Risk assessment, improved data needs

The evaluative process was used to assess lithium and boric acid for their potential to cause reproductive and developmental toxicity (J.A. Moore et al. 1995b, 1997). The final product is a summary statement that integrates reproductive and developmental data and exposure data. The summary also makes a statement about the certainty of judgment about an agent s reproductive and developmental risk, and critical data needed for reducing uncertainty and improving the risk assessment. 

Fear is contradictory to continuous improvement and a broad regulatory approach is needed to address this challenge. Timely risk assessment, communication, information, and collaboration between CMC review and CGMP inspection functions will be essential components of such a regulatory approach. In addition, common data bases and information systems will be necessary. 

Because of the gaps in our present understanding of carcinogenesis and the paucity of human data for most chemicals, risk assessments for chemicals are generally more uncertain than risk assessments for radiation. Tb improve such assessments, there is pai ticular need for further development and validation of methods for extrapolation from animal data to man and further refinement of methods for evaluating variations in hmnan susceptibility and human exposure, e.g., biological markers. 

REACH has provided a structure in which a well-informed chemicals risk management can be developed. In particular, it creates a legislative and regulatory framework for all substances in which the procurement of data for making reasonably reliable risk assessment is possible. But on the other hand, as we have seen, it does not require the creation of such data for all substances for which it is needed. This should be no surprise. The deficiencies in the previous system of chemicals regulation were so large that it would be unrealistic to believe that they could be solved in one single reform. It is only to be expected that there should be scope for improvement. A discussion is needed that identifies the most important of the potential improvements of the system, and in this spirit we would like to propose three important issues for the further development of REACH. 

In this context it is important to improve the analysis of the extent to which sensitive organisms and ecosystems in such areas may need specific test methods and specific concern in environmental risk assessment of chemicals (Breitholtz et al. 2006a). In the future, it is therefore important to increase research efforts to elucidate potential consequences of varying physical and chemical environmental factors for toxicity of a wide range of chemical substances, in order to develop tools for hazard identification and dose-response assessment that include scientifically well-based combinations of species, endpoints and environmental factors. The battery of endpoints to select from should, as far as possible, comprise population level data (Forbes and Calow 1999, Forbes et al. 2001, Breitholtz et al. 2006a), possibly obtained by using population models. 

There is a general agreement concerning the need to improve risk assessment of bioassay data through the incorporation of more information on mechanism of action of tumor induction. Investigative studies to identify the mechanisms of action of rodent epigenetic carcinogens may clarify the relationship between... 

Risk assessment deals with data acquisition, verification and monitoring. The collection of data is required and the need for additional data may be identified at any point in the process. Verification and monitoring can help to determine the overall effectiveness of the framework approach, provide feedback concerning any need for future modifications, help to evaluate the effectiveness and practicality of policy decisions and point out the need for new or improved risk assessment techniques (USEPA, 1992a). 

The National Academy of Sciences indicated in a recent study that the primary problem with risk assessment is the sparse data upon which decisions are based. Agriculture has an opportunity to meet the need for improved risk assessment and management through new research, new technology, and the use of that technology. 

In conclusion, validity assessments of the currently available QSARs in environmental sciences indicate that there are models for predicting some of the properties of potential contaminants reliably. If operated in accordance with their limitations, they can be valuable tools to fill data gaps. However, a complete coverage of all parameters and endpoints relevant to the hazard and risk assessment of chemicals has not yet been achieved. To improve and extend the techniques for estimating the intrinsic properties of environmental chemicals, experimental as well as theoretical aspects of these parameters need further elaboration. 

---