PUBLIC HEALTH ASSESSMENT
TERRY CREEK DREDGE SPOIL AREAS/HERCULES OUTFALL SITE
BRUNSWICK, GLYNN COUNTY, GEORGIA
APPENDIX F: PUBLIC COMMENTS AND ATSDR RESPONSES (continued)
Public Comments and Responses, Set #III (Outline)
III-1. Comment: Characterization of the Toxicology of Toxaphene- Some of the information regarding toxaphene toxicology and underlying studies might be incomplete or misleading in the Assessment.
Response: See the responses to specific comments in this category.
III-2. Comment: Analytical Methods in Evaluating Toxaphene in Seafood--It is recommended that the Assessment be modified to clarify which data were analyzed by which methods, and the relative strengths, weaknesses, and acceptability of these methods.
Response: See the responses to specific comments in this category.
III-3. Comment: Hazard Assessment of Toxaphene -- It is recommended that the Assessment be modified to provide the reader with further information on and clarification of the risk assessment conducted in the Assessment. It is also recommended that the Assessment be modified to incorporate the most current reliable data on toxaphene levels in fish.
Response: See the responses to specific comments in this category.
III-4. Comment: Overall Hazard Assessment - The data presented in the Assessment indicate the potential for adverse health effects stemming from exposure to toxaphene (from fish ingestion or any other pathway) is negligible. The Recommendations section of the Assessment should be modified to reflect these data. However, on the basis that community well sampling data has yet to be completed, a classification of "indeterminate health hazard" for the Terry Creek Site is supportable. Once these data are available, and demonstrate no need for concern, the classification should be modified to "no apparent health hazard" with respect to toxaphene and the Terry Creek Site.
Response: Major human PCC intake is still from biota or fish. This is the case in the Great Lakes, in the seal tissues in Eskimo food, and in the fish near this site. The concentrations of PCC in fish atthis site are in ppm levels while the concentrations in water are in ppb levels.
III-5. Comment: Historical and Non-Site Related Information --Some clarification of the Assessment may allow readers to more easily distinguish between historical data and current data and their relative impacts on public health. Further, the potential for the reader to be confused between public health concerns stemming from the Terry Creek site and those caused by possible contamination from other sources around the Brunswick area is of concern. It is recommended that the document be modified such that non-site related issues and concerns are clearly identified as such.
Response: See the responses to specific comments in this category.
III-6. Comment: Factual Inaccuracies in the Assessment - Several factual inaccuracies in the assessment that should be corrected have been identified.
Response: See the responses to specific comments in this category.
III-7. Comment: VII. Assessment Recommendations - Based on the comments for the above, some modifications to the Assessment's recommendations may be warranted.
Response: See the responses to specific comments in this category.
Public Comments and Responses, Set #III (Details):
III-1. Comment: Characterization of the Toxicology of Toxaphene
III-1-A. Comments: Developmental Toxicity
III-1-A-a. Comment: In its conclusions regarding the potential health implications of toxaphene at the Terry Creek site, the Assessment suggests that toxaphene may act as a developmental toxicant at levels to which the local population is exposed through consumption of fish and shellfish. This implication is questionable for several reasons, including the following.
*** Comment: The Assessment's statements are inconsistent with ATSDR's 1996 Toxicological Profile (Profile) for Toxaphene. The Profile [at page 40] states that "no major anatomical defects in rat or mouse fetuses were reported at doses ranging from 0.05 to 75 mg/kg body weight/day (Allen et al., 1983, Chernoff and Carver 1976, Chernoff and Kavlock 1982, Crowder et al. 1980, Kavlock et al 1982, Kennedy et al 1973, and Olson et al. 1980)".
Response: The main clause in the cited Profile sentence was omitted in this comment. The whole sentence in the Profile is the following:
"Although no major anatomical defects in rat or mouse fetuses were reported at doses ranging from 0.05 to 75 mg/kg/day ( Allen et al. 1983 .... and Olson et al. 1980), behavioral effects were reported in the offsprings of rats at doses as low as 0.05 mg/kg/day (Olson et al. 1980)."
The reported behavioral effects in the profile are consistent with the potential developmental effects suggested in the health assessment. In this case, functional end points of adverse effects (i.e., retarded maturation) are more sensitive than the structural end points (e.g., change in relative liver weight) in the neonatal rats.
The Profile explained this sentence further in the text on page 41:
"Exposure to toxaphene may also alter normal behavioral development. Delayed righting reflex development has been reported for rats following prenatal exposure to 6 mg/kg/day of toxaphene . . . Behavioral alterations have been described for juvenile rats after perinatal exposure to 0.05 mg/kg/day toxaphene (Olson et al., 1980). During early development, pups from all three treatment groups showed retarded maturation in the swimming test compared to the controls. However, all groups displayed mature swimming behavior by postnatal day 16."
***Comment: The reference dose[*(1)] apparently used as the basis of the screening level of 0.001 mg/kg/day toxaphene in the Assessment was derived by USEPA so as to be protective of health endpoints that might occur at lower doses (more sensitive endpoints) than those at which the profile indicates developmental toxicity might occur. Thus the screening level used in the Assessment (0.001 mg/kg/day) should be protective of any unexpected developmental effects from low level exposures to toxaphene. Therefore, special emphasis should not be placed on the need to protect against developmental effects above and beyond the screening level.
Response: The screening value is ATSDR's intermediate Minimum Risk Level, not the assumed 1993 RfD value of USEPA.
USEPA's reference dose RfD value was not used in this assessment because when ATSDR was preparing this edition of the public health assessment, the RfD was unavailable in the IRIS database. The assumed old RfD value of 0.00025 mg/kg/day in the 1993 document mentioned in this comment was not used either. ATSDR has not established an MRL (Minimum Risk Level) for chronic-duration oral exposure to toxaphene because a suitable NOAEL or LOAEL value could not be identified in the available literature (p. 66 in the Profile). A chronic oral MRL is ATSDR's counterpart to USEPA's RfD.
ATSDR does not follow USEPA's practice of employing an uncertainty factor of 10 for the extrapolation of a chronic RfD from toxicological data of intermediate studies. However, an intermediate oral MRL of 0.001 mg/kg/day for an intermediate exposure period (15-364 days) was established by ATSDR. Should ATSDR adopt the USEPA RfD methodology of extrapolating from intermediate durations to chronic durations the ATSDR MRL would be 0.0001 mg/kg/day.
The estimated PCC dose at this site (0.002 mg/kg/day) exceeded ATSDR's intermediate oral MRL of 0.001 mg/kg/day. Therefore, the potential PCC exposure at this site was further evaluated, using additional toxicological criteria.
***Comment: The Profile does note the same behavioral impairments (swimming ability and righting reflex seen in Olson et al., 1980) referenced in the Assessment, but the Profile also notes the critical fact that after postnatal day 14 there were no differences in the results of these behavioral tests between the control and toxaphene-treated populations.
Response: The profile stated that all groups (treated and control groups) displayed mature swimming behavior by postnatal day 16, without mentioning the righting test. Figure 1 in the Olson paper also showed that the control rats already reached the full score of 4.0 on postnatal day (PND) 15, while the toxaphene-fed rats caught up and reached score 4.0 on PND 16. There was a 1-day delay in the maturation of swimming ability in the toxaphene-fed rats. The Olson paper also stated that toxaphene-fed animals were inferior to control animals in righting ability on Day 14, and Day 15.
The sentence in the health assessment was reworded as following. The rat pups showed inferior righting ability and retarded maturation in the swimming test during their early development.
***Comment: The Profile does not identify the Olson study as the definitive developmental study for toxaphene (as is implied in the Assessment) nor does it give the Olson study any particular distinguishing emphasis.
Response: The Olson study was listed under the section of "126.96.36.199. Developmental Effects" on page 40 of the profile. It was also listed in the "Developmental"column (see 49r, the number corresponds to figure entries or key to figure) in Figure 2-2: "Level of Significant Exposure to Toxaphene-Oral" (page 31). In Table 2-2, with the same caption for Figure 2-2, the Olson study was also listed under the "Developmental" heading (page 26). The health assessment will mention the perinatal exposure (i.e., in the period around the birth).
The Olson study was used in establishing one of the two important health guidelines in the profile for toxaphene. The Olson study was the supporting study, besides the principal study, used to derive an intermediate oral minimum risk level (MRL) of 0.001 mg/kg/day. To derive the MRL, the LOAEL oral dose was not only divided by the uncertainty factor of 100 but also divided by an additional modifying factor of 3 because of the Olson study. The principal study and the supporting study and a note for the intermediate MRL was cited in pages 22 and 29 of the profile. The Olson study was the basis for the modifying factor (See Appendix A Minimum Risk Level Worksheet, pages A-5 and A-6 of the profile) used in the derivation because toxaphene could alter offspring behavioral and functional development.
The profile provided an entire paragraph to explain the use of the Olson study for the modifying factor of 3. The text of the paragraph (p.66 of the profile) is listed below.
"Additionally, perinatal exposure to toxaphene in Holtzman rats for 47 days (approximately gestational day 17 through postnatal day 40) impaired swimming ability on postnatal days 10, 11, and 12 (Olson et al. 1980). The results of that study indicate that toxaphene has the potential to alter offspring functional and behavioral development. For that reason, an additional modifying factor of 3 was included in the MRL derivation to take into account the potential for toxaphene to affect the developmental nervous system."
III-1-A-b. Comment: Further review of the Olson study itself reveals further shortcomings in the Olson study design that raise additional questions about the significance of the study's findings (and ATSDR's subsequent conclusions). Specifically,
*** Comment: The study revealed no effects of toxaphene on pup body weight or maze learning and memory tests.
Response: The health assessment focuses on the end points of health effects that are relevant to the protection of public health at this site. Assessors look for relevant information to be used in identifying useful public health actions. Comprehensive information (such as body weight change) belongs to the text in the profile and the original papers in the reference list of the profile. In fact, 10 studies on body weight change were cited with the other 48 studies in the Significant Exposure Level table (Table 2-2) of the profile.
At the test dose level for technical toxaphene in this study, many other less sensitive endpoints of health effects might not be observed. They are not the major information for this public health assessment. Other relevant effects include the memory loss in the rats treated with a toxaphene component, Toxicant A (or Toxic Fraction A), at a dose level of 2 ug/kg/day. This end point is relevant to this health assessment. The body weight change is not an important endpoint for this assessment.
The Olson study is the only available study that found adverse behavior effects at this low dose level. The limited data suggested that developmental toxicity is a valid and relevant endpoint from which the assessors base the health assessment toxicity assessment. Until such time that a definitive examination of the developmental toxicity of toxaphene is performed, developmental toxicity must be considered one of the most sensitive endpoints of toxaphene toxicity and be used accordingly in the assessment.
*** Comment: The Olson study observed the "righting reflex" in exposed and control pups after an air drop on post natal days (PND) 14 and 15. However, the World Health Organization (WHO) has stated that the righting reflex in pups is unlikely to be developed until PND 17.5 (WHO 1984). Thus, the significance of results observed prior to full development of this reflex is questionable.
Response: The Olson paper was published in the peer-reviewed journal of Archives of Environmental Contamination and Toxicology. The study was included in the profile, which was published in 1996 after formal processes of review and comments. These comments included input from the chemical industry. The retarded righting ability was evident in the toxaphene-fed animals.
In addition, Table 6 of the 1984 WHO Environmental Health Criteria 30, eight kinds of reflex development in rats were listed, including free fall reflex (page 61). Table 6 simply listed PND 17.5 as the average appearance time for the righting reflex without making any comment on the application of this postnatal age. The WHO 1984 simply compiled the 17.5 PND from the reference of Smart JL and Dobbing J. 1971, Vulnerability of developing brain. II. Effects of early nutritional deprivation on reflex ontogeny and development behavior in the rat. Brain Res. 28:85-95.
On page 57 of the 1984 WHO book, the timing for assessing reflex and other behavioral functions were mentioned as follows. "Ideally, behavioral functions should be assessed as they are initiated, as well as during their development, maturation, and decline. . . Most behavioral testing is done on adolescents or young adult animals, with the exception of reflex and motor development tests, which are performed in early neonatal life."
*** Comment: The Olson study observed swimming ability in exposed and control pups at PND 7 through PND 17. However, the eyes of the pups, as is normal for rats, did not open until PND 14 and 15 in both the control and exposed groups. The WHO has stated that "tests of swimming ability are not conducted during the pre-weaning period" (i.e., before PND 21) (WHO, 1984). The test, therefore, seems to have assessed an acute stress response rather than any particular behavioral or developmental measure. By PND 14, as the eyes began to open, there was no difference in the swimming ability between the toxaphene-dosed and the control groups.
Response: The two sentences, in the entire paragraph on the swimming test of the WHO 1984 book, are copied below. "Physical strength is assessed using activity wheels or swimming ability. However, these tests are not practiced during the pre-weaning period." No reference was cited to give specific supporting information for this statement. No information was given to show how the physical strength was measured by the swimming ability. The WHO 1984 book cited neither the Olson study nor the methodology (Dorcey 1972 and Shapiro 1970) cited in Olson's study.
The WHO book used swimming ability to assess the "physical strength," while the Olson paper use swimming ability to judge " retarded maturation." There is no detailed description of swimming ability in the WHO book to demonstrate that the WHO book method is the same test as the swimming score described in the 1980 Olson study.
This comment did not provide a reason why tests should be done after PND 21. Is the purpose of this test for the development, maturation, or decline of swimming ability? The animals acquired full score (4.0) on PND 16. After PND 21, the opportunity for testing the development of swimming ability is gone.
III-1-A-c. Comment: Conclusion: The Assessment's conclusions regarding the Olson study are highly determinative because this study is the basis for the Assessment's recommendation that pregnant and lactating women reduce their fish consumption to avoid the adverse developmental effects of toxaphene ingestion. It is very important to insure consistency between ATSDR's toxicological profile and the Assessment and avoid presenting incorrect information about potential risks. Therefore, the Assessment should clarify that no anatomical developmental defects were noted as a result of exposures to a wide range of toxaphene concentrations. At the very least it should be noted that implications of the early results (before PND 14) of the Olson study are equivocal. Consequently, the Assessment should omit its recommendation that pregnant women avoid fish consumption because of concerns about toxaphene's potential developmental effects.
Response: As mentioned in the responses to the various points in this section, the Assessment is consistent with the profile. The structural and functional deficiencies are not necessarily detected at the same time. End points of adverse effects have different levels of sensitivity. Behavior deficits can precede pathological changes. The assessment recommended that pregnant and lactating women avoid consuming a large quantity of seafood harvested in this area. This precaution will protect the developing nervous system of the fetus and infant from the potential harmful effects of PCCs.
III-1-B-a. Comment: The Assessment correctly concludes, "there is no adequate human data to evaluate the carcinogenicity of toxaphene." (p. 25). However, the text continues with a seemingly confounding statement indicating the availability of two human studies that link toxaphene to anemia and chromosomal aberrations. The document does not include a specific citation for these studies. ATSDR's 1996 Toxicological Profile for Toxaphene describes only one human study indicating genotoxic or carcinogenic effects in humans. The Assessment should include a specific citation to allow interested readers to obtain further information.
Response: The conclusion of this paragraph (p. 25 in the Assessment) is that "toxaphene was classified as a probable human carcinogen." This conclusion was supported by the sufficient animal evidence and the "no or lack of" adequate human data. The lack of "adequate" human data was explained by the sentence that "Only two human studies were available and they link toxaphene to anemia and chromosomal aberation" in the Assessment. This is the standard language found in the National Toxicology Program20. The following two sentences in the NTP 2001 are omitted from the text in the Assessment: "Two cases of aplastic anemia were associated with dermal exposure to a toxaphene-lindane mixture. In another study, an increased frequency of chromosomal aberation was observed in the lymphocytes of workers exposed to toxaphene." The NTP cited IARC V.20 1979 for inadequate human data. There are two human studies on this topic in a recent review: Environ. Health Perspectives, 107: 115-144. These two articles are Brown LM et al. 1990, Cancer Res. 50: 6585-591; and Cantor KP et al.1992, Cancer Res. 52: 2447-455.
III-1-B-b. Comment: A review of the ATSDR Toxicological Profile and other existing literature suggests that the two studies to which the Assessment may be referring are Samosh 1974 and U.S. USEPA 1978. The Samosh study is the only human genotoxicity study cited in the Profile. Neither of these studies provides proof of a link between toxaphene and chromosomal aberrations. The source of the reference to anemic effects is unclear and should be included.
Response: The chromosomal aberrations citation was Samosh 1974 and the anemic effects was cited in the NTP, 2001, and in the review paper of de Gues et al., 1999, Environ. Health Perspectives 107:115-44.
*** Comment: The Samosh 1974 study reported an increase in the number of lymphocyte chromosomal aberrations as compared to unexposed individuals as a result of exposure to polychlorocamphene, a Russian toxaphene-like compound. However, polychlorocamphene as manufactured in Russia has not been reliably shown to be toxaphene (Hayes, 1982). Additionally, in the 1996 Toxicological Profile on Toxaphene, ATSDR states that "it is unclear whether the chromosomal aberrations observed in the lymphocytes of these women were directly attributable to the (toxaphene) exposure." (ATSDR, 1996, p15). Given these uncertainties, this study of a different compound is of limited relevance to a health assessment for toxaphene.
Response: Toxaphene is not a single compound. Instead, it is a mixture of polychlorinated camphenes, which has hundreds of possible compounds. The polychlorinated camphenes (or toxaphene) products were produced by many manufacturers in Europe, Russia, and USA. In recent years, many products with similar properties to toxaphene are referred to as toxaphene. While the toxaphene-like products might have slightly different compositions than the Hercules Incorporated product, they share many common components. Therefore, the toxicological data form Samosh 1974 is relevant to the protection of public health at this site.
The assessment mentioned the connection of toxaphene to the chromosomal aberrations and stated nothing about the cause-effect relationship. The uncertainty on the cause-effect relationship (p. 15) is consistent with the statement of "no adequate human data" in the Assessment.
*** Comment: The U.S. EPA 1978 study monitored workers who applied toxaphene to various crops for chromosomal aberrations in lymphocytes. The results indicated no increase in the incidence of aberrations in the lymphocytes of toxaphene-exposed individuals. Thus, any reference or allusion to this study as evidence of adverse genotoxic effect stemming from toxaphene exposure would be inaccurate.
Response: The assessment did not cite this Draft Report, as indicated in the reference list for the comments.
*** Comment: Author of a leading toxicology text, Dr. John Doull (M.D., Ph.D.) has stated that "there are no reports of adverse effects from chronic exposures to toxaphene. The remarkably low chronic toxicity of toxaphene is due in part to its rapid excretion from the human body . . . the threshold limit value (TLV) committee of the American Conference of Government and Industrial Hygenists (ACGIH) has not classified toxaphene as carcinogenic in either animals or humans..." (Doull, 1996).
Response: The statement in the text is too general to find out the dose levels or end points for the adverse effects. The profile listed many lowest-observed-adverse-effect-levels for chronic exposures in test animals. The classification of carcinogenicity for toxaphene was determined by four agencies: IARC (possibly carcinogenic to humans), USEPA (probably carcinogenic to humans), NTP, DHHS (resonably anticipated to be a carcinogen) and NIOSH (potential occupational carcinogen). The NIOSH 1992 classification provides adequate guidelines for occupational safety although the ACGIH does not establish any classification for toxaphene.
*** Comment: Dr. Frank Mitchell (D.O., M.P.H.), former Chief Medical Officer of the ATSDR, has also stated that "no adverse chronic human health outcomes have been shown at any level . . . toxaphene has not been detected in human adipose tissue, which is due to its very short half life and complete metabolism. There is thus no opportunity for the compound to be stored . . . no tissue deposits would be available to cause later potential effects . . . considering how long toxaphene has been used (well over 30 years) and the millions of people who had contact with the material, it is unlikely that any of these effects (chronic effects) will be seen in humans" (Mitchell, 1996).
Response: This comment was cut and pasted from various parts of Dr. Mitchell's letter (i.e., p.8 paragraph 3 and 6, p.9). The general comment did not provide the specific topics in Dr. Mitchell's letter. Dr. Mitchell reviewed human exposure to very low levels of toxaphene in soil samples (e.g., 1.6-1.8 ppm, 4.7, 4.8 ppm) from the area near the Hercules plant. This situation is very different than the contaminated food fish with up to 26 ppm of PCCs at the Terry Creek site. Of the 123 soil samples analyzed for toxaphene levels, 59 were below 0.17 ppm, 105 are below 0.6 ppm, and only two are above 7 ppm. The two samples from the same property (2810 Hopkins) had 10 and 14 ppm toxaphene. As for the toxaphene level in soils, Dr. Mitchell stated that very low levels of toxaphene are present in the area. Dr. Mitchell reviewed the soil data in that area, not the fish data at Terry Creek site.
Although PCC was not found in human fat, PCC congeners were reported in the fat of human breast milk in Scandinavian countries and Nicaragua (the Profile). Some PCC components, Parlar 26 and Parlar 50, were found in human serum at very low concentration in some U.S. cities. These PCC components at part per trillion levels were analyzed with advanced methodology.
III-1-B-c. Comment: It should also be noted that in a series of in vitro mutagenicity tests on four toxaphene congeners (a heptachlorobornane, an octachlorobornane, and two nonachlorobornanes) it was determined that these compounds were not mutagenic (Steinberg et al., 1998).
Response: The finding that four congeners were not mutagenic in this in vitro study does not necessarily mean that none of the other hundreds of congeners are not mutagenic. It should also be noted that at least a Hercules staff (Dr. FK Kinoshita) is among the list of authors for this Steiberg 1998 paper. This comment missed an important finding in the cited report: Toxaphene was mutagenic in the TA100 strain at concentrations of 2,500, 5,000, and 10,000 microgram /ml.
As cited in the profile(P. 73, 75, & 76), toxaphene has been found to be genotoxic with the Ames test for mutagenicity in the bacteria Salmonella tiphimurium.(Hooper, NK et al., 1979 , Science, 205:591-93; and Mortelmans, K et al., 1986, Envir. Mutagen (Supl. 7):1-119.) Toxaphene increases the frequency of sister-chromatid exchanges of chromosomes in a cultured cell line derived from human lymphoid cells (Sobti, RC, et al 1983, Arch Toxicol 52:221-31.)
III-1-B-d. Comment: Conclusion: Available information indicates that there are no reliable studies that demonstrate that toxaphene exposure in humans results in chromosomal aberrations.
Response: The Health assessment did not mention any "reliable studies" as human evidence to support the carcinogenicity classification of toxaphene. Instead, the Health assessment stated that "there were no adequate human data" to evaluate the carcinogenicity of toxaphene. This is the standard language of the NTP report. The carcinogenicity classification of toxaphene is almost entirely supported by sufficient animal evidence.
III-2. Comment: Analytical Methods in Evaluating Toxaphene in Seafood
III-2-A-a. Comment: The Assessment specifically states that people in the Terry Creek region are most likely to be exposed to toxaphene through seafood. This conclusion is not consistent with the data presented in the Assessment.
The method employed by USEPA and the Georgia Environmental Protection Division (Ga EPD) for fish tissue analyses for toxaphene was USEPA method 8080. The 8000 series methods taken from the USEPA's Solid Waste analysis method book are designed to determine the extent of contamination of specific chemicals in Superfund sites (or sites where analysis is required under RCRA/CERCLA). These methods have been time-tested and utilized by laboratories throughout the United States.
Method 8080 is the specific organochlorine multianalyte method of which toxaphene is a listed analyte. USEPA Methods 8080 and 8081 (the capillary GC column sister-method of 8080), when used with approved, validated sample extraction and cleanup methods, provide sufficient sample preparation so that experienced analysts can detect and quantify toxaphene. Additionally the use of the ECD detector provides the necessary sensitivity to determine if low levels of toxaphene are presen[*(2)].
Response: Method 8080, 3rd ed., did not list fish tissue as one of its matrices. The matrices listed in Table 2 are groundwater, soil, sludge, and non-water miscible waste. Method 8080, Organochlorine Pesticides and PCBs, was dated September 1986. Fish tissue was not mentioned in the August 14, 1997 protocol either. The "Procedures for the determination of Toxaphene" was intended to be used by USEPA Region IV and Hercules for determining toxaphene in soil, water, and sediment samples in the Brunswick, GA area. Even the recent (5/16/2000) protocol by F.J. Carlin, Jr., H.L. Revells, and Danny L. Reed did not list fish samples as a matrix. The soil, sediment, and water samples from Brunswick, GA are the intended matrices for the "Application of the Methods for the Determination of Toxaphene in Environmental Media." ATSDR is unaware of any document that provides justification for the use of this August 14, 1997 protocol for fish samples.
Moreover, new versions of the Method 8081 using modern capillary columns are now available. The official version of Method 8081A, December 1996, did not containe the language of the "last four peaks only" which was retained in the January 1995 version of the Method 8081A. The newer version of the Method 8081B, Revision 2, January 1998, did not retain the language either.
III-2-B. Comment: All results presented in the fish analyses discussed in the Assessment met the required recovery and precision criteria expressed in method 8080. Although lacking the inherent specificity of techniques such as SIM-MS, ECD (as used as part of Method 8080) still provides the highest sensitivity of any gas chromatographic detector.
Response: The precision of the four-peak-in-back-half method could have been high for the Method 8080 with packed column back in September 1986. At that time, the packed column in the gas chromatography produced only several peaks (peaks u, v, w, x, y) in the back half of the chromatogram of technical grade toxaphene. In the packed column chromatogram, the use of four-peak-in-back-half method was precise because four out of the five usable peaks in the back-half were included in the calculation. This method produced estimated concentrations comparable to that calculated from the total area method. The precision of the four-peak-in-back-half method, however, was lost when the high resolution capillary column was introduced into modern gas chromatography instruments.
The capillary column used in Method 8081 produced scores of peaks in the back-half of the standard toxaphene chromatogram. The chemist can pick one of the many four-peak sets. Some of the four-peak-in-back-half set can indicate toxaphene concentrations comparable to the concentration estimated from the total area method. The other sets indicate different concentrations. The four-peak-in-back-half method of the packed column time, which was carried over to the January 1995 version of the Method 8081A, was purged out of the text in the December 1996 version of Method 8081A, the approved, official USEPA method. The new Method 8081B of January 1998 did not carry the four-peak-in-back-half method either. Nevertheless, the outdated four-peak-in-back-half method is adapted as the sole method in the August 14, 1997 protocol intended to be used by the USEPA Region IV, GA EPD, and Hercules.
It is unclear why the total area method was abandoned and the four-peak-in-back-half method was adapted in the August 14, 1997 protocol. No evidence was provided to demonstrate that there is significant interference from DDT in front half of the GC chromatogram of toxaphene in the fish samples at this site. At 26 ppm toxaphene level in a mullet sample, only a fraction of 0.08 (2.1 ppm) of PCB and non-toxaphene organochlorine pesticides (OCPs) was found. The level of PCBs/OCPs was a fraction (0.11) of the toxaphene level (i.e., 5.3 ppm ) in a croaker sample at this site (Keith A. Maruya, June 16, 2000, Analysis of Toxaphene in fin- and shell fish from Terry Creek , Glynn County, GA .)
Regarding the comment to the Carline and Hoffman reference, ATSDR clarified this point in its Health Consultation of December 17, 2000. The following paragraph is the text from that health consultation.
The method of the latter-half-peaks underestimated the concentration of a toxaphene standard, i.e., Ultra Scientific standard in methanol, when its distribution of components shift to the front half of the GC chromatogram. Both the total area method and the latter seven peak method accurately calculated the toxaphene level when "neat" or regular toxaphene standard was analyzed against the Hercules standard. On the other hand, when the Ultra Scientific toxaphene standard is prepared in methanol, its distribution of components shifted to the early eluting GC peaks. As a result, the 5.00 ppm of toxaphene standard in methanol was accurately estimated as 5.19 ppm with total area method but the concentration was underestimated to be as low as 1.42 ppm when it was calculated with the method of latter seven peaks32.
The toxaphene peaks in killifish and sediment at this site was also shifted to the front half of the GC chromatogram as shown in the Phase I study by the Skidaway lab26.
III-2-C. Comment: However, ATSDR chose to verify the detection/quantitation of toxaphene in its samples by employing Dr. M.A. Saleh of Texas Southern University. Dr. Saleh's review of the chromatograms (using 12 peaks) found no quantitative differences greater than 1 ppm for any of the theoretical values (below LOQ) of toxaphene reported by the Ga EPD. Dr. Saleh's re-calculation of the results for the fish sample extracts validates the results obtained by GA EPD using Method 8080/8081. His calculated results do not change any conclusions drawn by GA EPD, and they corroborate the results of the original analysis. He also provided confirmation that, at low levels, even qualitative assessments are difficult. But he further confirmed that if the samples contained any toxaphene they contained no more than trace levels.
Response: There are PCC residues, the components of PCC and their degradation products, in historical fin fish samples and recent (1997) samples at this site. The residential killifish had an average concentration of 50 ppm PCC in 1970s. The concentrations of PCC up to 27 ppm still remained in the killifish in March 1997, as measured by the total area method 25. In finfish for human consumption, the three sampling efforts between 1995 and 1997, produced inconsistent data for the toxaphene levels. In 1995, Hercules found no toxaphene in the composite sample of spotted sea trout27. In the four fin fish samples of February-March 1997, USEPA found PCC in all samples in the range of 1.6 to 3.9 ppm25. In the fish samples of April 1997, GA EPD found no toxaphene in all of the 38 finfish samples, as mentioned in its draft release in 199828.
The USEPA found PCC with the total area method in chemical analysis. The method was cited in the report as the Laboratory Operations and Quality Control Manual (USEPA, Region IV, 1990) and the Method for Chemical Analysis of Water and Waste (USEPA 1983, EPA-600/4-79-020). Method 8080 or 8081 was not found in the text of this formal report of July 199725.
The GA EPD draft release of 1998 (no date) contained mainly the hard copies of chromatograms and some raw numbers and a memorandum. No formal format-like method, materials, and result sections are found. No text was found to mention whether the August 14, 1997 protocol was used for the analysis. No information on the procedures for homogenization, extraction, and clean up of the sample was found in the 5.5 inches of these loose pages in the draft. Among the materials is the chemist's memorandum of January 23, 1998, which explained the clutter in the chromatograms of the fin fish samples. The chemist actually found significant levels of unidentified components, clutter for lack of a better term, with the retention times within the chromatogram envelope of toxaphene. These components of PCC were ignored and the qualitative determination of no toxaphene detection was made for the following reason: "A match for toxaphene was called if the retention times and peak ratios in the latter half of the chromatogram matched an analytical standard of toxaphene. No hits were found in any of the fish samples." In short, the PCC residues are not toxaphene, only the technical grade toxaphene is bona fide toxaphene and it is not found in the fish samples.
Dr. Saleh reviewed gas chromatograhic data for the total toxaphene level in fish samples. He used the hard copies of the chromatograms for the task. He requested the electronic files of these chromatograms but they were unavailable from GA EPD at the time. The detailed information on sample preparation such as the homogenization, extraction (e.g., Soxlet extraction or sonication extraction), solvent used (e.g., acetone, ethylene chloride) were not available in the package. ATSDR also has no ideas about these details. The August 14, 1997 protocol was formally provided to ATSDR on April 20, 2000 by USEPA, far behind Dr. Saleh's review time of February 9, 1998.
By June 16, 2000, GA EPD again analyzed the April 1997 fish samples. Advanced methods with GC-ECD and GC-MS were used at the Skidaway lab to measure individual PCC congeners and their sum as total PCCs. PCCs were found in all 38 fin fish samples with PCC concentration up to 26 ppm. ATSDR will assess these data in a later health consultation.
III-2-D. Comment: In conclusion, both the original analyses using the EPA standard 8080 method, as well as the reanalysis by Dr. Saleh, show that, after numerous samplings of seafood in mid-1990s, no levels of toxaphene above 1 ppm have been found in seafood in Terry and Dupree Creeks[*(3)].
Response: The concentration of the toxaphene in fin fish can be zero, or 3.9, or 26 ppm depending on who did the testing, how the testing was done, and what definition of toxaphene or PCCs is used.
III-2-E. Comment: Despite the apparent validity of the data generated using Method 8080 and Dr. Saleh's reanalysis, the Assessment recommends that "due to interference from other chlorinated compounds in the fish samples and the uncertainty they cause in the toxaphene estimates, sensitive and specific methods, such as negative chemical ionization mass spectrometry (GC-NCIMS) are recommended for the evaluation of toxaphene in fish and sediment"(p28). Such systems are highly specialized, available to only a handful of laboratories, and require highly trained operators in order to produce data of integrity. The applicability, acceptability, or practicality of this method for analyzing toxaphene in fish is questionable given the following:
- The GC-NCI-MS method has not been standardized or validated.
- The criteria for defining the quantitation peaks for toxaphene by GC-NCI-MS have not been defined.
- It is unclear whether the methods described above would be acceptable to the USEPA and the State of Georgia for use in monitoring the samples from Brunswick, GA.
- It is unclear whether the levels of compounds detected in the fish samples analyzed by Dr. Saleh were high enough for quantitation by GC/MS.
- The reference in Appendix 4 of the Assessment (Dr. Saleh's report) to the isolation and identification of 20 isomers of toxaphene further highlights the inherent problems associated with the technique of using GC/MS. A prerequisite of proper quantitation using GC/MS, in particular GC-NCI-MS, is the availability of pure standards for determining response factors. Thus, if only 20 isomers are available (out of 200 that have been detected, according to Dr. Saleh), the task of quantitation of toxaphene samples by GC/MS is, at best, quite formidable. (Note: 4 of these 20 isomers have undergone in vitro mutagenicity analysis -see Section I(B) of these comments.)[*(4)]
Response: The August 14, 1997 protocol is not strictly the Method of 8080 with packed column and a few peaks in the back half of the toxaphene chromatogram (USEPA 1986 version of Method 8080). The protocol is not the December 1996 version of Method 8081A either. The method in this protocol could be one of the major sources for the inconsistent fish data generated in the period of 1995 to 1998.
ATSDR added a reference on the GC NCI-MS method in the Profile of 1996 in response to a 1995 public comment from a member of the chemical industry. In fact, the method of GC-ECD and GC-ECNI-MS was successfully used by GA EPD and the Skidaway lab in producing the data of total and individual PCC components in the April 1997 fish samples at this site. These quality data were prepared on June 19, 2000 by Dr. Keith Maruya of the Skidway lab. This manuscript was subsequently published in peer-reviewed, open literature in September 200131.
III-2-F. Comment: The Assessment also recommends that "additional seafood sampling is needed to help assure residents that fish caught in this area are safe." The basis for this recommendation is unclear given the following:
*** The Assessment discusses three separate fish analyses over the last four
*** Between 1995 and 1997 there has been sampling and analysis of spotted sea trout, consumer fish (unspecified), mullet, sea trout and drum, and blue crab.
*** The data in Assessment Table 4B (p45) demonstrate that toxaphene was not detected in shellfish in April 1997. (Shellfish, however, are already banned for eating due to contamination from raw sewage.)
*** In the April 1997 study, 59 composites of finfish were taken. GA EPD concluded that these samples contained no toxaphene. Dr. Saleh concluded that they contained less than 1 ppm toxaphene.
Response: The fish data in 1995-1997 samples are inconsistent. The methods used greatly influence the data produced. In the only finfish sample of 1995, the toxaphene concentration was reported as zero. Then all of the four finfish samples of February/March 1997 were found to contain PCCs at concentrations from 1.6 to 3.9 ppm. One month later, all of the 38 or so finfish samples collected in April 1997 were found again to contain no toxaphene. These 38 finfish samples were recently analyzed further with an advanced methodology and all fish samples were found to contain PCCs up to 26 ppm in the bottom feeders.
Additional samplings will address the influence on PCC concentrations by seasonal change, and residence time at the site for the migrating fish species. The RI/FS sampling will assess the effect of the sediment removal action on the PCC residues in the fish. Additional sampling will offer an opportunity to use the new methodology to assure the data quality for use in protecting public health. Actually, GA EPD collected a new batch of fish samples in August 2001 and the chemical analysis are currently under way at the Skidaway lab.
Even the shellfish had detectable PCC as shown in the June 2000 data sets for the April 1997 samples. The actual numerical data are helpful for the assessment of the human exposure. The whole grass shrimp has 1.2 ppm of PCC even though the tail of the food shrimp has very low toxaphene levels. When data on the other parts of tissues such as eggs became available, the public may be advised to avoid eating some part of the crabs or shrimp.
III-2-G. Comment: Conclusion:The existing sampling data are sufficient (in both quality and frequency) to demonstrate that minimal (less than 1 ppm and zero according to GA EPD) toxaphene is present in Terry Creek finfish. The analytical methods with which these data were collected are in full accordance with USEPA-approved, standardized, and validated methods for the quantitation of toxaphene. The Assessment's conclusion that further analyses and sampling are required to determine whether toxaphene is present in fish at hazardous levels is unwarranted and should be removed from the Assessment.
Response: There were some details in the chemical analysis for the February-March, 1997 sampling. But the April 1997 samples were not well described in the draft release by GA EPD. If the August 14, 1997 protocol was used in the chemical analysis, the protocol is neither Method 8080 of 1986 nor the Method 8081A of December 1996 version. The fact is that the modified versions did not produce consistent data for the three batches of fish samples in the mid-1990s. If a capillary column has to be used, the approved Method 8081A has to be followed closely. Reliable data are important for the protection of public health, the PCC concentrations in the April 1997 fish should not be assumed as zero or less than 1 ppm. Recent data indicated that the concentrations can reach 26 ppm.
III-3. Comment: Hazard Assessment for Toxaphene
III-3-A. Comment: The Assessment concludes that people are most likely to be exposed to toxaphene near the Terry Creek Dredge Spoil Areas/Hercules Outfall Site through consumption of seafood. Although the Assessment's conclusions and methodology for assessing risks related to fish ingestion are equivocal (as described below), support for the following conclusions in the Assessment have a sound technical basis.
*** "With respect to the Terry Creek Dredge Spoil Area/Hercules
Outfall site, the air pathway is not likely to be a problem" (p22)
*** "The contaminated areas are sufficiently vegetated and the main dredge area is remote . . . it is unlikely that low level exposure to toxaphene in soils would cause harm to humans." (p22)
*** Exposure through contact or ingestion of water (in wells or through swimming) is expected to be low because of "the low solubility of toxaphene in water and its strong adsorption by soil/sediment particles" (p23). "Currently, surface water is not thought to pose a hazard for the occasional swimmer or water skier" (p.23).
Response: The three conclusions in these comments were not exactly what was written in the assessment. All the caveats in the assessment for these three statements were omitted in these comments. To be accurate, the top of the dike of the main dredge spoil area was bare, i.e., without vegetation, as shown in Figure 4.
The potential inhalation exposure and potential exposure during dredging of contaminated sediment were both mentioned in the assessment (p.22).Yet the potential inhalation exposure mentioned in the assessment (p. 22) was omitted in this comment. ATSDR did not have detailed, specific information on the airborne toxaphene concentrations at this site. In the daily preliminary reports during the removal action, USEPA stated that no detectable toxaphene was found in the air but provided no information on the detection limit or other details. ATSDR learned from a carbon copy of an USEPA memorandum in August 2, 2001 that the detection limit was 2.5 - 6.6 ug/m3; no supporting information for this value was provided. ATSDR will continue to assess the potential risk on various pathways when new data become available.
III-3-B. Comment: The Assessment raises the possibility that "leafy vegetables in residential gardens may be contaminated" (p18). However, this speculation is highly unlikely given the following:
*** Plant uptake studies conducted during toxaphene's initial registration indicate that toxaphene is not taken up by leafy vegetables or produce.
*** The Assessment appears to base this hypothesis on the condition that "the incidence of cordgrass (contamination) is applicable to leafy vegetables in contaminated garden soils" (p18). However, even if there were a relationship between cordgrass data and leafy vegetables, the cordgrass data cited by ATSDR is not current and unlikely to be relevant to current conditions.
*** Former FDA toxaphene residue tolerances for leafy vegetables were 7 ppm for most fruits and vegetables (40 CFR 180.138 [1993, 1986],58 FR 46087 , 36 FR 22556 [November 1971]), Toxaphene Status Report, USEPA, 1971)[*(5)]. Given the evidence indicating that toxaphene is not taken up by fruits and vegetables in any significant quantities, it is unlikely that these tolerance levels would be exceeded.
Conclusion regarding vegetables: Available data on toxaphene levels in soils and toxaphene uptake by fruits and vegetables (evidence indicates this is not likely) indicate that toxaphene exposure as a result of ingestion of garden vegetables is unlikely. The Assessment should be modified to reflect the fact that consumption of garden vegetables is unlikely to result in exposure to toxaphene at any level of concern.
Response: The potential accumulation of toxaphene in leaves and fruit (not the roots) in the green house testing can differ from the accumulation under the insitu exposure to the PCC in sandy sediments in the windy seashore. ATSDR recommended the testing of the garden produce only when the residents are concerned about their produce. The insitu data from the cord grass cannot be ignored, even though the greenhouse testing showed different results.
In the uptake testing performed at USEPA-Corvallis lab in 1980s, Craig McFarlane found that uptake was influenced by many factors such as air movement, hairs on leaf surface, and surface area of the plant. In addition, soil texture (clay vs. sand) and organic matter content affected bioavailability of toxaphene in soils. There is little organic matter (and clay minerals) in the sandy beach to bind PCC on soil particles. Therefore, PCC can remain in the water phase where it can be absorbed and transported into plants under certain conditions. Toxaphene (tagged with Cl-36) in seawater was picked up by smooth cord grass in an environmental chamber. The toxaphene was distributed mainly in the roots, and to a lesser extent, in the stems and leaves5. In the marsh, toxaphene was accumulated in the leaves. Windy conditions could lead to increased transpiration which could in turn increase water uptake. Many local gardens have sandy soils and are exposed to strong winds.
The tolerance of toxaphene residues was 0.1-7.0 ppm for a peeled banana at the low end, and leafy produce at the other end of the scale. In 1993 the USEPA revoked the tolerance for toxaphene, and no importation of toxaphene-contaminated agricultural produce is allowed into the U.S.
The Assessment recommends testing if a resident has a concern about produce safety.
III-3-C. Comment: The Assessment concludes that the expected daily dose of toxaphene (resulting from ingestion of fish) will exceed the ATSDR screening level. This conclusion hinges on several key variables, including the following:
A. The expected toxaphene concentration in fish tissue,
B. The expected fish consumption rate, and
C. The selection of a screening level.
Although the Assessment provides the values selected for each of these variables, it includes no information on the basis for the selection or how variations in these factors could affect the outcome of the assessment. In the interest of providing a more informative and complete public health document, this information should be included.
Response: The assessment is based on the data available at the time of its preparation in December16, 1998. Quality data such as fish data became available after that time. ATSDR will use the recent data in the update version of this public health assessment.
III-3-C-a. Predicted Toxaphene Concentration in Fish Tissue
The Assessment estimates that the daily dose of toxaphene to a fisherman in Terry Creek would be 0.002 mg/kg/day as result of fish consumption. This estimate is based on an assumed 3.9 ppm toxaphene in edible fish, which appears to be the highest single value in any recent analysis of edible fish. The justification for use of this single data point is questionable given the following:
*** Comment: In March 1997 sampling by GA EPD and USEPA, 3.9 ppm of "potential toxaphene constituents"--not toxaphene--were identified in one sample. Table 4A (p44) identifies these concentrations as "JN" or "not enough proof to positively identify the compound as toxaphene."
Response: This was the only quantitative number (maximum level in four finfish) that was available when this health assessment was prepared. The toxicity of the PCC mixture is distributed over various fractions of the PCC residues. The PCC residues, other than the technical toxaphene can be harmful to human health. Therefore, toxaphene, its weathered residues, and the toxic components are all relevant to public health.
Recent data on the 1997 fish samples indicated both the food fish and killifish have PCC concentrations up to 26 or 27 ppm, respectively, using a different analytical method for quantification.
*** Comment: In April 1997 GA EPD concluded, "toxaphene was not detected" in fish samples. Because of an apparent concern about the significance of the non-detects and what it termed "potential toxaphene constituents" found in recent surveys, ATSDR contracted with Dr. Saleh to conduct a reevaluation of the April 1997 data "to estimate the total toxaphene concentrations." Saleh "concluded that toxaphene is present at concentrations below 1 ppm" in all of the April 1997 fish samples [emphasis added] (p17).
Response: Dr. Saleh conducted the review of hard copies of the PC chromatograms. The GA EPD's draft release reviewed is not a complete report. As mentioned earlier, the draft release provided neither the information on the use of the August 14, 1997 protocol nor a description of sample extraction or clean up.
In June 16, 2000, the samples were analyzed using a better methodology, and the concentration of PCC was found to be up to 26 ppm. ATSDR will assess the new data later in the updated health consultation.
*** Comment: GA EPD's conclusion that the constituents found at the method detection limits of 0.1 - 1.0 ppm could not be determined to be toxaphene and that "clutter" cannot be defined as toxaphene for risk assessment purposes should be supported. Even if, however, ATSDR were to base its risk assessment on Saleh's findings, his study suggests a maximum concentration of 1 ppm toxaphene in fish, and possibly none at all. Saleh concluded that 45 of the 57 samples (79%) analyzed by Georgia's Environmental Protection Division contained less than 0.5 ppm toxaphene. The 57 samples included both fish and shellfish, all of the shellfish samples (19) were <0.2 ppm, 26/38 finfish samples were <0.5 ppm in Saleh's analysis.
Response: The issue of the clutter was resolved in the reanalyses done with the concentrations of individual congeners determined by the Skidaway lab and the GA EPD. The PCC concentration is neither below detection limits nor 1 ppm. It was found up to 26 ppm.
*** Comment: Saleh's analyses suggest that a highly conservative assessment should be based on a maximum fish tissue concentration of perhaps 1 ppm, but certainly not 3.9 ppm. Based on a 1 ppm fish tissue concentration, the predicted dose would equal 0.0004 mg/kg/day (less than 1/2 of ATSDR's proposed screening level). Based on a 0.1 ppm fish tissue concentration, the predicted dose would equal 0.00004 mg/kg/day (approximately 95% less than the ATSDR screening level.)
Response: Under the worst case conditions, the maximum fish concentration could be as high as 26 ppm. ATSDR will adjust the predicted doses when the update of this document is prepared.
*** Comment: The Assessment should be modified to utilize the most reliable and accurate data available. These data (from Ga EPD and Saleh) suggest minimal (less than 1 ppm), if any, toxaphene in the resident fish population. Even at the maximum hypothesized tissue concentration (1 ppm) the expected intake from fish consumption is predicted to be far below the proposed screening level. Additionally a concentration of 1 ppm, or 3.9 ppm for that matter, is below the previous FDA action level of 5 ppm for toxaphene in fish (55 FR 14359, FDA, 1986). FDA withdrew this action level because, after the cancellation of the registered uses of toxaphene, the agency found only an occasional low level (<0.02 ppm) of toxaphene in fish (58 FR 60859)[*(6)].
Response: The available quantitative data generated in the 1995-1998 period is used in this Health assessment. In the upcoming health consultation, ATSDR will assess the available new data such as those of June 16, 2000, generated by the GA EPD and the Skidaway lab.
*** Comment: Conclusion: The Assessment should be modified to indicate that even at the maximum hypothesized tissue concentration (1 ppm), the expected intake from fish consumption is predicted to be far below the proposed screening level.
Response: The maximum concentration was 3.9 ppm at the end of 1998 when this document was prepared for public comment. The actual concentration could be 26 ppm. The fish data collected in other seasons will also be assessed.
III-3-C-b. Comment: Expected Fish Consumption Rate
*** Comment: The hazard assessment is based on the fish consumption rate of a hypothetical fisherman of 0.028 kg fish/day. This rate is greater than four times the average consumption rate (0.0065 kg/day) of "fish and shellfish from estuarine and fresh waters by the general U.S. population." (USEPA, 1993, p 5-6). USEPA indicates a consumption rate of 0.014 kg/day for the average consumption rate of fish and shellfish from marine, estuarine, and fresh waters by the general U.S. population. "However, this Assessment is intended to focus on the Terry Creek area, a non-marine source. Clearly, if consumers ingest fish from a specific source such as Terry Creek at a lower rate, they are less likely to exceed the hazard level.
Response: The consumption rate of 30 grams per day is based on the rate of one fish meal (210 gram) a week. During the Memorial Day Holiday (5/28/2001), one ATSDR staff (S.C.Tsai) made an unofficial trip to the site. He talked to the anglers near the old bridge and the pier in the park. One angler on the pier said that they go fishing every day, and fish once or twice a week in the area near the site. The pail contained several croakers and a whiting. So for some, the level of 30 grams/day could underestimate the fish consumption rate. The intake rates of recreational marine fish for sport anglers ranged from 24.2 to 70.6 g/person /day pending on ethic groups (USEPA 1996, Exposure Factors Handbook.)
*** Comment: Conclusion: The sampling data presented in the Assessment indicate that even fishers (consuming fish at a rate of 0.028 kg fish/day) would not exceed a level of concern for toxaphene. Further, the Assessment should include language clarifying that residents who consume fish from Terry Creek at a lower rate than fishers would be expected to have an even greater margin of safety with respect to toxaphene.
Response: No data are available for the actual consumption rate. The standard assumptions are used to identify the potential exposure for the protection of public health.
III-3-C-c. Comment: Selection of a Screening Level
*** Comment: Because the Assessment relies upon exceedance of a screening level to determine the possibility of health concerns, the selection of this level is highly determinative. As pointed out previously, at present, the document contains no explanation of the origin of this screening level (0.001 mg/kg/day) or its underlying assumptions. It is recommended that this information be included in the document.
Response: This guideline is ATSDR's intermediate oral Minimum Risk Level (MRLi) as documented in the profile1. It is listed on page 22 in the table 2-2 within the reference of Chu et al., 1986. More explanation is provided on page 29, footnote c of the same table. Detailed descriptions can be found on Appendix A, Minimal Risk Level Worksheet, Pages A-5 and A-6 of the profile. The uncertainty factor of 100 and a modifying factor of 3 were used in the derivation of this MRLi.
*** Comment: A review of existing USEPA documents suggests that the screening value is derived from USEPA's Guidance for Assessing Chemical Contaminant Data For Use in Fish Advisories (USEPA, 1993). ATSDR's value appears to be based on the screening value for toxaphene in fish of 3 ppm (USEPA, 1993) as follows:
- ATSDR Screening Level (mg/kg/day) = (Fish tissue concentration * Consumption Rate)/ Body Weight
- ATSDR Screening Level = (3 ppm toxaphene in fish * 0.028 kg fish/day)/70 kg consumer
- ATSDR Screening Level = 0.001 mg/kg/day
Response: ATSDR's screening level is the intermediate oral MRL of ATSDR. It is not derived from any of USEPA's guidance.
*** Comment: Although the 3 ppm "screening value" is a published value, it is important to note that it is based on an assumed reference dose for toxaphene of 2.5*10-4 mg/kg/day. This reference dose was previously listed by USEPA's Office of Pesticide Programs but "was never submitted for verification" (USEPA, 1993, p 5-11). There is currently no oral RfD listed for toxaphene as part of USEPA's Integrated Risk Information System (IRIS) database. The Assessment should be modified to acknowledge that the screening level is based on an assumed reference dose and that no accepted value currently exists.
The IRIS database for toxaphene is currently under review (63 FR 75). Based upon a review of the existing toxicological database, Hercules has submitted an alternative RfD of 7*10-4 mg/kg/day to OSWER/USEPA under a pilot program for revising IRIS records. If this alternative RfD were used in USEPA's calculation of the screening value, the value would be raised from 3 ppm to 7.5 ppm toxaphene in fish. Further, if ATSDR were to use a screening value of 7.5 ppm, as opposed to 3 ppm, the hazard level would be raised from an allowable 0.001 mg/kg/day to 0.003 mg/kg/day. At this screening level and a fish tissue concentration of 3.9 ppm toxaphene, even fishers consuming 0.028 kg fish/day would not be expected to exceed the screening level.
Response: ATSDR's intermediate oral MRL level is related to neither USEPA's proposed RfD of 1993 nor Hercules proposed alternative RfD. No RfD value was used in this Assessment.
*** Comment: It is also important to note that page 9 of the Assessment, states that, "ATSDR's screening level for toxaphene is based on non-cancer effects in animal testings. . . ATSDR added safety margins . . . to protect the sensitive human population, such as infants and pregnant women." However, as discussed previously, toxaphene has not caused adverse effects in developing or very young animals and the reference dose was deemed by USEPA to be protective of sensitive endpoints and populations. Based on this information, it is expected that the reference dose upon which the Assessment screening level is based would be protective of sensitive populations such as fishers or pregnant or lactating women.
Response: The health guideline used in this Assessment is the intermediate oral MRL as documented in the profile1. ATSDR has an intermediate oral MRL of 0.001 mg/kg/day for exposure of 15-364 days. ATSDR also has an acute oral MRL of 0.005 mg/kg/day for exposure duration of 14 days or less. Nevertheless, ATSDR has not derived a chronic oral MRL (the counterpart of USEPA's RfD) because a suitable chronic NOAEL or LOAEL value could not be identified in the available literature (see page 66 of the Profile.) The intermediate MRL of 0.001 mg/kg/day for intermediate exposure is not expected to be overprotective because it is derived from intermediate duration studies and not chronic studies. Exposures at Terry Creek are chronic. If ATSDR were to adapt USEPA's methodology of extrapolating from intermediate to chronic studies, the MRL would be 0.0001 mg/kg/day (only 2.5 times less than USEPA's proposed RfD and 20 times below the potential exposure dose at this site.)
ATSDR does not develop any fish screening levels. Total oral doses from food, water, and soil ingestion were pooled and the combined doses at the site were compared to the oral MRL. The site doses are also compared to the cancer risk guideline for assessing the potential cancer hazard.
The fishers mentioned in this comment are not a sensitive population. The sensitive populations include young, the old, the pregnant, and the sick.
*** Comment: Conclusion: Based on the discussion above (and the discussion of toxaphene's toxicology in Section I of these comments), the basis for the Assessment's recommendation that fishermen and pregnant and lactating women should avoid consumption of fish in Terry Creek is not justified and should be omitted.
Response: ATSDR recommends that pregnant and lactating women avoid exposure to contaminated fish.
III-3-C-d. Comment: Cancer Assessment
*** Comment: Ga EPD concluded that there are no detectable levels of toxaphene in the Terry Creek fish sampled in April 1997 and thus the cancer risk attributable to toxaphene consumption from fish ingestion would be negligible. However, as described in III(A) of these comments, the Assessment's assumptions about the expected concentration of toxaphene in the fish greatly impact the dose and hazard predictions. These assumptions also affect the Assessment's predictions of potential cancer risks associated with fish consumption as follows.
Response: The fish data of 3.9 ppm remains in the Assessment. The non-detectable data from GA EPD for the April 1997 fish samples has been replaced with more specific, accurate, numerical data. ATSDR will assess the new data later in a upcoming health consultation.
- The Assessment calculates a cancer risk of 2 in 1,000 based on a fish concentration of 3.9 ppm and ingestion rate of 0.028 kg fish/day.
- If the Assessment was based on the highly conservative assumption of 1 ppm toxaphene in fish (rather than 3.9 ppm), and a cancer potency factor[*(7)]of 1.1 /(mg/kg/day), the predicted cancer risk would be decreased to 4 in 10,000.
Response: The April 1997 fish samples, assumed to have non-detectable (or >1ppm) level of toxaphene, were analyzed again using sensitive and specific methodology. PCC levels, up to 26 ppm, were found in all finfish samples and all shellfish samples.
- An expert panel, chaired by Dr. Jay Goodman with participation from Drs. David Brusick, William Busey, Samuel Cohen, and Thomas Starr, convened on December 11, 1997 to re-evaluate the cancer potency factor for toxaphene. The panel recommended the potency factor be reduced from the current 1.1/(mg/kg/day) to 0.11/(mg/kg/day). At an upper-bound estimate fish concentration of 1 ppm toxaphene, and a cancer potency factor of 0.11/(mg/kg/day),(this slope factor has been submitted to USEPA/OSWER based on new information), the predicted cancer risk would drop to 4 in 100,000 (Peer Review Report 1998).
- In the more likely scenario that the toxaphene fish tissue concentration is less than or equal to 0.1 ppm, and assuming a cancer potency factor of 0.11/(mg/kg/day), the predicted cancer risk would drop to 4 in 1,000,000 if 0.028 kg of fish were consumed daily.
Response: The current cancer potency factor in USEPA's IRIS database is 1.1 /(mg/kg/day). The most current data indicated that PCC levels up to 26 ppm were found in the April 1997 finfish samples.
*** Comment: While there are no published studies describing carcinogenicity of toxaphene in humans, there is a long history of human exposure to toxaphene through occupational and agricultural exposures with no documented long term health effects. Dr. John Doull and Dr. Frank Mitchell have submitted formal reports which concur with the above statements. A summary of Dr. Doull's and Dr. Mitchell's comments are included in these comments in Section I(B).
Response: See the last two paragraphs in Comment III-1-B-b for responses about Dr. Mitchell's evaluation.
Comment: Suggested modifications to the Assessment (p24 and 25) to address the issues discussed are included in Part Two of these comments.
Conclusion: The Assessment's predictions regarding cancer frequency resulting from consumption of fish containing 3.9 ppm toxaphene is not based on the best available data. The data indicate that there is less than 1 ppm toxaphene (no toxaphene according to GA EPD) in Terry Creek fish. The cancer risk assessment analysis should be revised to reflect these data. The cancer risk assessment should also acknowledge the proposed revision to the cancer potency factor.
Response: The cancer potency factor of 1.1/(mg/kg/day) is the current guideline in USEPA's IRIS database. The 3.9 ppm was taken from the only quantitative data set in recent sampling efforts (1995-1998) when this document was published for public comment in December 1998. The GA EPD's former data of >1 or 0 ppm was replaced with a set of more accurate data. PCC in finfish was up to 26 ppm in the report of June 16, 2000 by GA EPD and the Skidaway lab. ATSDR will assess the new data in an upcoming health consultation.
III-3-C-e. Comment: Respiratory Concerns
The major adverse health effect cited in the Assessment, as reported by community members, is "asthma or respiratory problems" (p 7). However, the document also states that "most residents did not think their respiratory problems were related to the contaminated dredge spoil"(p 7). The Assessment should be clarified to indicate that, as written, it does not contain sufficient data to indicate whether the prevalence of respiratory illnesses near the site is statistically significant or whether the incidence of respiratory problems noted is above what would normally be expected in a Southeast-central locale. For example, the Assessment's statement that "approximately 26% of local residents reported that they or their family members have asthma" (p 28) has not been validated or demonstrated to be statistically significant (note this quotation refers to 7 members of the sampled population -- 26% of 27 people.) It should also be noted that there are more than 4000 residents living within a 1- mile radius of the site (p6).
Response: This is simply a record of the community concerns.
III-4. Comment: Conclusions about Hazard Level at Terry Creek Site
*** Comment: Given the repeated sampling of toxaphene fish tissue concentrations, the sophisticated reanalyses of these data by Dr. Saleh, and the lack of evidence indicating pathways likely to result in significant exposure to toxaphene, ATSDR should consider the Terry Creek site to pose No Apparent Public Health Hazard with respect to the site-specific contaminant, toxaphene. However, it is recognized that information about potential toxaphene concentrations in the community well located on Terry Creek Drive is currently incomplete and further testing is underway.
Response: The major toxaphene concern is still the contaminated fish and biota. In recent years, the fish data is inconsistent even in the same batch of fish samples analyzed using different methodologies. With the PCC concentrations up to 26 ppm in food fish and 27 ppm in rough fish at this site, the focus should be placed on the food fish, not the well water. GA DEP and the Skidaway lab are analyzing a new batch of fish samples collected in August 2001.
*** Comment: The data presented in the Assessment indicate that the potential for adverse health effects stemming from exposure to toxaphene (from fish ingestion or any other pathway) is negligible. The Recommendations section of the Assessment should be modified to reflect these data. Neverthless, on the basis that community well sampling data has yet to be completed, a classification of "indeterminate health hazard" for the Terry Creek Site is supportable. Once these data are available, and demonstrate no need for concern, the classification could be modified to "no apparent health hazard" with respect to toxaphene and the Terry Creek Site.
Response: The PCC level in fish is still the focus. The fish samples of April 1997 had PCC concentrations up to 26 ppm. GA DEP and the Skidaway lab are analyzing a new batch of fish samples recently collected in different season. Using one analytical method in the RI/FS should provide quality data for assessing the status of fish contamination after the removal action. Further, the toxicity of the PCC mixture in fish is another focus for further clarification.
III-5. Comment: Non-Site Related Information and Historical
There is particular concern about the potential for the reader to be confused between public health concerns stemming from the Terry Creek Site and those caused by possible contamination from other sites around the Brunswick area. It is very important that this distinction be clearly identified throughout the Assessment. Specific recommendations in this regard are included as Part Two of these comments.
It is also important to distinguish historical data from current data. Discharges and levels of contamination measured many years ago are not relevant to any public health risks posed by the Terry Creek Site today. The Assessment should clarify that historical data are not indicative of current conditions. Specific recommendations in this regard are included in Part Two of these comments.
Response: Historical data provided a starting point for contamination levels and the basis for the extent of contamination decrease over the decades. These historical data are especially important in view of the inconsistency among fish data reported during 1995-1998. Contaminants from unknown sources can come from any potential sources including . Some of the contaminants might be used as solvents in the processes for toxaphene production.
III-6. Comment: Factual Inaccuracies in the Assessment
There are miscellaneous factual inaccuracies in the Assessment that should be corrected. These inaccuracies relate to the following issues:
*** Comment: Nomenclature used in the Assessment to describe compounds that are not known to be toxaphene. Throughout the Assessment, compounds other than toxaphene are referred to by a variety of terms including "components of toxaphene," "organic constituents similar to those found in toxaphene," etc., and sometimes even referred to as toxaphene itself. Unless a compound is definitely characterized as toxaphene by a validated laboratory methodology, the Assessment should refer to these compounds as "organochlorine compounds." There is insufficient scientific basis to speculate that these unidentified compounds (particularly those identified in fish) are now, or once were, technical toxaphene. At most, reference should only be made to the fact that ATSDR believes some of these organochlorine compounds could be toxaphene breakdown products.
Response: In the United States, the definition of "technical toxaphene" was patterned after the Hercules Incorporated product, marketed under the trademark name of "Toxaphene." In recent years, Hercules Incorporated has essentially let the name of toxaphene lapse into the public domain so that many products with similar properties are referred to as toxaphene (see the Profile). As much as possible in this document, however, the toxaphene will be referred to as polychlorinated camphenes (PCC).
In the June 2000 report by Skidaway lab and the GA EPD, 23 individual toxaphene congeners in the April 1997 fish samples were determined, and total toxaphene concentration up to 26 ppm were reported.
In the profile (1), Toxaphene has many synonyms: campheclor, chlorinated camphene, polychlorocamphene, chlorocamphene, octachlorocamphene, toxafeen (Dutch), polychlorocanthene (former USSR). Organochlorine compounds, the suggested name in this comment, is not among the synonyms.
The suggested name of "organochlorine compounds" represents a very large group of chemicals including solvents (e.g., chloroform, PCE), insecticides (e.g., DDT), herbicides (e.g., 2,4-D), PCBs, and Dioxins. The smaller group of organochlorine insecticides include three major families: DDT-analogs, BHC, and cyclodiene-compounds. Toxaphene is related to cyclodiene-compounds, which include chlordane, heptachlor, aldrin, dieldrin, and endrin.
- References to Hercules operations and practices.
- References to toxaphene composition, usage, and properties.
- References to reported health effects or potential exposures or both in the Terry Creek area.
Response: These three categories of comments provide no specific issue to be addressed.
III-7. Comment: Recommendations in the Assessment
Based on the comments provided above, some modification of the Assessment's recommendations could be warranted (specific textual suggestions are included as Part 2 of these comments). These modifications fall into five general categories:
*** Comment: Recommendations that imply or state there is insufficient data on toxaphene concentrations in fish to conduct a hazard assessment should be modified to state that sufficient data are available and indicate that hazard levels are not expected to be exceeded (Recommendations 1 and 3). Similarly, the recommendations should reflect the fact that sampling conducted by Method 8080 is an approved and valid procedure and that alternatives such as GC-NCI-MS are not yet validated or approved for toxaphene (Recommendation 2).
Response: The inconsistent nature of the toxaphene levels in the fish samples of the 1990s make the first result of April 1997 data (28) insufficient for use in this health assessment. The same batch of fish samples, when re-analyzed, had much higher toxaphene levels (26 ppm) than the non-detectable or 1 ppm reported in the first testing. The August 14, 1997 protocol is not in fact Method 8080 of September 1986 with packed column. The approved, updated method is actually the December 1996 version of the Method 8081A.
The GC-ECD/ GC-NCI-MS has been successfully used in the analysis of the April 1997 fish samples from this site and quality data were reported in June 2000 by the Skidaway lab. The GC-MS method is also listed in the profile.
*** Comment: Recommendations that pertain to areas outside of the Terry Creek Dredge Spoil Site or to contaminants that are not known to be linked to the site should clearly distinguish the relationship (if any) between the location or contaminant in the recommendation and the Terry Creek site (Recommendations 4, 9, and 11).
Response: ATSDR will indicate that the recommendation for #9 and #11 are for general areas of Brunswick.
*** Comment: Recommendations based on inaccurate factual references should be omitted (Recommendations 5 and 8).
Response: ATSDR deleted the recommendation # 8 to test the soil borne toxaphene in the yards that might receive the Hercules compost .
*** Comment: Recommendations relating to the need for sampling of yards--out of concern for contamination of produce by toxaphene--should be re-evaluated in light of the low potential for toxaphene to be taken up by produce (Recommendation 6).
Response: The cord grass data at the windy, sandy beach areas cannot be ignored, even though the root uptake is unlikely under greenhouse conditions. Testing, however, is only recommended if a resident is concerned.
*** Comment: The recommendation that the site be considered an indeterminate health hazard should be modified to note that only the community well data at the site is indeterminate. Once the well data collection is completed (and the results are acceptable) the site should be characterized as no apparent health hazard with respect to toxaphene (Recommendation 7).
Response: The focus is still on the fish contamination. The fish data of April 1997 fish samples were analyzed twice. The data quality of GA EPD's first chemical analysis in 199828 was not as good as the second analysis carried out at the Skidaway lab (Maruya K. Analysis of toxaphene in fin- and shellfish from Terry/Dupree Creek, Glynn County, Georgia. Prepared for GA EPD. 2000). The new data from the Skidaway lab was published in peer-reviewed, open literature31.
References for the Comments in Set # 3:
[ATSDR] Agency for Toxic Substances and Disease Registry. 1996. Toxicological profile for toxaphene (Update). Atlanta: US Department of Health and Human Services.
[ATSDR] Agency for Toxic Substances and Disease Registry. 1999. Public health assessment for Terry Creek dredge spoil area. Atlanta: US Department of Health and Human Services.
Doull J.1996. Letter to Wallace E. Harrell of Gilbert, Harrell, Gilbert, Summerford and Martin, and J. Kevin Buster of King & Spaulding regarding response to request to review material related to potential exposures to toxaphene. January 11th.
[USEPA] US Environmental Protection Agency. 1971. Toxaphene status report. Washington, D.C. p 104-6.
[USEPA] US Environmental Protection Agency. 1978. Occupational exposure to toxaphene. Human Health Effects Monitoring Branch, Benefits, and Field Studies Division, OPP, OTS. Draft report. Washington, D.C.
[USEPA] US Environmental Protection Agency. 1993. Guidance for assessing chemical contaminant data for use in fish advisories. Volume 1. Office of Water. USEPA 823-R-93-002. Washington, D.C.
40 CFR Sect.180,183 (1993, 1986). Toxaphene; tolerances for residues.
[FDA] Food and Drug Administration. 1986. Food and Drug Administration compliance policy guide, Chapter 41 - Pesticides. Number 7141.01, Attachment B12. Washington, D.C.
[IRIS] Integrated Risk Information System. 1998. Announcement of 1998 Program; Requestfor Information. 63 FR 75 (January 2, 1998).
Toxaphene; Revocation of Tolerances. Final Rule. 58 FR 46087 (September 1, 1993).
Revocation of action levels for residues of toxaphene in food and feed. 58 FR 60859 (November 8, 1993).
Action levels for residues of certain pesticides in food and feed. 55 FR 14359 (April 17, 1990).
Toxaphene; tolerance for residues. 36 FR 22556 (November 25, 1971).
Hayes WJ Jr. 1982 Toxaphene and related material (Chapter 6: Chlorinated hydrocarbon insecticides). From:Pesticides studied in man. Baltimore: Williams &Wilkins; p 260-64.
Mitchell Frank L. 1996. Letter to Wallace E. Harrell of Gilbert, Harrell, Gilbert, Summerford and Martin, and J. Kevin Buster of King & Spaulding regarding response to request to review material related to potential exposures to toxaphene. January 11th.
Olson KL, Matsumura F, Boush GM. 1980. Behavioral effects on juvenile rats from perinatal exposure to low levels of toxaphene, and its toxic components, Toxicant A and Toxicant B. Arch Environ Contam Toxicol 9(2):247-57.
Samosh LV. 1974. Chromosome aberrations and character of satellite associations after accidental exposure of the humans to polychlorocamphene. Tsitol Genet 8:24-7.
Steinberg M, Kinoshita FK, Ballantyne M.1998. Mutagenicity studies with toxaphene congeners. Organohalogen Compounds 35:243-46.
Toxaphene Peer Review.1998. Recommendations on the cancer potency factor. April 10, 1998.
[WHO] World Health Organization. 1984. Principles for evaluating health risks to progeny associated with exposure to chemicals during pregnancy. Geneva, Switzerland. p. 177.
This section acknowledges the presence of a copy of public comments providing ATSDR with suggested revisions for the text of the health assessment. These suggested revisions have been evaluated and addressed in the text as appropriate.
[*1] The assessment does not specifically reference the source of the screening level and underlying RfD. A review of the literature suggests that it is derived from a RfD of 2.5E-4 from a 1993 USEPA document titled, "Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories - Volume 1."
[*2] The assessment's statement that "the method used to calculate total toxaphene from chromatograms also affects the estimated concentration" (p17) is a misrepresentation of the conclusions expressed in the Carlin and Hoffman reference (reference #32). The differences in the results of analyses calculated among the different reference standards for toxaphene were intended to illustrate the effects of different GC profiles among the standards.
[*3] It is unclear what the difference is between the methods used to obtain results reported as "CLP Pesticide Analysis" and "USEPA Toxpahene Task Force Analysis" as described in Table 1, Appendix 1.
[*4] In several places through the Assessment it is stated that USEPA is working on improving an analytical method using GC-MS. It is unclear whether this is a validated method and if so when it will be published.
[*5] According to a 9/30/93 memo written by the Associate Commissioner for Regulator Affairs for the FDA, "FDA's monitoring data have shown in recent years the absence of detectable toxaphene residues in those food and feed commodites for which FDA had established action levels…Accordingly, the agency decided that its action levels for toxaphene were no longer needed for FDA pesticide enforcement programs."
[*6] According to a 9/30/93 memo written by the Associate Commissioner for Regulator Affairs for the FDA, "FDA's monitoring data have shown in recent years the absence of detectable toxaphene residues in those food and feed commodites for which FDA had established action levels…Accordingly, the agency decided that its action levels for toxaphene were no longer needed for FDA pesticide enforcement programs."
[*7] This value is included in the current IRIS record.