Use of the benchmark dose approach in risk assessment Guidance of the Scientific Committee

Considering the need for transparent and scientifically justifiable approaches to be used when risks are assessed by the Scientific Committee and the Scientific Panels of EFSA, the Scientific Committee was requested by EFSA to assess the existing information on the utility of the benchmark dose (BMD) approach, as an alternative to the traditionally used NOAEL approach, and to make recommendations on whether EFSA should use the BMD approach and under which circumstances this use would be appropriate. The Scientific Committee was also asked to provide some guidance on how to use the BMD approach for analysing dose-response data from experimental studies, and to look at the possible application of this approach to data from observational epidemiological studies. Finally, the Scientific Committee was asked to advise on whether the selection of appropriate uncertainty factors are needed when using the BMD approach for deriving the Reference Point. Traditionally, when experimental animal data are used for risk assessment of substances in food, which are not genotoxic and carcinogenic, the No-Observed-Adverse-Effect-Level (NOAEL) and/or the Lowest-Observed-Adverse-Effect-Level (LOAEL) for the critical effect of a substance, forms the Reference Point for deriving health-based guidance values, such as an Acceptable Daily Intake (ADD. However, while this approach may utilise qualitative information, it does not use the data available in a quantitative way. In contrast, the BMD approach makes extended use of the dose-response data from studies in experimental animals or from observational epidemiological studies to better characterise and quantify potential risks. EFSA has not systematically used the BMD approach so far, although some EFSA Scientific Panels have been applying the BMD approach occasionally. Moreover, experimental data from which a BMD has been calculated are submitted from time to time. EFSA's Scientific Committee has also proposed in a previous Opinion to use the BMD approach for deriving the Reference Point for the estimation of Margins of Exposure for substances that are both genotoxic and carcinogenic. After comparing the strengths and weaknesses of the BMD and NOAEL approaches for deriving Reference Points for risk assessment, the Scientific Committee concludes that the BMD approach is a scientifically more advanced method to the NOAEL approach for deriving a Reference Point, since it makes extended use of available dose-response data and it provides a quantification of the uncertainties in the dose-response data. Using the BMD approach also results in a more consistent Reference Point, as a consequence of the specified benchmark response. Whilst software is currently available for BMD analysis, the Scientific Committee notes that this is evolving and significant developments are anticipated in the medium term, e.g. model averaging, analysis of continuous data. The Scientific Committee also considers that when toxicity test guidelines, e.g. OECD guidelines, are revised, specific aspects of the BMD approach are taken into account. Health-based guidance values derived using the BMD approach can be expected to be as protective as those derived from the NOAEL approach, i.e. on average over a large number of risk assessments. Therefore the default values for uncertainty factors currently applied remain appropriate and there is no need for any additional uncertainty factor. The Scientific Committee does not consider it necessary to repeat all previous evaluations using the BMD approach, because, on average, the BMD and NOAEL approaches give comparable results. Where refinement of previous risk assessments is considered necessary, for instance where the human exposure is close to the ADI, application of the BMD approach would be of particular value. The BMD approach is applicable to all chemicals in food, irrespective of their category or origin, e.g. pesticides, additives or contaminants. The BMD approach is of particular value for i) situations where the identification of a NOAEL is uncertain, ii) providing a Reference Point for the Margin of Exposure in case of substances that are both genotoxic and carcinogenic, and iii) dose-response assessment of observational epidemiological data. In the short term, the EFSA Scientific Panels and Units are strongly encouraged to adopt the BMD approach to situations such as those described above. In the longer term, the Scientific Committee anticipates that the BMD approach will be used as the method of choice for the determination of the Reference Points for deriving health-based guidance values and margins of exposure. To that end, recognising that there are practical considerations regarding its introduction and wider use in EFSA, and recognising that its application requires a level of expert judgement and modelling expertise, the Scientific Committee proposes that training in dose-response modelling and the use of the software be offered to experts in the scientific Panels and EFSA Units. The Scientific Committee would then like to review the implementation, experience and acceptability of the BMD approach in EFSA's work in two years time.