Genetic Predictors of Chemotherapy Toxicity in Breast Cancer

Institution: University of California, San Francisco
Investigator(s): Deanna  Kroetz , Ph.D. -
Award Cycle: 2012 (Cycle 18) Grant #: 18IB-0027 Award: $99,998
Award Type: IDEA
Research Priorities
Detection, Prognosis and Treatment>Innovative Treatment Modalities: search for a cure



Initial Award Abstract (2012)

In the attempt to provide breast cancer patients with “individualized therapy”, there has been little attention to understanding the genetic basis for the common debilitating side effects from chemotherapy. Paclitaxel and adriamycin/cyclophosphamide are both effective therapies in the treatment of breast cancer, but are also associated with significant and in some cases long-lasting toxicity. This can lead to suboptimal dosage and an increased risk of recurrence, especially in the adjuvant setting. Women having high response rates and persistent toxicities, such as sensory peripheral neuropathy and ovarian suppression, experience significant effects on the quality of life. The large degree of interpatient variability in these side-effects from treatment suggests that genetic variation may influence a patient’s risk for toxicity.

We will test whether differences in a patient’s DNA sequence are associated with the risk of developing either paclitaxel-induced sensory peripheral neuropathy or with the development of neutropenia and menopause during treatment with adriamycin/cyclophosphamide. These studies will be carried out using DNA samples collected from women with primary breast cancer enrolled on a Phase III clinical trial (CALGB 40101), which tests whether single agent paclitaxel is as efficacious as adriamycin/cyclophosphamide for disease-free survival. Toxicity data was collected during the treatment trial and this data will be used for the current studies. Genotyping data is also available for >500,000 markers scattered across the human genome. Statistical methods will be used to identify if any of these genetic markers predict the risk of developing paclitaxel or adriamycin/cyclophosphamide toxicity. Replication of our findings will be carried out using similar datasets from the use of these drugs in the primary and metastatic setting.

To date, pharmacogenetic studies of drug toxicity in breast cancer treatment have looked at single genes that are thought to be important since they control the level of drug within the body. In most of these studies, only small numbers of patients were evaluated and few, if any, of these findings have been replicated in a different patient population. Our study is innovative because it is taking an unbiased, genome-wide approach coupled with state-of-the-art “next generation sequencing” technology to identify novel genetic markers. The identification of novel genetic biomarkers will inform our understanding of the molecular basis of the toxicities, so these can be targeted in future studies.




Final Report (2015)

Chemotherapy-induced toxicity is a serious issue for women treated for breast cancer. Drug toxicity can often result in reduced dosage and even discontinuation of effective agents. In addition, the toxicities significantly influence a patient’s quality of life. The overall goal of the studies completed during this funding period was to identify genetic markers that could predict a patient’s risk of developing chemotherapy-induced toxicity.

The study was largely completed although the analysis of the sequencing data for Aim 1 is still ongoing. We were also not able to complete a genetic association analysis of chemotherapy-induced ovarian suppression because there was not follow up data available to accurately phenotype the patients for this toxicity. We were faced with several logistical obstacles which significantly delayed our progress on this project. The first delay was getting the DNA samples identified and ready for next generation sequencing. This was a result of budget related reductions in the staff at the CALGB Pathology Core Laboratory who are responsible for the sample requests and at the CALGB Statistical Center who we work closely with to curate the study phenotype data. As a result, the samples for Aim 1 were not ready for sequencing until 18 months after the start of the study. The phenotype data curation for the neutropenia analyses were similarly delayed at the Statistical Center. Despite these barriers we were still able to complete most of the proposed studies. The neutropenia analysis was completed in 759 patients of European descent and no significant genetic markers were identified for either chemotherapy-induced neutropenia or baseline neutrophil counts. This may truly reflect a non-significant contribution of genetics to this toxicity or be a statistical power issue.

The next generation sequencing for FGD4, FZD3 and EPHA5 genomic loci in 92 patients with serious toxicity at low doses of paclitaxel and the same number of patients who developed no peripheral neuropathy during six cycles of paclitaxel is complete and the analysis of these data is currently ongoing. We expect to identify novel markers in these genomic regions that are associated with paclitaxel-induced peripheral neuropathy. We plan to carry out functional studies to understand the role of FGD4, FZD3 and EPHA5 and the variants identified in this current study on peripheral nerve function. We will also validate these findings using other populations of breast cancer patients treated with taxanes. A long term goal of these studies is to identify genomic markers that can be used clinically to predict chemotherapy-induced toxicity and to use the genomic data from these studies to understand the molecular basis for these common toxicities.