Data-derived uncertainty factors: Boric acid (BA) as a case study

There is growing support for the use of data-derived uncertainty factors. I n recent years, risk assessments of boric acid have been performed by sever al well-respected organizations, including IEHR, ECETOC, IPCS, and WHO. For each, the pivotal study was a developmental toxicity study in rats with a no-observed-adverse-effect level (NOAEL) of 55 mg BA/kg/day. These risk ass essments employed reduced uncertainty factors in the range of 25 to 60 for boric acid, because available pharmacokinetic data for boric acid reduced u ncertainty in evaluating the overall data base with this compound. However, a limitation of previous risk assessments was the absence of specific data on the renal clearance of boric acid in pregnant rats and pregnant women. New data has demonstrated that when renal clearance was normalized to body weight (ml/min/kg), pregnant rats cleared boric acid at a rate roughly thre e times greater than pregnant women. In addition, the boric acid specific a llometric relationship was determined from the log-log plot of clearance vs . body weight. Based on the new renal clearance data, it was estimated that pregnant women and rats would have the same AUC when pregnant women are gi ven 30% of the boric acid dose given to pregnant rats. In addition, the ren al clearance of boric acid among pregnant women varied by a factor of about 2. Therefore, boric acid-specific data on renal clearance in pregnant wome n and rats supports reduced interspecies and intraspecies pharmacokinetic u ncertainty factors of approximately 3 and 2, respectively. Further, growing evidence of the essentiality in animals, combined with consistency of effe cts among species in toxicity studies, suggests a reduced pharmacodynamic u ncertainty factor is appropriate for boric acid. Total uncertainty factors in the range of 22 to 44 are scientifically justified for this compound. An acceptable daily intake of 1.25 to 2.5 mg BA/kg/day is estimated by applyi ng an uncertainty factor of 22 to 44 to the NOAEL of 55 mg BA/kg/day. Data- derived uncertainty factors should be used whenever possible, and they shou ld be determined and applied in a consistent manner. Ultimately, estimates based on target tissue dose-adjusted relationships should offer a better ap proach to risk assessment.