Science Inventory

Shape and Steepness of Toxicological Dose-Response Relationships of Continuous Endpoints

Citation:

Slob, W. AND Woodrow Setzer. Shape and Steepness of Toxicological Dose-Response Relationships of Continuous Endpoints. CRITICAL REVIEWS IN TOXICOLOGY. CRC Press LLC, Boca Raton, FL, 44(3):270-297, (2014).

Impact/Purpose:

This re-analysis of a large number of historical dose-response data for continuous endpoints indicates that an exponential or a Hill model with four parameters both adequately describe toxicological dose-responses. Future analysis of an individual dataset could be combined with similar datasets (same endpoint) for other chemicals: subjecting such a combined dataset results in considerably smaller confidence intervals. This approach will be particularly useful for weak datasets (e.g. few doses, much scatter).

Description:

A re-analysis of a large number of historical dose-response data for continuous endpoints indicates that an exponential or a Hill model with four parameters both adequately describe toxicological dose-responses. The four parameters relate to the background response, the potency of the chemical, the steepness of the curve, and the maximum response. No exceptions were found for the datasets considered, which were selected on the sole criterion that they might provide information on the shape of the dose-response (e.g. more than the usual number of doses were tested). The datasets re-analysed related to a variety of endpoints (e.g., body/organ weights, AChE inhibition, blood parameters, micronuclei counts, cell proliferation, time-to-tumor), to both in vivo and in vitro studies of various types (e.g., repeated dose, developmental, cancer bioassay, LLNA test, genotoxicity test). Among endpoints, the parameter reflecting maximum response could be quite different, in concordance with existing experience (based on direct observations of dose-response data). Among chemicals dose-response shapes for a given endpoint/study type were found to be homogenous in the in vitro studies, while a mild variation in the steepness parameter appeared to be present in the in vivo studies. However, the latter observed variation is inflated by experimental noise in the data, and will be in reality smaller. The general conclusion that dose-response shapes are homogenous and simple seems to contradict current views among toxicologists and risk assessors. One of the reasons is that current views are based on visual inspection of dose-response data (or fitted models), while visual shapes of curves are misleading, and, in an absolute sense, have no meaning at all, as shown in the discussion section. Our findings have various practical consequences. The selection of the model(s) in the BMD approach for continuous endpoints appears to be easily answered, by selecting a four-parameter exponential or Hill model (or both). This implies that the BMD approach relying on an assumed model is not an issue. Further, model uncertainty can apparently be fully captured by uncertainty in the model parameters. The habit of using constraints on the model parameters to prevent “infinite” slopes at dose zero in fitting a model appears unjustified, both from the results of our comprehensive data-analysis, and for theoretical reasons. Instead, re-analysis of large numbers of dose-response datasets, such as in this paper, provides information on realistic ranges of parameter values that could be used as parameter constraints in future individual datasets. Or, any future analysis of an individual dataset could be combined with similar datasets (same endpoint) for other chemicals: subjecting such a combined dataset results in considerably smaller confidence intervals. This approach will be particularly useful for weak datasets (e.g. few doses, much scatter). In addition, this approach may open the way to use fewer animals in future studies.

URLs/Downloads:

http://informahealthcare.com/doi/abs/10.3109/10408444.2013.853726   Exit

Record Details:

Record Type: DOCUMENT (JOURNAL/PEER REVIEWED JOURNAL)
Product Published Date: 03/01/2014
Record Last Revised: 02/24/2015
OMB Category: Other
Record ID: 298572

Organization:

U.S. ENVIRONMENTAL PROTECTION AGENCY

OFFICE OF RESEARCH AND DEVELOPMENT

NATIONAL CENTER FOR COMPUTATIONAL TOXICOLOGY