Conclusions and Recommendations

The major conclusions and research suggestions identified during the workshop are described in this chapter. No attempt was made to prioritise the research topics during the workshop but the participants were asked to do this prior to this report being finalised.
Conclusions
A key message is that the uncertainty/error in measured or estimated water solubilities translates directly into uncertainty/error in the calculated chemical activity.  In the absence of empirical data for chemicals of interest, it is strongly recommended that multiple estimation software and/or other techniques (e.g., polyparameter Linear Free Energy relationships; ppLFERs) be utilised to assess the uncertainty in the model output.  Note that while a high level of agreement between various estimates increases the level of confidence, and confers precision in the predicted values, it does not necessarily guarantee accuracy.
Chemical Activity Concept.
The workshop participants concluded that there were a number of opportunities and challenges facing the chemical activity concept.
The opportunities include:
Chemical activity is a more insightful and relevant metric of chemical exposure than concentration because concentrations are media-dependent while activity applies to all media, allowing exposure and toxicity to be expressed on a common basis.
Activity provides a good metric for characterising baseline toxicity for single non-polar organic chemicals and mixtures of non-polar organic chemicals.
Activity data are useful for discriminating between nn-polar narcosis, which occurs at activities between 0.01 and 0.1, and excess toxicity, which occur at activities less than 0.01.
Activity can also be used to identify poor quality data (McCarty et al., 2013) such as toxicity data from experiments where dosing concentrations were above the solubility of the chemical in the exposure medium (e.g., toxicity reported at an activity, a > 1), and exposure data from experiments subject to background contamination.
The application of activity to describe the toxicity of mixtures of non-polar organic chemicals represents a novel tool in chemical risk assessment that can be particularly useful in addressing chemical risks in real world environments.
The challenges are:
Translation from concentration to activity is crucial in studies where existing data are converted into the chemical activity space. However, this translation can be challenging and can add error to measurement error.
Improved communication of the activity concept is a major issue and will be central to future application and impact. Communication of the activity approach to a non-scientific audience may not be easy. Whether a broader acceptance of the chemical activity framework can be achieved might also be a matter of semantics. How can chemical activity be communicated in a comprehensible way and become widely accepted?
Suggestions for communicating the chemical activity concept to a wider audience may require adopting alternative terms that similarly convey the concept of chemical activity, such as “percent of saturation” or “fractional solubility” both of which would also reflect the output from Equation 1.  The following suggestions were also proposed:
An online tool, such as an “activity calculator” provided as an Excel file or interactive website. Such a calculator was used in a SETAC short course on the application of chemical activity, and can be made available for application to existing and new chemicals. A copy of the calculator will be made available on the following website: http://www.rem.sfu.ca/toxicology/models/.
The characterisation of the error and/or uncertainty in chemical activity needs to be better understood and quantified.
It is of critical importance to emphasise that at all times the domain of applicability for the activity approach be carefully defined.  Current knowledge would limit the applicability domain to non-polar organics log KOW > 2 and predictions of toxicity to MOA 1 and possibly MOA 2 (Baseline toxicity).
It was also concluded that:
Current ‘chemometer-based’ methods are preferable for hydrophobic chemicals relative to conventional methods. However, credibility needs to be enhanced by improved communication.
The domain of applicability needs to be carefully defined, and the limitations stated:
It works for those chemicals for which training sets exist.
A number of methods for measuring activity have been identified:
passive samplers for more hydrophobic pollutants,
head-space approaches for volatile chemicals,
utilising concentration data and dividing them by liquid solubility.
The conversion from concentration to chemical activity by the use of partition coefficients and calculated activity coefficients can be inaccurate.
There is a need to identify and clearly communicate the domain of applicability for various conversion methods.
Classification of chemicals
AOPs and biologically based assays will be useful to differentiate further MOA classes 3 and 4, which may then benefit from interpretation using a chemical activity approach to provide improved bioavailability metrics;
Verhaar classes offer a simple approach to differentiate between expected specific, reactive and narcotic MOAs for acute exposure with some recommendations (see below) if the chemical domain is covered;
In specifically acting and reactive classes a secondary consideration is to assess the target of action and consider if chemical activity can be applied  in a useful relevant manner;
Chemical activity seems best applied to MOA 1 and 2 with significant research effort necessary to see to what extent the application can be broadened to MOA 3 and 4;
Initial analyses of HC5 values have shown that recalculating HC5 into chemical activities leads to constant “threshold of toxicological concern”. This concept is interesting and could be useful in risk assessment;
New approaches in classification based on “omics” data are promising. The variability in omics data is presently high and the regulatory applicability is still limited. Large scale efforts with an important bio-informatics component are needed to overcome the current limitations and it is further recommended that efforts should be focused on increasing the applicability of omics for identifying MOAs of untested chemicals (including assigning chemicals to the narcosis MOA) and specifically on sub-lethal effects and species differences.
Future Research
Suggestions for future research were separated into three themes, (i) the chemical activity concept, (ii) application of the chemical activity approach and (iii) classification of chemicals. The topics listed in the following text were identified as areas where further research could prove important.
Chemical Activity Concept
QSARs should be re-evaluated in terms of chemical activity.
Insights into the validity of the assumption that the “cytotoxic burst” phenomenon is an in vitro analogue to baseline toxicity. Concentrations associated with the “cytotoxic burst” could be expressed as chemical activity to test the hypothesis that these concentrations would be equivalent to activity in the 0.1-0.01 range. Second, it would be useful to apply structure-based mode of action classification schemes to the Toxcast chemical library and examine the agreement (or lack thereof) between chemical structure-based identification of putative baseline (MOA 1,2) toxicant and biologically-based identification of baseline toxicants as based on the cytotoxic burst analysis.
Examination of the correlations between chemical activity and potency of ToxCast chemicals in specific assays and identify those for which a strong relationship exists. One could then examine the localisation and function of those targets in more detail and begin to investigate whether there is a scientifically-plausible theoretical basis on which to expect that activity based predictions would have value for predicting chemical potency against those targets.
Study of the applicability of the concept of chemical activity to chronic toxicity data, exercises to analyse MoA specificities of acute to chronic relationships for consistent data sets, e.g. zebrafish early life stage assay (FELS, OECD guideline 210, 2013) would be helpful.

Classification of chemicals
Refinement of the Verhaar classification scheme by:
Subcategorisation of the major classes;
Updating of chemical information;
Including information about the target site and target environment;
Extending the chemical domain.
Create a separate activity to improve existing classification schemes for non-narcotic chemicals.
Develop evidence to support the application of chemical activity to chronic toxicity for non-polar and polar narcotics.
Test the applicability of chemical activity in deriving threshold of toxicological concern (TTC).
Expand the application of chemical activity concept for mixtures.
Omics efforts should be focused on increasing the applicability for identifying MOAs of untested chemicals (including assigning chemicals to the MOA 1) and specifically on sub-lethal effects and species differences.

Application of the activity approach
Convert and compare critical body burdens to activity. The critical body burden concept has many similarities with the chemical activity approach, and it seems thus useful to relate and contrast data and results from both approaches. As an example, an activity-based conversion of narcosis critical body residue (CBR) data showed that there appeared to be some questionable CBRs in the selected set of baseline toxicant data (McCarty et al., 2013).
Apply the activity approach to data-rich chemicals. The chemical activity approach can be used to convert many different data sets into one “currency”, which then can (i) provide a basis for comparisons of data from different areas, (ii) help in utilisation of more existing data and (iii) facilitate the process towards an overview of the entire data basis for potential assessments and management actions.
Apply activity to monitoring data sets, which (i) is expected to give better data for heterogeneous and biological media and (ii) will help to connect measurements between environmental compartments.
Apply the activity concept and activity ratios in order to prioritise and guide monitoring chemicals.
Activity-based species sensitivity distributions (SSDs). Toxicity tests that were conducted at controlled chemical activity (i.e., via passive dosing) have been published recently. Determining the SSDs from such studies is expected to give an improved estimate of actual sensitivity distributions, since differences in exposure conditions between test methods, which normally confound the distributions, are largely accounted for.
Develop an activity calculator and make it publicly available.
Evaluate the agreement between computational and measured data. There are fundamental differences between calculating and measuring chemical activities, and it was found important to distinguish and to compare these two different ways to obtain activity data.
Use chemical activity to characterise and predict mixture toxicity, which was identified as one of the most important applications of the chemical activity framework during the initial presentations and the discussions.