Several microtubule-targeted drugs are commonly used to treat breast cancer, but for many patients they do not work. The long-term goal of this research is to accurately predict which patients will benefit from microtubule-targeted drugs including paclitaxel, docetaxel, vinorelbine, eribulin, and ixabepilone. Multiple lines of evidence from our laboratories and others support the idea that chromosomal instability (CIN) is the key feature of cancer governing response to paclitaxel and other anti-microtubule drugs. The central hypothesis is that breast tumors with the highest levels of pre-existing CIN are most sensitive to the enhanced CIN caused by microtubule-targeted therapies. Our preliminary data show that paclitaxel causes CIN due to multipolar spindles in patient tumors, that similar concentrations of other anti-microtubule drugs cause multipolar spindles in cultured cells, and that CIN measured by interphase FISH correlates with taxane response in metastatic breast cancer.
Aim 1 will determine whether clinically useful microtubule poisons universally induce multipolar spindles. Paclitaxel, docetaxel, eribulin, vinorelbine, and ixabepilone will be tested for effects on mitotic spindle morphology and function in cell models, mouse models, and in samples obtained from human breast cancer in patients receiving these treatments as single agents as part of the standard of care.
This aim will thereby determine whether these microtubule-targeted drugs have similar or disparate biologic effects on cancer.
Aim 2 will determine which types and degrees of CIN confer sensitivity to diverse microtubule targeted agents. Four models of CIN will be used to generate specific mitotic defects including multipolar divisions, polar chromosomes, lagging chromosomes, and chromosome bridges at defined rates, and these will be tested for sensitivity to microtubule-targeted drugs in multiple models. Patient-derived primary organoid breast cancer cultures with defined mechanisms of CIN will be tested in parallel.
Aim 3 will establish a standardized method to quantify CIN to use as a biomarker in human breast cancer. The four CIN models will be used to compare proposed methods to quantify CIN including interphase FISH, bulk DNA and RNA sequencing, and digital karyotypes from low-pass single-cell DNA sequencing. We anticipate that this will provide a basis to accurately infer CIN from the thousands of sequenced tumors for which data is publically available. These measures of CIN will also be evaluated for their ability to predict taxane response in metastatic breast cancer patients, employing archived tumor samples, to verify ability to predict response to paclitaxel. The work is significant because it will advance our knowledge of the mechanism of widely used cancer drugs as well as how CIN, a common feature of tumor biology, affects response to these agents. It ensures clinical relevance by incorporating both models and human samples in each aim. Ultimately the knowledge gained will allow for accurate prediction of patients who will and will not benefit from widely used treatments, and thereby has the potential to address the ongoing problem of overtreatment and ineffective treatment of cancer.
The proposed research is relevant to cancer because it provides a mechanistic basis to predict tumor response to microtubule-targeting drugs, which are widely used classes of anti-cancer drugs. This work will provide a basis to improve public health by limiting exposure of cancer patients to ineffective and toxic therapies.