Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 1
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels BACKGROUND In 1991, the U.S. Environmental Protection Agency (EPA) and the Agency for Toxic Substances and Disease Registry (ATSDR) asked the National Research Council (NRC) to provide technical guidance for establishing community emergency exposure levels (CEELs) for extremely hazardous substances (EHSs) pursuant to the Superfund Amendments and Reauthorization Act of 1986. In response to that request, a subcommittee of the NRC Committee on Toxicology prepared a report titled “Guidelines for Developing Community Emergency Exposure Levels for Hazardous Substances” (NRC 1993). That report provides step-by-step guidance for the derivation of CEELs for EHSs. In 1995, EPA, several other federal and state agencies, and several private organizations, academia convened an advisory committee—the National Advisory Committee on Acute Exposure Guideline Levels (AEGLs) for Hazardous Substances (referred to as the NAC)—to develop, review, and approve AEGLs (similar to CEELs) for up to 400 EHSs. AEGLs developed by the NAC have a broad array of potential applications for federal, state, and local governments, and for the private sector. AEGLs are needed for prevention and emergency response planning for potential releases of EHSs, either from accidents or as a result of terrorist activities. THE CHARGE TO THE SUBCOMMITTEE The NRC convened the Subcommittee on Acute Exposure Guideline Levels to review the AEGL documents approved by the NAC. The subcommittee members were selected for their expertise in toxicology, pharmacology, medicine, industrial hygiene, biostatistics, risk assessment, and risk communication. The charge to the subcommittee is to (1) review AEGLs developed by the NAC for scientific validity, completeness, and conformance to the NRC (1993) guidelines report, (2) identify priorities for research to fill data gaps, and (3) identify guidance issues that may require modification or further development based on the toxicological database for the chemicals reviewed. This interim report presents the subcommittee’s comments concerning the NAC’s draft AEGL documents for 15 chemicals: toluene, xylenes, ammonia, bromine, aniline, methyl ethyl ketone, hydrazine, iron pentacarbonyl, phosphine, chlorine trifluoride, ethyleneimine, propylenimine, allyl alcohol, ethylene oxide, and nickel carbonyl.
OCR for page 2
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels COMMENTS ON TOLUENE At its August 31-September 2, 2004, meeting, the subcommittee reviewed the AEGL document on toluene. The presentation was made by Sylvia Talmage of Oak Ridge National Laboratory. The subcommittee recommends a number of revisions. The subcommittee will review the revised AEGLs draft at its next meeting. General Comments The TSD appears too long. It should be condensed. The consideration to use PBPK modeling is appropriate. After review by NAC, the NAS/COT Subcommittee on AEGLs will review this approach. Some specific points to the PBPK presentation are listed below under Specific Comments. In a recent publication (Tanaka et al., 2003. J. Med. Sci. 49:129-39), 19 symptoms related to CNS and autonomic nervous system are reported to occur upon exposure to low (15.3-31.5 ppm) concentrations of toluene. It is recommended to evaluate whether these include some symptoms which may serve as starting points for AEGL derivation and to include this study in Section 2.2.1 (page 20, line 22). It would be desirable and hopefully possible for the justifications of the AEGL values to be more explicit. Thus, on one hand, the use of 200 ppm for AEGL-1 (page 82, line 12) appears too conservative. This exposure level is a threshold for altered performance of an extended series of certain complex psychophysiological tests by humans. If the AEGL-1 values are to be based on CNS depression, the 15-min 300 ppm NOAEL of Baelum et al. (1990) could be used in conjunction with PBPK modeling to extrapolate the shorter and longer exposure periods. Once a decision is reached about factoring exercise into derivation of AEGLs, this can also be accommodated by PBPK modeling. On the other hand, an intraspecies uncertainty factor (UF) of 1 was used to derive AEGL-1, yet insufficient scientific support appears to be provided to justify a reduction of the standard UF of 10 to 1 (as if there were zero intraspecies uncertainty left). The text says that the “preponderance of data as a weight of the evidence consideration indicates that an 8-hr exposure to 200 ppm would be without adverse effects for the general population” (page 7, lines 33-34). The discussion in the text (page 81, Section 5.3 Derivation of AEGL-1) is even more vague. The study used as the basis for deriving the AEGL-2 values is weak and not supported well by the studies cited in the text. The primary study is by Gamberale and Hultengren (1972), where people were exposed to 700 ppm toluene for 20 min. It says that at this level, “only a very subtle effect on the CNS was observed during this short exposure” (page 83, lines 5-6). Given the numerous studies on toluene, it is surprising that this is the best study that could be found to derive AEGL-2.
OCR for page 3
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Also the number taken as the basis for the AEGL-2 (page 85, lines 15-18) merits reconsideration. The 20-min 700-ppm exposure cannot be considered a true 20-min exposure. As noted in line 17, it immediately followed successive 20-min 100-, 300- and 500-ppm sessions. PBPK modeling should be used to establish what the subjects’ blood level would have been after such a four-part regimen. This blood level would serve as the basis from which to extrapolate (by PBPK) to other time periods to predict the magnitude of exposures that would be required to produce the same blood level. In addition, there was considerable loading of blood and tissues with toluene before Gamberale and Hultengren’s (1972) 20-min, 700-ppm exposure. Near steady-state was therefore likely achieved during this 20 min of exposure (page 86, lines 22-27). Thus, reduction of 30-min AEGL-2 to 570 ppm is probably excessive. PBPK modeling should give a more accurate value. Also, the basis for “slightly lowering” the 1 hr AEGL-2 value from 570 to 510 ppm is not well stated or clear. Why 510 and not 500 ppm? Why such precision to a value judgment? The only reason given is that the “steady-state in the blood and brain may not be reached at the 30 min time point” (page 86, lines 25-26). An intraspecies UF of 1 was used to derive AEGL-2. Again, little specific support is provided to justify moving away from a default UF of 10. The text says only that the “observed effects are below the definition of an AEGL-2 (which should always be the case since the AEGL values are defined as the concentration above which it is predicted that…), and this value is conservative as the exposure followed a 20-min exposure to 500 ppm,” (which bears no relation to the expected magnitude of intraspecies variation) (page 8, lines 24-25). The supporting analyses in both cases (AEGL-1 and AEGL-2) appear quite arbitrary as opposed to being consistent, scientifically logical, and defensible. This is apparent in the analysis of the supporting data at higher concentrations of AEGL-1 and AEGL-2. In both cases, the minimum alveolar concentration (MAC) for volatile organics in humans is said to vary by about 2-3-fold, though no citation is provided (page 8, lines 2-3; page 77, lines 1-7). In the case of AEGL-1, the high exposure was divided by a factor of 3 to show that there was close agreement with the selected AEGL-1 value. In the case of AEGL-2, the same argument is given, but in this case the higher concentration is divided by a factor of 2, apparently because it gives a better agreement with the selected AEGL value than dividing it by 3 would. The decision to divide by 3 versus 2 provides no insight or justification for moving away (let alone for a quantification of how much to move away) from a default intraspecies UF of 10. An interspecies UF of 1 was used to derive AEGL-3, even though a study of rats was used to evaluate exposures. The support given to justify a UF of 1 appears insufficient. The Executive Summary argues that a UF of 1 is “sufficient because toluene uptake is more rapid and tissue toluene levels are higher in the more rapidly respiring rodents than in humans” (page 9, lines 30-31). That factor may be sufficient to justify using a UF of 3, but it is not sufficient to justify
OCR for page 4
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels using a UF of 1. The text where this is discussed is confusing and needs to be rewritten to clarify the rationale for selecting both the intra- and interspecies UFs (see page 88, lines 3-9). It is recommended that in the Executive Summary a statement that the AEGL-3 values are all greater than 10% of the lower explosive limit be included. Specific Comments The following comments are in regard to the PBPK-Modeling Based Derivation of AEGL Values for Toluene, preliminary draft, August 21, 2004. Some background should be given on why PBPK was used. Will this tool be used for other chemicals? What is the point of the time to steady-state statement on page 2 and then showing a simulated time to steady-state later (Figure A-13)? Clarify and discuss in the text. Were dosimetrics measured under steady state conditions for each exposure scenario? One deficiency in the report is on the details of how the simulations (of the NOAELs) were carried out. If the NAC did not evaluate the blood concentration levels at steady state or for multiple days (or weeks), what are the implications? Single exposure, non-steady state? The subcommittee believes that the review of blood/air partition coefficients and the list of values in a table format are appropriate. This should also be carried out for metabolic constants such as KM, Vmax, and first-order rate constants used in published human and rodent PBPK models. This will provide more insight into the selection of metabolic constants for use in this exercise. Is there a lung compartment? Clarify the lung blood volume in Table A-2. Figures do not show up in black and white print very well. It would help to show more figures that demonstrate the effect of exercise on the model-predicted blood and breath concentration levels, such as those in Figure A-12. The subcommittee agrees that the early time points with blood and breath can be problematic without going to a more complicated description of the lung. Exhaled breath can be problematic, even for longer time points, in part, because of methods used to collect breath samples. The subcommittee emphasizes the blood as the important dosimetric in comparison to exhaled breath. Sensitivity analysis: The subcommittee usually looks for a 1:1 correspondence between model parameter change (1%) and the change in the outcome of interest (blood concentration of toluene). If this is the case, everything over a sensitivity coefficient value of 0.1 (absolute value) would be sensitive. Elaborate on sensitivity analysis and how it is important for the modeling papers. Perhaps, for the few most sensitive parameters, running simulations and
OCR for page 5
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels demonstrating the effect on model-predicted toluene concentrations would be beneficial in gaining acceptability of the model. Again, to better understand exactly what was done with the simulations for the AEGL values, it would be good to specify the details. Comment on Figure A-14, where the standard approach (ten Berge et al.) may underpredict the blood concentration levels at rest and overpredict with exercise relative to PBPK model predictions. What are the implications of using modified CxT calculations vs. PBPK? For AEGLs, how many blood concentration measurements are there? Derivation of “n”: When using an AEGL-3 effect for the derivation of AEGL-2, for completeness, state that the mechanisms are the same. If this is not true, then per the SOP use the default values of 1/3. Page 6, lines 5-6. Range 0.16 to 100 ppm? (see page 14, line 23). Page 8, lines 2-3; page 77, lines 1-7. A specific reference is needed for the statement repeated several times in the text that “among humans, the minimum alveolar concentration (MAC) for volatile anesthetics typically varies by about 2-3 fold.” Page 9, lines 16-17. The statement, “because of the long-term exposures to 800 ppm in the supporting studies of von Oettingen et al. (1942) and Carpenter et al. (1944),” is incomplete. Page 11, lines 12-16. Unclear: a) “the primary use is for production of chemicals” and b) “accounts for about 14%.” Do these refer to toluene or solvents in general? Page 11, lines 30-34. Move this paragraph to begin at line 6, and make a summary statement of this paragraph at line 5 on page 6. Page 11, line 31. Since toluene vapor (saturated) is only slightly (about 10%) more dense than dry air (see calculation below), it is rapidly dispersed with normal eddy currents. It may therefore be advisable to limit the warning by adding, “and in still air (confined space) may travel”) after “The vapor is heavier than air.” Calculation: change in density with toluene vapor concentration (compared with dry air) is 1 mole at NTP (normal temperature and pressure, 25 °C and 760 mm Hg) = 24.45 L MW dry air = 28.96 gm/mole MW toluene = 92.14 gm/mole
OCR for page 6
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels At 30,000 ppm toluene = 30,000 of a million molecules (i.e., 3% are toluene) Similarly, at 20,000 ppm toluene = 20,000 of a million molecules (i.e., 2% are toluene) At the saturated toluene vapor pressure of 36.7 mm Hg: Page 12, line 24. Define the term “low.” As written, the statement implies that toluene release into ambient air results in a “euphoric community.” Page 13, line 3. Adverse effects on liver, kidneys, lungs, and heart are limited to acute and chronic exposures to very high vapor concentrations. Page 13, line 7. Bruckner and Warren (2001) have also recently reviewed toluene toxicity. Page 13, line 22. Delete the speculation, “and thereby provides a built-in safety mechanism.” The conclusion suggests that toluene abuse at 10,000 ppm can be considered safe. Page 13, line 34. Probably section 3.1.1 is meant. Page 14, line 16. The statement refers to the total range, that is, to 0.16-100 ppm (see lines 18 and 23).
OCR for page 7
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Page 14, line 27; page 15, line 2. It should be pointed out that solvent abusers repeatedly inhale anesthetizing concentrations on a daily basis during most of their waking hr. Page 14, line 30. “…metabolic acidosis, often with “anion gap…”. Is “…often with an increase in anion gap…” what does this mean? Page 15, lines 21-23. The discussion of the exposure of the workers is unclear. Which and how many employees were grouped according to concentration of toluene? Page 15, line 20 to the end. The Wilson (1943) study is weak due to the highly suspect concentrations. These concentrations were measured using a combustible gas indicator (CGI). With an LEL of 1.4% v/v, 200 ppm is only 1.4% LEL. This is typically much lower than the measuring range of the CGI (generally > 5% and more reliably 10%), and that is with today’s technology. There is also no indication of confounding chemicals. The article states it was commercial toluene. A CGI measures all combustibles. Further, there is no indication the CGI was calibrated to toluene. In fact, this is unlikely because most CGIs are calibrated to a combustible gas (methane, propane, etc.). In summary, the reported atmospheric concentrations are suspect at best. Apply an appropriate disclaimer. Page 15, lines 37-38. Why is it stated that the results of this study are troubling,” when it has been stated previously in the document that toluene was contaminated with benzene during these early years? Page 16, lines 20-27. Details are needed on exposure for this study (Ukai et al., 1993). If they are not available, say so. Page 17, lines 1-12. Is there anything that can be stated about the sampling and analysis (for example, NIOSH analytical method, charcoal tubes with GC/FID, or GC/MS analysis)? It would be helpful to know something about these. Page 19, line 16; page 20, lines 5-6. How does the subcommittee know that there were no permanent or persistent effects? Was a follow-up study done that led to this conclusion? Page 22, Table 2 (cont’d), 2nd column. Add that the workers, exposures ranged from 9 to 25 years. Page 33, lines 18-19. Add and evaluate Svensson et al. (1992), Am. J. Ind. Med. 22:99-107, 1992, study which found the opposite. Page 35, lines 6-10. Use new IARC evaluation (1999) that concludes “there is no evidence that toluene is a potential human carcinogen based on animal studies.” Page 36, line 7. It is preferable to state that exercise results in “increased,” rather than “maximum,” uptake.
OCR for page 8
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Page 36, lines 22-27. Include some of the developmental effects from page 31, lines 35-37, in this summary. Page 36, line 32. Drop “two” because there are several. Page 40, lines 21-22. Should the value of 1,000 ppm for 2 weeks be higher? Page 49, line 20. Should 80,000 be 8,000 ppm? Page 51, line 6. What was the duration of the exposure in the rabbit study? Page 54, lines 32-33. The major metabolites.. are nongenotoxic while the minor metabolite o-cresol is clearly genotoxic. Page 55, line 25. Add, “while the great majority of tests showed no genotoxicity of toluene, some were positive (reviewed in IARC Monograph 1999).” Page 56, line 18. Add, “the IARC (1999) concluded that there is evidence suggesting lack of carcinogenicity of toluene in experimental animals.” Page 57, line 35. Systemic uptake of inhaled toluene is dependent upon cardiac output, as well as respiratory rate. Both increase with exercise. Page 59, line 15. Which species? Page 60, line 38. Replace IIC11 and IIE1 with 2C11 and 2E1. Page 61, line 1. Add that CYP2B1 is induced by toluene concentrations as low as 500 ppm and is important for the formation of the genotoxic o-cresol (Wang et al., 1993). Biochem. Pharmacol. 46:413-9. The Wang et al. (1993) reference should be added in the reference section. Page 61, line 4. Add the information on the human CYPs responsible for toluene metabolism (Nakajima et al. 1997, given in the list of references, but the information is not given in the text). Page 61, lines 14-16. This sentence is awkward and needs to be rewritten (what are “times of metabolites”? rates of formation? times of persistence?). It seems to be saying that there is a wide variability among humans in their metabolism of toluene. If this is right, then doesn’t that contradict the general reasoning used to support the use of a UF of only 1 for intraspecies variations? Page 64, lines 18, 22-25. Did Carlsson (1982) actually report blood levels in mg/kg? Page 69, lines 11, 27, 31, 33; page 72, line 25. Since the parent molecule and several toluene metabolites appear in the circulation after toluene exposure, please indicate which compound(s) was (were) measured here.
OCR for page 9
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Page 72, line 9. General statement or referring to toluene? Page 72, line 11. Exposure to toluene? Page 73, line 1. It is more accurate to state that toluene produces CNS “depression” or “inhibition” rather than “toxicity.” Page 73, lines 7-8. It is not clear whether the last three values are tissue:air or tissue:blood partition coefficients. Also, one value appears to be lacking. Page 73, line 12. The paper by DeJongh and Blaauboer (1996) is not included in the references. Did these authors compare their model predictions with empirical data to validate their model? Page 75, lines 26-29. Xylene(s) would be expected to be somewhat more potent CNS depressants than toluene, because the additional methyl group makes xylene(s) more lipid soluble. Page 75, lines 38-39. The higher respiratory rate and cardiac output of mice produce greater systemic uptake of toluene. This should result in substantially greater CNS depression in mice than in rats. These species differences, however, are offset to some degree by more rapid toluene metabolism by the mouse. Page 77, line 16; page 78, line 4. It would be worthwhile to point out the relative merits of PBPK modeling vs. the ten Berge et al. (1986) approach for time scaling. Page 78, lines 23-34. It should be related here that toluene and a number of other VOCs are competitive metabolic inhibitors, as they are oxidized by some of the same P450 isozymes. The net effect is an increase in the blood and tissue (for example, brain concentration) levels of each parent compound (despite some increase in exhalation) and an increase in the degree and duration of CNS depression. Page 80, line 13. Are subtle manifestations of CNS depression indicative of “neurotoxicity” or reversible “inhibition”? Page 81, lines 21-22. What is meant by “neurobehavioral effects were subtle and reversible, also below the definition of an AEGL-1”? Clarify. Page 81, lines 24-25. It is incorrectly stated that 700 ppm was a NOAEL in the study of Gamberale and Hultengren (1972). Their subjects exhibited a decrease in perceptual speed, as well as apparent decreases in simple and complex reaction times at this exposure level. Page 81, lines 24-30. It is not clear what is meant by the statement that “the concentrations would effectively approach a doubling value, i.e., 400 ppm in the former study and 600 ppm during the latter due to exercise.” If exercise produces an approximate two-fold
OCR for page 10
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels increase in toluene uptake/blood concentration levels, exercising subjects would only have to inhale half as much (that is, 100 and 150 ppm) in order to exhibit the same blood concentration levels as the sedentary subjects inhaling 200 and 300 ppm. Page 81, lines 26-27. It is not accurate to state that exercise in the studies of Astrand et al. (1972) and Baelum et al. (1990) takes into account the stress that may occur during an emergency situation. Although an emergency would likely involve increased exercise, stress would also likely result in increased release of catecholamines. Toluene can sensitize the myocardium to catecholamines, but predisposition to arrhythmias requires inhalation of ≥5,000 ppm toluene. Page 81, lines 31-32. What is meant by, “the preponderance of data as a weight-of-evidence consideration indicates that an 8-hr exposure to 200 ppm would be without an effect that exceeds the definition on and AEGL-1”? Clarify. Page 81, line 38. “Dividing the 700 ppm.” Explain to what study “the 700 ppm” refers. Page 82, lines 5-7. Increases in blood and brain toluene concentration levels are asymptotic once near-steady-state is reached. It would be better to say that these increases are “relatively modest” rather than “minimal.” Page 82, lines 21-25. It is true that the systemic uptake (and CNS depression) of/by toluene is (are) greater in rodents than in humans. One cannot measure subjective complaints (for example, headache, dizziness, irritation) in rodents. Our measures of more pronounced manifestations of CNS depression in rodents are insensitive. Thus, rodent toluene NOAELs are often considerably higher than corresponding human NOAELs. Page 85, line 13. In light of the foregoing, “difficult” should be replaced by “not possible.” Page 88, lines 6-8. Another important factor that contributes to lower blood toluene concentration levels in humans is their lower blood:air partition coefficient (PC). Although Gargas et al. (1989) do not include human and rat blood:air PCs for toluene in their Table 8, the PCs for F-344 rats are higher for benzene and o- and m-xylene. Page 88, lines 21-24. It appears preferable to use PBPK modeling for time scaling from 2 to 4 and 8 hr. Page 88, lines 26-27. Near-steady-state is reached within 60 min in rats. Minor Points Page 6, line 8-9. Remove the bolding. Page 13, line 18-22. Remove the bolding. Page 16, line 27. Typo: should be “essentially.”
OCR for page 11
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Page 31, line 37. Delete “fetal effects.” Page 32, line 2. Insert to read, “other chemicals or drugs (especially ethanol)…” Page 68, line 26. The abbreviation PBTK should be defined the first time it appears in the text. PBPK is used on pages 72 and 73. Choose one of these abbreviations for sake of consistency. Page 72, lines 13-16. The blood concentration values of Benignus et al. (1998) in mL/L should be stated in mg/L in the text, so they may be readily compared to the blood concentrations reported in mg/L by other investigators. Page 74, line 34. Typo: should be demyelinization. Page 76, line 15. It is not clear here whether the volume of the blood sample was “smaller than usual” or whether the third dog was smaller than the other two dogs studied by von Oettingen (1942). Page 81, lines 34. What is meant by, “Although these concentrations do not approach gross CNS effects”? Page 85, line 15. Delete the phrase, “of the general population.” As written, the conclusion implies that community exposures to 700 ppm are “appropriate”? Page 97, line 21. Complete the author names in accord with NRC style directions. This entry appears to be a repeat of that at line 18, and one of the Donald references can be deleted. COMMENTS ON XYLENES At its August 31-September 2, 2004, meeting, the subcommittee reviewed the AEGL document on xylenes. The document was presented by Claudia Troxel of Oak Ridge National Laboratory. The subcommittee recommends a number of revisions. The subcommittee will review the revised xylene AEGLs at its next meeting. Overall Comments Of major importance is the fact that the AEGL-2 and AEGL-3 values are greater than 10% of the lower explosive limit. This fact should not be hidden in the footnotes to Table 4 (page 42) and the footnote in the Executive Summary on page xi. The fact that the xylene AEGLs represent an explosion hazard should be highlighted on line 8 of the Executive Summary (page ix) and should appear in the first sentence of Section 8. Since the authors included results of rodent behavioral testing with xylene, some explanation of the relevance of those data to the AEGLs is in order here. The absence of any comment as
OCR for page 46
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels some variability is still expected (e.g., in repair), so that a full UF of 10 is not needed. Again, this is consistent with the SOPs (page 90, section 184.108.40.206.4), but not well articulated. The rationale for decreasing the interspecies and intraspecies UFs from 10 appear to be the same for both factors. Section 8.3. The statement at line 17 is in direct contradiction to that made at lines 20-23. How is it that AEGL-2 values can be derived (page 20) when Section 8.3 states, “The data for deriving the AEGL-2 value was less than adequate (line 27)?” Appendix B. The text ignores the Dow report (Kilian 1973, see above comment to page 5, lines 21-22). COMMENTS ON PROPYLENIMINE At its August 31-September 2, 2004, meeting, the subcommittee reviewed the AEGL document on propylenimine. The document was presented by Kowetha Davidson of Oak Ridge National Laboratory. The subcommittee recommends the following revisions. General Comment The document has been improved substantially; however, minor problems remain. Major Comment The rationale for selecting a relative potency of 5 for AEGL-2 needs to be clearly explained. Specific Comments Section 2.2, line 30. The statement on line 31 is not entirely correct in that duration of human ethyleneimine exposure has been described on at least one occasion. It is worthwhile to note the report by Danehy and Pflaum (Ind. Eng. Chem. Res. 30:778, 1938) who found that a single exposure to ethyleneimine vapor for 2-3 min resulted in vomiting that was delayed some 3 hr after the incident. While no concentration data were presented, the nausea and vomiting continued for “a few days.” Page 4, line 27. Use and reference the IARC (1999) evaluation. Page 7, line 23; page 8, line 35; page 9, bottom line; page 10, line 5; page 10, line 34. No data concerning percutaneous toxicity associated with topical or airborne propylenimine are presented in the document. It appears that the NAC relied upon the ACGIH skin designation assigned in 1991. Quoting from ACGIH (1991), “Because propylenimine resembles ethylenimine in its physiologic action, a skin notation is also recommended.” As
OCR for page 47
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels written, the draft AEGL document suggests that airborne propylenimine contributes to systemic toxicity due to its absorption through intact human skin. No data are cited in the present document to support that contention. The erroneous statements listed above should be deleted from the document. Page 8, line 1. Add “by inhalation” after ethylenimine, since the toxicity difference on the skin is smaller than given here. Page 10, line 13. It is stated that a relative toxicity approach compared to ethylenimine was used to develop AEGLs 1 and 2; however, there is no AEGL-1. AEGL-3 is based upon a single concentration over time. Since the relative toxicity was developed on mortality data, it would seem the relative potency approach would work best for AEGL-3. Based upon ethylenimine what is the AEGL-3? Page 10, line 40. No references to ACGIH (1991) or ACGIH (1997) appear in the page 12 bibliography. COMMENTS ON ALLYL ALCOHOL At its August 31-September 2, 2004, meeting, the subcommittee reviewed the AEGL document on allyl alcohol. The document was presented by Claudia Troxel of Oak Ridge National Laboratory. The subcommittee recommends the following revisions. A revised draft should be reviewed by the subcommittee at its next meeting. Overall Comments The NAC used an interspecies UF of 1 in deriving an AEGL-3 based on data from an animal study. The rationale provided for this determination is that “these data suggest little difference between species in response to allyl alcohol exposure” (page vii, line 28; page 17, lines 25-26). However, no data are provided in the Executive Summary to support this claim. The text (page 17, Section 4.3 Species) discusses some data, but it is not sufficient to conclude that all species (including humans) respond similarly to the effects resulting from exposure to allyl alcohol. The data discussed in Section 4.3 are mostly lethality data, and no data on humans were presented that are comparable to the animal data. In addition, the text states that “the lethality data summarized in Table 5 lack LC50 values suitable for direct comparisons of species sensitivity” (page 17, lines 19-20). In addition, the data presented on nonlethal effects come from a study in which all the animal data were grouped together such that the reader cannot determine which specific effects occurred in which specific species. The text states that these results were “discussed in general terms for all species” (page 9, line 33). For these reasons, selecting an interspecies UF of 1 for AEGL-3 may not be justified, and a UF of 3 could be used to derive AEGL-3.
OCR for page 48
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels The argument for selecting an intraspecies UF of 3 for AEGL-3 is weak and not scientifically based; the values would be “inconsistent with available empirical data” (page vii, line 33). Inconsistency between the results and other established values is not sufficient reason to alter the UFs. It is illogical to make a scientific judgment about what the UF should be based on the data and available information, and if the end result values seem inconsistent with other values, go back and adjust the UFs. The UFs should remain the same and then, if there is a strong reason to change the resulting numbers, an adjustment should be made. There needs to be a solid scientific basis for moving away from the default value of 10. This should not be done in order to “make the numbers work.” The basis for selecting an intraspecies UF of 3 for AEGL-1 is that “irritants are not likely to vary greatly among individuals” (page vii, lines 11-12). No data were presented to support this statement in the Executive Summary or in the text. While it may be true, it needs to be discussed and documented here or by reference to the exact section of the SOP. The statement should be precise in discussing nasal irritation, not ocular or other forms of irritation. Does this protect the individual from other adverse health effects as well? Are children more sensitive to irritants than adults? General Comments Explain how the NAC justified a 30 min AEGL-3 nearly 7 times greater than the IDLH. At a minimum, the text on page 22, line 3, should provide some rationale for the marked difference—perhaps by providing a one- or two-sentence critique of the 20-ppm IDLH for this material. The authors should refrain from statements like those made on page 10 (“Thus, the NAC inferred…”), page 17 (“The NAC committee recommended…”), and page 23 (“The NAC recognizes…”). As written, the text begs the questions, if NAC recognizes the potential carcinogenicity of allyl alcohol, why don’t the AEGL values recognize that conclusion accordingly? As written, it is not clear why the experimentally derived n = 0.8 in Section 4.4 was not used for time scaling since page 94 of the SOP lists TCE as one example of a substance with n = 0.8. There is nothing in Section 2.7 of the SOP that states empirical n values < 1.0 shall be assumed equal to the default n value of 1; SOP page 103 states, “The lowest value of n was 0.8 and the highest value of n was 3.5.” Therefore, additional justification for n = 1 (page 20, lines 28-29) is needed unless the empirical n = 0.8 is used in time scaling. As written, lines 2-11 on page 21 are not understandable. On line 1, it is not clear whether the “default” to which the author refers relates to n or possibly to the various uncertainty factors. This section should either be deleted or rewritten. In discussion of the intraspecies UF on page 20, lines 15-17, there is no consideration of those with clinical, bacterial, or other infections. Given that allyl alcohol-treated rats pretreated with bacterial endotoxin experienced enhanced hepatic damage as compared to rats given ally
OCR for page 49
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels alcohol alone (Sneed et al. 1997), it is not clear that the justification provided for the 3-fold intraspecies UF can be supported. The regression analysis yielded n = 0.78, and the NAC rounded it to 1. As discussed at the meeting, rationale must be provided for why the value was changed (not rounded) based upon the data. Reference should be made to the appropriate section in SOP to support treatment of the data. Specific Comments Page vii, line 8, Executive Summary. What is meant by “the model”? Clarify the intent of this phrase. Page 3, Table 2. Fix spacing typo in 5th column. Page 9, Section 3.2.2., lines 2-14. This paragraph is redundant because earlier text describes the same study (page 6 Lines 16-38). Is it necessary to describe both studies in detail? Should these results only be included in the lethality section since some animals died, although others, at different exposure levels, survived. Page 9, Section 3.2.2., lines 15-20. This paragraph (lines 15 to 20) is redundant. Same comment as above. Page 9, lines 31 to 33. The comment that the “effects reported by the study authors were not separated by species (dogs, guinea pigs, rabbits [Section 3.2.1.], and rats), but were discussed in general terms for all four species” should be included in the discussion of the same data on page 8, lines 26-39. Page 15. The document will be improved if the mode of action (depletion of reduced glutathione [GSH], loss of protein sulfhydryls, and peroxidation of lipid membranes) was explained. It is important to explain that studies with inhibitors of alcohol dehydrogenase (such as 4-methylpyrazole) abolish allyl alcohol-induced hepatotoxicity and that pretreatment with inactivators of Kupffer cells (such as gadolinium chloride) can attenuate allyl alcohol-induced hepatoxicity. As written, the text leaves the reader with the impression that not much is known about the mode of allyl alcohol action when this is really not the case at all. Page 16, lines 8-9. The statement, “there was no delay in the appearance, development, or disappearance of the measured irritant response,” does not follow the previous four lines on the role of acrolein in allyl alcohol toxicity. Page 17, Section 4.3. This section should include a discussion on potentially susceptible populations, especially children. As written, the AEGL document does not consider whether the proposed AEGL values would be protective of children or those with compromised medical conditions.
OCR for page 50
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Page 17, lines 5-12. It is not clear from the text here whether allyl alcohol is considered a reactive or a nonreactive alcohol. Page 17, line 36. What does allyl amine have to do with the present document on alcohol? Delete. Page 18, lines 18-19. Should this sentence state “nasal irritation is not likely to vary greatly among individuals” (as opposed to irritation in general)? Page 19, Section 6.2. This section states that “no single exposure inhalation study reported effects consistent with the AEGL-2 definition.” This may be true, but there were several single-exposure mouse studies that should be mentioned in this section. Page 20, lines 10-12. The comment is made here that AEGL-3 values were based on the “highest concentration causing mortality in mice, rats, and rabbits,” but it is not clear which species was used to derive the AEGL-3 values. As is stated in the next paragraph, there were deaths in these same species at higher exposures. The difference is that these higher exposures occurred for different periods of time—longer in the rabbit and shorter in the mouse (see Table 5). This should be noted in the text. The reason for selecting the 200 ppm exposure values, as opposed to the 500 ppm exposure, to derive AEGL-3 needs to be made clear in the text. Page 21, line 2. Should this read AEGL-3 and not AEGL-2? Page 21, lines 6-7. What is meant by the statement, “The AEGL-2 values help to serve as a baseline: they are based on a multiple scenario in which rats exposed to 40 ppm for 7 hr/d exhibited reversible signs of irritation”? The current AEGL-2 is based on data from human exposure, not on animal data. To what does this statement refer? Page 21, lines 16-17. Why is this sentence necessary? Additional References Belinsky, S.A., T. Matsumura, F.C. Kaufman, and R.G. Thurman. 1984. Rates of allyl alcohol metabolism in periportal and pericentral regions of the liver lobule. Mol. Pharmacol. 25:158-164. Hormann, V.A., D.R. Moore, and L.E. Rikans. 1989. Relative contributions of protein sulfhydryl loss and lipid peroxidation to allyl alcohol-induced cytotoxicity in isolated rat hepatocytes. Toxicol. Appl. Pharmacol. 98:375-384. Maellaro, E., A.F. Casini, B. Del Bello, and M. Comporti. 1990. Lipid peroxidation and antioxidant systems in the liver injury produced by glutathione depleting agents. Biochem. Pharmacol. 39:1513-1521.
OCR for page 51
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Przybocki, J.M., K.R. Reuhl, R.G. Thurman, and F.C. Kaufman. 1992. Involvement of nonparencyhmal cells in oxygen-dependent hepatic injury by allyl alcohol. Toxicol. Appl. Pharmacol. 115:57-63. Sneed, R.A., S.D. Grimes, A.E. Schultze, and P.E. Ganey. 1997. Bacterial endotoxin enhances the hepatotoxicity of allyl alcohol. Toxicol. Appl. Pharmacol. 144:77-87. COMMENTS ON ETHYLENE OXIDE At its August 31-September 2, 2004, meeting, the subcommittee reviewed the AEGL document on ethylene oxide. The document was presented by Kowetha Davidson of Oak Ridge National Laboratory. The subcommittee recommends the following revisions. A revised draft should be reviewed by the subcommittee at its next meeting. General Comments The subcommittee understands the argument that dose to the tissue in rats is likely to be higher than in humans. This should have been succintly stated for the non-toxicologist who may read this TDS. The subcommittee is not convinced of the correctness of the inter- and intra-species UFs used. They seem low and not sufficiently justified. Is the ossification issue expected to be the same in rats and humans? The subcommittee believes the answer is no. If no, the inter-species UF should be higher. The use of 100 ppm (from the Snellings et al. 1982a study) as a starting point for the development of the AEGL-2 value needs a convincing argument saying that (consistent with the definition of the AEGL values as “concentrations above which it is predicted that…”) the effects seen at this concentration are not serious health effects and those at the next higher concentrations are (or else the proposed values should be changed). Specific Comments The reasons for choosing the older and analytically less refined study (Jacobson et al. 1956) as the basis for deriving the AEGL-3 values (instead of Nachreiner 1991, 1992) should be explained better. The addition of the argument that this leads to more conservative values may be useful. It is stated that the reason for using the Jacobson study is that it is a clear dose response, and if this is the case, showing the plots may be helpful. Page 1, 2nd paragraph. The major use of ethylene glycol is not in the production of antifreeze, but rather as an intermediate in the production of polyesters (IARC monograph 1994).
OCR for page 52
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Pages 43-45. It should be reevaluated whether the contribution of glutathione S-transferase to the total disposition of ethylene oxide is really only 10%-20%, and whether at the same time, the contribution of glutathione S-transferase polymorphism(s) to ethylene oxide toxicity is really significant. Could it be that the two statements refer to different concentrations of ethylene oxide (10%-20% contribution at high saturating concentrations; significant contribution to toxicitiy at low, nonsaturating concentrations)? Page 44, lines 11-14. It is probably not accurate to say that EtO is metabolized by hydrolysis. The hydrolysis of EtO is reportedly not catalyzed by epoxide hydrolase or any other enzyme. It would be better to replace “metabolized by hydrolysis” with “hydrolyzed.” Similarly, the word “metabolism” in line 4 should be replaced with “clearance.” Page 45, line 5. Change “adduct level” to “adducts” at the end of the line. Page 45, lines 6-9. Specify which GST isozyme the authors are referring to here. Page 46, lines 1-5. Have any more recent papers been published that address the mechanism(s) of EtO-induced neurotoxicity? Page 47, Section 4.4.3 and Figure 1. The text should point out that the data base for using n = 1.2 to extrapolate from 4 hr to other time points is slim but considered to be better than just using a default factor. Page 49, lines 4-6. It is stated in paragraph 1 that “the reproduction study of Snellings et al. (1982b) cannot be used because it is difficult to attribute the observed effects to a single exposure to EtO.” This reasoning appears inconsistent with that utilized in paragraph 2 to justify basing the AEGL-2s on the developmental study of Snellings et al. (1982a). Page 49, lines 5-17. It is correctly stated that there would not be an accumulation of EtO upon daily exposure to the chemical. There could, however, be an accumulation of EtO adducts to proteins, RNA, and DNA in the fetus. This is an argument against use of a repeated-dose study’s results to derive AEGL-2 values. It should be pointed out here that EtO is a direct-acting alkylating agent that has been shown to be a reproduction toxicant. The mention of two anticancer agents at line 9 may imply that the next mentioned agent, chlorpromazine, is one of these. Chlorpromazine’s major clinical applications have been as a tranquilizer and a skeletal muscle relaxant may be preferable to name the two anticancer agents. Despite the foregoing, manifestations of EtO’s fetotoxic effects in a substantial number of studies are limited to growth retardation. EtO has not been shown to cause fetal malformations (that is, be a teratogen). Modest delays in ossification of bone(s) are consistent with ongoing/prolonged nutritional deficits or chemical effects that retard cell growth and replication. Adverse developmental outcomes of a single/acute dose of toxicant would more likely result in fetal malformation(s). Therefore, the AEGL-2 values derived by use of the repetitive dosage study of Snellings et al. (1982a) should be supported by results of another investigation. Division of the AEGL-3 values by 3 yields
OCR for page 53
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels AEGL-2 values that are similar to those based on the repeated-dose study of Snellings et al. (1982a). Page 49, last 3 lines; page 50, line 1. An interspecies UF of 3 is advocated despite (a) the aforementioned simulations of Fennell and Brown (2001) showing comparable blood EtO concentrations in rats and humans inhaling EtO at 100 ppm for 4 hr; and (b) the statement that the potential mechanism of toxicity is not expected to differ across species. Is the magnitude of alkylation/toxicity known to be dependent upon the peak blood EtO concentration or the area under the blood concentration versus time curve (AUC)? The PBPK modeling (see their Figure 6) of Fennell and Brown (2001) indicates that the post-exposure rate of elimination is slower in humans. This implies that a 100-ppm exposure would result in a larger AUC in humans than in rats. Page 50, lines 2-8. The reasoning appears to be illogical. Change to a more logical deduction or conclusion or improve the wording (or omit). Page 50, lines 8-13. It should be noted here that both Muller et al. (1998) and Farmer et al. (1996. Environ. Health Perspec. 104(Suppl. 3):449-452) found no effect of glutathione S-transferase M1 genotype on levels of HEV hemoglobin adducts in a human study population. Muller et al. did find a 2-fold higher HEV level in persons deficient in the T1 isozyme. Page 50, lines 13-15. The question of whether EtO is passed from the mother to the fetus is critical to determining the propriety of using the results of the Snellings et al. (1982a) study as the basis for deriving AEGL-2s. EtO has such a short half-life and is so reactive that it might not be expected to cross the placental barrier and reach the fetus in amounts adequate to retard its growth. No study was located in which a pregnant animal was dosed with EtO and her fetuses analyzed for EtO or HEV adducts. Tavares et al. (1994) and Farmer et al. (1996) did find elevated HEV hemoglobin adducts in the newborns of smoking mothers. Cigarette smoke contains substantial concentrations of ethylene, but very small amounts of EtO. Ethylene is very lipid soluble and, therefore, readily crosses the placenta. Ethylene is metabolized by hepatic cytochrome P450s in the fetus to EtO. Thus, ethylene rather than EtO may be responsible for Tavares et al. finding of elevated neonatal HEV hemoglobin adducts. Alternatively, the mother metabolizes some of the ethylene she inhales to EtO (though ethylene metabolism is slow in humans [Filser et al. 1992], followed by possible transfer to the fetus. Thus, there is considerable uncertainty about transplacental passage of EtO. The findings of Tavares et al. (1994) are not strong supporting evidence for such transfer. These investigators’ publication in Carcinogenesis (66:157-163, 1994) should be cited rather than their abstract (?) in Human Experimental Toxicology. Page 50, 2nd paragraph. If possible, PBPK modeling should be conducted to extrapolate from rats to humans, and from 6 hr to the shorter exposure periods. Such modeling sometimes results in lower short-term AEGLs than does the ten Berge et al. (1986) method. The ten Berge et al. paper should be included in the references if their method is utilized.
OCR for page 54
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Page 52, 2nd paragraph. PBPK modeling should be utilized if possible for time scaling and interspecies extrapolation in the AEGL-3 derivations. Minor Points Numbering of the pages is from 1-8 and then starts again with 1. When it is referred here to one of the pages 1-8 as they occur for the first time, this is marked by (1) in parenthesis following the page number. Page 6 (1). Title should be “Executive Summary” Page 7 (1), line 11 from bottom. Typos: “lower the. AEGL-3 values.” Page 3, line 22. Typo: add i to “distinct.” Page 3, line 23. Should this be “to be 1625 ppm”? Page 3, line 25. Change to “The derivation of the LOA.” Page 27, line 3. Capitalize “gd”; define at first occurrence, add to list of abbreviations, or write out in full. Page 43, lines 22-23. Change to “ethylene oxide metabolizing activity.” Page 43, lines 29-30. Change to “demonstrated that glutathione in various.” Page 43, line 3 from bottom. Do the three different depletion levels refer to the three different doses or to the three different organs? Page 49, line 13. Put “too” in front of “mild.” Page 49, line 22. Change to “The developmental study in rats by Snellings.” Page 49, line 3 from bottom. Typo: “3 for interspecies sensitivity and 3 for intraspecies.” Page 51, line 9 from bottom. Change to “being lower.” Page 51, line 5 from bottom. Put “it” between “because” and “presented.” COMMENTS ON NICKEL CARBONYL At its August 31-September 2, 2004, meeting, the subcommittee reviewed the AEGL document on nickel carbonyl. The document was presented by Robert Young of Oak Ridge National Laboratory. The subcommittee recommends the following revisions.
OCR for page 55
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels General Comments The document has improved substantially; however, relatively minor problems remain. While page ii, lines 26-30, address the developmental toxicity of inhaled nickel carbonyl in relation to maternal health status, page 15 and page 19, line 5, provide no corresponding statement. It is worthwhile to repeat the data described on page 22, lines 15-16, at this point. On page 25, what is the meaning of the phrase, “the total uncertainty adjustment of 10 is weighted towards the uncertainty in individual sensitivity to nickel carbonyl exposure”? Page 23 suggests that information is not adequate to assess rigorously an intraspecies UF. In that respect, what is the meaning of this page 25 conclusion? Specific Comments Page ii, lines 27-28. What data are available on the maternal health status of the rat dams? If none are available, this should be so stated. Page 16, Table 12. While the table is titled, “Maternal Lethality and…,” no data on maternal deaths are included in the table. Page 16, lines 23-41; page 17, lines 4-5. The discussion should indicate which groups experienced maternal deaths. To what does the statement, “increased mortality in some treatment groups,” refer? Refer to Table 13, line 4. Page 16, line 35; page 17, line 22. In general, data on prenatal deaths are presented as either resorptions (early embryonic demise) or as late fetal deaths rather than the number of live fetuses per implantation site. Page 17, line 1. As written, it is not clear whether the total numbers of malformations (that is, the numbers of malformations per fetus) or an increase in the total numbers of abnormal fetuses is the intent. Page 17, line 7. It is not clear whether the “parenteral” administration includes intravenous nickel injection. It is probably wise to state the precise route (for example, intraperitoneal injection), since bioavailability is complete after intravenous injection. Page 17, line 16. To what does the table entry “0.16 AEGL” refer? Page 22, line 16. Into which body cavity (thoracic, peritoneum) was the serous hemorrhage observed? Page 22, lines 13-20. Refer the reader to (a revised) Table 12.
OCR for page 56
Twelfth Interim Report of the Subcommittee on Acute Exposure Guideline Levels Page 22, lines 28-29. This is an incomplete sentence. What happened to the 9 of 19 and 9 of 14 rat dams? Include a description of maternal toxicity. If that parameter was not presented by Sunderman, indicate that. Page 22, lines 35-37. The text reads as speculation. If no empirical data are available for direct support of this NAC conclusion, it should be deleted. Page 24, lines 15-17. The text should mention the 80% mortality in the hamster dams and the fact that the hamsters were afflicted with serous hemorrhage (page 22, line 16). Page 24, line 42. Does this sentence refer to a general practice or to a specific section of the SOP?
Representative terms from entire chapter: