Courant Institute of Mathematical Sciences
New York University
February 9, 2016
EPA + Math = fewer estrogen receptor pathway disrupters
This is the story of how a 5-day workshop at the American Institute of Mathematics produced a relatively simple mathematical model (a variational method with regularization for an inverse problem), which -- with a great deal of additional effort -- has led to a big improvement in how the Environmental Protection Agency (EPA) evaluates chemicals as possible estrogen receptor pathway disrupters. This ER model was published in Toxicological Sciences in November 2015 (vol 148 no. 1, pp. 137-154).
The Endocrine Disruptor Screening Program (EDSP) of the EPA has the goal of helping insure that people and wildlife are not exposed to chemicals that inadvertently perturb the endocrine system, which could lead to harmful effects including developmental and reproductive problems, increased risk of certain kinds of cancers, or significant decreases in populations of ecological species such as fish, frogs or birds. These chemicals may arise from pharmaceuticals, food additives and contaminants, or air and water pollutants.
The EPA has now stated that it intends to allow companies to use the results of our ER pathway model to replace data from two low-throughput in vitro tests and the uterotrophic (animal-based) assay. There are ~10,000 chemicals that are required to be tested in these three assays. Using the standard methods, less than 60 have been evaluated in the last several years, while with our model, we now have data from >1000 on the EDSP list, with another 1000 likely to be evaluated in the next year. The EPA is already in the process of using a variant of the ER model to do a similar evaluation of androgen receptor (AR) activity.