1 15 2 https://www.nal.usda.gov/exhibits/speccoll/files/original/dd1ebcd642f6ebd4f7666ada03f31002.pdf 4ef0bc9b3f43aa98047bf3086cd5e4ea PDF Text Text Item D Number °5494 Author Mukerjee, Debdas Corporate Author u RBPOrt/ArtlCiO TltlB Res D Not Scanned ' S. Environmental Protection Agency (EPA), Office of Ponse to Key Issues Raised by Public Review Comments on Health Assessment Document for Polychlorinated Dibenzo-p-dioxins, External Review Draft Journal/Book Title Year Month/Day Color Number of Images D ° EPA-600/8-84-014A Friday, March 15, 2002 Page 5494 of 5571 �RESPONSE TO KEY ISSUES RAISED BY PUBLIC REVIEW COMMENTS ON HEALTH ASSESSMENT DOCUMENT FOR POLYCHLORINATEO DIBENZO-&-DIOXINS EPA-600/B-84-014A, May 1984 External Review Draft Prepared by U.S. Environmental Protection Agency Office of Health and Environmental Assessment Office of Research and Development Cincinnati. Ohio 45268 Project Manager Debdas Mukerjee November 1984 �1. Comment: This 1s a concise, well-written review and evaluation of published Information on 2,3,7,8-tetra-CDD, on 1,2,3,7,8-pentaCDD and two hexa-CDD Isomers (1,2,3,6,7.8- and 1,2,3,7,8,9HxCOO). It will be a most valuable reference work. However, from this reviewer's perspective 1t appears regrettable that the focus of the document was kept so narrow, given the wide range of potentially d1ox1n-contam1nated Industrial chemicals, herbicides and other pesticides, antimicrobials, and drugs Identified 1n the 1980 U.S. EPA "D1ox1ns" document (EPA 600/2-80-197), and the published and unpublished studies available on 2,7-d1COD, l,3,7-tr1-CDO, 1,3,6,8-tetra CDO, 1,3,7,9-tetra COO and octa-CDO. Some mention should also have been made of chlorinated azoxybenzenes etc. It might have been very useful to have at least an addendum with a comprehensive listing of additional relevant published references on these compounds. Response: This document has been prepared at the request of the Office of Air Quality Planning and Standards (OAQPS). The OAQPS selected the congeners and Isomers for discussion 1n this document. 2. Comment: 2.1. Summary. Regulatory limits for permissible levels of 2,3,7,8-TCDD 1n 2,4,5-T, fenoprop, and 1n fish and other food sources, as well as similar limits for dloxlns 1n 2,4-0 and chlorophenols, set by the U.S. and other countries might have been mentioned. For recent examples, see e.g. U.S. EPA Position document 4 on wood preservatives; Agriculture Canada Memorandum to Registrants No. R-l-216 and Trade Memorandum T-l-233 on 2,4-D which set byproduct "production limits* for 2,7-d1-, l,3.7-tr1- and 1,3,6.8/1,3.7,9-tetra COO (see Ref 1,2 attached), an Agriculture Canada Memorandum on chlorophenols also proposes production limits on HxCOO. Response: Regulatory limits pertaining to the chemicals discussed 1n this document have been dealt with 1n Chapter 13. 3. Comment: Page 2-3, last paragraph. Should spell out the actual 1050 values of 1157-5051 vg/kg for the hamster. Response: The final draft of this document will reflect this. 4. Comment: Page 2-6, last continued paragraph. read 1 ng/kg/day. Should be corrected to Response: The final draft of this document will reflect this. 5. Comment: 2-2, line 8. "highest levels" suggest Including numerical value so the reader can understand what 1s considered a high level. Response: Since there 1s great variation 1n the range of the levels 1t 1s felt not to give the levels 1n the summary. However, detailed figures have been given 1n Table 6-3 and 6-4. 0039K -1- 11/20/84 �Comment: 2-3, lines 18-19. Change to ". ,. some of the most acutely toxic compounds . for male guinea pigs being 0.6-2.1 vg/kg ...." Response: Since 2,3,7,8-TCDD Is acutely as well as chronically one of the most toxic compounds known, It Is felt not to change the statement as 1t currently appears 1n HAD. However, It will be Indicated fhat this 1s the lowest 1050 value. 7. Comment: 2-3, line 22. Although 2,3,7,8-TCOO 1s highly toxic 1n all species tested — The chemical may be toxic 1n all .spedes but 1s not highly toxic 1n all species. The spedes differences observed for TCOD should be given more Importance 1n discussions rather than the toxldty. Response: 2,3,7,8-TCDD 1s toxic at 0.6 iig/kg dose for the guinea pig (the most sensitive species) and for the hamster (the least sensitive species) toxldty can be expressed at 5.5 mg/kg, a dose that Is low 1n comparison with most other chemicals. Accordingly, the statement as 1t appears 1n HAD correctly reflects the current understanding of 2,3,7,8-TCOO toxldty. 8. Comment: 2-3, line 24. state signs. Characteristic symptoms — more correct to Response: The final draft of HAD will reflect this correction. Comment: 2-4, line 1-6. A statement Indicating that none of the responses observed are 1n and of themselves the cause of death. Response: This 1s just the summary. The text explains this. 10. Comment: 2-4, line 7. Herbicide should be singular 1f the reference 1s to 2,4,5-T which 1s the only herbicide to contain 2,3,7,8TCDD as a contaminant. Response: 2,4,5-T: (2,4,5-Trlchlorophenoxy) acetic add and Sllvex: 2-(2,4,5-Tr1ch1orophenoxy) proplonlc add are herbicides that contain 2,3,7,8-TCOO as contaminant. No change 1s warranted. 11. Comment: with exposure leading In. 2-4, lines 9-10. Change to some cases to ..." remove "porphyrla cutanea tarda". Response: Since In animals 2,3,7,8-TCOO can Induce porphyrla and people exposed occupationally to 2,3,7,8-TCOD and other chemicals have developed porphyrla cutanea tarda (PCT), 1t would not be considered prudent to remove PCT as one of the symptoms of 2,3.7,8-TCOO exposure. 12. Comment: 2-4, line 7-15. The paragraph should be reconstructed to Indicate that signs of toxldty observed 1n man exposed to chemicals containing TCDD often reflect the signs of toxldty of the chemical as observed 1n animals, but the only signs associated with TCDD are skin lesions. 0039K -2- 11/20/84 �Response: Ep1dem1olog1c studies and case reports suggest that signs and symptoms 1n addition to chloracne are associated with 2,3,7,8-TCDD exposure; It 1s felt that the paragraph should remain as 1t appears now. 13. Comment: 2-4, lines 23-24. Certainly does not reflect the presentation 1n Chapter 10. Surprising that Section 2.3. 'Needs for Future Research" does not have mutagenldty testing listed. Response: The purpose of mutagenldty testing 1s to determine whether the chemical should be elected for animal bloassay., Since 2,3,7,8-TCDD 1s a proven animal carcinogen 1t will not be for the best Interest for a regulatory Agency that further short term tests should be done and will not add further guidance for regulating 2,3,7,8-TCDD. However, to determine the mechanism of cardnogenesls by 2,3,7,8-TCDD and other mutagenldty studies, DNA adduct, defective DNA repair, slster-chromatld exchange studies can be suggested. 14. Comment: 2-5, lines 19-20. These chemicals are highly toxic to certain animal species. Differences 1n spedes response should be made 1n these discussions. Response: Same as for Comment 7. 15. Comment: 2-5, line 27. Fetus of certain spedes are sensitive to TCDD but a generalized statement should not be made. Response: The words "1n the species studied* this sentence. will be added to 16. Comment: 3.3. Analytical Methodology. Cochrane et al. (1981) might have been cited In connection with the Identification of tetra-CDD 1n chlorophenoxy herbicides. (This reference 1s however cited 1n 4.3.1.2., p. 4-12.) For adipose tissue, Stanley et al. (1983), Ryan and Williams (1983), Schecter et al. (1983) or their later full publications might be cited (Refs. 3,4,5 attached). Response: Such revisions will be made In the final draft of the document. 17. Comment: The Table 4-2 In the chapter on Environmental Sources lists major producers. The table 1s outdated. Response: The title of the table will be changed to: "Locations of Companies that have been major producers and formulators of chlorophenols and their derivatives*. 0039K -3- 11/20/84 �18. Comment: The HAD 1s Incorrect to characterize the trace chemistries of fire report as limiting precursors to "many chlorinated hydrocarbons*. ... Dow has recently published a paper by Nestrlck and Lamparskl 1n showing that burning of untreated wood, the unlocated nature of the wood carefully documented, produces dloxlns. Another unpublished study by Lamparskl et al. found levels of dloxln 1n mirorganlte (municipal sludge sold as fertilizer), collected 1n 1933 1n Milwaukee and sealed until 1982, to be roughly equivalent to dloxln levels found 1n 1982 mHorganlte sample. The year 1933 predates production of chlorinated phenols and many other of the narrow 11st of supposed precursors. Response: No detailed reference of the paper by Nestrlck and Lamparskl have been supplied with the comment. EPA will contact these authors to get these papers. J.M. Czuczwa and R.A. Kites (1984), 1n a presentation entitled "Trends 1n sediment case" at the 4th International Symposium on Chlorinated 01ox1ns and Related Compounds, discussed about their studies on the presence of PCDD and PCOF preserved 1n the sediments from the Great Lakes Including the sediment core from S1sk1w1t Lake, Isle Royale. The sediment that came from S1sk1w1t Lake was used because H received only atmospheric Inputs. In all cases the authors detected the flux of PCOO and PCOF, which began at about 1940. When this "1940 horizon* was compared with combustion trends 1n the last century, the authors found evidence that the combustion of synthetic chlorinated organic chemicals 1s the primary so'urce of PCDD and PCDF. Furthermore, the authors responded that the flux of PCDD and PCDF to three Swiss Lakes, where combustion has been extensive during the last century, Increased only after the development of the chlorinated organic chemical Industry. The trace chemistries of fire as proposed by the scientist from Dow Company as the source of PCDD Into the environment remains still a controversial Issue. Accordingly, no change 1n HAD 1s warranted. 19. Comment: 4.3.1.1., p. 4-5. Legend to Figure at bottom of page should read: 1,2,4,5-tetrachlorbenzene, not trlchlorobenzene. Response: The final draft will have this correction. 20. Comment: p. 4-6. Reference could be made to Canadian chlorophenol production figures (Jones, 1981, 1984; Refs. 6,7 attached) and Interdepartmental Committee on Toxic Chemicals 1983 (Ref. 8 attached). Response: These references will be Included 1n the final draft. 0039K -4-- 11/20/84 �21. Comment: 4.3.1.4. The dlphenylether pesticide TOK contains 2,7-d1COO; the same 1s true for dlcamba. Response: Since these chemicals are beyond the scope of this document, this Information will not be added 1n the document. 22. Comment: 4.5.2. The presence of 1,3,6,8- and 1,3,7,9-TCDO 1n eplgeal parts of plants might also have been due to contamination by herbicides arfd other pesticides containing TCDDS (Yamag1sh1 et al.. 1981). Response: The above statement will be Included In this section. 23. Comment: 4.4. For additional relevant Information (1981, 1984) (Refs. 6,7). see also Jones Response: These references will be Included In this section. 24. Comment: 7-18, last line. Remove "heavily* vague and the degree of contamination 1s not a question. Response: The final draft will reflect this suggestion and the word "heavily* will be removed. 25. Comment: 8-1, paragraph 1. "... may depend on the species or strain examined." Major differences exist with some strains tested. Response: The final draft will reflect this change. 26. Comment: 8-5, lines 1-4. Poorly worded — Change to "In general. ... and guinea pigs (remove words) required a specific ..." Response: The final draft will reflect this change. 27. Comment: 8-60, paragraph 2, line 1. "Chronic exposure to 2,3,7,8-TCOD has probably occurred 1n ..." Current wording Infers presently occurring and that 1t 1s positively known that the workers were exposed to 2,3,7,8-TCDD. Response: The final draft will reflect this change. 28. Comment: 8-60, paragraph 2, Line 3. "Chloracne Is generally the first. Remove "normally* — a poor word 1n this usage. Response: The final draft will reflect this change. 29. Comment: 8-60, paragraph 2. lines 6-7. a) Remove "porphyrla cutanea tarda" — there 1s not an adequate reference for this effect given. Also, see attached paper being submitted for publication that discusses this 1n detail. b) Also, change 'associated with" to "reported 1n Individuals that may have had" ... 0039K -5- 11/20/84 �Response: a) Same as Comment 11. b) The final draft will read as ..."reported 1n Individuals that have had" ... 30. Comment: The relnteroretatlon of the NTP-chron1c tox1c1ty study of 2.3.7.8-TCDD 1s unwarranted (Page 8-3). The HAD concludes a NOEL was not established 1n the NTP study, because of "toxic hepatitis* 1n the lower and middle doses. This conclusion 1s unwarranted. First, the use of the category "toxic hepatitis" 1s non-specific and Inappropriate. Second, the authors of the NTP study found that NOEL was established for "toxic hepatitis" at the low and middle dose levels. Third, the data do not support the HAD conclusion. For male mice, the Incidence rates for "toxic hepatitis" was 0-4% for the various control groups, 2% for low, 0% for the middle dose and 72% for the high dose. In female mice, the Incidence rates for toxic hepatitis were 0, 2, 4 and 72% for the control, low, middle and high dose groups. These data do not come close to suggesting an effect at the low and middle doses. A closer look at the diagnostic categories for liver toxldty Included In the NTP report gives little Indication of any dose response for liver degenerative, necrotlc or Inflammatory change at the middle or low dose levels. Response: Toxic hepatitis describes general toxic effect 1n the liver. Table 03 1n page 172 of NTP gavage study document describes non-neoplast1c lesions 1n vehicle control and dosed groups of male mice. This table Indicates that the Incidence of toxic hepatitis 1n vehicle control was 1(4%), low dose 5(10%), mid dose 3(60%) and high dose 44(88%). The number of mice with lesions of the liver study used and the Incidence of toxic hepatitis are enumerated 1n Table 11 (page 53) of the NTP gavage study. Toxic hepatitis Incidence data of this study discussed 1n HAD concur with NTP's data. The only error detected 1n HAD 1s the Incidence for control group. The correct figure 1s 1/73 Instead of 0/73. This correction will appear In the final draft. 31. Comment 0039K The HAD also concluded that a NOEL was not established 1n the mouse study by Toth et al. Nuch more Information 1s required on the alleged Incidence of amyloldosls 1n the lowest dose, especially on the spontaneous Incidence rate of amyloldosls 1n the strain of mice used, and the thoroughness of his topathologic examination 1n a very sketchlly reported study. -6- 11/20/84 �Response: The concomitant control group had an Incidence of adverse * effect of 0. Although the Incidence 1n the low dosed group was only 5/44, there appears to be a dose response relationship which supports the conclusion that the effects observed 1n this group were treatment related. 32. Comment: 8-62, line 10. Add after "... to 2,3,7,8-TCOO and other cornpounds ()." t Response: The final draft will be changed accordingly. 33. Comment: 8-63, lines 2-3. If the results were not significant/ do not Imply that aberrations were more frequent. Change to "Lymphocyte aberrations of exposed Individuals when compared to values of "healthy* Individuals showed no significant differences." Response: The sentence reflects the raw data In the study and clearly Indicates that the elevation of lymphocyte aberrations was not significant. No change Is required. 34. Comment: 8-63, next to last sentence. "... was believed to be minimal." This should be removed; the exposure level was unknown as were the levels of exposure for other people In the area. To state that these people were "minimal" without a base for comparison Is unfounded. Response: This sentence reflects the observation by the authors on the work habits of the clean-up crew. No change 1s required. 35. Comment: 8-64, paragraph 2. The most notable results from the Missouri episode was that these children were exposed to levels that killed horses, small pets and birds and none of the children died and 1n fact all recovered. This Information should be brought out here. Response: There 1s no data on the exposure of different species and comparison between species here. Since such data are not available, no change 1n HAD Is warranted. 36. Comment: 8-64 to 8-65, Poland study. The description should note that Poland examined for PCT but did not find any. Response: We will add 1n the results of the study by Blelberg et al. (1964), which originally described PCT 1n the group of workers. Further 1t will be stated that on re-examination after 6 years, Poland et al. (1971) could not detect PCT 1n any of the workers of the plant. 37. Comment: 8-65, lines 8-9. Remove sentence and Insert "All eight workers had decreased levels of high-density Upoproteln cholesterol levels and elevated total cholesterol levels. Response: The final draft will reflect change. 0039K -7- 11/20/84 �38. Comment: 8-67, 5th line from bottom. Change to "... (forest Industries) who may have been exposed ..." Currently reads present tense, this should be past tense. Response: The final draft will reflect this change. 39. Comment: 8-68, lines 2-5. Concerning sensitivity, the statement can be made that "To date the data for humans Indicates that they are not a sensitive species to the effects from TCOO." Response: Data 1s Inadequate to conclude that humans are less sensitive. 40. Comment: 8-71, 8.3.1.2., line 10. animals which ..." Add "... biologic responses In. Response: Since many of these signs and symptoms have been observed 1n humans as well as In animals, no change 1s warranted. 41. Comment: 8-71. 8.3.1.2., lines 14-15. Item "4) PCT"; this should be removed as noted from previous comments. Response: Same as Comment 11. 42. Comment: 8-72, paragraph 2, lines 5-9. "... an excellent correlation between ... 3 different Hems? Need to explain. Response: Since the correlation 1s between toxIcUy and AHH Induction, the statement should be kept as 1t reads now. 43. Comment: 8-73, lines 1-4. ... and again a correlation between 3 different Items? A correlation 1s between 2 Items; also need to explain the "degree" 1s this numerical or judgmental? Response: The word "between" on line 1 of the 8-73 will be changed to "among". 44. Comment: 8-83, 8.4.2., lines 2-3. Change to "... lead to chloracne. Other symptoms reported are altered ... abnormalities, (remove porphyrla cutanea tarda). Response: It 1s considered that separating chloracne from other symptoms does not further the clarification of clinical observations and hence this change Is not warranted. 45. Comment: 8-83, 8.4.2., last line. Change to "... 2,3,7,8-TCDD subside or 1n the case of chloracne may persist..." Response: There 1s evidence that signs other than chloracne may persist for many years. No change 1s warranted. 0039K ., -8- 11/20/84 �46. Comment: 8.1.1.2., 8.1.1.5.2., 8.3.3. The findings of Thunberg et al. (1984) on 2,3,7,8-TCDD and vitamin A storage are of Interest here (Ref. 9). Response: This reference will be Included 1n sections 8.1.1.2., 8.1.1.5.2. and 8.3.3. 48. Comment: 8.2. Five cases of heavy chloracne In workers exposed to 2,4,5-T herbicide mixtures were described by Londono (1966), Ned. cutanea 3: 225-232. A number of cases which have been observed 1n South American workers producing pentachlorophenol or Us salts (E. Astolfl, personal communication). In addition, chloracne may be caused by tetrachloroazoxybenzenes, a possibility which could lead to presumption of exposure to 2,3,7,8-TCDO. Response: The Londono (1966) paper will be Included 1n Section 8.2. EPA will contact E. Astolfl to get the data on South American workers for Inclusion 1n this section. 49. Comment: 9-18, line 3. Remove "single*. Response: The final draft will reflect this change. 50. Comment: 9-18, line 15. It appears that the HAD 1s Inferring that an Increase 1n extra ribs Is a teratogenlc effect. This 1s not so. Response: Since there 1s difference of opinion among the teratologlst on this Issue, 1t 1s felt that the statement should be kept as 1t reads now. 51. Comment: 9-18, lines 18-20. Confusing, rewrite "... was an Increase (p<2.05) 1n total soft-tissue anomalies from the control (0/87) to 3/78, 2/33 and 2/28 1n the 0.1, 0.25 and 0.5 tig/kg treatment group, respectively. Response: Will be changed as follows: "There was an Increase 1n total soft-tissue anomalies from 0/87 to 3/78, 2/33 (p<0.05) and 2/28 (p<0.05) 1n the control, 0.1, 0.25 and 0.5 pg/kg groups, respectively. But there was no significant Increase for the Incidence where specific soft-tissue anomalies were concerned." 0039K -9- 11/20/84 �52. Comment: The NOEL 1n the Hurray et al. reproduction study 1s .001 mq/kq/dav. This was the conclusion of the authors of the study. In addition, the EPA Science Advisory Panel specifically addressed this Issue, and agreed with Hurray et al., that .001 iig/kg/day was the NOEL. The statistical revaluation of Nlsbet and Paxton contained a number of errors, as demonstrated durlrtg the cross-examination of Paxton during the 2,4,S-T cancellation hearings. The Nlsbet and Paxton reevaluatlon focussed on statistical aspects of the dat£, and failed to adequately consider biological factors which aid trained lexicologists 1n making judgments on the existence of NOELs. Response: HAD adequately reflects al. (1979), SAB (U.S. (1982). It 1s clearly by SAB Indicates that study. 53. Comment: the conclusions derived by Murray et EPA, 1979b) and Nlsbet and Paxton Indicated 1n the HAD that the review 0.001 »g/kg/day 1s a NOEL 1n this TCDD has shown to be teratogenlc only 1n mice, not In rats or other species. Pages 2-4, 9-1 and 9-34 should be changed to reflect this. Response: During the peer review process this Issue was discussed extensively and 1t was felt by the panel that 2,3,7,8-TCDD 1s teratogenlc In mice, rats, rabbits and ferrets. 54. Comment: Page 9-19. The section pertaining to the study by Huscarella et al. (1982) 1n ferrets 1s based on only sketchy details 1n a short abstract and has not been peer reviewed. If this study 1s to be considered the complete experimental data base should be evaluated. Response: HAD clearly Indicates that this study was reported 1n an abstract form. A detailed paper on this observation Is not yet available. 55. Comment: The HAD acknowledges that the Alsea study has been criticized. Although mentioning three "unpublished11 critical reviews (the Alsea study has also not been published) EPA failed to mention there were at least 15 other reviews critical of the Alsea study, an no favorable reviews. In discussing the ep1dem1olog1cal evidence stemming from the accidental contamination of Seveso, Italy, the HAD notes several factors — mainly Involving the then existing Italian system for reporting birth defects — complicating the task of conducting accurate studies. Nevertheless, It 1s certainly reassuring that no Increase 1n reproductive problems have been seen In a population receiving the exposure to 2,3,7,8-TCDD many orders of magnitude higher than could be expected from exposures In the environment. This should be noted 1n the HAD. 0039K -10- 11/20/84 �Response: The HAD reports the available data and 1t 1s considered that the reader will be able to derive appropriate conclusion from this presentation. Consequently, no changes are warranted. 56. Comment: 9-23, lines 1-3. The Oregon study 1s too flawed to use as a reference here, the New Zealand study and Australian studies are both questionable. Thus, the first sentence should start "An association ..." Response: The strength and weaknesses of Individual studies are discussed 1n the text of this section. No change 1s warranted. 57. Comment: 9-28, next to last line. "In any event..." This should be removed, 1t Implies that the statement 1s questionable. Response: This change will be made 1n the final draft. 58. Comment: 9-30, lines 4-5. This sentence needs to be rewritten. Implies that there should be an effect. It Response: These lines will be changed as follows: "There are several Inadequacies 1n these studies which might make them Insensitive 1n detecting reproductive effects." 59. Comment: 9-31, lines 12-13. This sentence should be rewritten, 1t also Implies that a positive result Is expected and that a mistake was made 1n the attempt. Response: This sentence reflects the conclusion derived by the author. No change Is warranted. 60. Comment: 9-32, lines 10-15. With seven counties 1n Michigan showing the Increase 1n cleft palate 1t 1s Inappropriate to try and Implicate that 2,3,7,8-TCDD may have played a part 1n this Increase. [This whole section on Michigan 1s poorly written and extremely slanted to try and make TCDO appear as a bad actor.] Response: This portion reflects the conclusion derived by the Michigan Department of Public Health (1983a). 61. Comment: 9-34, line 4. Change sentence to "For mice a MED of ..." Response: This change will be made 1n the final draft. 62. Comment: 9-34, paragraph 2, lines 1-2. This 1s not a true statement, TCDD has not consistently produced teratogenlc effects 1n all strains of rats. Response: The word "consistently* will be deleted. 63. Comment: 9. At least one tetrachlorozaoxybenzene 1s also teratogenlc, and might act as a confounding factor 1n studies on health effects of pesticides (Hassoun et al., 1964, Ref. 10). 0039K - -11- 11/20/84 �The possible contribution of dloxlns In 2,4t-D and 2,4,5-T to teratogenlc and other adverse reproductive effects was evaluated 1n NRCC (1984, Ref. 11). A fu-11 report on the Health Protection Branch teratogenlclty experiments with 2,4-0 and 2,4,5-T 1s found In Khera and McKlnley (1972), Toxlcol. Appl. Pharmacol. 22: 14-28; the reference to Khera et al. (1971) cited on p. 15-43 In the Review DRaft \s to a preliminary report only. Response: The final draft will have the above discussions. 64. Comment: In connection with "Agent Orange" the already released U.S. Vietnam veteran studies (Ref. 13) and the Australian Vietnam Veterans birth defects study (Ref. 14) could be mentioned. Response: These studies will be Included 1n the final draft. 65. Comment: 9-34, paragraph 3, line 3. Again the assumption Is being Implied that TCOD does cause "a teratogenlc response" 1n these species. [These types of statements Implying effects that have not been noted or are questionable at best must be rewritten (removed) from the HAD. Science must be kept accurate, but the "act" of presenting the Information must be strictly controlled.] Response: The word "demonstrate" will be changed to "evaluate". 66. Comment: 9-34, paragraph 4, line 3. Change to "although two studies have shown a questionable association between ... other studies have not." The use of the word "failed" again Implies that the negative studies were flawed and that a positive response should have been noted. [This 1s not good reporting of the facts and 1s very biased.] Response: The word "failed" will be changed to "not". 67. Comment: 9-35, last sentence. Again, a one-sided sentence that Implies humans will show teratogenlc effects. Response: The sentence Is self-explanatory. No change 1s warranted. 68. Comment: It Is stated In the External Review Draft of the Health Assessment Document for Polychlorlnated D1benzo-p-D1ox1ns that "Schwetz et al. (1973) demonstrated that HxCDD (Isomers not specified) was both fetotoxlc and teratogenlc when administered to pregnant rats at 100 mg/kg on days 6-15 of gestation." This Information should not stand alone. Statements should be added to the report detailing the differences between fetotoxldty and teratogenlclty and making a clear distinction between the fetotoxlc and teratogenlc effects of HxCDD at the administered doses. It should be stated In the 0039K -12- 11/20/84 �report that HxCOO 1s fetotoxlc and teratogenlc to rats only at high doses, and man 1s unlikely to be exposed to these high doses. Schwetz et al. (1973) administered 0.1, 1.0, 10 or 100 mg HxCOD/kg/day (purity >99X, two unspecified Isomers In ration of 89:11) In corn o11:acetone (9:1) solutions by gavage to groups of pregnant Sprague-Oawley rats on days 6 through 15 of gestation.* On day 21 of gestation, the pregnant rats were killed and necropsies performed, and the fetuses Were examined. A dose-related decrease 1n maternal weight gain was observed during gestation; at gross necropsy, evidence of maternal toxlclty was seen only 1n rats receiving 100 mg HxCDD/kg/day. The Incidence of early resorptlons did not Increase with dosage of HxCDO but there was a significant Increase 1n late resorptlons. The 10 and 100 mg HxCDD/kg/day doses were highly fetotoxlc during late gestation and surviving fetuses were of significantly decreased weight and length. The Incidences of fetal soft tissue and skeletal anomalies were significantly Increased with the 100 mg HxCOO/ kg/day dose; the Incidences of cleft palate, subcutaneous edema, split vertebral centra, and split sternebrae were significantly greater 1n the treated group that 1n the controls. Subcutaneous edema occurred with a significantly greater Incidence 1n the 1.0 and 10 mg HxCOD/kg/day dosage groups than 1n the controls. Fetal anomalies were not significantly Increased 1n the fetuses exposed to 0.1 mg HxCDO/kg/day. Schwetz et al. concluded that a 100 mg HxCDO/kg/day dose administered by gavage on days 6 through 15 of gestation was both teratogenlc and embryotoxlc to rats. EPA has previously reviewed the Schwetz et al. (1973) study 1n Position Document 1 ("PD-1") which Initiated In 1978 a rebuttable presumption against registration ("RPAR") review of pentachlorophenol under the Federal Insecticide, Fungicide and Rodentldde Act. In PD-1, EPA stated that HxCDD was both fetotoxlc and teratogenlc. AWPI solicited Dr. Schwetz' views regarding EPA's reliance upon his 1973 work as evidence that HxCDO 1Includes fetotoxlc and teratogenlc effects. Or. Schwetz comments were as follows: Regarding (HxCOD), a statistically significant Incidence of a malformation, cleft palate, was observed 1n rats given 100 yg (HxCDD)/kg/day on days 6 through 15 of gestation (Schwetz et al., 1973). In Section II-A-2-b of the Position Document 1, the Working Group states that "They found statistically significant Increases over the controls 1n all of the teratogenlc parameters observed at 100 tig/kg." The only finding In this study Indicative of a teratogenlc event was the Induction of cleft palate. The other findings, Including dilated renal pelvis, subcutaneous edema, and abnormal vertebral development are evidence of 0039K -13- 11/20/84 �embryotoxlclty but not teratogenUUy. The Agency has correctly concluded that a (HxCDO) content of pentachlorophenol does not represent a teratologlc hazard to humans. Response: The document reflects the conclusions derived by the authors. 69. Comment: HxCDD (Isomers not specified) 1s fetotoxlc and teratogenlc to rats only at extremely high doses; man 1s unlikely to be exposed to these high doses. Response: Not enough exposure assessment on HxCDD has been done to derive a conclusion as stated 1n this comment. No change 1s warranted 1n the document. 70. Comment: 10-1, line 21. According to the paper, the concentrations were less than 2-3 ug/ml. Response: This 1s correct. this comment. The document will be changed according to 71. Comment: 10-1, line 23. Some Interpretation should be given to accepting a positive response at IX survival. To 11st this on Table 10-1 without an Indication of some doubt does not reflect good scientific judgment. Response: The following statement will be Inserted on page 10-1, two lines up from the bottom after "colonies/surviving cells,*: •This positive response 1s questionable because of the extremely high toxldty observed." Also, the *+• will be changed 1n Table 10-1 under TA1532 for the Hussaln et al. (1972) study to "QR". 72. Comment: 10-3, lines 19-22. The values for the duplicate sample run at 2 tig/mi should be given as well as a criteria raised to judge the results. The results at 4 yg/mi should also be discussed 1n relationship to the final conclusion. One might judge these results questionable rather than positive. Response: The following sentences will be Inserted after the sentence ending with "cell survival." on line 23, page 10-3: A duplicate sample resulted 1n an 82X decrease In survival and a mutation frequency of 34xlO~*. These results Indicate that the reproduclbHUy of the assay may not have been perfect, but both results are well above the control value of 2.2x10~*. A dose-response relationship was not observed, Indicating that the' results at 2 yg/mt are only suggestive of a positive response. In addition, the positive results were obtained at a concentration of 2,3,7,8-TCDD (2 iig/mi) that was well above solubility 1n water (0.2 pg/l), which also casts doubt on the significance of the positive result. 0039K -14- 11/20/84 �73. Comment: 10-S. The references to Bronzettl et al. should be checked. I believe the reference for 10-6 should be to the 1980 work not 1983. Response: The correct reference Is Bronzettl et al. (1982) and this reference will be corrected. 74. Comment: 10-6, lines 17-20. The discussion of Rogers et al. (1982) should Include the result that: No significant mutation was noted In ouabaln or cytoslne arablnoslde selective systems. Response: "The discussion of Rogers et al. (1982) will Include "No significant mutation was noted 1n ouabaln or cytoslne arablnoslde selective systems." Comment: 10-6, lines 26-28. The studies referred to are NTP government sponsored results and they cannot be evaluated because procedures used to obtain the data cannot be found. In view of the Importance of this data 1t would seem a better Interpretation than this could be made. 75. Response: Procedures used 1n this experiment are not available to EPA. 76. Comment; The National Cancer Institute (NCI) data for hexachlorodlbenzo-p-d1ox1n (HxCOO) was seriously flawed and cannot be considered "sufficient" evidence of carc1nogen1c1ty. Even under the NCI's analysis, the data failed to show a significant Increase 1n malignant tumors. Response: Prollferatlve hepatic lesions 1n the liver of female Osborne- Hendel (OH) rats gavaged with HxCDO for 2 years have been examined or reviewed Independently by the National Toxicology Program (NTP) pathologlsts and Drs. Squire, Schueler, Haberman, and Hlldebrandt. The latter pathologlsts (Squire, Schueler, Haberman, Hlldebrandt) all Indicate In their review findings that the total Incidence of neoplastlc hepatic lesions were markedly reduced from that reported by the original NTP pathologlsts. Further, all pathologlsts agree that HxCDO Induces significant hepatotoxldty. It Is not uniformly agreed, however, what the exact nature of some of these prol1ferat1ve hepatic lesions represent, I.e., nonneoplastlc vs. neoplastlc. The association of hepatocellular carcinoma 1n man with long-term toxic liver Injury (e.g., ethyl alcohol, aflatoxln) 1s well established. Animals, Including the rat, respond to prolonged hepatotoxldty 1n a similar manner, even though some species differences exist. Therefore, 1n a 2-year rodent study (virtually a lifetime) with a compound that produces marked hepatotoxldty, 1t would be reasonable to expect an Increased Incidence of neoplasla 1n the liver of dosed animals. While Dr. Squire states that many hepatic nodules In this study were difficult to characterize as being neoplastlc, 1t must also be emphasized that many nodules were 0039K -15- 11/20/84 �equally difficult to characterize as non-neoplast1c. Dr. Squire also states that microscopic features of drrhotlc or regenerative nodules often appear more bizarre than neoplast1c lesions. Since bizarre hlstomorphology was not reported by the latter pathologlsts (Squire, Schueler, Haberman, Hlldebrandt) to be associated with the HxCDO-1nduced prollferatlve lesions, this might further point to their neoplastlc as opposed to regenerative nature. It 1s apparent that the histomorphologic separation of honneoplastlc proHferatlve hepatocellular lesions from/benign neoplastlc hepatocellular lesions 1n the rat Is often unclear, especially 1n the case of toxic hepatic Injury. There appear to be more similarities than differences between the lesions which does not always allow for a distinct or categorical diagnosis. In view of this, no more definite evidence 1s available to conclude that these lesions are non-neoplastlc than there 1s to conclude that they are neoplastlc. Whether the lesions are neoplastlc or not, most would agree that they are Induced by HxCOO, they are proHferatlve, and many are likely to be neoplastlc or would become neoplastlc 1f allowed to progress. The review of all data accumulated to this point suggests that the Incidence of hepatic neoplasms, although probably lower than that originally reported by NTP, 1s still Increased above historical controls 1n female rats and constitutes a tumor1gen1c response Induced by HxCOO 1n rats. Or. Squire states In his January 9, 1984, letter that 'the HxCOD bloassay conducted by the NCI provides only a weak hepatocardnogenU response 1n female rats and mice "and that" one could change my wording to limited evidence, according to the International Agency for Research on Cancer (IARC) criteria." Even 1f one were to use the pathology data for male and female rats from Or. Squire's June 29, 1983, report, which had the lowest Incidence figures of any reviewing pathologist, the combined Incidences of hepatocellular carcinoma and adenomas 1n high-dose females 1s still statistically significant by Fisher's exact test when compared to vehicle controls and yields a borderline response In male rats (I.e., females—7/50 In high dose vs. 1/75 1n vehicle controls for a P value of 0.006; males—3/49 1n high dose vs. 0/75 In vehicle controls for a value of 0.059). In either case, the data were also positive for dose-trend. In mice, the only neoplastlc changes reported by NTP were Increased Incidences of hepatocellular adenomas and carcinomas 1n males and females. Using the original NTP data, the combined Incidences of hepatocellular adenomas and carcinomas were statistically significant by Fisher's exact test 0039K -16- 11/20/84 �1n both male and female mice (I.e., males—24/48 1n high dose vs. 15/73 1n vehicle controls for a P value of 7.33xlO~«; females—10/47 1n high dose vs. 3/73 1n vehicle controls for a P value of 0.004). Even 1f one were to use the pathology data for female mice from Dr. Squire's June 29, 1983 report, the combined Incidences of hepatocellular carcinomas and adenomas 1s still statistically significant 1n female mice (I.e., 8/48 In high dose vs. 3/74 1n vehicle controls for a P value of 0.02). In either case, the data were also positive for dose-trend. ; It can thus be concluded that the 1:2 mixture of 1,2,3,6,7,8and 1,2,3,7,8,9-HxCOO was carcinogenic In animals as Indicated by a statistically significant dose-related Increased Incidence In liver tumors 1n both rats and mice as was stated before (memo dated January 3, 1984, from B. Haberman and S. Bayard/CAG to J. BelUn/OSW), even 1f one were to use Dr. Squire's Incidence figures from his June 29, 1983, report. Historically, the Carcinogen Assessment Group (CAG) has been using a modified IARC criteria scheme for weighing the evidence for carc1nogen1c1ty of various compounds. This scheme considered the combined Incidences of certain benign and malignant tumors for purposes of carcinogen evaluation where there was evidence of a tumor progression, such as Is seen 1n the liver. IARC (vol. 33, page 13, 1984) states with regard to the term cardnogenlclty that "the commonly accepted meanIng 1s the Induction of various types of neoplasms or of a combination of malignant and benign tumours." Further, IARC (Monograph Supplement 2, page 69, 1980) states that "Chemical agents that markedly Increase the Incidence of benign tumours are now viewed with almost as much suspicion as potential human hazards as they would have been If the Induced tumours had been malignant." Using the U.S. Environmental Protection Agency (EPA) Proposed Guidelines for Carcinogen Risk Assessment (draft 1984), the experimental animal evidence for HxCOD cardnogenldty would be classified as "sufficient" (I.e., Indicated by an Increased Incidence of combined malignant and benign tumors 1n multiple species). The overall evidence would then place HxCDO 1n Group 62—a probable human carcinogen, since there 1s Inadequate evidence from human studies and sufficient evidence from animal studies. 77. Comment: The NCI data on HxCOO were severely flawed by the presence of tetrach1orod1benzo-p-d1ox1n as an Impurity 1n the test material, which may have seriously affected the results. The use of corn oil gavage may have Introduced other confounding factors. Response: The HxCOO material containing the mixture of 1,2,3,6,7,8hexachlorod1benzo-p-d1ox1n and 1,2,3,7,8,9-hexachlorod1benzop-d1ox1n had an Impurity of 0.09X; this 1s partially attMb0039K -17- 11/20/84 �uted to 2,3,7,8-tetrachlorod1benzo-p-d1ox1n, which 1s one of the four TCDO Isomers. The following table shows both observed and calculated liver tumor response for HxCOD and a 0.09X TCOD contaminant respectively. Since the calculated response for the 0.09% TCDD are so very low, It 1s reasonable to conclude that the Impurity 1n the test did not contribute significantly to the observed carcinogenic response for HxCDO. 78. Comment: The use of c'orn oil gavage may have Introduced other confounding factors. Response: In the Koclba et al. TCDD rat study the compound was given 1n diet, whereas, 1n the NTP TCDD rat study, the compound was administered by gavage using the corn oil as vehicle. In both of these studies TCDD produced a clear positive carcinogenic response; therefore, 1t appears that corn oil has no obvious effect on the Induction of carcinogenic response. HxCOD, which 1s structurally related to TCDD, also produced a positive carcinogenic response 1n rats when the compound was administered by gavage using the corn oil as vehicle. The Carcinogen Assessment Group (CAG) considers that the observed carcinogenic response may not be mediated by corn oil. 79. Comment: The selection of liver tumor data for the HxCDO unit risk estimate 1s Inappropriate. The EPA used the male rat liver tumor data from the NTP study, and female rat liver tumor data from a re-evaluation by Hlldebrandt to calculate the unit risk for HxCDD. However, there 1s no mention of the 1983 Squire re-evaluation of the NTP study, although the EPA saw fit to use the Squire re-evaluation of the Koclba TCDD study as the most 'sensitive* evaluation of TCOO oncogenesls. This highly selective use of data by the EPA 1s unscientific and Inappropriate. Response: The reasons for selecting the rat liver tumor data for the unit risk estimation of TCDD and HxCDD was discussed 1n Health Assessment Document for PolychloMnated Olbenzo-pdloxlns dated May 1984. The EPA's selection of hlstopathologlc data 1s based on sound scientific judgment, and not based on who read the hlstopathologlc slides. 80. Comment: Koclba et al. (1978) TCDO study. There was also a significant decrease Incidence of endometrial hyperplasla, subcutaneous mammary tumors, pituitary adenomas, adrenal medullary hyperplastlc nodules, pheochromocytomas, pancreatic adnar adenomas; all should be discussed. (This document 1s 'chronically* one-sided and stresses the positive response data.) Response: The CAG agrees that the Koclba et al. (1978) study on 2,3,7,8-tetrachlorod1benzo-p-d1ox1n shows a significant decreased Incidence of endometrial hyperplasla, subcutaneous mammary tumors, pituitary adenomas, adrenal medullary hyperplastic nodules, pheochromocytomas, pancreatic, and adnar 0039K -18- 11/20/84 �Liver Tumor Response for HxCOO (Observed) and TCOD Contaminant (Calculated) Animal 0.09% TCDD Contaminant Oosea (pg/kg/wk) Liver Cancer Response Calculated 95% Upper UmH 4/48b 0.0045 TCOO has shown no effect 1n NCI study 18/50C 0.0045 0.02/50d 5 24/48e 0.0045 0.20/48d 10 10/47f 0.009 0.22/476 Liver Cancer HxCOD Dose (vg/kg/wk) ' Response Observed Rat (OH) Male 5 Female Mouse (B6C3F1) Male Female a it Is assumed that all of the contaminant 1s 2,3,7,8-TCOO. NTP reviewed. c Re-evaluat1on by Heldebrandt (see table 11-34). Based on response In NCI 2,3,7,8-TCDO study-see table B-10. e . . . . see table B-ll. .... see table B-12. b 0039K -19- 11/20/84 �adenomas. These results were not described In the Health Assessment Document because these findings have no bearing on the clear positive carcinogenic evidence observed by Kodba et al. 81. Comment: HxCDO and TCDD cause tumors through non-genetic mechanisms, and, therefore, should be considered "weak carcinogens.* Response: Currently no data exist on mutagenldty of HxCOO and available Information concerning the mutagenldty of TCDD Is Inconclusive. HxCDD and TCOD both have been shown/ to be carcinogenic 1n long-term cancer bloassay studies. Although TCDD has been shown to be a promoter for rat liver cancer, 1t also Induces a carcinogenic response 1n lung, tongue, and nasal turblnate for which a promotion effect 1s not demonstrated. The CAG considers that evidence 1s Insufficient for suggesting the mechanism of carclnogenesls for HxCDO and TCDD. The relative carcinogenic potencies of HxCDD and TCDO are presented 1n Table 11-36, Health Assessment Document. Relative Carcinogenic Potencies Among 54 Chemicals Evaluated by the Carcinogen Assessment Group. Quantitatively, In terms of low-dose response, 2,3,7,8-TCDO and the 1:2 mixture of 1,2,3,6,7,8- and 1,2,3,7,8,9-HxCDO rank respectively, as the most potent and second most potent carcinogens that the CAG has evaluated. 82. Comment: The negative Ott study of an Industrial cohort 1s discounted because dloxln exposures were "exceedingly low" although such exposures were likely to be as high as some of those 1n reported positive studies and are actually higher than dloxln exposures to the general population. Response: As with most studies by Gerald Ott, this study 1s a cohort mortality study of a very small number of workers (204) that reported altogether only 11 deaths and as such 1s unlikely to provide enough power, given a latent period of 15-20 years, to detect even two or three fold excess risk of the relatively rare STS. This 1s true also with respect to total cancer In workers followed for the same length of time. (Only one of Ott's eleven deaths was due to cancer). 83. Comment: A British government report critical of the Hardell studies 1s discussed as "overly optimistic*. Response: The British government report entitled "Advisory Committee on Pesticides; Report on Advisory Committee on Phenoxy Add Herbicides* was criticized by the Carcinogen Assessment Group (CAG) because of Us too willing dismissal of the evidence of STS 1n the Hardell and Erlkssen's studies while concurrently accepting as valid without critical comment the non-positive studies although questions have been raised concerning limitations 1n these studies. The British Report concluded that no evidence exists to alter their earlier conclusion that 0039K -20- 11/20/84 �formulations of phenoxy add herbicides and related wood preservatives as "presently cleared" are safe and may continue to used. Such as a blanket endorsement of the use of a substance for which sufficient questions have been rafted concerning Us health Implications, seems a bit prematures' 84. Comment: The sentlon on cancer epidemiology Is unfairly weighted 1n favor of the Reported positive studies, especially Hardell. Response: We have made every effort to present all the data from the ep1dem1olog1c studies 1n a careful and unbiased manner based upon all available Information at the time. Advantages and limitations of each study were presented In a balanced manner. Where It appears that the author made an extra effort to explain or measure the Impact of potentially limiting factors 1n his study such as was done by Hardell 1n his case control study utilizing color cancer controls (Hardell, L. 1981), we felt obligated to report on his efforts. It appears that Hardell took extra precautions to show that the presence of certain potential confounders could not have caused the statistically significant excess of STS and non-Hodgk1ns lymphomas among phenoxy add and/or chlorophenol exposed persons. However, we do agree that selective recall 1s still resent to some extent 1n the study. Observer bias could be kept under control by the researchers and 1t appears evident Hardell took suitable measures to eliminate 1t. On the other hand, selective recall would have been much more difficult to control. A credible argument could be made that persons with STS would be more likely to connect their cancer somehow with exposure to phenoxyacetlc add and/or chlorophenols than would non-STS persons because of the publicity and attention given to that possibility by the news media and the medical community. However, 1t does not seem likely that with the statistically significant risks ranging from a low of 5.3 to a high of 6.6 In Hardell's study, selective recall alone could have produced the significant risks seen. 85. Comment: Hardell Included the seven sarcomas found 1n his observation study 1n the first STS case control study. This tends to make the results a "self-fulfilling* prophecy and 1s scientifically unsound. Response: This argument has no meaning. The literature 1s replete with studies very similar In Initial development. There 1s nothing unscientific about 1t. If a cluster of cases of an unusual type of cancer occurs In a given area, 1t 1s only natural that epidemiologists would want to know If the Incidence of this cancer differs markedly from the Incidence 1n the background population from which It came, hence, 1t may be necessary to conduct a rigorous epidemlologic Investigation which would determine 1f a significant excess risk of that type of cancer exists. If 1t appears that this risk 1s 0039K -21- 11/19/84 �associated with exposure to a particular substance such as the mentioned herbicides, then separate Independent studies should be accomplished to confirm or deny the association. 66. Comment: A recent publication by Dr. Marilyn Fingerhut of the NatUna! Institute of Occupational Safety and Health reports that of the seven sarcomas found 1n cohort studies of Industrially exposed workers from Dow and Monsanto Chemical only two of the seven could be possibly associated with dloxln exposure, three of the workers were found to have had no occupational exposure to dloxlns, and that 2 of the 4 purported sarcomas were found to be carcinomas. Or. Flngerhut also pointed out difficulties 1n classifying sarcomas and relying on death certificate classifications of sarcomas. Furthermore, since the number of sarcomas have been reduced from seven to two (and one would have been classified as a skin cancer), these cohort studies do not supply confirmatory evidence on dloxln as a cause of soft tissue sarcoma (STS). Two cohort studies of (relatively) heavily dloxln exposed workers did not show a significant elevation for cancer mortality. Response: First, the CAG could not have Included a critique of the Flngerhut paper prior to the May 1984 draft of the polychlorinated d1benzo-p-d1ox1n document because Us existence was not known at the time. However, a review of this paper will be Included 1n the document before It 1s published 1n final form. We have had an opportunity to review this paper subsequently and 1t does not support the commentors claim. What 1t Indicates 1s that of the seven soft tissue sarcomas alluded to above that were pathologically shown to be soft tissue sarcomas, a subsequent review by the Armed Forces Institute of Pathology and a review by one of the authors of the Flngerhut paper confirmed five of the seven as soft tissue sarcomas while the remaining two were determined to be carcinomas although the subtypes differed In three. In terms of occupational exposure, Dr. Flngerhut proposed a strict definition of exposure as follows: a record must exist somewhere that shows an assignment to either a 2,4,5-T department or to a trlchlorophenol department at sometime 1n the past. If such a record did not exist then the Individual would not have been considered to have a confirmed exposure. Four of the seven who had a confirmed exposure In this manner were also members of cohorts that had been studied previously while the remaining three who could not be confirmed as having been assigned to any 2,4,5-T department or trlchlorophenol department. The latter three were not Identified as having been part of a study previously but were case reports of Johnson et al. (1981) and Moses et al. (1981). Individuals who were members of a study cohort might be*expected to have better documentation of exposure, based upon an employment record, than would cases turning up 1n a medical prac0039K -22- 11/19/84 �tlce. However, Or. Mngerhut did point out that of these 3 cases, one worked 32 years 1n production, clerical, truckdriving, and maintenance jobs, In a chemical manufacturing site which produced trlchlorophenol and 2,4,5-T, the sec'ond worked two and one half years as a production worker *n' a plant that made 2,4,5-T, while the last was a production and maintenance worker for 29 years at the same facility as the former. « One must question the usefulness of a classification.scheme that relies on documentation of an assignment to a specific area of a plant as proof of exposure to dloxln without real evidence substantiating that exposure. While at the same time assignment to all other areas of the same plant 1s considered Insufficient evidence of exposure although nothing 1s offered to substantiate the presence or lack of exposure to 2,3,7,8-TCDD 1n either case. It 1s Ironic that 1n most occupational ep1dem1olog1c studies, employment at a plant where the agent Is produced or found11 1s generally considered enough to call such a person "exposed and thus Included 1n a cohort for study. On the other hand, 1f Or. Mngerhut's definition were retrospectively applied to the already small occupational cohorts from which the first four STSs came, even two of these relatively rare STSs might probably constitute an excessive risk 1n the much smaller cohorts circumscribed by her definition. With respect to the difficulties Involved 1n the classification of sarcomas and problems with death certificate classifications, this topic 1s discussed 1n detail on pages 11-62 and 11-63 In the HAD. 87. Comment: The Environmental Protection Agency (EPA) discounts the negative case control study of agricultural workers In New Zealand. The EPA takes Smith to task for falling to confirm exposure data but does not analyze the same weakness 1n the Hardell exposure data. Smith's exposure data 1s more reliable than Hardell1s exposure data and uses sounder methodology than the Hardell studies. Response: The EPA disagrees with the commentor regarding his characterization of our Interpretation of the Smith study. First, the authors data was not completely negative. Smith found elevated but nonsignificant relative risks of exposure ranging from 1.3 1n Individuals who were "probably exposed" for a minimum of 5 days not In the previous 10 years prior to cancer registration, to a high of 1.6 1n Individuals "probably exposed" for a minimum of 1 day not In the previous 5 years prior to cancer registration. No power calculations are provided with either of these statistics. However, the author concluded that his findings did not support the hypothesis that exposure to phenoxyacetlc add* herbicides causes STS which CAG faithfully reported on Page 11-77 of the HAD. But, as we stated 1n our critique of his study, 0039K -23- 11/19/84 �documentation of actual exposure 1s not that good although we did say that his documentation was at least as good as that of Harden 1n his study. In fact Smith reported the possibility that TCDO contamination may be lower In the flew Zealand study compared to that of the Swedish studies. ,* • Smith's patients were considered "exposed* 1f they had definite, probable, or possible exposure to phenoxyacetlc adds through spraying or hand contact. Smith said that the actual chemical was Identified only 1n some Instances. They concluded In all remaining situations that 1f the'member sprayed "gorse" or "blackberries" this was tantamount to potential exposure to phenoxyacetlc adds. Such designations are not supported by concrete evidence of actual exposure to phenoxyacetlc add and consequently to the dloxln Impurities within. Another problem with the Smith data 1s the short period of time that one had to have "sprayed" the phenoxyacetlc herbicides to be considered as an "exposed* person I.e., 5 days or 1 day. Perhaps a longer exposure time requirement might have helped to Insure better specificity of exposure to the herbicide. Furthermore, the time Interval of 10 years and/or 5 years from exposure to registration may not have been long enough to allow latent effects to become evident. If very few persons 1n New Zealand, especially among the 2000 professional sprayers alluded to by Smith, received their potential exposure more than 15 years prior to registration than not enough time has elapsed for the detection of the relatively rare STS. Both studies use case control methodology and both have no evidence of actual real exposure to dloxln contaminated phenoxyacetlc adds. It Is difficult to Imagine that the Smith study has "sounder methodology* than the Hardell study. Both have limitations but Hardell seems to have paid more attention to potential confounders and the degree of contribution of each. Comment: Hardell failed to find a predominate type of STS from the wide variety of soft tissue sarcomas which 1s contrary to evidence for the known carcinogens. Response: As was discussed 1n the EPA document, STS may be classified at several different sites. Furthermore, they may not have their origin 1n only predominate mesenchymal tissue but may also arise In non-mesenchymal tissues as well. However, In terms of Identifying a predominate type of STS that may be affected, 1t may be that all or only certain types of STS may be subject to an Increased tumorlgenlc response". Furthermore, exposure to a carcinogen may not be specific to one site only but may affect several sites I.e., asbestos has 0039K -24- 11/19/84 �been shown to cause lung cancer Independent of the risk of mesothellomas and vinyl chloride has been found to be associated with Increased risks of brain cancer and lung cancer as well as anglosarcomas of the liver. 89. Comment: Chapter 14. This chapter does not clearly present major concerns nor criteria used to determine what 1s or 1s not discussed. No mention of the observation of species difference 1s made and this would seem to be an Important consideration In trying to make a human health hazard assessment. Again, however, no criteria are presented and fortunately no conclusion or health hazard assessment appear to have been made. Response: This chapter concurs with the approach followed 1n other Health Assessment Documents. No change 1s warranted. 90. Comment: Given the many sources of combustion 1n the country, the listing and regulation of dloxln under Section 112 of the Clean Water Act would be unworkable should EPA mistakenly find that our exposure to dloxln presents a hazard. Response: It does appear that dloxln 1s formed and released to the atmosphere when wastes containing chlorinated organUs are combusted at certain temperatures. However, the listing of dloxln under Section 112 of the Clean A1r Act would not dictate that all sources of combustion would have to be regulated; combustion sources can be regulated by source category and all source categories need not be Included for regulation after a listing under Section 112. At the time of listing as a hazardous air pollutant, OAQPS would have determined all sources and done dispersion modeling to determine which source categories are the most Important, what technologies might be most appropriate for mitigation of atmospheric release, and what the economic factors Involved would Indicate. A decision on source category regulation would be based on these and other factors. The slope factor for the dose response data (q-|*) for Inhalation exposure has been determined to be 3.3xlO~5 (pg/m'j. If this 1s multiplied by maximum annual ground levels of dloxln In air, OAQPS anticipates that maximum human lifetime risk can be on the order to 10~4 risk In same area. It 1s a risk management decision as to whether 1CT* risk represents a hazard, but this would be the level that Dow would be referring to as the one where EPA would •mistakenly" find that air exposure 1s a hazard. 91. Comment: The workshop cannot be considered an Independent review of EPA's work. The panel was selected by EPA, and the proceedings were orchestrated and dominated by Agency personnel. 0039K -25- 11/19/84 �Response: The peer review panel consisted of Independent scientists of International repute, most of whom have contributed extensively on dloxlns Issue. Considering this the peer review workshop reflects an Independent review of the document. 0039K -26- 11/I9/84 � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 195 Folder The folder containing the original item. 5494 Series The series number of the original item. Series VIII Subseries I Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Mukerjee, Debdas Description An account of the resource <strong>Corporate Author: </strong>U. S. Environmental Protection Agency (EPA), Office of Health and Environmental Assessment, Office of Research and Development, Cincinnati, Ohio Date A point or period of time associated with an event in the lifecycle of the resource 1984-05-01 Title A name given to the resource Response to Key Issues Raised by Public Review Comments on Health Assessment Document for Polychlorinated Dibenzo-p-dioxins, External Review Draft ao_seriesVIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 192 Folder The folder containing the original item. 5423 Series The series number of the original item. Series VIII Subseries I Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Mukerjee, Debdas Description An account of the resource <strong>Corporate Author: </strong>United States Environmental Protection Agency (EPA), Office of Research and Development, Environmental Criteria and Assessment Office, Cincinnati, Ohio Title A name given to the resource Letter from Debdas Mukerjee to Alvin L. Young, July 7, 1983 ao_seriesVIII