Western blotting and related to Herceptin and lapatinib response in the cell line panel.
 Proliferation was completely inhibited in the parental BT-474 cells when treated with Herceptin; this is in contrast to the conditioned BT-474 cells (BT-H) that show only a 1.15 fold decrease in proliferation when exposed to Herceptin. This effect was also observed in conditioned SKBR3 cells (SK-H). Both the conditioned BT-474 and SKBR3 cells also showed reduced response to Herceptin compared to their parental cells when grown in soft agar (colony number after 3 weeks: BT-H 12.0% increase vs BT474 27.1% decrease; SK-H 5.5% decrease vs SKBR3 65.3% decrease). Interestingly, both the resistant and parental cells were equally sensitive to lapatinib. This suggests that the reduced response observed in these cell lines is specific to Herceptin. In the panel of HER2 overexpressing cell lines, relatively low responses to Herceptin were observed in the SUM225, SUM190, HCC1569, MDA453, HCC1419, HCC1954 and JIMT-1 cells. However, only the SUM225, HCC1419, HCC1954 and JIMT-1 cells were found to be Herceptin resistant by both cell count and colony formation assay. Both the SUM225 and HCC1419 cell lines were sensitive to lapatinib (IC50: 85 nM and 40 nM respectively), whereas the HCC1954 and JIMT-1 cells showed a low level of sensitivity to lapatinib (IC50: 383 nM and 2.603 µM respectively). Neither response to Herceptin nor lapatinib correlated with HER2 protein levels in the panel of cell lines. However, Herceptin response showed a weak association with loss of PTEN protein (p = 0.0177).
 We have identified, from a large panel of HER2 overexpressing cell lines, models of both acquired and de novo Herceptin resistance. This is the first study to use two independent methods to validate Herceptin resistance and confirms that these cells genuinely have an altered response to the drug and can therefore be used as reliable models to study the biology of acquired and de novo Herceptin resistance.