Health Assessment Document for Diesel Emissions - NSCEP | US ...

1 though specific particle surface area, a factor related to cancer potency, is increased following
2 extraction.
3 In summary, the use of rat data for quantitating human lung cancer risk from exposure to
4 DE results in considerable uncertainty. There is some doubt concerning the validity of the rat
5 model for this purpose, although proof is still lacking. The greatest uncertainty concerns low-
6 dose extrapolation of lung cancer risk. Information to date indicates that DE induces lung cancer
7 by more than one means, including particle overload-induced pathology, carcinogenic organics
8 present on the particle, and oxygen free radicals. While all these methds are likely to function
9 at high doses, particle effects are likely to be absent or minimal at low exposures, resulting in a
1 0 high degree of uncertainty regarding the shape of the dose-response curve.
11· Despite these uncertainties, rat data are still useful for comparison with other estimates,
12 especially those derived using human epidemiology data. The greater risk estimates derived using
13 epidemiology data, for example, suggest that even though rats are the only laboratory species in
14 which lung cancer can be regularly induced by DE and other fine particulate matter, they may be.
15 less sensitive than humans to nonoverload mechanisms induced by DE at low concentrations.
16 Unfortunately, limitations on animal numbers do not allow testing this possibility.
17
18 11.3.3.3. Dose-Response Estimates Based on the Biomarker Approach
19 This approach provides reasonably good estimates of lung cancer risk in spite of the fact
20 that B(a)P may constitute a relatively small traction of the carcinogens present in combustion and
21 pyrolysis products of coke ovens, hot pitch gas productions, etc. Risk estimates were also based
22 on well-documented lung cancer rates in the occupationally exposed groups: On the other hand,
23 while predictions are good for the pollutants tested, the particles present, unlike diesel particles,
24 generally lack an insoluble carbon core. As noted below, adsorption to an insoluble particle core
25 may influence risk estimates. Estimates of cancer risk will also vary based on B(a)P
26 concentration.on the particle. The variability in B(a)P concentration among different DE sources
27 and its effect on cancer potency have not been evaluated.
28
29 11.3.3.4. Dose-Response Estimates Based on the Comparative· Potency Approach
30 In this method, the potency of diesel DPM extract is compared with other combustion or
31 pyrolysis products, coke oven emissions, roofing tar, and cigarette smoke condensate for which
32 epidemiology-based unit risk estimates have been developed. Comparisons are made using
33 short-term tests such as skin painting, mutation, and mammalian cell transformation. The ratio
34 ofthepotency ofDPM extracts to each of these agents is then multiplied by their unit risk
35 estimates to obtain the unit risk for DE. Because no new studies have been carried out since the
2/1/98 11-21 DRAFT--DO NOT CITE OR QUOTE