Exposing our ignorance: EPA study reveals barren exposure data landscape

Jennifer McPartland, Ph.D., is a Health Scientist.

This past November, EPA scientists published a sobering paper, “The exposure data landscape for manufactured chemicals,” in the journal Science of the Total Environment.  The paper reveals how little systematic information we have about human and environmental exposures to the thousands of chemicals in use today.

The aim of the study was “to define important aspects of the [chemical] exposure space and to catalog the available exposure information for chemicals being considered for analysis as part of the U.S. EPA ToxCast screening and prioritization program.”  Its conclusion:  “The results suggest that currently available exposure data are insufficient to provide the evidence base required to inform risk assessment and public health decision making.”  Not good, but not surprising.  Read on for more detail.

As I’ve described in a series of earlier blog posts, EPA has been investing in new toxicology platforms like ToxCast and Tox21 to better evaluate the potential hazards of the thousands of chemicals for which we have little or no data.  On the exposure side of the risk equation, EPA is developing complementary programs like ExpoCast, which seeks to “ensure that the required exposure science and [relevant] computational tools are developed.”  An early goal of ExpoCast is to improve publicly available information on chemical exposures.  The “gap analysis” presented in this recent paper is a necessary step toward realizing that goal.

The authors focused their analysis of exposure data availability to chemicals in EPA’s Aggregated Computational Toxicology Resource (ACToR) database.  This is likely EPA’s most comprehensive chemicals database, containing varying – and I do mean varying! – levels of information on 547,088 chemicals.  EPA notes that the ACToR database, which aggregates information from several data repositories, holds “essentially all publicly available information on chemical identity, structure, physical-chemical properties, in vitro assay results and in vivo toxicology data.”

The researchers classified the available types and sources of exposure data into the following categories:  production/import volumes and processes; consumer product ingredients and usage; environmental release; environmental media concentrations; and human exposure/biological monitoring.

I won’t attempt to describe here all of the findings of this paper, but here are some highlights:

  • 8 million chemicals are “commercially available” world-wide.
  • 100,000 chemicals have been inventoried as available in U.S. Commerce:  82,000 chemicals regulated under the Toxic Substances Control Act (TSCA), 8,600 food additive chemicals, 3,400 cosmetic ingredients, 1,800 pharmaceuticals, and 1,000 pesticide active ingredients.
  • Fewer than 22,000 of the total nearly 550,000 ACToR chemicals – a minuscule 4% – have anysort of exposure-related data.  These include:
    • the roughly 15,000 chemicals reported under EPA’s periodic TSCA reporting system (formerly, the Inventory Update Reporting (IUR) rule, now the Chemical Data Reporting rule) at least once since 1986;
    • 2,485 food additive chemicals; and
    • 203 additional chemicals identified as chemicals of concern for children by various authoritative bodies.
  • Of the 100,000 ACToR chemicals having at least some toxicity data, fewer than 20% have any exposure-related information, and even for these chemicals, that information is extremely limited (e.g., production volume).
  • Even production volume information is available for fewer than 15,000 (less than 3%), of the ACToR chemicals.
  • Use information is available for only about 7,500 (1.4%) of the ACToR chemicals, and for only 2,400 (17%) of the IUR-reported chemicals.
  • Truly minuscule numbers of the ACToR chemicals have any data on their concentrations in environmental media: 1,153 chemicals in water, 788 chemicals in soil, 720 chemicals in food, 670 chemicals in air, and 390 chemicals in people (through biomonitoring).
  • The vast majority (nearly 90%) of ACToR chemicals that have any type of exposure-related information have only one type of such information (i.e., only production volume or use information or presence in food or water concentration, etc.).
  • Of the 700 ACToR chemicals that have so far been identified as of concern for children, exposure-related information is available for only 185 of them.

If we are serious about basing decisions about chemicals on risk, then we need adequate information on both hazard and exposure.  This paper shows just how far away from that goal we are.

So where does EPA plan to go from here?  On the hazard side of the equation, the ToxCast and Tox21 programs hold the promise of quickly generating useful toxicity data on many thousands of chemicals.  Given the high-throughput nature of these assays—automated small-volume tests done in multi-well plates—such an endeavor is possible.

But the possibility of developing useful exposure information on a comparable scale and pace is far more limited.  This is especially true for critical information such as understanding the exposures resulting from the use of products.  For this, among other things EPA will need to charge chemical manufacturers, importers, and processors of chemicals with providing far more exposure-relevant data.

To achieve this, we will need enhanced policies that give EPA the authority and resources it needs to systematically collect and process the necessary information.   For industrial chemicals, that clearly means we need reform of the Toxic Substances Control Act (TSCA).  The authors of the paper I’ve reviewed here make this clear:

[I]f enacted, proposed reform of U.S. legislation regulating potentially toxic chemicals would likely require chemical manufacturers to provide a minimum set of data, including more detailed data on exposure potential, for large numbers of chemicals. Such data requirements hold the promise of greatly expanding the universe of chemicals for which useful exposure-related information exists.

EPA has made some positive steps recently to gather more exposure (as well hazard) information on chemicals even while awaiting more authority from Congress to develop a truly effective system.  For example, as noted earlier, EPA has revamped the Inventory Update Rule (IUR) to create what is now called the Chemical Data Reporting (CDR) system (see this blog post for more information on the CDR).  The agency is also is moving ahead on prioritizing chemicals for risk assessment.  Just this month, using a process explained here, EPA identified eighty-three “work plan” chemicals for which it intends to complete risk assessments.  EPA selected these chemicals in part because they are among the few chemicals that have sufficient hazard and exposure information to allow for risk assessment.  EPA will develop risk assessments for seven of these chemicals in fiscal year 2012.

Admittedly relatively small-scale efforts, given the magnitude of the overall task, but important and much more than the Agency has attempted in the past.

This entry was posted in Emerging Testing Methods, EPA, Health Policy, Regulation and tagged , , , , , , , , . Bookmark the permalink. Both comments and trackbacks are currently closed.

One Comment

  1. Chloe
    Posted March 16, 2012 at 2:31 pm | Permalink

    Great post. Thanks for sharing this data. There is always more to learn!