BACKGROUND: Resistance to lapatinib generally develops in approximately half of patients within one year of initiating treatment. When lapatinib is withdrawn, there is often a rapid progression of disease which is associated with high mortality rates. This “flare” phenomenon has also been observed upon discontinuation of other tyrosine kinase inhibitors (TKIs) in lung cancer, renal cell carcinoma, and gastrointestinal stromal tumors.

METHODS: Using the HER2+ cell lines SKBR3 and BT474, we developed both in vitro and in vivo models demonstrating lapatinib resistance and the tumor “flare” phenomenon. Parental HER2+ cell lines were gradually exposed to increasing doses of lapatinib (100nM to 5uM) to develop a lapatinib-resistant form which was ultimately maintained in 1uM of lapatinib. Lapatinib-”released” cell lines were developed by removing lapatinib exposure from resistant cell lines and allowing the cells to grow for at least two weeks. Cell lines were grown in vitro using traditional monolayer cell culturing techniques and mammosphere technology. A comprehensive analysis of parental, lapatinib-resistant, and lapatinib-released cell lines was performed using microscopy, protein/phosphoprotein analysis, cell cycle, cancer stem cell markers by FACS, and invasion assays. Parental cells served as the control in all experiments. In vivo models were performed by injecting 10,000 cells of parental, resistant, and released cell lines in the mammary fat pads of SCID mice. Tumors were surgically resected after approximately sixty days and volume recorded. All experiments were repeated three times and calculated for statistical significance.

RESULTS: Cell cycle analysis and proliferation assays demonstrate that lapatinib-resistant and released cell lines continue to proliferate despite the addition of 1uM lapatinib. Released cells also demonstrate less of a response to retreatment with lapatinib. Tumor volume of lapatinib-released cell lines were significantly larger than parental and resistant counterparts [parental: 39.9mm3, SD 9.48; resistant: 71.8mm3, SD 62.33; released: 943.4mm3, SD 100.1] and this difference was statistically significant (p < 0.01). After three series of passages, mammosphere forming efficiency (MFE) was higher among lapatinib-released cell lines [parental: 1.03%; resistant: 1.93% released: 7.23%] and was statistically significant (p <0.01). Lapatinib-released mammosphere phenotypes appeared larger and less organized than the parental and resistant counterparts. MFE did not correlate with ALDH expression by FACS, and expression varied based on cell line. Western blot analysis revealed loss of E-cadherin among lapatinib-released cells which was not demonstrated in parental or resistant counterparts. Invasion assays demonstrated increased migration among resistant and released cell lines.

CONCLUSIONS: We demonstrate the first in vitro and in vivo models of the tumor “flare” phenomenon using HER2+ breast cancer cell lines. This model demonstrates that lapatinib-resistant cells released from the exposure to lapatinib are propelled into a unique and more aggressive phenotype. This model has potential to elucidate new mechanisms of resistance to TKI therapy and provide novel preclinical data that may help in the understanding of disease progression.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P4-08-07.