Toxicologically relevant benchmark responses for benchmark dose analysis of (continuous) hematological endpoints

Introduction A critical step in applying the benchmark dose (BMD) approach for choosing a point of departure (reference point) is to choose an appropriate benchmark response (BMR). For continuous toxicity endpoints, a BMR is defined as the maximum change in response which is considered non-adverse. It is recognized that BMRs are endpoint-specific, and that a default BMR of 5% for continuous endpoints, while expedient, may result in over-conservative or under-protective health-based guidance values. Methods In this study, we proposed toxicologically relevant BMRs for continuous endpoints by applying the Effect Size (ES) theory. Accordingly, BMRs were calculated for hematological endpoints such as various red blood cell (RBC)-related endpoints, white blood cell (WBC) count, and platelet count, based on the analysis of a large number of historical hematological datasets retrieved from the Chemical Effects in Biological Systems database of the National Toxicology Program (NTP). Results The analyses showed that the within-group variance ( var ) of the RBC endpoints is generally lower than that of the platelet count, while the var of the WBC count is the largest of the endpoints analyzed. Discussion On the basis of these analyses, endpoint-specific BMRs of 5% change for RBC-related endpoints, 10% for platelet count, and 20% for WBC count are recommended. These recommendations are based on rodent data intended to inform human risk assessment associated with repeated (long-term) exposure to chemicals. They are subject to reevaluation when additional data become available. The results of the analyses confirmed the assumption of a lognormal distribution for these continuous endpoints.