Retention Modeling of Diesel Exhaust Particles in Rats and Humans

Transcription

1 Retention Modeling of Diesel Exhaust Partiles in Rats and Humans C. P. Yu and K. J. Yoon Department of Mehanial and Aerospae Engineering, State University of New York at Buffalo, Amherst, NY Inludes the Commentary of the Institutets Health Review Committee Researh Report Number 4

2 HEALTH EFFECfS INSTITUTE 1- [ ~ The Health Effets Institute (HEI) is a nonprofit orporation founded in 198 to assure that objetive, redible, high-quality sientifi studies are onduted on the potential human health effets of motor vehile emissions. Funded equally by the U.S. Environmental Protetion Ageny (EPA) and 28 automotive manufaturers or marketers in the United States, HEI is independently governed. Its researh projets are seleted, onduted, and evaluated aording to a areful publi proess, inluding a rigorous peer review proess, to assure both redibility and high sientifi standards. HEI makes no reommendations on regulatory and soial poliy. Its goal, as stated by former EPA Administrator William D. Rukelshaus, is "simply to gain aeptane by all parties of the data that may be neessary for future regulations." The Board of Diretors Arhibald Cox Chairman Carl M. Loeb University Professor [Emeritus), Harvard Law Shool William. Baker Chairman [Emeritus), Bell Laboratories Health Researh Committee Rihard Remington Chairman University of Iowa Foundation Distinguished Professor of Preventive Mediine and Environmental Health, University of Iowa Joseph D. Brain Ceil K. and Philip Drinker Professor of Environmental Physiology, Harvard University Shool of Publi Health Leon Gordis Professor and Chairman, Department of Epidemiology, Johns Hopkins University, Shool of Hygiene and Publi Health Curtis C. Harris Chief, Laboratory of Human Carinogenesis, National Caner Institute Health Review Committee Arthur Upton Chairman Chairman, Department of Environmental Mediine and Diretor, Institute of Environmental Mediine, New York Shool of Mediine Bernard Goldstein Professor and Chairman, Department of Environmental and Community Mediine, University of Mediine and Dentistry of New Jersey, Robert Wood Johnson Medial Center Gareth M. Green Assoiate Dean for Eduation, Harvard Shool of Publi Health Millient W. P. Higgins Assoiate Diretor for Epidemiology and Biometry, National Heart, Lung and Blood Institute Herbert Rosenkranz Chairman, Department of Environmental and Oupational Health, Graduate Shool of Publi Health, University of Pittsburgh Offiers and Staff Andrew Sivak President and Treasurer Rihard M. Cooper Corporate Seretary Judith Zalon Lynh Diretor of Administration and Finane Kathleen M. Nauss Diretor for Sientifi Review and Evaluation Jane Warren Diretor of Researh William F. Busby, Jr. Senior Staff Sientist Brenda E. Barry Staff Sientist Aaron J. Cohen Staff Sientist Maria G. Costantini Staff Sientist Bernard Jaobson Staff Sientist Debra A. Kaden Staff Sientist Martha E. Rihmond Consulting Staff Sientist Donald Kennedy President, Stanford University Walter A. Rosenblith Institute Professor [Emeritus), Massahusetts Institute of Tehnology Roger. MClellan President, Chemial Industry Institute of Toxiology John W. Tukey Senior Researh Statistiian and Donner Professor of Siene Emeritus, Prineton University Mark J. Utell Professor of Mediine and Toxiology, University of Rohester Shool of Mediine Robert M. Senior Professor of Mediine and Diretor, Respiratory and Critial Care Division, The Jewish Hospital at Washington University Medial Center James H. Ware Dean of Aademi Affairs and Professor of Biostatistis, Harvard University Shool of Publi Health Mary C. Williams Professor of Mediine [Cell Biology), Boston University Shool of Mediine W. Kent Anger Speial Consultant to the Committee Assoiate Diretor for Oupational and Environmental Toxiology, The Oregon Health Sienes University Ann Y. Watson Consulting Staff Sientist Debra N. Johnson Controller L, Virgi Hepner Publiations Manager Gail V. Allosso Assistant to the Diretor of Administration and Finane Andrea L, Cohen Editorial Assistant Robin A. Cuozzo Aounting Assistant Jean C. Murphy Researh Assoiate Mary-Ellen Patten Senior Administrative Assistant Hannah J. Protzman Administrative Assistant Joye L. Speers Seretary to the President Carolyn N. White Administrative Assistant Charisse L. Smith Reeptionist Copyright 199 by Health Effets Institute. Printed at Capital City Press, Montpelier, VT. Library of Congress Catalogue No. for the HE! Researh Report Series: WA 754 The paper in this publiation meets the minimum requirements of the ANSI Standard Z [Permanene of Paper) effetive with Report Number 21, Deember 1988, and with Report Numbers 25 and 26 exepted. Reports 1 through 2, 25, and 26 are printed on aid-free oated paper.

3 Attainment of Study Objetives The Model: Results of Calulations and Preditions Remaining Unertainties and Impliations for Future Researh Conlusions Referenes Introdution Regulatory Bakground Sientifi Bakground Justifiation for the Study Study Objetives Tehnial Evaluation HEALTH REVIEW COMMITTEE'S COMMENTARY Health Effets Institute Predited Burdens in Humans Parametri Study of Retention Model Disussion and Conlusions Aknowledgments Referenes Appendix A. Kineti Equations for Diesel Soot and Partile-Assoiated Organis and Their Solutions Appendix B. Transport Rates of Diesel Soot and Partile-Assoiated Organis in Rats Appendix C. Transport Rates of Diesel Soot and Partile-Assoiated Organis in Humans About the Authors Publiations Resulting from This Researh Abbreviations Abstrat Introdution Speifi Aims Methods... 2 Partile Model for Clearane Study Retention Model and Kineti Equations Solutions to Kineti Equations Derivation of Transport Rates of Diesel Soot in Rats Derivation of Transport Rates of Partile-Assoiated Organis in Rats Method of Extrapolation to Humans Results Simulation of Rat Experiments Comparison Between Rats and Humans INVESTIGATORS' REPORT C. P. Yu and K. J. Yoon Retention Modeling of Diesel Exhaust Partiles in Rats and Humans TABLE OF CONTENTS Researh Report Number 4

4

5 1 The retention model of diesel exhaust partiles for rats was extrapolated to humans of different age groups, from birth to adulthood. To derive the transport rates for the human model, the mehanial learane from the alveolar region of the lung was assumed to be dependent on the speifi partiulate burden on the alveolar surfae. The redution in the mehanial learane in adult humans aused by exposure to high onentrations of diesel exhaust was found to be muh less than that observed in rats. The redution in hildren was greater than that in adults. For learane by dissolution, the transport rates were assumed to be the same for humans and rats. We ombined the retention and deposition models for diesel exhaust partiles to ompute the aumulated mass of diesel soot and the assoiated organis in various ompartments of the human lung under different exposure onditions. The lung burdens of both diesel soot and the assoiated organis were found to be muh higher in humans than in rats for the same period of exposure beause of the higher partile intake and slower learane rate in humans. The redution in learane aused by exessive lung burdens would not our in humans if the exposure onentration were kept below.5 mg/m 3. Also, it was found that for the same exposure, the lung burden per unit of lung weight was higher in hildren and reahed a maximum at about five years of age. These results are of use in assessing the health risk of exposure to diesel exhaust partiles. INTRODUCTION Diesel-powered motor vehiles provide onsiderably higher fuel eonomy and redued exhaust emissions of arbon monoxide and hydroarbons than do equally performing gasoline engines. However, they also produe signifiantly more partiulate matter. These partiles onsist prinipally of a ombustion-generated arbonaeous ore on whih various amounts of solvent-extratable organi ompounds are adsorbed. Some of these ompounds are arinogens and mutagens (Shuetzle 1983). The health effets of exposure to diesel exhaust have been of publi onern for many years and ontinue to be, given the potentially inreased use of diesel engines in future forms of transportation. Long-term exposure of animals to high onentrations of inhaled diesel exhaust onduted at different laboratories have shown that there is an aumulation of diesel soot (Chan et al. 1981, 1984; Griffis et al. 1983; ABSTRACT The objetive of this study was to predit the lung burden in rats and humans of diesel exhaust partiles from automobile emissions by means of a mathematial model. We previously developed a model to predit the deposition of diesel exhaust partiles in the lungs of these speies. In this study, the learane and retention of diesel exhaust partiles deposited in the lung are examined. A diesel partile is omposed of a arbonaeous ore (soot) and adsorbed organis. These materials an be removed from the lung after deposition by two mehanisms: (1) mehanial learane, provided by muoiliary transport in the iliated airways as well as marophage phagoytosis and migration in the noniliated airways, and (2) learane by dissolution. To study the learane of diesel exhaust partiles from the lung, we used a ompartmental model onsisting of four anatomial ompartments: nasopharyngeal, traheobronhial, alveolar, and the lung-assoiated lymph node ompartments. We also assumed a partile model made up of material omponents aording to the harateristis of learane: (1) a arbonaeous ore of about 8 perent of partile mass, (2) slowly leared organis of about 1 perent of partile mass, and (3) fast-leared organis aounting for the remaining 1 perent of partile mass. The kineti equations of the retention model were first developed for Fisher-344 rats. The transport rates of eah material omponent of diesel exhaust partiles (soot, slowly leared organis, and fast-leared organis) were derived using available experimental data and several mathematial approximations. The lung burden results alulated from the model showed that although the organis were leared at nearly onstant rates, the alveolar learane rate of diesel soot dereased with inreasing lung burden. This is onsistent with existing experimental observations. At low lung burdens, the alveolar learane rate of diesel soot was a onstant, equal to the normal learane rate ontrolled by marophage migration to the muoiliary esalator, whereas at high lung burdens, the learane rate was determined prinipally by transport to the lymphati system. 1 Correspondene may be addressed to Dr. C. P. Yu, Depa1tment of Aerospae and Mehanial Engineering, State University of New York at Buffalo, Room 314, Engineering East, Amherst, NY Retention Modeling of Diesel Exhaust Partiles in Rats and Humans. P. Yu 1 and K. J. Yoon INVESTIGATORS' REPORT

6 Heinrih et al. 1986; Wolff et al. 1987; Strom et al. 1988), formation of DNA adduts (Wolff et al. 1986), and inidene of tumors (Brightwell et al. 1986; Ishinishi et al. 1986; Iwai et al. 1986; Stober 1986; Mauderly et al. 1987) in the lungs of rats. No arinogeni effets of diesel exhaust were observed in hamsters (Heinrih et al. 1982, 1986), and onfliting results were reported in mie (Orthoefer et al. 1981; Pepelko 1982; Heinrih et al. 1986). Controlled experiments of diesel exposure have not been onduted on humans. Epidemiologial studies on the health effets of diesel exhaust are inonlusive. A study by the British Medial Researh Counil, whih evaluated London bus workers, laimed that diesel exhaust posed no serious threat to publi health (Waller 198). However, other studies onduted more reently (Harris 1983; Garshik et al. 1988) appear to show a definite relationship between exposure and lung aner risk. Despite all these unertainties, it is lear that the potential hazards of diesel exhaust are diretly related to diesel onentration and duration of exposure. Beause exposure to the gas phase of diesel exhaust does not result in tumor formation (Heinrih et al. 1986), work has foused on the effets of the partiulate phase. Central to any risk assessment of the partiulate phase is knowledge of the amount of inhaled partiles deposited in the lung during an exposure and the subsequent fate of the partiulate-phase omponents after deposition. Mathematial models have often been used to omplement experimental studies of the deposition and learane of inhaled partiles in the lungs. They not only enhane our understanding of the exposure-dose-response relationship for rats, but also provide a quantitative basis for extrapolating the data to humans. The deposition of diesel partiles in laboratory speies and humans was investigated using mathematial models in a previous study (Yu and Xu 1987b; see also Yu and Xu 1986, 1987a; Xu and Yu 1987). The predited deposition results in animals from this study were verified by atual experimental data. Although modeling studies on learane and retention have been onduted previously by many investigators (for a review see Morrow and Yu 1985; Oberdorster 1989), few studies dealt speifially with diesel exhaust (Soderholm 1982; Strom et al. 1988). In these studies, efforts were made to simulate the measured aumulations of diesel soot in rat lungs at high inhaled onentrations using mathematial models. The learane of the partile-assoiated organis from the lung was not addressed. Beause the organis ontain ertain polyyli aromati hydroarbons that are mutageni (Lewtas 1983; Brooks et al. 1984) and arinogeni (El Bayoumy et al. 1984), there is a ritial need to establish a omplete retention model that simulates the transport and removal of both diesel soot and the assoiated organis in the lung. 2 Retention Modeling of Diesel Exhaust Partiles in Rats and Humans SPECIFIC AIMS The speifi aims of this study were to develop a mathematial retention model of diesel exhaust partiles (DEPs) 2 in the lungs of rats on the basis of available experimental data, and to extrapolate this model from rats to humans for preditive uses. The model would onsider two speial features of learane for DEPs: (1) the omposition ofthe partile, that is, arbonaeous ore and the assoiated organis, and (2) the redution of the learane rate of diesel soot at high lung burdens. The retention model developed in this study, in onjuntion with the deposition model of DEPs developed earlier, would then be used to alulate aumulations of different material omponents of DEPs in various anatomial ompartments of the lung during exposure. METHODS PARTICLE MODEL FOR CLEARANCE STUDY Diesel exhaust partiles are irregularly shaped aggregates with a mass median aerodynami diameter (MMAD) of approximately.2 11m. The adsorbed organis generally aount for 1 to 3 perent of the partile mass. The exat size distribution of DEPs and the amounts and speifi omposition of the adsorbed organi ompounds depend on many fators inluding engine design, fuels used, engine operating onditions, and thermodynami proesses that our during exhaust. Extensive reviews of the physial and hemial harateristis of DEPs were made, respetively, by Amann and Siegla (1982) and Shuetzle (1983). To develop a mathematial model that simulates the deposition and learane of DEPs in the lung, appropriate partile models haraterizing a diesel partile must first be introdued. In the deposition study, we employed an equivalent sphere model for diesel soot with different aerodynami, diffusional, and intereptional diameters to simulate the dynamis and deposition of DEPs by various mehanisms (Yu and Xu 1987b). In the present learane study, we assumed that a diesel partile was omposed of three material omponents aording to their harateristi learane rates: (1) a arbonaeous ore of approximately 8 perent of partile mass, (2) adsorbed organis slowly leared from the lung of about 1 perent of partile mass, and (3) adsorbed organis quikly leared from the lung aounting for the remaining 1 perent of partile mass. The presene of two disrete organi phases in the partile model was suggested by observations that the removal of the 2 A list of abbreviations appears at the end of this report for your referene.

7 3 A,HB A,TB (i) I H rt (i) A,HG (i) _ (i) T B A, AT A,AB (i) A,TG (i) (i) ALB A (i) ra (i} A,AL Figure 1. Compartmental model of diesel partile retention. H = head; T ~ traheobronhial; A ~ alv~olar; ],. ~ lung-assoiated lymph nodes; B ~ blood; G ~ gastrointestinal trat; A,(') terms are the transport rates for the ompartments indiated in the subsript; and r(') terms are the mass deposition rates for the ompartments indiated in the subsript. obviously the most important ompartment for long-term retention studies. However, for short-term onsideration, retention in other lung ompartments may also be signifiant. The presene of these lung ompartments and the two outside ompartments in the model therefore provides a omplete desription of all learane proesses involved. In Figure 1, rw, rffj, and r~l are, respetively, the mass deposition rates of DEP material omponenti (i = 1 [ore], 2 [slowly leared organis], and 3 [fast-leared organis]) in the head, traheobronhial, and alveolar ompartments; and A.~y represents the transport rate of material omponent i from ompartment X to ompartment Y. Let the mass fration of material omponent i of a diesel partile be fi. Then (i) L G rh = firh' (1) (i) ry = firt, (2) (i) ra = fira ' (3) where rh, ry, and ra are, respetively, the total mass deposition rates of DEPs in the H, T, and A ompartments, determined from the equations partile-assoiated organis from the lung exhibited a biphasi learane urve (Sun et al. 1984; Bond et al. 1986). This urve represents two major kineti learane phenomena: a fast phase of organi washout with a half-time of a few hours and a slow phase with a half-time that is a few hundred times longer. The detailed omponents involved in eah phase of the learane are not known. It is possible that the fast phase onsists of organis that are leahed out primarily by diffusion mehanisms while the slow phase might inlude any or all of the following omponents: (1) organis that are "loosened" before they are released; (2) organis that have beome interalated in the arbon ore, whih impedes release; (3) organis that are assoiated for longer periods of time due to hydrophobi interation with other organi phase materials; (4) organis that have been ingested by marophages, and as a result, effetively remain in the lung for a longer period of time due to metabolism by the marophage-metabolites formed may interat with other ellular omponents; and (5) organis that have diretly ated on ellular omponents, for example, by forming ovalent bonds with DNA to form adduts. The above distintion of the organi omponents is largely mehanisti, and it does not speifially imply the atual omponent nature of the organis adsorbed on the arbonaeous ore. However, this distintion is neessary in appreiating the dual-phase nature of DEPs. For aerosols made of pure organis, suh as benzo[a]pyrene (BaP) and nitropyrene (NP), in the same size range of DEPs, Sun and olleagues (1984) and Bond and assoiates (1986) observed a nearly monophasi learane urve. This might be explained by the absene of interalative phenomena (2) and of hydrophobi interation imposed by a heterogeneous mixture of organis (3). The measurement of a pure organi might also neglet that quantity that has beome intraellular (4) or ovalently bound (5). RETENTION MODEL AND KINETIC EQUATIONS Diesel exhaust partiles are removed from the lung by two prinipal mehanisms: (1) mehanial learane, provided by muoiliary transport in the iliated airways, marophage phagoytosis, and migration in the noniliated airways, and (2) learane by dissolution. Under normal irumstanes, diesel soot is removed by mehanial transport, and the partile-assoiated organis are removed by dissolution. To study the transport and removal of DEPs from the lungs, we used a ompartmental model onsisting of four anatomial ompartments: the nasopharyngeal or head (H), traheobronhial (T), alveolar (A), and lung-assoiated lymph node (L) ompartments, as shown in Figure 1. In addition, we used two outside ompartments, the blood (B) and gastrointestinal trat (G). The alveolar ompartment in the model is C. P. Yu, K. J. Yoon

8 Retention Modeling of Diesel Exhaust Partiles in Rats and Humans rh = C(TV)(RF)(DF)H, rt = C(TV)(RF)(DF)T, ra = C(TV)(RF)(DF)A. In equations 4 through 6, C is the mass onentration of DEPs in the air, TV is the tidal volume, RF is the respiratory frequeny, and (DFJH, (DFJT, and (DFJA are, respetively, the deposition frations of DEPs in H, T, and A ompartments over a respiratory yle. The values of (DFJH, (DFJT, and (DFJA, whih vary with partile size, breathing onditions, and lung arhiteture, were determined from our deposition model (Yu and Xu 1987b). The differential equations for m ~, the mass of material omponent i in ompartment X, as a funtion of exposure time t an be written as Head (H) Traheobronhial (T) dmyj I dt = ryj - "-YJmYJ - J..fj)BmYJ, (7) dm[ti)ldt = r(ti) + J..Ul m[il - J..Ul m[il - J..Ul m[il (8) AT A TG T TB T ' Alveolar (A) dm~l I dt = r~l - /..~~m~l - "-~im~l - "-~km~l, (9) Lymph nodes (L) dm[illdt - J..Ul m[il - J..[il m[il L - AL A LB L Equation 9 may also be written as where dm[illdt _ r[il _ J..Ulm[il A - A A A ' (4) (5) (6) (1) (11) J,.[iJ - J..Ul + J..Ul + J..Ul A - AT AL AB (12) is the total learane rate of material omponent i from the alveolar ompartment. The total mass of the partile-assoiated organis in ompartment X is the sum of m~l and m~l, and the total mass of DEPs in ompartment X is equal to m - m(1l + m(z) + m(3l X- X X X (13) The lung burdens of diesel soot (ore) and organis are defined, respetively, as m( 1 l - m(ll + m( 1 ) (14) Lung - T A ' mfu~~) = m~l + m)fl + m~l + m~l. (15) Beause the learane of diesel soot from ompartment T is muh faster than from ompartment A, m~l~ m~l a short time after exposure, and equation 14 leads to m(ll "'m(1) Lung- A (16) Solutions to equations 7 through 1 an be obtained one all the transport rates /..~yare known. When A~y are onstant, whih is the ase with linear kinetis, equations 7 through 1 will have solutions that inrease with time at the beginning of exposure but eventually saturate and reah a steady-state value. This is the lassi retention model developed by the International Commission on Radiologial Protetion (ICRP) (1979). However, experimental data have shown that when rats were exposed to DEPs at high onentrations for a prolonged period, the diesel soot aumulated in various peribronhial and subpleural regions in the lungs and the long-term learane was impaired (Chan et al. 1981, 1984; Griffis et al. 1983; Oberdorster et al. 1984; Heinrih et al. 1986; Wolff et al. 1987; Strom et al. 1988). No suh hange was observed regarding muoiliary learane (Wolff et al. 1987). The redution in the ability to lear partiles from the deep lung at high lung burdens was also observed for other insoluble partiles (Ferin and Feldstein 1978; Vinent et al. 1987; Muhle et al. 1988; Strom et al. 1989), as shown in Figure 2. This is alled the overload effet. Although the real ause of this effet is presently unknown, Morrow (1988) postulated that it was due to a derease in alveolar marophage mobility aused by an exessive number of partile-laden ells, as well as by the volumetri inrease of ell size due to phagoytized partiles, rather than by the diret toxi effets of the partiles. A funtional relationship between the alveolar learane rate and the lung burden has been derived mathematially on the basis of this hypothesis (Yu et al. 1989). The ontinuous buildup of DEPs in the lung at high lung burdens during a prolonged exposure annot be predited by the lassi ICRP model. A revised model, designed spe- "'..2 "' "' "' "' "' v X.1 t> {) Vb.b, 'ij)i,q >- A : 'i1 i5 "' a "' "' [> :::.,V [> "' 2<:.5 v -<. r:+ v.2.1 <l <l [> "' "' [>. [> (1) ma (mg) Figure 2. Comparison of A.~l as a funtion of main rats between DEPs and other insoluble partiles based on the data from Chan and oworkers (1981), + Chan and oworkers (1984),.6. Griffis and assoiates (1983), -f-oberdorster and olleagues (1984), Wolff and assoiates (1987), T Strom and olleagues (1988) for DEPs; and data from Ferin and Feldstein (1978),!::, Vinent and oworkers (1987), X, [>, 'i7, <> Muhle and assoiates (1988), and <l Strom and olleagues (1989) for other insoluble partiles. 4

9 5 For a onstant r~l, equation 17 has a few simple solutions at limiting ases. At the beginning of an exposure, A.~lm~l in equation 17 is muh smaller than rfll and the solution of equation 17 takes the form of (18) where m~lo is the value of m{l) at t =. Equation 18 represents a linear buildup of m~l with time. This solution prevails for a long period of time if the soot onentration is high, a result observed in many experiments onduted at high levels of exposure (for example, Wolff et al. 1986). After longer periods of exposure, however, A.~lm~l eventually reahes the value of r~l and equation 17 has a steady state solution of m{l) - r~l A - A,{l), Aoo (19) where A.~l"' is the value of A.~l as time approahes infinity. In this ase, partile intake is balaned by removal due to learane. The time required to reah the steady state inreases with the exposure onentration. The general solution to equation 17 for onstant r~l at any time, t, an be obtained by the separation of variables to give (2) If r~l is an arbitrary funtion of t, equation 17 must be solved numerially suh as by a Runge-Kutta method (Press et al. 1989). One m~l is found, the other kineti equations 7 through 1 for both diesel soot and the partile-assoiated organis an be solved readily, sine they are linear equations. The solutions to these equations for onstant rfj), r~l, and r~l are given in Appendix A. The transport rates A.i{)y in the retention model formulated above need to be determined from experimental data. For instane, if the data for mi{.j in every ompartment during an exposure are known, A.i{)y an be determined from this information using the retention model. Up to the present time, a omplete set of data is not available for both diesel soot and the partile-assoiated organis from a single exposure experiment. In this study, the transport rates of diesel soot and organis for rats were derived separately from the best data available to date. Beause these data did not provide information for all transport rates, several approximations were used in the derivation. DERIVATION OF TRANSPORT RATES OF DIESEL SOOT IN RATS The transport rates of diesel soot in rats were derived using the lung burden and lymph node burden data from ifially for DEP removal from the lung, was proposed by Soderholm (1982) and improved by Strom and assoiates (1988). This model subdivided the alveolar region into two ompartments on the basis of the physiology of learane. The first ompartment was assoiated with mobile, phagoyti marophages and was alled the marophage ompartment, and the seond was identified by slowly moving, lustered, partile-laden marophages and, therefore, was alled the sequestering ompartment. Although both ompartments eliminate DEPs to the lymph nodes, the marophage ompartment also eliminates DEPs via muoiliary transport in the traheobronhial tree to the gastrointestinal trat and to the sequestering ompartment. Through the laborious task of fitting the alulated results from their model with experimental data, Strom and oworkers (1988) were able to find a set of transport rates between various ompartments that give the best data fit. However, the uniqueness of their solution annot be guaranteed beause of the larger number of transport rates involved. The inreased number of transport rates also prevents a ready extrapolation of the model to humans. In the retention model proposed in Figure 1, the buildup of lung burdens and the introdution of overload effets were simulated mathematially using a single ompartment for the alveolar region of the lung. The effet of sequestration was aounted for by a set of variable transport rates A.~~. A.~L and A.~l from this ompartment suh that A.~~. A.~i, and t..w are funtions of rna, whih is the total mass of DEPs in ompartment A. Without the hypothetial sequestering ompartment in the model, the kineti equations for transport of material (soot and organis) are onsiderably simplified, and are more readily used for interspeies extrapolation. The transport rates A.~land A.~i in equations 7 through 1 an be determined diretly from experimental data on lung and lymph node burdens, and A.~~ and A.~k an be determined from equation 12. SOLUTIONS TO KINETIC EQUATIONS Equation 11 is a nonlinear differential equation of m~l with known funtion of A.~l. For diesel soot, this equation beomes dm( 1 ljdt - r{l) - A,( 1 l(m )m{l) A -A AAA (17) Beause learane of the partile-assoiated organis is muh faster than learane of diesel soot, m~l and m~l onstitute only a very small fration of the total partile mass (less than one perent) after a long exposure, and we may onsider A.~l as a funtion of m ~) alone. Equation 17 is then redued to a differential equation with m~l as the only dependent variable. C. P. Yu, K. J. Yoon

10 6 Figure 5. Experimental and predited llmph node burdens of rats exposed to DEPs at a onentration of 6. mg/m for 1, 3, 6, or 12 weeks. The solid line represents the predited burdens during exposure and the dashed lines represent those during postexposure. Partile harateristis and exposure pattern are explained in the text. The symbols represent the experimental data from Strom and olleagues (1988). Figure 3. Experimental and predited lung burdens of rats exposed to DEPs at a onentration of 6. mg/m 3 for 1, 3, 6, or 12 weeks. The solid line represents the predited burdens during exposure and the dashed lines represent those during postexposure. Partile harateristis and exposure pattern are explained in the text. The symbols represent the experimental data from Strom and olleagues (1988). Time (Week) 65 Time!Week) 12wk." 6~-----o s Oi Q) 'E ::J!Il Q) " z.<:::. E.3' -~ ;1, _ 12 wk I dl Oi 2 s 16 Q) :; "!Il w CJ) ::J...J ;,?, Q) 'E ::J!Il CJ) ::J...J " Q) -~ 6 m 4 E z 2 26 Time (Week) 3wk <!> --- "' _1:'!_k_ Figure 4. Experimental and predited normalized lung burdens of rats exposed to DEPs at a onentration of 6. mg/m 3 for different exposure spans. The dashed lines are the predited burdens during postexposure. The symbols represent the experimental data from Strom and olleagues (1988). signifiantly over the postexposure period for eah of the four postexposure experiments, but the differenes between the initial postexposure lung burdens were muh larger. Under this assumption, equation 17 an be written as dm( 1 ljdt = -A (m(1l )ml 1 l and the solution to equation 21 is A A Ap A ' (21) m~l / m~k = exp[- AA (m~k)(t - tp)], (22) where tp is the time at whih the postexposure period starts and m~k is the alveolar burden of diesel soot at t 65 a single experiment by Strom and olleagues (1988). Beause other investigators (Griffis et al. 1983; Chan et al. 1984; Heinrih et al. 1986; Wolff et al. 1986) did not measure burden outside of the lung, data from these studies are not suffiient to determine the transport rates of diesel soot from ompartment A to the other ompartments in the retention model. In the experiments of Strom and assoiates (1988), male Fisher-344 rats were exposed to diesel exhaust diluted to a nominal mass onentration of 6 mg/m 3 for 2 hours/day and 7 days/week for anywhere from 3 to 84 days. Animals were killed for as long as a year after exposure in order to obtain lung and lymph node burdens that would eluidate the mass dependeny of partile retention. The experimental data showed that the lung burden aumulated in a nearly linear manner for the first 12 weeks of exposure, but approximated an exponential deline during the postexposure period, as shown in Figure 3. The magnitude of deline, however, dereased with the initial postexposure lung burden. To make the learane more apparent, lung burdens were normalized by their initial postexposure values and were plotted as the perentage of retained mass in the lungs versus postexposure time, as shown in Figure 4. The lymph node burdens during postexposure were also measured and found to be strongly dependent on the initial postexposure lung burden (Figure 5). To derive the expressions for A~l and A~i as funtions of m~l from these data, we assumed a learane model during postexposure in whih A~l and A~i, depended only on the value of m~l at the beginning of postexposure, m~k The reason for this assumption is that the m~l did not derease Retention Modeling of Diesel Exhaust Partiles in Rats and Humans

11 C. P. Yu, K. J. Yoon tp. Equation 22 was used to fit the individual data in Figure 3 by means of a nonlinear regression proedure (Dixon 1985}. This proedure minimized the funtion 4(Yi - Yi) 2!yf where Yi is the lung burden and Yi is th~ alulated lung burden. The best statistial fits of the data resulted in four different values of J...~l, eah orresponding to a given value of m~b shown in Table 1. The funtional relationship between J...~l and m~b an be desribed by an exponential funtion of the form J...~l = a exp [- b(m~bl J + d day - 1, (23} where m~b is expressed in milligrams, and the onstants a, b,, and d were found to be.12,.11, 1.76, and.86, respetively, from the four sets of values of J...~l and m~b During exposure, equation 23 is modified by replaing m~b with m~l, beause J...~l depends on the instantaneous value of m~l. It follows that J...~l =.12 exp[ -.11(m~l)1.76] +.86 day- 1. (24} Although derived at high levels of exposure, equation 24 an be utilized under any exposure ondition. For the speial ase of low levels of exposure, m~l should remain low and is limited to the steady-state value of r(1) m(1) A _ - A A,(1), (25} AD where A. ~) =.129 day - 1 is the value of J...~l obtained from equation 24 as m~l -+. In this ase, the overload ondition will never be reahed. To derive an expression for A.~i as a funtion of m~l from the data for lymph node burden shown in Figure 5, we used the solution of equation 1 for diesel soot during postexposure. This solution is m(1) - exp(- A,(1) t)[ rt :~,( 1 ) m(1) exp(a.(1) t)dt + m(1) L - LB Jt AL A LB Lp p ( ) exp(a.fstp)], (26} where mn is the lymph node burden of diesel soot at t = tp. Table 1. Values of J...~l Derived from the Diesel Soot Retention Experiment of Strom and Colleagues (1988} mvl Ap (mg) X X X X 1-4 As in the derivation of J...~l, we again assumed that A.~i was only a funtion of m~b beause m~l did not vary appreiably during postexposure. After substitution for m~l in equation 22, equation 26 beomes m(1) A.(1) m(1) m(1l exp[ A.(1) (t t )] Ap AL L - Lp - LB - P - A, (1) _ A, (1) A LB [exp(- J...~l(t - tp)) - exp(- A.~hrt - tp))]. (27} There are no data urrently available on the transport of diesel soot from various lung ompartments to blood. Beause of the small partile size, it is oneivable that the soot partiles may penetrate the alveolar wall and enter the blood stream. Assuming this, we have A.~k = A.~1 = onstant. (28} In addition, the mehanial transport of diesel soot from ompartment A to ompartment T was assumed to stop ompletely at very high m~l, that is, lim A.~~ =. m~l-+oo Then, from equation 12, we obtain A. (1) = lim (A. (1) _ A. (1) _ A. (1) ) AB ( 1 ) A AT AL ma-+oo With the use of equation 28, equation 27 beomes m(1) A,(1) (1) (1) (1) _ Ap AL ml - mlp exp(- J...AB(t - tp)] - - (1) (1) J...A - J...AB [exp(- J...~l(t - tp)) - exp(- A.~k(t - tp))]. (29} (3} (31} Again, as for J...~l, an exponential funtion was used to approximate the relationship between A. ~i and m~l, and we have (32) Therefore, A. ~k = d - 8. Equation 31 was used to ompute a, ~. y, and 8 from the best fits offour sets of data points for lymph node burden at different mn, shown in Figure 5, following the same proedures used in determining J...~l. The values of a, ~, y, and 8 were found to be -.68,.46, 1.62, and.68, respetively. Thus, and during exposure A.~k = d - 8 =.18 day - 1, (33} A.~i = -.68 exp [-.46(m~l)1.6Z J +.68 day- 1. (34} Substituting equations 24, 33, and 34 into equation 12, we obtain 7

12 8 A(3J - 4A(3J AB - AL' (43) whih are the relationships suggested by the ICRP model (1979) for soluble partiles, we obtain from equations 4 and 41 Lung Burden per Gram of Lung (mg/g) Figure 6. Dependene of dimensionless learane rate :1.~ 1 I :\.~ 1 ~ on m~l per gram of lung. The solid line represents the theoretial results from equation 24. The symbols are the data points from various soures. The lung weight is assumed to be 1.5 g. and where A( 1 J day - 1 is the value of A( 1 J at m( 1 J-+ ATO - ' AT A alulated from equation 35. If we further assume that A(2J - 4A(2J AB - AL' (42) (41) A(3) - A(3) + A(3) + A(3) AO - ATO AB AL and (39) (4) At low lung burdens, we have (2) - (3) - (1) - (1) ~~,AT- ~~,AT- ~~,AT- ~~,ATO' A(2) - A(2) + A(2) + A(2) AO - ATO AB AL \l Griffis et al o Chan et al Heinrih et al f\. Wolff et al o Strom et al (37) (38) A(2) - A(2) + A(2) + A(2) A - AT AB AL A~l = A~~ + A~k + A~i. and DERIVATION OF TRANSPORT RATES OF PARTICLE-ASSOCIATED ORGANICS IN RATS The learane rates of partile-assoiated organis for rats were derived from the retention data of Sun and olleagues and m~l at the time at whih the postexposure period >. >!') "' :::'- i::o::t: AB,<: ~ m~l (mg) Figure 7. Variation of transport rates :1.~ 1,!Jl, :1.~ 1 ]., and :~.!;1 with m~l. (1984) for BaP and the data of Bond and oworkers (1986) for NP adsorbed on diesel partiles. The results of these measurements may be written in the following mathematial form: AL xi) AT -}.1l _f_ 3 - exp(- A~lt), f2 + h (36) starts, and where m~b and m~b are, respetively, the values of m)fl The derived expressions for A~l, A~i, A~~. and A~k are plotted in Figure 6. It is apparent that the major omponent of A( 1 J is A( 1 l when m( 1 J is smaller than 4 mg and that A( 1 l is dominated by A~i when m~l is larger than 12 mg. A AT A A An alternative approah to deriving the expressions for A~i and A~~ is to assume that A~k = (perfetly insoluble) instead of using equation 29. Beause A~k onstitutes only a very small fration of A~J, this approah would hange the lung burden alulation only slightly (less than 1.5 perent). Thus, the urrent approah is adequate and also offers a more general desription of the learane proess. To further illustrate the dependene of alveolar learane rate A~J on partiulate mass burden in the lungs, we plotted A~J;A~l versus m~l per gram of lung in Figure 7, where, again, A~Jo is the value of A~J for the limiting ase of m~l-+. This permitted omparison between the alveolar learane rate given by equation 24 and rates obtained from other studies of diesel soot retention (Griffis et al. 1983; Chan et al. 1984; Heinrih et al. 1986; Wolff et al. 1986) beause only the ratio A~lfA~Jo was measured in some ofthese studies. Figure 7 shows that all data are onsistent in that A~J;A~l dereases with inreasing m~l, but there are substantial differenes in magnitude among different studies. A~~ = A~l - A~i - A~k.12 exp [-.11(m~l) 1 76 J +.68 exp[ -.46 (m~l) 1 62 ] day- 1. (35) Retention Modeling of Diesel Exhaust Partiles in Rats and Humans

13 C. P. Yu, K. J. Yoon and A(2) ~ (A(2) - A(2) ) AB - 5 AD ATD A.(3J ~ (A.(3J - A.(3J ) AB - 5 AD ATD (44) (45). h h d. ' (2) '(3) (2) ' (3) Equatwns 37 t roug 45 etermme "AT "AT A AB "AB A.( 2 l and A.( 3 ) for given measured values of A.( 2 ) and A.( 3 l AL' AL AD AD The transport rates of partile-assoiated organis in DEPs were determined in the manner desribed above. Beause the A~JD and A~JD of all the organis adsorbed on the partile have not been measured experimentally, the mean of the transport rates of the partile-assoiated BaP and NP measured by Sun and assoiates (1984) and Bond and olleagues (1986) were used as the representative transport rates of the organis. This led to and A.~JD =.288 day- 1, (46) The final results for partile-assoiated organis are and (47) "-~k = 4A.~i =.129 day - 1, (48) "-~k = 4A.~k = day - 1, (49) '(2) - '(3) - '(1) "AT - "AT - "AT.12 exp[ -.11(m~l) 1 76 ] +.68 exp[ -.46(m~l)1.6 2 ] day- 1, (5) A~) exp[ -.11(m~l) 1 76 ] +.68 exp[ -.46(m~l)1.6 2 ] day- 1, (51) A,~l = exp[ -.11(m~l) 1 76 ] +.68 exp[ -.46(m~l)1.6 2 ] ~ 15.7 day- 1. (52) Other approximations for the transport rates of partileassoiated organis are and ' (2) - '(2) - ' (2) - '(2) "HB - " TB - "LB - "AB ' ' (3) - ' (3) - ' (3) - ' (3) "HB - " TB - "LB - "AB (53) (54) The remaining transport rates in our model were not ritial to the model development and were estimated from the data of Chan and olleagues (1981) and the ICRP model (1979). These are and "-We = 1.73 day -l, = 1,2,3 (55) A.(j}e =.693 day- 1, i = 1,2,3. (56) Appendix B lists all the transport rates of diesel soot and partile-assoiated organis derived for rats. It should be emphasized that exept for the values of A~JD' A.~lD' "-We and A~t whih are given by equations 46, 47, 55, and 56, respetively, all other A~y were derived using experimental data. The derivations were based on reasonable assumptions that, in any ase, do not ontribute signifiantly in magnitude to the value of A~l. For simulating the existing experiments and prediting the results of lung burden under new onditions, the retention model presented above will prove to be valuable. The struture of the retention model is designed to permit ready modifiations that an aommodate newer, more detailed, ompartmental data and analyze additional transport rates. METHOD OF EXTRAPOLATION TO HUMANS Experimental data on the deposition and learane of DEPs in humans are not available. In order to estimate the lung burden of DEPs for human exposure, we must extrapolate the retention model from rats to humans. The extrapolation method onsists of two steps. The first step is to determine the deposition rates r~ for humans. This work was previously reported (Yu and Xu 1987b). In the seond step, we derived a set of transport rates (A~y) for humans. Although it is known that many physiologial time onstants, suh as respiratory frequeny and metaboli rate, are bodysize dependent, the transport rates of partile-assoiated organis are believed to be insensitive to body size. The lung learane rate of inhaled lipophili ompounds was shown to be dependent only on their lipid/water partition oeffiients and to be independent of speies (Shanker et al. 1986). In ontrast, the transport rates of diesel soot were speies dependent. Differenes in the alveolar learane rates of insoluble partiles at low lung burdens among speies were observed in numerous previous studies (for example, Snipes et al. 1983; Bailey et al. 1985a,b; Snipes and MClellan 1985). Respetive retention half-times ranged from about 5 to 1 days in rats, mie, and hamsters to several hundred days in dogs, guinea pigs, and humans. The reason for suh a large interspeies differene is not yet understood. The number of respiratory bronhioles, partile deposition pattern, learane pathway length, and alveolar marophage number and mobility may all ontribute to suh a differene. Assuming that diesel soot is leared from the human lung as other insoluble partiles are, we obtained a value of "-~h =.169 day -l for humans using 9

14 (6) W = T/(1 + T), for T ~ 56 days, where T is the age of the rat measured in days. To test the auray of our model, the transport rates listed in Appendix B were used in equations 7 through 1 1 Table 2, Ratio of Pulmonary Surfae Areas Between Humans and Rats (S) for Different Ages of Humans Age (years) s to alulate the retention of diesel soot in the rat lung, and these values were ompared with the data on lung burden and lymph node burden obtained by Strom and olleagues (1988). The partile deposition rates rh, rr, and ra were omputed from the deposition model for rats (Yu and Xu 1987b). A partile size of.19 f..!m MMAD and a geometri standard deviation, ag, of 2.3 (as used in Strom's experiment) were used in the alulation. The respiratory parameters for rats were based on their weight and were alulated using the following orrelations of minute volume, respiratory frequeny, and growth urve data. Minute Volume =.9W (m 3 /min) (58) Respiratory Frequeny = 475W~ 3 (L/min) (59) in whih W is the body weight (in grams) as determined from the equation the data from Bailey and assoiates (1982). This is about 7.6 times less than the value of A~b observed for rats. There are, as yet, no data available on the hange in alveolar learane due to exessive lung burdens in humans. Although human exposure to environmental diesel exhaust is not likely to result in lung overload, it is desirable to derive relationships between the transport rates and lung burdens in order to determine the exposure onditions under whih overload might our. From equation 24, it is seen that A~l for rats onsists of two terms, marophage-mediated mehanial learane and learane by dissolution. The first term depends on the lung burden m~l, whereas the seond term does not. To extrapolate this relationship to humans, we assumed that the mehanial learane term for humans varied with the speifi partiulate dose to the alveolar surfae in the same proportion as in rats, while the dissolution learane term was speies independent. This assumption resulted in the following expression for A~l in humans: A~l =!!._ exp[ -b(m~lfs)] + d (57) p where P is a onstant derived from the human-rat ratio of the alveolar learane rate at low lung burdens, and Sis the ratio of the pulmonary surfae area between humans and rats. Equation 57 implies that rats and humans have equivalent degrees of biologial response in the lung to the same speifi surfae dose of inhaled DEPs. In equation 57, S ould have been taken to be the ratio of marophage numbers in the respetive speies' lung fields for extrapolation. However, the number of marophages is highly variable and no data reliably quantify suh numbers. The value of P was obtained by letting A~l -+ A~b =.169 day~ 1 as m~l -+ in equation 57. This led to P = Also, we found S = 148 using data from the anatomial lung model ofyeh and Shum (198) for rats and the Weibel (1963) model for human adults. For humans 25 years of age or younger, we assumed the same value for P, but S was obtained from the lung model for young humans (Yu and Xu 1987b). The values of S for different ages are shown in Table 2. The expressions for other transport rates that have a lung burden-dependent term were extrapolated from rats to humans in a similar manner, and the onstant terms remained unhanged. Appendix C shows all the transport rates derived for the human model used in the retention study. RESULTS SIMULATION OF RAT EXPERIMENTS Retention Modeling of Diesel Exhaust Partiles in Rats and Humans

15 Time (week) Figure 9. Calulated lung burdens in humans of different ages exposed via nose breathing to DEPs at 6 mg/m 3 for 12 weeks at 2 hours/day and 7 days/week. Dashed lines are the orresponding burdens in rats Time (day) Figure 8. Comparison between tbe alulated lung retention (solid line) and tbe experimental data (data points witb error bars) obtained by Sun and oworkers (1984) for the partile-assoiated BaP in rats. 5 2 s >2 Q) "' 15 :::J []) C> 3 1 ~ 6 _Q 2 Q) 4 : 25 8 Soot 3 1 used in the derivation of the transport rates in the retention model. The omparisons provide, nonetheless, a hek of the auray of the deposition model and ensure that all approximations used in the derivation were reasonable. The ultimate suess of the model will depend on future experimental validation. COMPARISON BETWEEN RATS AND HUMANS To gain an understanding of the differenes in the lung and lymph node burdens between rats and humans under the same exposure onditions, we alulated these quantities in humans of different ages exposed via nose breathing for a 12-week period of exposure to DEPs, followed by a postexposure period under the same partile onditions and exposure pattern desribed in Figures 3 through 5. The transport rates used in the alulation for humans were those listed in Appendix C and the deposition rates rh, rr, and ra were omputed from the deposition model (Yu and Xu 1987b). Figures 9 and 1 show, respetively, the lung burden and lymph node burden results. The dashed lines are the orresponding alulated burdens in rats that we plotted in Figures 3 and 5. The lung burdens and lymph node burdens of both diesel soot and the assoiated organis were muh larger in humans than in rats. This is attributed to the large minute volume of DEPs inhaled by humans. Figures 9 and 1 also show that human adults have the highest burdens, and that a derease in the various burdens ours with dereasing age beause of the derease in minute ventilation. PREDICTED BURDENS IN HUMANS Extensive alulations were performed to predit the Equation 58 was obtained from Mauderly's data (1986) for rats ranging in age from three months to two years, equation 6 was obtained from the data of Strom and olleagues (1988), and equation 59 was determined from the best fit of the experimental deposition data. Figures 3, 4, and 5 show the alulated lung burden of diesel soot (m~l + m~l), the normalized lung burden, and the lymph node burden, respetively, for the experiment by Strom and assoiates (1988) using animals exposed to DEPs at 6 mg/m 3 for 1, 3, 6, and 12 weeks; exposure in all ases was 7 days/week and 2 hours/day. The solid lines represent the alulated aumulation of partiles during the ontinuous exposure phase and the dashed lines indiate alulated postexposure retention. The agreement between the alulated and the experimental data for both lung and lymph node burdens during and after the exposure periods was very good. Comparison of the model alulation and the retention data of partile-assoiated BaP in rats obtained by Sun and olleagues (1984) is shown in Figure 8. The alulated retention is shown by the solid line. The experiment of Sun and assoiates onsisted of a 3-minute exposure to diesel partiles oated with 3 H-benzo[a]pyrene ( 3 H-BaP) at a onentration of 4 to 6 J-Lglm 3 of air, followed by a postexposure period of more than 25 days. The fast and slow phases of 3 H-BaP learane half-times were found to be.3 day and 18 days, respetively. These orrespond to A.t;( =.385 day - 1 and A, ~) = 23.1 day - 1 in our model. Figure 8 shows that the alulated retention is in exellent agreement with the experimental data obtained by Sun and oworkers (1984). The good agreement between the alulated results and the experimental data for both diesel soot and partileassoiated BaP is not surprising beause these data were C. P. Yu, K. J. Yoon

16 Retention Modeling of Diesel Exhaust Partiles in Rats and Humans 5 4 ;;.s <:: ~ 3 :; OJ Q) " z 2... E,., 25 -' 1 8 Soot Soot /a.,~ ;; a;.s E.8 ::::.6 ~ : ro (f) (J) Time (week) Figure 1. Calulated lymph node burdens in humans of different ages exposed via nose breathing to DEPs at 6 mg/m 3 for 12 weeks at 2 hours/day and 7 days/week. Dashed lines are the orresponding burdens in rats Time (year) Figure 12. Calulated traheobronhial burdens in human adults exposed to DEPs at.1 mg/m 3 for up to 1 years at 24 hours/day and 7 days/week. lung burdens of diesel soot and the assoiated organis under different exposure onditions. The alulations were based on nose breathing for humans of different ages at the normal tidal volume and breathing rate desribed previously (Yu and Xu 1987b). The partile onditions used in the alulation were, again,.2 ~m MMAD with ag = 2.3, and the mass frations of the strongly and weakly bound organis were eah 1 perent (f 2 = f3 =.1). 5 4 _3 (J).s (f) 2 Soot Organis.8.6 a;.s : "' " ro.4 f? Burdens in Various Anatomial Compartments 1.2 To gain an overall understanding of the interompartmental transport of diesel soot and the assoiated organis, we alulated the burdens of these materials in eah anatomial ompartment for human adults, as shown in Figures 11 through 14. The exposure onditions for the al- OL L ~------~------~------~ Time (year) Figure 13. Calulated alveolar burdens in human adults exposed to DEPs at.1 mg/m 3 for up to 1 years at 24 hours/day and 7 days/week Soot Organis.8.6 a; ;;.s E ::::.3 ~ : ro (f).4 f? ;; a;.s E ::::4.15 ~ Soot : ro (f) f? Organis Time (year) Figure 11. Calulated head burdens in human adults exposed to DEPs at.1 mg/m 3 for up to 1 years at 24 hours/day and 7 days/week Time (year) Figure 14. Calulated lymph node burdens in human adults exposed to DEPs at.1 mg/m 3 for up to 1 years at 24 hours/day and 7 days/week. 12

17 13 was not notieable for exposure pattern C. The dependene of lung burden on the soot onentration is aused by the redution of the alveolar learane rate at high lung burdens disussed above. Effet of Age on Lung Burden To illustrate the effet of age on lung burden, we alulated the lung burden of diesel soot and the assoiated organis per unit onentration per unit lung weight. The lung weight data at different ages were those reported by Snyder and olleagues (1975). The exposure pattern used in the alulation was 24 hours/day and 7 days/week for a period of one year at the two soot onentrations,.1 mg/ m 3 and 1 mg/m 3. Figures 17 and 18 depit the results. On a unit lung weight basis, the lung burdens of both soot and the organis were funtions of age, and the maximum lung '[ 6 Ol E ' ; E soo " 4 Q) 5 3 (.) ~ "E 2 Q) ro ~ Ol 1 -' 1 mg/m' Time (year) Figure 15. Calulated lung burdens of diesel soot per unit exposure onentration in human adults exposed ontinuously to DEPs at two different onentrations,.1 mg/m 3 and 1. mg/m 3. Exposure patterns are (a) 24 hours/day and 7 days/week, (b) 12 hours/day and 7 days/week, and () 8 hours/day and 5 days/week. (A) (B) (C) The results of the lung burden per unit onentration are shown in Figures 15 and 16. The exposure patterns used in the alulation were (A) 24 hours/day and 7 days/week, (B) 12 hours/day and 7 days/week, and (C) 8 hours/day and 5 days/week, simulating environmental and oupational exposure onditions. The results show that the lung burdens of both diesel soot and the assoiated organis reahed a steady-state value during exposure. Due to differenes in the amount of partile intake, the steady-state lung burdens per unit onentration were the highest for exposure pattern A and the lowest for exposure pattern C. Also, inreasing soot onentration from.1 to 1 mg/m 3 inreased the lung burden per unit onentration. However, the inrease 'E Cis E ' ; E ~4 ~ ~ 3 (.) ()?2 Q) "E ro ~ g'1 ~ -' 1 mg/m'.1 (A).1 (B).1 (C) ol l l ~ ~ Time (year) Figure 16. Calulated lung burdens of tbe partile-assoiated organis per unit exposure onentration in human adults exposed ontinuously to DEPs at two different onentrations,.1 mg/m 3 and 1. mg/m 3. Exposure patterns are (a) 24 hours/day and 7 days/week, (b) 12 hours/day and 7 days/week, and () 8 hours/day and 5 days/week. ulations were 24 hours/day and 7 days/week at a soot onentration of.1 mg/m 3. The organi burdens in the head, traheobronhial, alveolar, and lymph node ompartments all reahed their respetive steady-state values during exposure. The time required to reah the steady state varied from a few days for the head and traheobronhial ompartments, to a few months for the alveolar and about one year for the lymph node ompartment. The mass aumulation of organis in the head and traheobronhial ompartments at the steady state is less than one perent of that in the alveolar ompartment. Clearly, the alveolar ompartment is most important for magnitude and for long-term learane. For diesel soot, the aumulations in the head, traheobronhial, and alveolar ompartments also showed saturation with time, but the time to reah saturation in the traheobronhial and alveolar ompartments was muh longer beause of the slower partile learane in the alveolar ompartment. Again, the soot burdens in the head and traheobronhial ompartments at the steady state were smaller than one perent of the alveolar burden. In the lymph node ompartment, the soot burden inreased with time during the first 1-year period of exposure, as shown in Figure 14, beause of the extremely small learane rate from this ompartment. Eventually, however, this burden would approah a steady state after 5 to 6 years of ontinuous exposure. Effets of Exposure Pattern on Lung Burden Lung burdens of diesel soot and the assoiated organis were alulated for different exposure patterns at two soot onentrations,.1 mg/m 3 and 1 mg/m 3, for human adults. 7 C. P. Yu, K. J. Yoon

18 Retention Modeling of Diesel Exhaust Partiles in Rats and Humans burdens ourred at approximately five years of age. Again, for any given age, the lung burden per unit onentration was slightly higher at 1 mg/m 3 than at.1 mg/m 3. Effet of Soot Conentration on Lung Clearane The inrease in lung burden per unit onentration due to the redution in the lung learane rate at high soot onentrations in humans (shown in Figures 15 through 18) is further illustrated in Figures 19 and 2, where the normalized alveolar learane rate, A.~l/A,~l, is plotted versus soot onentration. Figure 19 depits this result for human adults at the end of a ontinuous exposure of 1, 5, and 1 years for two exposure patterns: (1) 24 hours/day and 7 days/week and (2) 8 hours/day and 5 days/week. The derease of A,~l/A,~lo with onentration varied with exposure pattern and total time of exposure. Beause of a higher partile intake into the lung, a ontinuous exposure pattern of 24 hours/day and 7 days/week resulted in a faster deline of the learane rate with inreasing onentration than the oupational exposure pattern of 8 hours/day and 5 days/week. A longer exposure period also led to a greater redution in the learane rate. Figure 19 also shows that when the soot onentration was below.5 mg/m 3, the learane inhuman adults was not affeted regardless of the length of exposure. Figure 2 shows the dependene of A,~l/A,~lo on onentration for humans of different ages at ontinuous exposure for one year. The redution in learane was more pronouned at young ages. However, below a onentration of.5 mg/m 3, redution did not our at any ages..1.q ".8 1l 8.::::: E" 1.6 ClC) "iii E :S:""= g>~ => E.4 _,~_~ " ::; "' []).2 C) :J _J ol ~------~ l L Age (year) Figure 17. Calulated lung burdens of diesel soot per gram of lung per unit exposure onentration in humans of different ages exposed ontinuously for one year to DEPs of two different onentrations,.1 mg/m 3 and 1. mg/m 3, for 24 hours/day and 7 days/week. s~,..(.9.8 2<(,..( Soot Conentration (mg/m') Figure 19. Normalized learane rate of diesel soot, A.~J~~S~. versus soot onentration in human adults at the end of a ontinuous exposure of 1, 5, and 1 years. Solid lines represent an exposure pattern of 24 hours/day and 7 days/week and dashed lines represent 8 hours/day and 5 days/week. ~.4 ~.3 ~1 C) C) "iii E :s: 'a.2 ~ _,~_~ " "' :5.1 []) C) :J _J.9 2~.8,..( 2<(,..( Age (year) Figure 18. Calulated lung burdens of the partile-assoiated organis per gram of lung per unit exposure onentration in humans of different ages exposed ontinuously for one year to DEPs of two different onentrations,.1 mg/m 3 and 1. mg/m 3, for 24 hours/day and 7 days/week Soot Conentration (mg/m') Figure 2. Normalized learane rate of diesel soot, A.~J /A.~ 1 ~, versus soot onentration in humans of different ages at the end of a one-year ontinuous exposure of 24 hours/day and 7 days/week. 14

19 15 MMAD (,Urn) Figure 23. Calulated lung burdens in human adults versus MMAD for exposure to DEPs at.1 mg/m 3 for 1 years at 24 hours/day and 7 days/week. Parameters used in the alulation are ag ~ 2.3, fz ~.1, j 3 volume ~ 5 m 3, respiratory frequeny ~ 14 min-"; and Weibel's lung model, and lung volume ~ 3,2 m 3. ~.1; tidal Age (year) Figure 22. Calulated lung burdens of the partile-assoiated organis per unit exposure onentration in humans at two onentration levels,.1 mg/m 3 and 1. mg/m 3, for exposure to DEPs over a life span, from birth to adulthood, at 12 hours/day and 7 days/week. ol l ~ ~ jo Cl E.8 il.6 g.4.2 ; "' Organis J 2 ~ ~ 4 () Q) "E :J 2 m Ol :J J 1.2 Soot PARAMETRIC STUDY OF RETENTION MODEL The retention model of DEPs in humans, presented above, onsists of a large number of parameters that harater~ze the size and omposition of diesel partiles, the struture and dimension of the respiratory trat, the ventilation onditions of the subjet, and the learane half-times of diesel soot and the partile-assoiated organis. Any single or ombined hanges of these parameters from their normal values in the model would result in a hange in the predited lung burden. We onduted a parametri study of the retention model to investigate the effets of eah individual parameter on alulated lung burden in human adults. The exposure pattern hosen for this study was 24 hours/day and 7 days/week for a period of 1 years at a onstant soot onentration of.1 mg/m 3. The effets of eah individual parameter on the aumulation of diesel soot and the assoiated organis in the lung at the end of exposure are summarized below. Effet of Partile Charateristis The parameters that govern the size distribution of DEPs are MMAD and Og Figures 23 and 24 show, respetively, the effets of these parameters on lung burden for varying MMAD (.1 to.3 J.lm) and og (1.2 to 4.6). Inreasing MMAD led to a redution in the lung burden of both diesel soot and the assoiated organis, while inreasing Og produed the opposite effet. The hanges in lung burden in both ases were aused by hanges in the partile intake rate into the lung during exposure. Other partile parameters that affet lung burden are the mass fration of the partile-assoiated organis, f 2 + f 3, Lung Burdens for Exposure over a Life Span The lung burdens of diesel soot and the assoiated organis were also alulated for exposure over a life span, from birth to adulthood, for a ontinuous exposure pattern at onentration levels of.1 mg/m 3 and 1 mg/m3. Lung burdens per unit onentration are depited in Figures 21 and 22. The aumulation of both diesel soot and the assoiated organis in the lung inreased with age from birth to about 23 years of age, and dereased slightly to a steadystate value beyond this age. At a soot onentration of 1 mg/m 3, the maximum lung burden of diesel soot was found to be about 8 mg and the assoiated organis approximately 6 mg. ~ Ol 1 ~ 8 Ol E 6 ~ ~ 8 4 Q) " :; m 2 Ol :J J 1 1 mg/m' 15 Age (year) Figure 21. Calulated lung burdens of diesel soot per unit exposure onentration in humans at two different onentrations,.1 mg/m 3 and 1. mg/m 3, for exposure to DEPs over a life span, from birth to adulthood, at 12 hours/day and 7 days/week. C. P. Yu, K. J. Yoon

20 16 Figure 26. Calulated lung burdens in human adults versus f 3/(f 2 + f 3) for exposure to DEPs at.1 mg/m 3 for 1 years at 24 hours/day and 7 days/week. Parameters used in the alulation are MMAD ~.2 ~m, "g ~ 2.3, f 2 + f 3 ~. 2; tidal volume ~ 5 m 3, respiratory frequeny ~ 14 min- 1 ; and Weibel's lung model, and lung volume ~ 3,2 m 3. Effet of Lung Volume and Lung Struture For a given lung struture, the lung burden of DEPs was affeted by lung volume in two speial ways: (1) the deposi- tion effiieny or partile intake into the lung dereased with lung volume, and (2) the overloading effet of learane was redued as the lung volume inreased. Although the seond effet was not appreiable at a soot onentration of.1 mg/m 3, both effets led to a smaller lung burden at large lung volume, as shown in Figure 29. At a given lung volume, the use of different lung models also led to different lung burden preditions. The lung burden results of diesel soot and the assoiated organis as a funtion of the exposure time are shown, respetively, in o; E 6 5 F ~ 4 :;5 3 (/) 2 1 fs /{ h+fs ) and the mass ratio of fast-leared organis to the total organis, f 3/(f 2 + f trated the effets of these parameters are shown, respetively, in Figures 25 and 26. Figure 25 shows that when the organis mass fration, j 2 + j 3 3). The lung burden results that illus, inreased, the lung burden of diesel soot dereased and that of the organis inreased. Figure 26 shows that for a onstant organi mass fration j 2 + j 3 =.2 with variable / 2 and j3, the lung burden of diesel soot remained onstant and the organis burden dereased as the ratio of j 3/(f2 + j 3 ) inreased. Effet of Ventilation Conditions The hanges in lung burden due to variations in tidal volume and respiratory frequeny are depited in Figures 27 and 28. Inreasing any one of these ventilation parameters inreased the lung burden, but the inrease was muh smaller with respet to respiratory frequeny than to tidal volume. This small inrease in lung burden was a result of the derease in deposition effiieny as respiratory frequeny inreased, despite a higher total amount of DEPs inhaled. The mode of breathing has only a minor effet on lung burden; swithing from nose breathing to mouth breathing does not produe any appreiable hange in the amount of partile intake into the lung (Yu and Xu 1987b). All lung burden results presented in this study are for nose breathing f2 + fs Figure 25. Calulated lung burdens in human adults versusf 2 + f 3 for exposure to DEPs at.1 mg/m 3 for 1 years at 24 hours/day and 7 days/week. Parameters used in the alulation are MMAD ~. 2 ~m, rg ~ 2.3, f 2 ~ f 3 ; tidal volume ~ 5 m 3, respiratory frequeny ~ 14 min_,; and Weibel's lung model, and lung volume ~ 3,2 m a, Figure 24. Calulated lung burdens in human adults versus "g for exposure to DEPs at.1 mg/m 3 for 1 years at 24 hours/day and 7 days/week. Parameters used inthealulationaremmad ~.2 ~m,f 2 ~.1,f3 ~.1; tidal volume ~ 5 m 3, respiratory frequeny ~ 14 min_,; and Weibel's lung model, and lung volume ~ 3,2 m Organis o; E :::-4 Soot o;.s en 1.5 (.) (/) l"' : o o;.s.8 en : (.) o.6 l"' o; E ~4 (/) Retention Modeling of Diesel Exhaust Partiles in Rats and Humans

21 Time (year) Figure 3. Calulated lung burdens of diesel soot in human adults for four different lung models for exposure to DEPs at.1 mg/m 3 for up to 1 years at 24 hours/day and 7 days/week. Parameters used in the alulation are MMAD ~.2 ~m. o ~ 2.3, 8 f 2 ~.1, f 3 ~.1; tidal volume ~ 5 m 3, respiratory frequeny ~ 14 min_,; and lung volume ~ 3,2 m Respiratory Frequeny (min ') Figure 28. Calulated lung burdens in human adults versus respiratory frequeny in breaths min- 1 for exposure to DEPs at.1 mg/m 3 for 1 years at 24 hours/day and 7 days/week. Parameters used in the alulation are MMAD ~.2 ~m. o ~ 2.3, 8 f 2 ~.1, f 3 ~.1; tidal volume ~ 5 m 3 ; and Weibel's lung model, and lung volume ~ 3,2 m Weibel 1963.s Olson et al. 197 C>4 " "' "E 3 : td Ol " ;.s B : "'.6 2'.4 Organis ; E 4 ::-3 (/) 2 5 Yeh and Shum Hansen and Ampaya models, the burdens from the model of Olson and assoiates were the lowest, and the results from the other two models were intermediate. These differenes an be attributed to the different deposition effiienies ofdeps in these lung models. Effet of Transport Rates Transport rates have an obvious effet on the retention of DEPs in the lung after deposition. Beause we were mainly onerned with the long-term learane of diesel soot and the assoiated organis, only the effets of two transport Figures 3 and 31, for four different lung models of human adults developed by Weibel (1963), Olson and olleagues (197), Hansen and Ampaya (1975), and Yeh and Shum (198). The lung model of Weibel is equivalent to the agedependent lung model evaluated at 25 years of age used earlier in this study. Major differenes between the other three lung models and Weibel's model were disussed previously (Yu and Xu 1987b). Figures 3 and 31 show that the lung burdens (of soot and the assoiated organis) alulated from the lung model of Hansen and Ampaya were the highest among all lung Figure 29. Calulated lung burdens in human adults versus lung volume in liters for exposure to DEPs at.1 mg/m 3 for 1 years at 24 hours/day and 7 days/week. Parameters used in the alulation are MMAD ~.19 ~m. o 8 ~ 2.3, f Z, ~.1, f 3 ~.1; tidal volume ~ 5 m3, respiratory frequeny ~ 14 min- ; and Weibel's lung model. Figure 27. Calulated lung burdens in human adults versus tidal volume in liters for exposure to DEPs at.1 mg/m 3 for 1 years at 24 hours/day and 7 days/week. Parameters used in the alulation are MMAD ~. 2 ~m. o ~ 8 2.3, f 2 ~.1, f 3 lung model, and lung volume ~ 3,2 m 3. ~.1; respiratory frequeny ~ 14 min-'; and Weibel's L -L ~ L ~ Lung Volume (liter) L ~ ~ L ~ Tidal Volume (liter) ;.s B : "'.6 2' Organis 4 2 ; Soot.s B : "' 2' ; 6.s (/) C. P. Yu, K. J. Yoon

22 18 peted, inreasing the multiple of A. ~lo redued the lung burden of diesel soot with pratially no hange in the organis burden (Figure 32), whereas just the opposite ourred when the multiple of A.~l was inreased (Figure 33). DISCUSSION AND CONCLUSIONS The retention model of DEPs presented above offers an indepth piture of material transport between various anatomial ompartments and the removal of diesel soot and assoiated organis from the lung. The most diffiult and ruial task in developing suh a retention model was the determination of the interompartmental transport rates for eah material omponent. Usually, it is the knowledge of the transport rates that ditates the struture and sophistiation of a model. The multiompartmental retention model that we propose in this study an be utilized for both fast and slow learane of diesel soot and assoiated organis from the lung. The value of this model lies in its ompleteness and versatility. It helps eluidate the system dynamis and provides a useful tool for simulation and predition. However, beause our major onern with DEPs was with their longterm aumulation in the lung, most quantitative results presented in this study entered on this speial ase. In the development of the retention model, we made a major effort to derive the transport rates assoiated with the alveolar ompartment using the limited experimental data available for rats. In our model, these alveolar transport rates, as well as the transport rates assoiated with the other ompartments, an be modified easily should additional animal or rates, A.~land A.~l. were studied. The variation in the value of A.~) from different studies was shown earlier, in Figure 7. This differene is equivalent to about a fator of 2 at low lung burden but an be as high as 3 to 4 at high burden. Beause the lung burden for human exposure is normally low, we used a multiple of.5 to 2 for the unertainty in A.~l to examine how suh variation affeted the lung burden. We also used the same multiple for the variation of A.~). Figures 32 and 33 show, respetively, the lung burden results for diesel soot and the assoiated organis versus the multiples of A.~) and A.~) used in the alulation. As ex s 6 C) (/) 4 2 L JQ.6 Soot Organis QL.8----~-----'1.L2----1~ ~ L8--~2Q Multiple of A~ en 1.5.~ Figure 32. Calulated lung burdens in human adults versus a multiple of A~l for exposure to DEPs at.1 mg/m 3 for 1 years at 24 hours/day and 7 days/week. Parameters used in the alulation are MMAD ~. 2 ~m. ag ~ 2.3, fz ~.1, f 3 ~.1; tidal volume ~ 5 m 3, respiratory frequeny ~ 14 min-'; and Weibel's lung model, and lung volume ~ 3,2 m 3..5 "' C) Multiple of A~ 1 Figure 33. Calulated lung burdens in human adults versus a multiple of A~l for exposure to DEPs at.1 mg/m 3 for 1 years at 24 hours/day and 7 days/week. Parameters used in the alulation are MMAD ~.2 ~m. ag ~ 2.3, f 2 ~.1, f 3 ~.1; tidal volume ~ 5 m 3, respiratory frequeny ~ 14 min- 1 ; and Weibel's lung model, and lung volume ~ 3,2 m 3. ~ ' ' ~------_J Time (year) Figure 31. Calulated lung burdens of partile-assoiated organis in human adults for four different lung models for exposure to DEPs at.1 mg/m 3 for 1 years at 24 hours/day and 7 days/week. Parameters used in the alulation are MMAD ~.2 ~m. ag ~ 2.3, f 2 ~.1, f 3 ~.1; tidal volume ~ 5 m 3, respiratory frequeny ~ 14 min_,; and lung volume ~ 3,2 m r ~ Organis.8 a;.s ~ 'C' "'.6 ~ Yeh and Shum 198 Weibel 1963 Olson et al. 197 a;.s "'.6 "E.4 :::l o:j C) :::l --' 1.4 Soot Hansen and Ampaya Retention Modeling of Diesel Exhaust Partiles in Rats and Humans

23 19 and the assoiated organis in this region of the lung onstitutes less than one perent of the total lung burden after a long exposure. 6. Exposure to diesel exhaust has no effet on muoiliary learane; however, the alveolar learane rate of diesel soot is redued as a result of high partile burden in the lung. For rats, the alveolar learane rate of diesel soot is redued by about 1 perent at a lung burden of 1 mg/g of wet lung and by approximately 95 perent when the lung burden exeeds 8 mg/g. At this point, marophagemediated learane is pratially nonexistent. 7. The learane of diesel soot from the alveolar region of the lung is due to its transport to the traheobronhial, lymph node, and blood ompartments. At low lung burden, transport to the traheobronhial region is dominant, whereas at high burden, transport is prinipally through the lymphati system. Several new onlusions may also be drawn from this model study: 1. When humans and rats are exposed to DEPs for the same period of time, the lung burdens of diesel soot and the assoiated organis are muh larger in humans than in rats. This is due to the higher partile intake and slower learane rate in humans. 2. During a ontinuous exposure, the lung burdens of diesel soot and the assoiated organis will eventually reah a steady state, even at high onentrations. The steady-state burden per unit onentration generally inreases with the onentration, but the inrease for diesel soot is muh larger than for the organis. At low onentrations, these inreases are negligible. 3. When hildren and adult humans are exposed to equal onentrations of DEPs, the redution in alveolar learane of diesel soot due to high lung burden is greater in hildren than in adults. For a ontinuous exposure of up to 1 years, the alveolar learane rate in adults is not affeted if the exposure soot onentration remains below.5 mg/m 3. This threshold onentration dereases slightly with age. 4. The aumulation of both diesel soot and the assoiated organis in the human lung varies with age due to the differenes in partile intake and learane rates. Per unit lung weight, the aumulations for a one-year ontinuous exposure reah a maximum at about five years of age. Finally, it must be realized that the lung burden of DEPs alulated from the retention model depends on a large number of parameters that inlude partile size and omposition, individual lung struture and breathing ondition, and exposure pattern and onentration. It is important that human data beome available. Furthermore, the partileassoiated organis defined in this study represent a general term that onsists of a mixture of organi ompounds with unertain proportions, inluding BaP and NP. It is oneivable that different organi ompounds have different transport rates. In our model, however, we used the mean of the transport rates of partile-assoiated BaP and NP as representative transport rates beause only these rates are known at this time. Our retention model an readily be extended to onsider the lung learane of eah speifi organi omponent if its transport rate is known. The redution of diesel soot at high lung burdens by mehanial learane was extrapolated from rats to humans in the model by assuming that the magnitude of response to partile loading is the same for any speies at a given speifi dose. This assumption was neessary beause no data presently exist on the relationship between mehanial transport rates and lung burden in humans. We also assumed that there were no speies differenes in the transport rates of the partile-assoiated organis, again beause of the lak of human data. Future measurements are alled for to larify all these points. The alulated lung burdens are onsistent with previous experimental observations and demonstrate the following: 1. When diesel partiles are deposited in the lung, the arbonaeous soot and the partile-assoiated organis are leared from the lung in different proportions and to different degrees. 2. The partile-assoiated organis of DEPs an be divided into two omponents aording to their learane halftime. The omponent with a short learane half-time of a few hours orresponds to the organis leahed out primarily by diffusion-driven mehanisms, whereas the other omponent has a learane half-time of a few hundred hours and inludes all those organis that are haraterized by more omplex interations with other omponents of the DEPs, the learane system, or the deposition surfae itself. 3. Diesel partiles deposited in the head and traheobronhial airways are quikly removed, prinipally by the ombined mehanisms of muoiliary transport and dissolution. The fast-leared organis are leared by dissolution in a matter of hours, while slowly leared organis are leared by muoiliary transport in a matter of days. 4. In the alveolar region of the lung, the removal of the partile-assoiated organis is ontrolled by dissolution, whereas marophage phagoytosis and migration are responsible for the removal of diesel soot. 5. Beause of their fast learane rates in the traheobronhial ompartment, the aumulation of both diesel soot C. P. Yu, K. J. Yoon

24 Retention Modeling of Diesel Exhaust Partiles in Rats and Humans these parameters are known aurately before reliable lung burden estimates an be made from the retention model developed in this study. ACKNOWLEDGMENTS We are grateful to Dr. P. E. Morrow of the University of Rohester for many disussions about the mehanisms of lung learane, and toy. K. Chen for arrying out some of the numerial work. REFERENCES Amann CA, Siegla DC Diesel partiles: What are they and why. Aerosol Si Tehnol 1: Bailey MR, Fry FA, James AC The long-term learane kinetis of insoluble partiles from the human lung. Ann Oup Hyg 26: Bailey MR, Fry FA, James AC. 1985a. Long-term retention of partiles in the human respiratory trat. J Aerosol Si 16(4): Bailey MR, Hodgson A, Smith H. 1985b. Respiratory trat retention of relatively insoluble partiles in rodents. J Aerosol Si 16(4): Bond JA, Sun JD, Medinsky MA, Jones RK, Yeh HC Deposition, metabolism and exretion of 1-[1 4 C]nitropyrene and 1-[ 14 C]nitropyrene oated on diesel exhaust partiles as influened by exposure onentration. Toxiol Appl Pharmaal 85: Brightwell J, Fouillet X, Cassano-Zoppi AL, Gatz R, Duhosal F Neoplasti and funtional hanges in rodents after hroni inhalation of engine exhaust emissions. In: Carinogeni and Mutageni Effets of Diesel Engine Exhaust (Ishinishi N, Koizumi A, MClellan RO, StOber W, eds.) pp Elsevier Siene Publishing Co., New York, NY. Brooks AL, Li AP, Duther JS, Clark CR, Rothenberg SJ, Kiyoura R, Behtold WE, MClellan RO A omparison of genotoxiity of automotive exhaust partiles from laboratory and environmental soures. Environ Mutagen 6: Chan TL, Lee PS, Hering WE Deposition and learane of inhaled diesel exhaust partiles in the respiratory trat of Fisher rats. J Appl Toxiol1(2): Chan TL, Lee PS, Hering WE Pulmonary retention of inhaled diesel partiles after prolonged exposures to diesel exhaust. Fundam Appl Toxiol 4: Dixon WJ, ed BMDP Statistial Software. University of California Press, Berkeley, CA. El-Bayoumy K, Heht SS, Sakl T, Stoner GD Tumorigeniity and metabolism of 1-nitropyrene in A/J mie. Carinogenesis 5: Ferin J, Feldstein M Pulmonary learane and hilar lymph node ontent in rats after partile exposure. Environ Res 16: Garshik E, Shenker MB, Munoz A, Segal M, Smith TJ, Woskie SR, Hammond SK, Speizer FE A retrospetive ohort study of lung aner and diesel exhaust exposure in railroad workers. Am Rev Respir Dis 137: Griffis LC, Wolff RK, Henderson RF, Griffith WC, Mokler BV, MCellan RO Clearane of diesel soot partiles from rat lung after a subhroni diesel exhaust exposure. Fundam Appl Toxiol 3: Hansen JE, Ampaya EP Human air spae, shapes, sizes, areas and volumes. J Appl Physiol 38: Harris JE Diesel emissions and lung aner. Risk Anal 3(2):83-1. Heinrih U, Muhle H, Takenaka S, Ernst H, Fuhst R, Mohr U, Pott F, StOber W Chroni effets on the respiratory trat of hamsters, mie, and rats after long-term inhalation of high onentrations of filtered and unfiltered diesel engine emissions. J Appl Toxiol 6: Heinrih U, Peters L, Funke W, Pott F, Mohr U, StOber W Investigation of toxi and arinogeni effets of diesel exhaust in long-term inhalation exposure of rodents. In: Toxiologial Effets of Emissions from Diesel Engines (Lewtas J, ed.) pp Elsevier Siene Publishing Co., New York, NY. International Commission on Radiologial Protetion Publiation 3, part 1. Limits for intakes of radionulides by workers. Ann ICRP 2. Ishinishi N, Kuwahara N, Nagase S, Suzuki T, Ishiwata S, Kohno T Long-term inhalation studies on effets of exhaust from heavy and light duty diesel engines on F344 rats. In: Carinogeni and Mutageni Effets of Diesel Engine Exhaust (Ishinishi N, Koizumi A, MClellan RO, StOber W, eds.) pp Elsevier Siene Publishing Co., New York, NY. 2

25 C. P. Yu, K. J. Yoon Iwai K, Udagawa T, Yamagishi M, Yamada H Longterm inhalation studies of diesel exhaust on F344 SPF rats: Inidene of lung aner and lymphoma. In: Carinogeni and Mutageni Effets of Diesel Engine Exhaust (Ishinishi N, Koizumi A, MClellan RO, Stober W, eds.) pp Elsevier Siene Publishing Co., New York, NY. Lewtas J Evaluation of the mutageniity and arinogeniity of motor vehile emissions in short-term bioassays. Environ Health Perspet 47: Mauderly JL Respiration of F344 rats in nose-only inhalation exposure tubes. J Appl Toxiol 6:25-3. Mauderly JL, Jones RK, Griffith WC, Henderson RF, MClellan RO Diesel exhaust is a pulmonary arinogen in rats exposed hronially by inhalation. Fundam Appl Toxiol 9: Morrow PE Possible mehanisms to explain dust overloading of the lungs. Fundam Appl Toxiol1: Morrow PE, Yu CP Models of aerosol behaviour in airways. In: Aerosols in Mediine: Priniples, Diagnosis, and Therapy (Moren F, Newhouse MT, Dolovih MB, eds.) pp Elsevier Siene Publishing Co., New York, NY. Muhle H, Bellmann B, Crentzenberg, Stober W, Kilpper R, Makenzie J, Morrow P, Mermelstein R Pulmonary deposition, learane and retention of test toner, Ti 2 and quartz during a long-term inhalation study in rats. Toxiologist 8:272. Oberdorster G Lung learane of inhaled insoluble and soluble partiles. J Aerosol Med 1(4): Oberdorster G, Green FHY, Freedman AP Clearane of 59 Fe 3 4 partiles from the lungs of rats during exposure to oal mine dust and diesel exhaust. J Aerosol Si 15: Olson DE, Dart GA, Filley GF Pressure drop and fluid flow regime of air inspired into the human lung. J Appl Physiol 28: Orthoefer JG, Moore W, Kraemer D, Truman F, Croker W, Yang YY Carinogeniity of diesel exhaust as tested in strain A mie. Environ Int 5: Pepelko WE EPA studies on the toxiologial effets of inhaled diesel engine emissions. In: Toxiologial Effets of Emissions from Diesel Engines (Lewtas J, ed.) pp Elsevier Siene Publishing Co., New York, NY. Press WH, Flannery BP, Teukolsky SA, Vettering WT Numerial Reipes (Fortran). Cambridge University Press, Cambridge, England. Shanker LS, Mithell EW, Brown RA Speies omparison of drug absorption from the lung after aerosol inhalation or intratraheal injetion. Drug Metab Dispos 14(1): Shuetzle D Sampling of vehile emissions for hemial analysis and biologial testing. Environ Health Perspet 47:65-8. Snipes MB, Boeker BB, MClellan RO Retention of monodisperse or polydisperse alluminosiliate partiles inhaled by dogs, rats, and mie. Toxiol Appl Pharmaal 69: Snipes MB, MClellan RO Retention of 134 C-labeled alluminosiliate partiles inhaled by dogs and guinea pigs: Simulation model projetions for humans. In: Inhalation Toxiology Researh Institute Annual Report , pp Inhalation Toxiology Researh Institute, Lovelae Biomedial and Environmental Researh Institute, Albuquerque, NM. Snyder WS, Nasset ES, Karhausen LR, Howells GP, Tipton IH Report of Task Group on Referene Man, pp Pergamon Press, Oxford, England. Soderholm SC Compartmental analysis of diesel partile kinetis in the respiratory system of exposed animals. In: Diesel Emission Symposium Proeedings. Doument EPA U.S. Environmental Protetion Ageny, Researh Triangle Park, NC. National Tehnial Information Servie, Springfield, VA. Stober W Experimental indution of tumors in hamsters, mie, and rats after long-term inhalation of filtered and unfiltered diesel engine exhaust. In: Carinogeni and Mutageni Effets of Diesel Engine Exhaust (Ishinishi N, Koizumi A, MClellan RO, Stober W, eds.) pp Elsevier Siene Publishing Co., New York, NY. Strom KA, Chan TL, Johnson JT Pulmonary retention of inhaled submiron partiles in rats: Diesel exhaust exposures and lung retention model. In: Inhaled Partiles VI (Dodgson J, MCallum RI, Bailey MR, Fisher DR, eds.) pp Pergamon Press, Oxford, England. Strom KA, Johnson JT, Chan TL Retention and learane of inhaled submiron arbon blak partiles. J Toxiol Environ Health 26: SunJD, WolffRK, KanapillyGM, MClellanRO Lung retention and metaboli fate of inhaled benzo[a]pyrene as- 21

26 Retention Modeling of Diesel Exhaust Partiles in Rats and Humans soiated with diesel exhaust partiles. Toxiol Appl Pharmaal 73: Vinent JH, Jones AD, Johnston AM, MMillan C, Bolton RE, Cowie H. 198Z Aumulation of inhaled mineral dust in the lung and assoiated lymph nodes: Impliations for exposure and dose in oupational lung disease. Ann Oup Hyg 31: Waller RE Trends in lung aner in London in relation to exposure to diesel fumes. In: Health Effets of Diesel Engine Emissions (Pepelko WE, Danner RM, Clarke NA, eds.) pp EPA-6/ U.S. Environmental Protetion Ageny, Offie of Researh and Development, Washington, DC. Weibel ER Morphometry of the Human Lung. Springer-Verlag, New York, NY. Wolff RK, Henderson RF, Snipes MB, Griffith WC, Mauderly JL, Cuddihy RG, MClellan RO Alterations in partile aumulation and learane in lungs of rats hronially exposed to diesel exhaust. Fundam Appl Toxiol 9: Wolff RK, Henderson RF, Snipes MB, Sun JD, Bond JA, Mithell CE, Mauderly JL, MClellan RO Lung retention of diesel soot and assoiated organi ompounds. In: Carinogeni and Mutageni Effets of Diesel Engine Exhaust (Ishinishi N, Koizumi A, MClellan RO, Stober W, eds.) pp Elsevier Siene Publishing Co., New York, NY. Xu GB, Yu CP Deposition of diesel exhaust partiles in mammalian lungs. Aerosol Si Tehnol 7: Yeh HC, Shum GM Models of human lung airways and their appliation to inhaled partile deposition. Bull Math Bioi 42: Yu CP, Chen YK, Morrow PE An analysis of alveolar marophage mobility kinetis at dust overloading of the lung. Fundam Appl Toxiol 13: Yu CP, Xu GB Preditive models for deposition of diesel exhaust partiulates in human and rat lungs. Aerosol Si Tehnol 5: Yu CP, Xu GB. 1987a. Predited deposition of diesel partiles in young humans. J Aerosol Si 18: Yu CP, Xu GB. 1987b. Preditive Models for Deposition of Inhaled Diesel Exhaust Partiles in Humans and Laboratory Speies Researh Report No. 1. Health Effets Institute, Cambridge, MA. APPENDIX A. Kineti Equations for Diesel Soot and Partile-Assoiated Organis and Their Solutions The differential equations for mfl and their solutions as a funtion of exposure time, t, an be written as (fori = 1, 2, and 3): Head (H) where dmhuljdt - r(i) - ~Yl m(il - j_(il m(il - H HG H HB H = ryj - Afj)mYJ (A.1) AYJ = AYJ + AYJs (A.2) myj = rfj)!ayj + (mfj) - rfj)!afj)) exp(- Afj)t) (A.3) Traheobronhial (T) where dm~l j dt = r~l + A~~m~l - A(j}m~l - A~km~l - r(il - A(ilm(il + A(i) mul (A.4) - T T T AT A AUl AUl + AUl T - TG TB (A.5) m~l = exp(- A~lt) 1: (r~l + A~~m~l) exp(a~)t)dt + m~6 (A.6) Alveolar (A) where dmauljdt - r(i) - A(i) m(il - A{i) m(i) - A{i) m(i) - A AT A AL A AB A - r{il - "(i)m(i) - A "'A A (A.7) Lymph nodes (L) (A.8) (A.9) dmfljdt = A~im~l - Af1mfl (A.1) mul exp(- AUl t) ft A(il mul exp(a(il t)dt + mul L LB J AL A LB LO (A.1~ where m)p is the mass of omponent i in X ompartment and r)p is the mass intake rate of omponent ito X ompartment alulated from a deposition model of DEPs (Yu and Xu 1987b). The total mass of the partile-assoiated organis in ompartment X is the sum of m~l and m~l, and the total mass of DEPs in ompartment X is equal to (A.12) 22

27 23 ABOUT THE AUTHORS C. P. Yu is Professor and former Chairman of the Department of Mehanial and Aerospae Engineering at the State University of New York at Buffalo. He reeived his Ph.D. from Purdue University in In 1972, Dr. Yu spent a sabbatial leave at the University of Essex, England, and worked with Dr. C. N. Davies on aerosol deposition in human airways. Dr. Yu's primary researh interests inlude the development of theoretial desriptions of aerosol deposition and the appliation of deposition models to various anatomial situations. K. J. Yoon obtained his B.S. in mehanial engineering from National Seoul University in He was awarded a Ph.D. in 1989 in mehanial and aerospae engineering, from the State University of New York at Buffalo for his work in mathematial modeling of lung learane. His researh interests have been the deposition and learane of inhaled environmental pollutants and the risk analysis of human exposure. He is urrently with Hyundai Motor Corporation, Seoul, Korea. A.~l =.12 exp(-.11 m~ 75 ) +.68 exp(-.46 m~ 52 ) A.~l =.12 exp(-.11 m~ 75 ) +.68 exp(-.46 m~ 52 ) where ma ~ m~l is the partiulate burden (in milligrams) in the alveolar ompartment. APPENDIX C. Transport Rates of Diesel Soot and Partile-Assoiated Organis in Humans The values of A.~y (in day -l) that we adopted from the literature and used in the model alulation for humans are listed below: [i) - "-He , i = 1, 2, 3 (Chan et al. 1981; ICRP 1979) [i) - "-Te -.693, i = 1, 2, 3 (Chan et al. 1981; ICRP 1979) A.~ 1 =.169 (Bailey et al. 1982) A.~ 1 = A.~~ + "-~i + "-~k i = 1, 2, 3 A.~l =.12 exp( -.11 m~ 75 ) +.86 A.~Jo =.288 A. r;{ = 15.7 (Sun et al. 1984; Bond et al. 1986) A. (i) - 4/... [i) (Sun et al. 1984; Bond et al. 1986) AB- AL i = 2, 3 (ICRP 1979) The following values of A.~y were derived by extrapolating the results of rats: A.~~ = (1/P)[.12 exp(-.11(mais) 1 75 ) +.68 exp(-.46(m A IS) 1 52 )], i = 1, 2, 3 A.~i =.68[1 - (1/P) exp(-.46(ma/s) 1 5Z)] (1) - (1) - (1) - (1) - 18 II.HB - II.TB - II.LB - II.AB - (2) - (2) - (2) - (2) II. HB - II. TB - II. LB - II. AB - (3) - (3) - (3) - (3) II.HB-II.TB-II.LB-II.AB- A,lil _ A,lil + A,lil + f...ul A-AT AL AB A.~J = (1/P)[.12 exp( -.11(mA/S]1.7 5 )] +.86 A.~J = (1/P)[.12 exp(-.11(m A IS) ) +.68 exp(-.46(ma/s) 1 52 )] A.~l = (1/P)[.12 exp(-.11(ma/s) 1 75 ) +.68 exp( -.46(mA/S) 1 52 )] where ma ~ m~l is the partiulate burden (in milligrams) in the alveolar ompartment, P = 14.4, and Sis the pulmonary surfae area ratio given in Table 2. APPENDIX B. Transport Rates of Diesel Soot and Partile-Assoiated Organis in Rats The values of A.~y (in day - 1 ) that we adopted from the literature and used in the model alulation for rats are listed below: [i) "-He = 1.73, i = 1, 2, 3 (Chan et al. 1981; ICRP 1979) A.[/b =.693, A.~l =.129 A. ~) =.288 A.~lo = 15.7 i = 1, 2, 3 (Chan et al. 1981) (Strom et al. 1988) (Sun et al. 1984; Bond et al. 1986) (Sun et al. 1984; Bond et al. 1986) "-~k = 4 "-~i i = 2, 3 (ICRP 1979) The following values of A.~y were derived using the experimental data of lung burden and lymph node burden: A.~~=.12 exp(-.11 m~ 75 ) +.68 exp (-.46 m ~ 52 ), i = 1, 2, 3 "-~i =.68[1 - exp( -.46 m~ 52 )] A.lj)B = A.% = /...~1 = A~B =.18 A.lj)B = A(f1 = /...~1 = A.~k =.129 A.lj)B = A.(f1 = A.~1 = A.~k = C. P. Yu, K. J. Yoon

28 24 Yu CP, Yoon KJ, Chen YK Retention modeling of diesel exhaust partiles in rats and humans. J Aerosol Med (in press). ABBREVIATIONS BaP benzo[a]pyrene DEPs diesel exhaust partiles 3 H-BaP [3H]benzo[a]pyrene MMAD mass median aerodynami diameter NP nitropyrene ICRP International Commission on Radiologial Protetion ag geometri standard deviation PUBLICATIONS RESULTING FROM THIS RESEARCH Yu CP, Morrow PE A nonlinear model of aerosol retention in the lung (abstrat). Amerian Assoiation for Aerosol Researh Annual Meeting, Seattle, WA, September Yu CP, Morrow PE, Chan TL, Strom KA, Yoon KJ A nonlinear model of alveolar learane of insoluble partiles from the lung. Inhalation Toxiol Premier Issue 97-1Z Yu CP, Chen YK, Morrow PE An analysis of alveolar marophage mobility kinetis at dust overloading of the lungs. Fundam Appl Toxiol 13: Yu CP, Yoon KJ Predited lung burdens of diesel exhaust partiles in rats and humans (abstrat). Amerian Assoiation for Aerosol Researh Annual Meeting, Reno, NV, Otober Retention Modeling of Diesel Exhaust Partiles in Rats and Humans

29 HEALTH REVIEW COMMI'ITEE'S COMMENTARY Health Effets Institute INTRODUCTION A Request for Appliations (RFA 83-3), whih soliited proposals for "Dose of Airborne Pollutants to Target Tissues;' was issued by the Health Effets Institute (HEI) in the summer of In response to the RFA, Dr. C. P. Yu from the State University of New York at Buffalo submitted a proposal entitled "Preditive Models for Deposition oflnhaled Diesel Exhaust Partiles in Humans and Laboratory Speies:' This study was ompleted in June 1986, and was published as HEI Researh Report No. 1. In January 1986, Dr. Yu submitted to the HEI a renewal appliation entitled "Determination of Lung Dose of Diesel Exhaust Partiles." The HEI approved the two-year study, whih began in April Total expenditures for the two-year projet were $144,918. The Investigators' Report was reeived at the HEI in July 1989 and aepted by the Health Review Committee in April199. During the review of the Investigators' Report, the Review Committee and the investigators had the opportunity to exhange omments and to larify issues in the Investigators' Report and in the Review Committee's Commentary. The Health Review Committee's Commentary is intended to plae the Investigators' Report in perspetive, as an aid to the sponsors of the HEI and to the publi. REGULATORY BACKGROUND The U.S. Environmental Protetion Ageny (EPA) sets emissions standards for diesel engines and vehiles under Setion 22 of the Clean Air At, as amended in 199. Setion 22(a)(1) direts the Administrator of the EPA to "presribe (and from time to time revise)... standards appliable to the emission of any air pollutant from any lass or lasses of new motor vehiles or new motor vehile engines, whih in his judgment ause, or ontribute to, air pollution whih may reasonably be antiipated to endanger publi health or welfare:' Setion 22(a)(3)(A)(i) speifially direts the Administrator to set standards for the "emissions of arbon monoxide, hydroarbons, oxides of nitrogen and partiulate matter from lasses of heavy-duty vehiles and engines... " The EPA has taken a variety of regulatory ations with respet to diesel engines and vehiles under the authority given it by Setion 22(a)(1) and 22(a)(3)(A)(i) of the At. For instane, the EPA has set emissions standards for both heavy-duty and light-duty truks. These emissions standards initially are made appliable to all engines and vehiles produed in a given model year. Engines and vehiles of the same lass that are produed in sueeding years must also omply with these existing standards unless the EPA establishes a new set of standards. The EPA issued emissions standards for diesel-fueled heavy-duty engines and vehiles in 198 that speified limits for hydroarbons, arbon monoxide, and oxides of nitrogen appliable to heavy-duty engines and vehiles produed during the 1985 model year, and in 1985 added limits on partiulate matter emissions for the 1988 model year. The EPA also set emissions standards for the 1991 and 1994 model years. The EPA revised those standards most reently in With respet to light-duty truks, the EPA issued emissions standards for the 1985 model year in 198 that speified limits for the emission of hydroarbons, arbon monoxide, oxides of nitrogen, and partiulate matter. New standards were later promulgated for the 1987, 1988, 199, and 1991 model years. The EPA set standards appliable to the 1991 model year in 1988 and revised them most reently in The 199 Amendments to the Clean Air At inluded several provisions that deal with diesel engines and vehiles. Setion 22(a)(3)(B)(ii), as amended, requires that, beginning in 1998, all diesel-fueled heavy-duty truks not emit more than 4. grams per brake horse power-hour (g/bhphour). Setion 22(a)(3)(B)(ii) sets new emissions standards for oxides of nitrogen produed from diesel-powered heavyduty truks. Setion 292(j) authorizes the Administrator to promulgate regulations for arbon monoxide emissions from various lasses of vehiles when operated at old temperatures. Setion 219 requires the use of ertain lowpolluting fuels in urban buses in ities that have not met ertain emissions standards. Setion 231 requires the Administrator to oversee a study to determine whether or not ethanol and high erui rapeseed oil might be used as an "alternative to diesel fuel:' The development of models that estimate retention and predit lung burdens of diesel partiles an ontribute to an inreased understanding of the risks to humans from exposure to diesel engine exhaust. These models an ontribute also to informed deision-making with respet to standards under the Clean Air At. SCIENTIFIC BACKGROUND The health effets of diesel engine emissions are of on- 25

30 Health Review Committee's Commentary ~ El ern for several reasons, inluding the respirable size of the diesel exhaust partile, the genotoxiity of a number of hemials assoiated with the partile (Claxton 1983; Lewtas 1983), and the reent reports of pulmonary arinogeniity of diesel engine exhaust in rats (Brightwell et al. 1986; Heinrih et al. 1986; Ishinishi et al. 1986; Iwai et al. 1986; Mauderly et al. 1986). Also, several epidemiologial studies suggest an assoiation between hroni exposure to diesel engine exhaust and an inreased risk of lung and bladder aner in humans (Silverman et al. 1986; Steenland 1986; Garshik et al. 1987, 1988). It is important to note, however, that obtaining an aurate assessment of exposure has been a major limitation in the interpretation of epidemiologial studies. In addition to the diffiulties in estimating exposure to diesel engine exhaust, several partiulate agents, most notably igarette smoke, onfound the estimates. After reviewing the genotoxiity, arinogeniity, and epidemiologial data, the International Ageny for Researh on Caner evaluated diesel engine exhaust as "probably arinogeni to humans" (International Ageny for Researh on Caner 1989). In the United States, the National Institute for Oupational Safety and Health (1988) has reommended that whole diesel exhaust be regarded as a "potential oupational arinogen:' The major fous of onern with exposure to diesel engine exhaust has been arinogeniity. However, nonarinogeni histologi and ytologi effets also have been noted in animal studies. After subhroni and hroni exposure to diesel emissions, an inflammatory ell response ours (Mauderly et al. 1981; White and Garg 1981; Heinrih et al. 1986; Lewis et al. 1986; MClellan et al. 1986). Partile-laden marophages aumulate in the alveoli and peribronhial regions (Wiester et al. 198; Karagianes et al. 1981; Mauderly et al. 1981; White and Garg 1981; Pepelko 1982; Plopper et al. 1983; Heinrih et al. 1986). Hyperplasia of bronhiolar and alveolar type II epithelial ells, as well as thikening of alveolar walls, have been reported (Wiester et al. 198; White and Garg 1981; Kaplan et al. 1982; Pepelko 1982; Plopper et al. 1983; Heinrih et al. 1986; Ishinishi et al. 1986; Lewis et al. 1986). Fibroti (Karagianes et al. 1981; Hyde et al. 1985; Heinrih et al. 1986; Lewis et al. 1986; MClellan et al. 1986) and emphysematous (Karagianes et al. 1981; Heinrih et al. 1986) lesions have been noted. The relationship of these observations to hroni lung disease remains to be determined. The overall assessment of the health effets of exposure to diesel engine exhaust involves the sum of numerous omplex variables, suh as the extent of exposure, target-tissue dose, toxiity of the original and metabolized agent or agents, and variations in host suseptibility. When relating ambient exposures to potential human health effets, deter- mining the dose of diesel partiulate matter to pulmonary tissues is ritial. Without measurements or aurate estimates of dose, it is diffiult to extrapolate findings from animal studies, evaluate exposure in epidemiologial studies, or assess individual variability. Airborne partiles of respirable size are inhaled and deposited onto respiratory trat surfaes (reviewed by Shlesinger 1988). Several physial, hemial, and biologial proesses at upon the deposited material to remove it from the lungs. In the airways, insoluble partiles are leared to the oropharynx primarily by the movement of surfae muus propelled by the iliated epithelial ells; this proess is termed muoiliary learane. Soluble partiles may dissolve and diffuse into the irulation. Both of these proesses are fairly rapid, and the majority of deposited material in the airways is removed within 24 to 48 hours. Partiles that reah the alveolar region of the lung are leared more slowly and by different pathways. Some partiles are taken up by alveolar marophages, whih, in turn, migrate onto airway surfaes and exit the lung by way of the muoiliary system. Alternatively, partile-laden marophages, as well as free partiles, may enter the pulmonary interstitium and make their way to the lung-assoiated lymph nodes. The learane half-life of material deposited in the alveolar region ranges from weeks to months, depending on the route of exit from the lung. If material reahes the pulmonary lymph nodes, the residene time inreases from months to years. Deposited partiles that are not leared from the lungs remain sequestered in pulmonary tissues. In healthy individuals exposed to small amounts of partiles, pulmonary learane mehanisms are usually effetive in removing the majority of the deposited material. However, if the lung is unable to lear partiles faster than the rate at whih they are deposited, the partiles aumulate; this phenomenon is termed partile overloading. The signifiane of partile overloading is that, with longer retention times, the dose of toxi ompounds to pulmonary tissues potentially inreases. In summary, deposition and learane are interrelated proesses, and if deposition exeeds learane, the balane is termed retention. These proesses influene the loal dose to target tissues, and hene, the toxi response to inhaled partiulate matter. Diesel exhaust partiles are omposed of a dense arbonaeous ore. Combustion-derived organi ompounds are adsorbed to this ore. It is assumed that the learane kinetis of diesel exhaust partiles from the respiratory trat are similar to those of insoluble partiles. After an aute exposure to diesel engine exhaust, the removal of partiles from rat pulmonary surfaes follows a biphasi pattern, indiating that muoiliary transport and marophage- 26

31 27 the uptake and metabolism of released organis by marophages and epithelial tissues have not been well haraterized. In addition to genotoxi mehanisms, nongenotoxi mehanisms may play a role in a tumorigeni response. After hroni exposure of rats to diesel engine exhaust, hyperplasia of bronhiolar and alveolar type II epithelial ells has been observed (Wiester et al. 198; White and Garg 1981; Pepelko 1982; Plopper et al. 1983; Heinrih et al. 1986; Ishinishi et al. 1986; Lewis et al. 1986). Inreased epithelial proliferation may inrease suseptibility to tumor indution, thus ontributing to the observed tumor inidene in hroni inhalation studies (reviewed by Ames and Gold 199). Therefore, the determination of tissue doses that indue ell renewal also may be relevant. Mehanisms that ause nonarinogeni effets are not well understood, but the inability of the lungs to lear deposited diesel partiles appears to play a role. The ellular hanges desribed above do not our at low partile onentrations. Suh hanges are assoiated with the redution of pulmonary learane and the aumulation of maro phages filled with ingested partiles. The relationships among partile overloading, marophage aggregation, and onditions suh as inflammation, ellular proliferation, fibrosis, or emphysema are not known. Thus, in order to assess the health risks from exposure to diesel engine exhaust, it is important to determine the retention time and sites of aumulation of the partiles and assoiated organi ompounds. Animal studies have foused primarily on the deposition, learane, and retention of diesel exhaust partiles, but have not measured diretly the dose of partiles to the ells and tissues of the respiratory trat (Chan et al. 1981; Griffis et al. 1983; Lee et al. 1983; Chan et al. 1984; Wolff et al. 1986). Data on the fate of ombustion-derived organis are not available, although studies on the disposition of organi ompounds experimentally adsorbed onto diesel exhaust partiles (Sun et al. 1984; Ball and King 1985; Bond et al. 1986) or other insoluble partiles (Wolff 1989; Wolff et al. 1989) have been onduted. No data on learane kinetis and retention, of either diesel partiles or assoiated organis, are available on humans. However, tissue doses an be estimated by onstruting dosimetry models that utilize experimental data obtained from animals. The auray of the model will depend, in part, on taking into aount the omplexity of the routes, rates, and mehanisms of learane. JUSTIFICATION FOR THE STUDY The health effets of inhaled partiles and gases from mo- mediated transport are likely to be the predominant learane mehanisms (Chan et al. 1981). Under onditions of high exposure levels, overloading has been demonstrated in rodents. A threshold onentration has not been determined, but learane from the alveolar region is impaired in studies in whih diesel exhaust partile onentrations are greater than 1 mg/m 3 (Vostal et al. 1982; Griffis et al. 1983; Chan et al. 1984; Wolff et al. 1987). Furthermore, the extent to whih alveolar learane is impaired depends on the initial partile burden: the greater the partiulate onentration, the slower the learane (Chan et al. 1984). The disposition ofthe adsorbed organis also needs to be onsidered (reviewed by Sun et al. 1988). For those adsorbed organi ompounds that remain assoiated with the arbon ore during learane, learane kinetis are the same as those of the partile. Clearane rates vary for those organi ompounds released from the partile. Dissoiated ompounds may pass into the irulation. Alternatively, they an be taken up by respiratory trat tissues or pulmonary marophages and be metabolized. Metabolism may lead either to ativation or degradation of the parent ompound. The determination of tissue doses of diesel exhaust partiles has important impliations when trying to eluidate mehanisms of toxiity. It is not known if the mehanisms operative at high exposure levels, whih are usually used in animal studies, also operate at low levels, whih may be more representative of human exposures. For example, several mehanisms have been proposed to aount for the arinogeni response observed in rodents. The deposition and learane of the adsorbed organis, some of whih are mutagens and arinogens, are affeted by their assoiation with the diesel partile (reviewed by Sun et al. 1988). The deposition sites and learane mehanisms are different, learane rates are slower, and retention times are longer for partile-assoiated organis than for aerosols of pure ompounds. Reently, onsiderable attention has been foused on the influene of partile overloading. The indution of respiratory trat arinomas in rats has ourred only with high exposure onentrations of 3.5 mg or more of diesel partiles/m 3. It has been suggested that under onditions of partile overload, the inreased retention of partileassoiated organi ompounds may inrease the bioavailability of genotoxi ompounds, thus inreasing the risk of a tumorigeni response (Heinrih et al. 1986). However, the atual bioavailability of adsorbed organi ompounds has not been determined. Their prolonged presene in the lungs may or may not be signifiant as long as they remain adherent to partile surfaes. The physial-hemial fators that govern desorption and subsequent bioavailability of genotoxi ompounds are poorly understood. In addition, 1- [ Health Effets Institute

32 m.rct Health Review Committee's Commentary bile soure emissions are of entral onern to the HEI. Considerable attention has been given to the evaluation of biologial responses aused by emission produts, but less effort has been devoted to the quantifiation of dose. However, without an aurate measurement or estimation of dose, desriptions of responses are of diminished value. In an earlier projet funded by the HEI, Yu and Xu (1987) made theoretial alulations of lung deposition to predit the amount and distribution of partiulate material deposited in the lungs during inhalation from a given airborne diesel exhaust partile onentration. However, tissue dose is determined not only by the distribution of deposition, but also by learane times and pathways of redistribution among ells and tissues. Thus, additional modeling researh was needed to desribe learane, retention, and total lung burdens. For the urrent study, Dr. Yu proposed to determine the lung burden of diesel engine exhaust partiles under various exposure onditions. Mathematial modeling first would be done in rats and then extrapolated to humans. The investigator proposed to develop a nonlinear model for learane kinetis, whih desribes learane as a funtion of lung burden over time, and then to develop a retention model for diesel exhaust partiles. Using this retention model and the deposition model developed in his previous study, Dr. Yu would then predit total lung burden of partiulate material. The HEI Researh Committee onsidered that the development of a reliable mathematial model of learane, taking into aount the nonlinear nature of learane proesses, was an important ontribution to dosimetry modeling. Estimates of diesel-partile burdens are neessary for the assessment of diesel engine exhaust exposure. Thus, the goals of Dr. Yu's proposal were onsidered worthy and, beause of his qualifiations and experiene, attainable. STUDY OBJECTIVES The objetive of this projet was to extend a previously developed model for deposition of inhaled diesel exhaust partiles (Yu and Xu 1987) to inlude learane kinetis and retention. The investigators sought to develop a mathematial retention model based on available experimental data in rats and then to extrapolate this model from rats to humans. A ritial feature of the theoretial model was the onsideration of the diesel exhaust partile as tripartite with regard to its learane harateristis. The desription of the partile inluded a arbonaeous ore (whih the investigators refer to as soot), a rapidly leared organi fration with a short retention time, and a slowly leared organi fration with a long retention time. The division of the organis-into slowly leared and rapidly leared omponents was based on the observations that the onentration of inhaled partile-assoiated organi ompounds in the lungs dereases along a onave, biphasi urve (Sun et al. 1984; Bond et al. 1986). A biphasi disappearane of the assoiated organis is onsistent with a single speies of organis passing though two ompartments on its way out of the lung. A biphasi disappearane urve is also onsistent with two distint speies learing independently at different rates. The physial transport of the partile was modeled in four anatomial ompartments: the nasopharyngeal, traheobronhial, alveolar, and lymph node regions. Equations were developed for the role of eah of these ompartments in the learane of the three omponents of the diesel exhaust partile. The model was based on experimental data olleted by Strom and oworkers (1988) in Fisher rats and was extended to humans. The retention model was based on exposures to diesel exhaust partiles at 6 mg/m 3 for 2 hours/day and 7 days/week. One developed, the retention model was ombined with the investigators' deposition model (Yu and Xu 1987), and burdens of diesel soot and assoiated organis were alulated. Exposure onditions of the model were varied, and their effets on estimated burdens were examined. The effets of diesel partile onentrations (.1 mg/m 3 and 1. mg/m 3 ), exposure pattern (ontinuous and intermittent), and age on lung burdens of diesel soot and assoiated organis were alulated. In addition, the effet of diesel partile onentration on lung learane, interompartmental transport, and the aumulation of lung burdens over time were evaluated. Finally, the investigators varied several parameters of the retention model to determine the effets on estimated lung burdens of diesel soot and assoiated organis. Partile harateristis (mass median aerodynami diameter, geometri standard deviation, mass fration of partileassoiated organis, and mass ratio of the rapidly leared organis), ventilation (tidal volume and respiratory frequeny), lung volume, lung struture (by using different lung models), and alveolar transport rate were varied, and lung burdens for adult humans were alulated. These parametri analyses used an exposure pattern of.1 mg/m 3 for 24 hours/day and 7 days/week for 1 years. TECHNICAL EVALUATION ATTAINMENT OF STUDY OBJECI'IVES The investigators developed a retention model for diesel 28

33 29 oated partiles were used for inhalation studies; therelease of the radiolabeled ompounds from the partile were assumed to predit the desorption of unlabeled ombustion-derived organi moleules. This assumption may not be valid beause adsorption energy dereases with overage. Diesel exhaust partiles ould ontain as many as eight layers of adsorbed moleules on their surfaes, ausing the adsorption energy of the ombustionderived moleules to hange dramatially depending on the number of layers. In addition, evidene suggests that diesel exhaust partiles, whih have heterogeneous surfaes, seletively retain polar moleules. These studies with radiolabeled moleules do not take into aount seletive adsorption and, hene, release. Thus, the release of ombustion-derived organis may not be analogous to the behavior of ompounds applied by passive adsorption. The investigators assumed that the amount of organis leared slowly from the lungs represents 1 perent of the total partile mass. Although the investigators suggested several mehanisms by whih organis would be removed from the lungs slowly over time, it should be pointed out that the amount of organis that are tightly bound to the arbon ore is probably only.1 to.2 perent of the total partile mass. Figures 25 and 26 of the Investigators' Report illustrate how the lung burden of soot and assoiated organis would be affeted if the total mass fration (f2 + f3 ) or the mass ratio of fastleared organis to total organis (f3/f2 + f3 ) were varied. The reader should refer to these figures in order to estimate lung burdens for different proportions of organis. 2. Exposure onditions. The retention model was based on one pattern of exposure onsisting of 6 mg of diesel soot/m 3 for 2 hours/day and 7 days/week, with the emphasis on obtaining a partile-overload ondition. The appliability of the model to short, low-level exposures or to intermittent peaks of high-level exposure, both of whih are more omparable to human exposure patterns, is unlear. Additional effort ould have been direted toward the appliation of the model to a wider range of exposure patterns. For example, Figure 15 shows that when human adults are exposed to diesel exhaust partiles at.1 or 1. mg/m 3, the aumulation of soot over time beomes disproportionately greater with the higher onentration as the intensity of exposure inreases. Also, Figure 19 shows that under intermittent exposure onditions over a 1-year period, the tolerated dose before the normalized alveolar learane rate starts to deline is five times greater than that under the ontinuous exposure onditions. These examples illustrate exhaust partiles and assoiated organis for rats and humans. This retention model was ombined with the deposition model developed earlier (Yu and Xu 1987}, and lung as well as lymph node burdens for diesel soot and assoiated organis were alulated. Thus, all the study objetives were attained. THE MODEL: RESULTS OF CALCULATIONS AND PREDICTIONS Most of the preditions reahed by the investigators are onsistent with the experimental results reported in previous studies on learane in animals. More speifially, alulations of the model showed that alveolar learane rates of diesel soot, but not organi ompounds, derease with inreasing lung burdens. At low lung burdens of diesel soot, muoiliary transport dominates alveolar learane, whereas at high lung burdens, material is leared to the lymphati system. On the basis of alulations of their model, the investigators arrived at four new preditions: 1. After exposure to diesel exhaust partiles for the same amount of time, lung burdens of diesel soot and assoiated organis are greater in humans than in rats. 2. During ontinuous exposure, the lung burdens of diesel soot and assoiated organis eventually reah a steady state. 3. After exposure to equal onentrations of diesel exhaust partiles, the redutions in alveolar learane of diesel soot from high lung burdens is greater in hildren than in adults. 4. The aumulations of diesel soot and assoiated organis, on a per-unit-of-lung-weight basis, reah a maximum at five years of age. On the basis of their alulations of dose and the resulting preditions, humans ould prove to be more suseptible than rats to toxi effets from diesel exhaust partiles, and hildren may be espeially more sensitive than adults. From the equations and assumptions presented in the Investigators' Report, the preditions drawn appear sound. In this type of mathematial modeling study, however, the preditions are driven strongly by the methods used. Therefore, there are several aspets of the study to onsider for refinement: 1. Charateristis of the partile. The desriptions of the three omponents of the diesel exhaust partile ould be improved. Experimental data for the partile-assoiated organis were derived from studies of radio labeled 1-nitropyrene (Bond et al. 1986} and benzo[a]pyrene (Sun et al. 1984}. In those studies, the radiolabeled ompounds were adsorbed onto diesel soot partiles, and then the ra ~~~ m

34 m r:ct : Health Review Committee's Commentary the value of looking at alternative exposure patterns. Thus, in addition to fitting existing data from experiments that emphasize lung burden and overload, assumptions of the model need to be varied further. 3. Model onstrution. A ritial aspet of the onstrution of the model is the parameter-fitting exerise onduted on the data from Strom and oworkers (1988). Beause many of the numerial parameters were derived from this exerise, the validity of the exerise is important. In most of the postexposure periods from Strom's experiment, alveolar soot mass (m~) delined very little below its initial level (m~kl Therefore, it is diffiult to asertain whether or not the alveolar learane rate (A~J) was a funtion of the dynamially hanging mass (m)fll or a funtion of the maximum mass (m~k whih remains a onstant) during the postexposure period. Beause using the maximum mass led to a simple exponential deline, it appears that it was hosen for onveniene. This hoie would be irrelevant if exponential delines fitted the data well, but in fat the fit was poor (see Figure 4). The fit might have been improved by allowing the alveolar learane rate to depend dynamially on the delining mass of soot. 4. Statistial reliability. The statistial reliability of the gathered, derived, or assumed parameters ould have been more thoroughly examined. Goodness-of-fit was not evaluated for the funtions fored through the data from Strom and oworkers (1988), on whih many of the model parameters were based. No reason was given for the unonventional 1/y 2 weighting that was used to fit these data. Furthermore, the fitted parameters were reported without standard errors (see Table 1). Similarly, all additional parameters derived from Strom's study or from other studies were reported without any indiation of statistial limits. The possibility remains that the propagation of error ould make some parameters of the model ill-determined, and thus of limited value. This aspet of variability has not been estimated by the alulation of standard errors, nor has its impat been assessed by sensitivity analysis. Without suh an assessment, it is diffiult to know if the model is well-determined on the basis of the data from whih it was onstruted. Thus, the stability, validity, and reliability of the preditions are unlear. 5. Extrapolation assumptions. The extrapolation from rats to humans represents the most interesting and valuable part of the study. The rat-to-human saling, however, was based on the unonfirmed assumption that mehanial learane varies with the speifi partiulate dose to alveolar surfae in the same proportion in humans and in rats. All other learane rates in humans were assumed equal to those in rats. These assumptions, however, are substantial ~nd require validation. 6. Model validation. Inomplete validation of the model is a onern. Although the investigators presented some independent data, some of the validation was irular, in that experimental data were used to onstrut the model and then were used to validate the model. The absene of abundant experimental data from whih the investigators ould have drawn emphasizes the need for suh studies. These limitations do not seriously detrat from the estimates and usefulness of the model; rather, they should be onsidered during future revisions of the model and for its use by others. The model generated reasonable qualitative behavior and quantitative results for various body burdens of soot and organi ompounds. No paradoxial or extravagant effets were apparent. The model yielded several interesting numbers, suh as the lifelong equilibrated soot aumulation in adult lungs (15 mg for eah mg/m 3 of hroni half-day exposure), the ratio of alveolar to traheobronhial burden (5:1 for organi ompounds, 4:1 for diesel soot), and the equilibration time for aumulating lung soot (five to ten years). Comparisons among learane rates after full-day exposures (24 hours/day and 7 days/week), half-day exposures (12 hours/day and 7 days/week), and work-day exposures (8 hours/day and 5 days/week), whih entail a total of 168, 84, and 4 hours/week, respetively, exhibited a ratio of 4:2:1. The lung burdens of diesel soot also exhibited an approximate ratio of 4:2:1 over a 1-year hroni exposure. The exeption to this alulation ourred when the effet of overload (that is, high onentrations and full-day exposures) added to the burden on a onentration-speifi basis. Assumptions and approximations are inevitable with an ambitious modeling exerise. The investigators took are to examine the effets that would have resulted if inorret values had been used for some of their simulations. For example, two ruial parameters, the alveolar soot learane rate (A~) fitted to the data from Strom and the slower alveolar organi learane rate (A~J) obtained by averaging two values from the literature, were perturbed to determine the effet on the lung burden of soot and organis. The response was approximately inversely proportional; halving the learane rate doubled the burden, and raising the learane to 3/2 of the unperturbed value lowered the burden by a fator of 2/3. One important parameter that was not perturbed was the saling fator for alveolar learane rate (P) between rats and humans. Extrapolation to humans was aknowledged as the "most tenuous" aspet of the model and is, therefore, a good andidate for sensitivity analyses. 3

35 31 on instantaneous levels of soot, but remains at the end-ofexposure level. Different parameters an be inserted into the model, making it dynami and useful. Values for lung burdens, whih are diffiult to obtain experimentally, an be alulated from the model. The estimations of dose and the resulting preditions both support previous findings in animal studies and address new and interesting information on the extrapolations of the animal studies to human exposure senarios and on the potential suseptibility of preshool hildren. Calulations of dose suggest that after repeated exposures, the aumulated lung burdens of diesel soot in humans would be higher than in rats, whih implies that humans may be more suseptible to toxi effets than are rats. Aording to the model, adults, who have larger lungs, aumulate more total soot than hildren in response to a given exposure. Conversely, the lungs of preshool hildren absorb more soot on a massspeifi basis than do tissues of adults. As with all models that lak data on many of the physiologial proesses involving learane, the validity of this model depends on the auray and ompleteness of the assumptions on whih it is based. Most of the onerns are with larifying the onstrution of the model and distinguishing between what was assumed and what was derived, so that the limits of the model's inferene an be more learly delineated. Current appliation of this model, however, must be approahed with great restraint. For example, the extrapolation of a model that assumes overload may not be appropriate for lower exposure onditions. With an understanding of the limitations of that assumption, the model alulation that shows a peak in lung burden per unit of lung weight in the age range of two to six years has important impliations for establishing standards for hildren based on data from adults. Confirmation of this estimation of dose, as well as others generated by the model, should be pursued and tested. In summary, the model represents a signifiant ontribution to the field and ould have a diret bearing on risk assessment as well as impliations for publi poliy. From a risk assessment standpoint, this model suffiiently desribes the potential tissue dose of inhaled partiles, but the model desription for the adsorbed organis may be too simplisti. For example, although the investigators reognized the importane of the binding of organi ompounds to maromoleules, they did not provide transport rates; there should be some experimental data on xenobioti metabolism that ould be inorporated into the model. One modified and validated, the model ould provide onsiderable insight into the dose to target tissues for diesel exhaust partiulate matter. Thus, the human health risks from inhaled diesel engine exhaust ould be better assessed. REMAINING UNCERTAINTIES AND IMPLICATIONS FOR FUTURE RESEARCH The model developed by Yu and Yoon remains to be fully explored. Some of the assumptions need testing. For example, the use of adsorbed radiolabeled ompounds as surrogates for ombustion-derived organis may not be valid; the organis forming the seondary oating of diesel exhaust partiles may not exhibit the same degree of adherene as those organis formed during ombustion. Also, some of the rat-to-human saling fators need validation. The extrapolations were based on lung size and lung surfae area; these assumptions ould be tested further. Although the model fouses on the influene of overload on lung burdens, this phenomenon is probably not relevant to most human exposures. The annual mean level of exposure to diesel partiulate matter, whih represents approximately 3.9 perent of the total suspended partiulate emissions, is estimated at less than 3. ~-tgfm3 (Carey 1987); this level is orders of magnitude less than the value on whih the model was onstruted. A greater range of exposure onditions, espeially those more losely approximating human exposure, should be onsidered. With intermittent or low-dose exposures, the lungs may lear deposited material adequately. The ultimate impat on tissue dose needs to be estimated under suh exposures. Finally, additional experimental data are needed to validate the model and its preditions. The model ould be used to suggest an appropriate experimental design for suh studies. In addition to validating the model, experimental studies ould be used to explore various omponents of the model. For example, with overload onditions the model ould be simplified by onsidering the alveolar region only; under suh irumstanes, the mehanisms of alveolar learane might be examined empirially. CONCLUSIONS The model desribed in this report addresses the pulmonary disposition of inhaled diesel exhaust partiles and the partile-assoiated organis. The model developed is well onsidered and exeuted, learly presented, and shows a high level of inventiveness and tehnial profiieny. The investigators took a novel approah in onsidering the alveolar ompartment as a single ompartment and the frational alveolar learane rate as a funtion of the existing burden, rather than as a onstant. They distinguished between learane during deposition, whih is a funtion of the instantaneous alveolar mass of soot, and learane following termination of exposure, whih no longer depends ~~~ 1- [ ~ ~

36 Health Review Committee's Commentary REFERENCES Ames BN, Gold LS Chemial arinogenesis: Too many rodent arinogens. Pro Natl Aad Si USA 87: Ball LM, King LC Metabolism, mutageniity, and ativation of 1-nitropyrene in vivo and in vitro. Environ Int 11: Bond JA, Sun JD, Medinsky MA, Jones RK, Yeh HC Deposition, metabolism and exretion of 1-[ 14 C]nitropyrene and 1-[ 14 C]nitropyrene oated on diesel exhaust partiles as influened by exposure onentration. Toxiol Appl Pharmaal 85: Brightwell J, Fouillet X, Cassano-Zoppi AL, Gatz R, Duhosal F Neoplasti and funtional hanges in rodents after hroni inhalation of engine exhaust emissions. Dev Toxiol Environ Si 13: Carey PM Air Toxi Emissions from Motor Vehiles. EPA Report No. EPA-AA-TSS-PA Offie of Mobile Soures, U.S. Environmental Protetion Ageny, Ann Arbor, MI. Chan TL, Lee PS, Hering WE Deposition and learane of inhaled diesel exhaust partiles in the respiratory trat of Fisher rats. J Appl Toxiol 1(2): Chan TL, Lee PS, Hering WE Pulmonary retention of inhaled diesel partiles after prolonged exposures to diesel exhaust. Fundam Appl Toxiol 4: Claxton LD Charaterization of automotive emissions by baterial mutagenesis bioassay: A review. Environ Mutagen 5: Garshik E, Shenker MB, Munoz A, Segal M, Smith TJ, Woskie SR, Hammond SK, Speizer FE A ase-ontrol study of lung aner and diesel exhaust exposure in railroad workers. Am Rev Respir Dis 135: Garshik E, Shenker MB, Munoz A, Segal M, Smith TJ, Woskie SR, Hammond SK, Speizer FE A retrospetive ohort study of lung aner and diesel exhaust exposure in railroad workers. Am Rev Respir Dis 137: Griffis LC, Wolff RK, Henderson RF, Griffith WC, Mokler BV, MClellan RO Clearane of diesel soot partiles from rat lung after a subhroni diesel exhaust exposure. Fundam Appl Toxiol 3: Heinrih U, Muhle H, Takenaka S, Ernst H, Fuhst R, Mohr U, Pott F, StOber W Chroni effets on the respiratory trat of hamsters, mie and rats after long-term inhalation of high onentrations of filtered and unfiltered diesel engine emissions. J Appl Toxiol 6: Hyde DM, Plopper CG, Weir AJ, Murnane RD, Warren DL, Last JA, Pepelko WE Peribronhiolar fibrosis in lungs of ats hronially exposed to diesel exhaust. Lab Invest 52: International Ageny for Researh on Caner, Working Group on the Evaluation of Carinogeni Risks to Humans IARC Monographs on the Evaluation of Carinogeni Risks to Humans: Diesel and Gasoline Engine Exhausts and Some Nitroarenes (Vol. 46). Results of the Working Group meeting, June 14-21, IARC, Lyon, Frane. Ishinishi N, Kuwahara N, Nagase S, Suzuki T, Ishiwata S, Kohno T Long-term inhalation studies on effets of exhaust from heavy and light duty diesel engines on F344 rats. Dev Toxiol Environ Si 13: Iwai K, Udagawa T, Yamagishi M, Yamada H Longterm inhalation studies of diesel exhaust on F344 SPF rats: Inidene of lung aner and lymphoma. In: Carinogeni and Mutageni Effets of Diesel Engine Exhaust (Ishinishi N, Koizumi A, MClellan RO, Stober W, eds.) pp Elsevier Siene Publishing Co., New York, NY. Kaplan HL, MaKenzie WF, Springer KJ, Shrek RM, Vostal JJ A subhroni study of the effets of exposure of three speies to diesel exhaust. In: Toxiologial Effets of Emissions from Diesel Engines (Lewtas J, ed.). Elsevier Siene Publishing Co., New York, NY. Karagianes MT, Palmer RF, Bush RH Effets of inhaled diesel emissions and oal dust in rats. Am Indust Hyg Asso J 42: Lee PS, Chan JL, Hering WE Long-term learane of inhaled diesel exhaust partiles in rodents. J Toxiol Environ Health 12: Lewis TR, Green FHY, Moorman WJ, Burg JAR, Lynh DW A hroni inhalation toxiity study of diesel engine emissions and oal dust, alone and ombined. In: Carinogeni and Mutageni Effets of Diesel Engine Exhaust (Ishinishi N, Koizumi A, MClellan RO, StOber W, eds.). Elsevier Siene Publishing Co., New York, NY. Lewtas J Evaluation of the mutageniity and the arinogeniity of motor vehile emissions in short-term bioassays. Environ Health Perspet 47: Mauderly JL, BensonJM, Bie DE, Henderson RF, Jones RK, MClellan RO, Mokler BV, Pikrell JA, Redman HC, Wolff 32

37 ~~~ m ,-----rn RK Observations on rodents exposed for 19 weeks to diluted diesel exhaust. In: Inhalation Toxiology Researh Institute Annual Report Report No. LMF-91. Inhalation Toxiology Researh Institute, Lovelae Biomedial and Environmental Researh Institute, Albuquerque, NM. Mauderly JL, Jones RK, MClellan RO, Henderson RF, Griffith WC Carinogeniity of diesel exhaust inhaled hronially by rats. Dev Toxiol Environ Si 13: MClellan RO, Bie DE, Cuddihy RG, Gillett NA, Henderson RF, Jones J, Mauderly JL, Pikrell JA, Shami S, Wolff RK Health effets of diesel exhaust. In: Aerosols: Researh, Risk Assessment, and Control Strategies (Lee S, Shneider T, Grant L, Vertek P, eds.). Lewis Publishing, Chelsea, MI. National Institute for Oupational Safety and Health Carinogeni Effets of Exposure to Diesel Exhaust. Department of Health and Human Servies. NIOSH Publiation No U.S. Government Printing Offie, Washington, DC. Pepelko WE EPA studies on the toxiologial effets of inhaled diesel engine emissions. In: Toxiologial Effets of Emissions from Diesel Engines (Lewtas J, ed.) pp Elsevier Siene Publishing Co., New York, NY. Plopper CG, Hyde DM, Weir AJ Centriainar alterations in lungs of ats hronially exposed to diesel exhaust. J Lab Invest 49: Shlesinger RB Biologial disposition of airborne partiles: Basi priniples and appliation to vehile emissions. In: Air Pollution, the Automobile, and Publi Health (Watson AY, Bates RR, Kennedy D, eds.). National Aademy Press, Washington DC. Silverman DT, Hoover RN, Mason TJ, Swanson GM Motor-exhaust-related oupations and bladder aner. Caner Res 46: Steenland K Lung aner and diesel exhaust: A review. Am J Int Med 1: Strom KA, Chan TL, Johnson JT Pulmonary retention of inhaled submiron partiles in rats: Diesel exhaust exposures and lung retention model. In: Inhaled Partiles VI (Dodgson J, MCallum RI, Bailey MR, Fisher DR, eds.) pp Pergamon Press, Oxford, England. Sun JD, Bond JA, Dahl AR Biologial disposition of vehiular airborne emissions: Partile-assoiated organi ompounds. In: Air Pollution, the Automobile, and Publi Health (Watson AY, Bates RR, Kennedy D, eds.). National Aademy Press, Washington DC. Sun JD, Wolff RK, Kanapilly GM, MClellan RO Lung retention and metaboli fate of inhaled benzo[a]pyrene assoiated with diesel exhaust partiles. Toxiol Appl Pharmaal 73: Vostal JJ, Shrek RM, Lee PS, Chan TL, Soderholm SC Deposition and learane of diesel partiles from the lung. In: Toxiologial Effets of Emissions from Diesel Engines (Lewtas J, ed.). Elsevier Siene Publishing Co., New York, NY. White HJ, Garg BD Early pulmonary response of the rat lung to inhalation of high onentrations of diesel partiles. J Appl Toxiol 1: Wiester MJ, Iltis R, Moore W Altered funtion and histology in guinea pigs after inhalation of diesel exhaust. Environ Res 22: Wolff RK Effets of adsorption of benzo[a]pyrene onto arbon blak partiles on levels of DNA adduts of rats exposed by inhalation. Toxiol Appl Pharmaal 97: Wolff RK, Henderson RF, Snipes MB, Griffith WC, Mauderly JL, Cuddihy RG, MClellan RO Alterations in partile aumulation and learane in lungs of rats hronially exposed to diesel exhaust. Fundam Appl Toxiol 9: Wolff RK, Henderson RF, Snipes MB, Sun JD, Bond JA, Mithell CE, Mauderly JL, MClellan RO Lung retention of diesel soot and assoiated organi ompounds. In: Carinogeni and Mutageni Effets of Diesel Engine Exhaust (Ishinishi N, Koizumi A, MClellan RO, Stober W, eds.) pp Elsevier Siene Publishing Co., New York, NY. Wolff RK, Sun JD, Barr EB, Rothenberg SJ, Yeh HC Lung retention and binding of [ 14 C]-1-nitropyrene when inhaled by F344 rats as a pure aerosol or adsorbed to arbon blak partiles. J Toxiol Environ Health 26: Yu CP, Xu GB Preditive Models for Deposition of Inhaled Diesel Exhaust Partiles in Humans and Laboratory Speies. Researh Report No. 1. Health Effets Institute, Cambridge, MA. 33

38

39

40

41 LIST OF HEI PUBLICATIONS fe[ ~~ ~ Speial Reports Title Gasoline Vapor Exposure and Human Caner: Evaluation of Existing Sientifi Information and Reommendations for Future Researh Automotive Methanol Vapors and Human Health: An Evaluation of Existing Sientifi Information and Issues for Future Researh Gasoline Vapor Exposure and Human Caner: Evaluation of Existing Sientifi Information and Reommendations for Future Researh (Supplement) Publiation Date September 1985 May 1987 January 1988 Researh Reports Report No. Title Prinipal Investigator Publiation Date 1 Estimation of Risk of Gluose 6-Phosphate Dehydrogenase-Defiient Red Cells to Ozone and Nitrogen Dioxide M. Amoruso August Disposition and Metabolism of Free and Partile-Assoiated Nitropyrenes After Inhalation J. Bond February Transport of Maromoleules and Partiles at Target Sites for Deposition of Air Pollutants T. Croker February The Metaboli Ativation and DNA Adduts of Dinitropyrenes F.A. Beland August An Investigation into the Effet of a Cerami Partile Trap on the Chemial Mutagens in Diesel Exhaust S.T. Bagley January Effet of Nitrogen Dioxide, Ozone, and Peroxyaetyl Nitrate on Metaboli and Pulmonary Funtion D.M. Drehsler-Parks April DNA Adduts of Nitropyrene Deteted by Speifi Antibodies J.D. Groopman April Effets of Inhaled Nitrogen Dioxide and Diesel Exhaust on Developing Lung J.L. Mauderly May Biohemial and Metaboli Response to Nitrogen Dioxide Indued Endothelial Injury J.M. Patel June Preditive Models for Deposition of Inhaled Diesel Exhaust Partiles in Humans and Laboratory Speies C.P. Yu July Effets of Ozone and Nitrogen Dioxide on Human Lung Proteinase Inhibitors D.A. Johnson August Neurotoxiity of Prenatal Carbon Monoxide Exposure L.D. Fehter September Effets of Nitrogen Dioxide on Alveolar Epithelial Barrier Properties E.D. Crandall Otober The Effets of Ozone and Nitrogen Dioxide on Lung Funtion in Healthy and Asthmati Adolesents J.Q. Koenig January Suseptibility to Virus Infetion with Exposure to Nitrogen Dioxide T.J. Kulle January Metabolism and Biologial Effets of Nitropyrene and Related Compounds C.M. King February 1988 Copies of these reports an be obtained by writing to the Health Effets Institute, 141 Portland Street, Suite 73, Cambridge, MA 2139.

42 Copies of these reports an be obtained by writing to the Health Effets Institute, 141 Portland Street, Suite 73, Cambridge, MA July 1988 August 1988 Otober 1988 Deember 1988 February 1989 February 1989 Marh 1989 November 1989 May 1989 July 1989 September 1989 September 1989 Otober 1989 November 1989 April199 Otober 199 July 199 Otober 199 Deember 199 Deember 199 January 1991 Marh 1991 M. Jaobsen R.K. Wolff G.J. Jakab S.M. Horvath A.K. Tanswell P.E. Morrow R.M. Rose HEI Multienter CO Study Team U. Mohr (U. Heinrih) J.J. MGrath J.M. Samet J.N. Evans J.L. Mauderly F.A. Beland R.C. Moon M.B. Shenker A.M. Jeffrey R.L. Verrier J.P. Farber D.A. Johnson J.A. Last J.S. Ultman 18 Respiratory Infetions in Coal Miners Exposed to Nitrogen Oxides 19 Fators Affeting Possible Carinogeniity of Inhaled Nitropyrene Aerosols 2 Modulation of Pulmonary Defense Mehanisms Against Viral and Baterial Infetions by Aute Exposures to Nitrogen Dioxide 21 Maximal Aerobi Capaity at Several Ambient Conentrations of Carbon Monoxide at Several Altitudes 22 Detetion of Pararine Fators in Oxidant Lung Injury 23 Responses of Suseptible Subpopulations to Nitrogen Dioxide 24 Altered Suseptibility to Viral Respiratory Infetion During Short-Term Exposure to Nitrogen Dioxide 25 Aute Effets of Carbon Monoxide Exposure on Individuals with Coronary Artery Disease 26 Investigation of a Potential Cotumorigeni Effet of the Dioxides of Nitrogen and Sulfur, and of Diesel-Engine Exhaust, on the Respiratory Trat of Syrian Golden Hamsters 27 Cardiovasular Effets of Chroni Carbon Monoxide and High-Altitude Exposure 28 Nitrogen Dioxide and Respiratory Infetion: Pilot Investigations 29 Early Markers of Lung Injury 3 Influene of Experimental Pulmonary Emphysema on Toxiologial Effets from Inhaled Nitrogen Dioxide and Diesel Exhaust 31 DNA Binding by 1-Nitropyrene and Dinitropyrenes in Vitro and in Vivo: Effets of Nitroredutase Indution 32 Respiratory Carinogenesis of Nitroaromatis 33 Markers of Exposure to Diesel Exhaust in Railroad Workers 34 Metaboli Ativation of Nitropyrene and Diesel Partiulate Extrats 35 Aute Effets of Carbon Monoxide on Cardia Eletrial Stability 36 Carbon Monoxide and Lethal Arrhythmias 37 Oxidant Effets on Rat and Human Lung Proteinase Inhibitors 38 Synergisti Effets of Air Pollutants: Ozone Plus a Respirable Aerosol 39 Noninvasive Determination of Respiratory Ozone Absorption: Development of a Fast-Responding Ozone Analyzer Marh 1988 V.M. Maher Studies on the Metabolism and Biologial Effets of Nitropyrene and Related Nitro-polyyli Aromati Compounds in Diploid Human Fibroblasts 17 Publiation Date Prinipal Investigator Researh Reports Report No. Title LIST OF HEI PUBLICATIONS 1 ( ~

43 THE HEALTH EFFECTS INSTITUTE: AN OVERVIEW fe[ The Health Effets Institute (HEI) is an independent nonprofit orporation that is "organized and operated... to ondut, or support the ondut of, and to evaluate researh and testing relating to the health effets of emissions from motor vehiles:' It is organized in the following ways to pursue this purpose. INDEPENDENCE IN GOVERNANCE The Institute is governed by a four-member Board of Diretors whose members are Arhibald Cox (Chairman of the Board), Carl M. Loeb University Professor (Emeritus) at Harvard University; William. Baker, Chairman (Emeritus) of Bell Laboratories and Chairman of the Board of Rokefeller University; Donald Kennedy, President of Stanford University; and Walter A. Rosenblith, Institute Professor (Emeritus), Massahusetts Institute of Tehnology. TWO-SECI'OR FINANCIAL SUPPORT The Institute reeives half of its funds from the United States government through the Environmental Protetion Ageny, and half from the automotive industry. Twentyeight domesti and foreign manufaturers of vehiles or engines ontribute to the Institute's budget in shares proportionate to the number of vehiles or engines that they sell in the United States. THE HEI RESEARCH PROCESS The Institute is strutured to define, selet, support, and review researh that is aimed at investigating the possible health effets of mobile soure emissions. Its researh program is developed by the Health Researh Committee, a multidisiplinary group of sientists knowledgeable about the omplex problems involved in determining the health effets of mobile soure emissions. The Committee seeks advie from HEI's sponsors and from other soures prior to independently determining the researh priorities of the Institute. After the Health Researh Committee has defined an area of inquiry, the Institute announes to the sientifi ommunity that researh proposals are being soliited on a speifi topi. Appliations are reviewed first for sientifi quality by an appropriate expert panel. Then they are reviewed by the Health Researh Committee both for quality and for relevane to HEI's mission-oriented researh program. Studies reommended by the Committee undergo final evaluation by the Board of Diretors, who review the merits of the study as well as the proedures, independene, and quality of the seletion proess. THE HEI REVIEW PROCESS When a study is ompleted, a final report authored by the investigator(s) is reviewed by the Health Review Committee. The Health Review Committee has no role either in the review of appliations or in the seletion of projets and investigators for funding. Members are also expert sientists representing a broad range of experiene in environmental health sienes. The Committee assesses the sientifi quality of eah study and evaluates its ontribution to unresolved sientifi questions. Eah Investigator's Report is peer-reviewed, generally by a biostatistiian and three outside tehnial reviewers hosen by the Review Committee. At one of its regularly sheduled meetings, the Review Committee disusses the Investigator's Report. The omments of the Committee and the peer reviewers are sent to the investigator, and he or she is asked to respond to those omments and, if neessary, revise the report. The Review Committee then prepares its Commentary, whih inludes a general bakground on the study, a tehnial evaluation of the work, a disussion of the remaining unertainties and areas for future researh, and impliations of the findings for publi health. After evaluation by the HEI Board of Diretors, the HEI Researh Report, whih inludes the Investigator's Report and the Review Committee's Commentary, is published in monograph form. The Researh Reports are made available to the sponsors, the publi, and many sientifi and medial libraries, and are registered with NTIS, MEDLINE, and Chemial Abstrats. All HEI investigators are urged to publish the results of their work in the peer-reviewed literature. The timing of the release of an HEI Researh Report is tailored to ensure that it does not interfere with the journal publiation proess.

44 1-H.HEALTH EFFECI'S INSTITUTE 141 Portland Street, Cambridge, MA 2139 (617} Researh Report Number 4 May 1991