Document VKKXm4g37b31NgrxDvVQ5JBMZ

ChmRisb A Division of McLaren/Hart Environmental Engineering The Courtland East Building 29225 Chagrin Boulevard Cleveland, OH 44122 (216) 464-6564 FAX (216) 464-6101 May 20,1994 Dr. Hasmukh C. Shah, PhD Chemical Manufacturers Association Vinyl Chloride Panel 2501 M. Street, N.W. Washington, DC 20037 Dear Dr. Shah, Thank you very much for your hospitality during our recent visit to Washington to present a seminar about our work on Vinyl Chloride risk assessments with PB-PK models. 1 am enclosing with this letter (1) an expense report since you were kind enough to offer to pay my travel expenses up to $600.00 (CMA contract VCRC-1.0-CONS REITZ), and (2) a hardcopy of the slides I used during my presentation. I took the liberty of reducing the slides by 50% and printing them two/ page. I think they are all still quite legible, but if you prefer I would also be happy to send you a full size copy. It was a real pleasure to meet you yesterday, and to interact with your Vinyl Chloride panel. If there are any occasion where McLaren/Hart could be of service to the CMA's Vinyl Chloride panel, we would be delighted to discuss these with you. Please make the expense check payable to me personally, since I charged all the expenses to my personal American Express Card, and please send the check to: Dr. Richard H. Reitz McLaren/Hart, ChemRisk Division 4105 Chelsea Ct. Midland, MI 48640. Page 1 CMA 119734 a Again, it was a real pleasure to meet you and I look forward to seeing more of you in the future. Sincerely, (f?S. 'ftfaot t> Richard H. Reitz, PhD, DABT Principal Health Scientist ^ ^ ^ i/c fr ^ -(n- ^ HhemRisk A Division of McLaren/Hart Environmental Engineering Page 2 CMA 119735 a 'vcvc'ea otc* H. Rtt, McUr*ft/H*t *{S1?)631-70d9 *13/23/94 10:39 AM 22/2 SEMINAR TOPIC #1 Predicting Cancer Risk irom Vinyl Chloride Exposure with a Physiologically-Based Pharmacokinetic Model. Kiciiard H. Reitz McLaren/Hart, ChemRisk Division Physiologically-based pharmacokinetic (PB -PK) models have been proposed as tools for facilitating the extrapolation of animal cancer tests to humans. For instance, several of us (Reitz, Andersen, Gargas, Cleweliydeveloped a risk assessment for methylene chloride which used PB -PK modeling to extrapolate between species, across different routes of exposure, and from high dose to low dose. This ride assessment predicted considerably less rlsk\to humans than earlier procedures, and the PB -PK risk assessment was consistent with epidemiology studies for methylene chloride (which are generally accepted as negative). Because the epidemiology studies for methylene chloride are negative (as the PB -PK model predicted they would be), they do not/provide an opportunity for comparing cancer rates in humans with predictions derived from the PB -PKmodeL However, there is one industrial compound (vinyl chloride, VC) which has been shown to produce measurable Increases in thfe Incidence of liwff'ahgiosarcoma in both rats and humans. In order to determine whether risk assessments based on PB -PK adjustments of dose adjustments are too conservative, not conservative enough, or about right, we have developed a PB -PK model for VC and predicted the risk to humans based on rat studies. The VC model for rats arid humans was developed from several in vivo and la vitro data sets. Both the rat arid human models were then validated with Independent studies of Vc metabolism in the appropriate species (i.e. the validation studies were not used in development of die model Itself). Then PB -PK model was combined with the multistage model of Crump & Howe (GLOBAL83) to predict the incidence of liver angiosacroma produced In humans. The human studies included more than 12,000 workers exposed to VC occupationally over a period of several decades. The risk assessment based on the PB -PK model predicted considerably less risk than would be calculated by default (nonpharmacokinetic) procedures. However, the incidence of angiosarcoria actually observed In human subjects was still considerably lower than predicted by the PB -PK/multistage model, suggesting that if the results obtained with VC are typical, risk assessments combining PB -PK and multistage modeling likely overpredict, rather than underpredict, human risk. CMA 119736 (171631-70*9 lE 3/23/W 10:39 AM 22/2 SEMINAR TOPIC #1 Predicting Cancer Risk irom Vinyl Chloride Exposure with a Physiologically-Based Pharmacokinetic Model. Richard H. Reitz McLaren /Hart, ChemRisk Division Physiologically-based pharmacokinetic (PB -PK) models have been proposed as tools for facilitating the extrapolation of animal cancer tests to humans. For instance, several of us (Reitz, Andersen, Cargas, Qe\yeh) developed a risk assessment for methylene chloride which used PB -PK modeling to extrapolate between species, across different routes of exposure, and froin high dose to low dose. This risk assessment predicted considerably less rlsk\to humans than earlier procedures, and the PB -PK risk assessment was consistent with epidemiology studies for methylene chloride (which are generally accepted as negative). Because the epidemiology studies for methytene chloride are negative (as the PB -PK model predicted they would be), they do notprovide an opportunity for comparing cancer rates in humans with predictions derived from the PB -PK model. However, there is one industrial compound (vinyl chloride, VO which has been shown to produce measurable Increases in thfe incidence of osarcoma in both rats and humans. In order to determine whether risk assessments based on PB -PK adjustments of dose adjustments are too conservative, not conservative enough, or about right, we have developed a PB -PK model for VC and predicted the risk to humans based on rat studies. The VC model for rats arid humans was developed from several in vivo and In vitro data sets. Both the rat arid human models were then validated with independent studies of Vc metabolism in the appropriate species (i.e. the validation studies were not used in development of the model Itself). Then PB -PK model was combined with the multistage model of Crump & Howe (CLOBAL83) to predict the incidence of liver angiosacroma produced in humans. The human studies included more than 12,000 workers exposed to VC occupationally over a period of several decades. The risk assessment based on the PB -PK model predicted considerably less risk than would be calculated by default tnonpharinacokinetic) procedures. However, the incidence of angiosaicorha actually observed in human subjects was still considerably lower than predicted by the PB -PK/multistage model, suggesting that if the results obtained with VC are typical, risk assessments combining PB -PK and multistage modeling likely overpredict, rather than underpredict, human risk. CMA 119737