Position Paper for Trifluoroiodomethane (CF₃I)

Prepared By:

Harvey Clewell

ICF Incorporated

Greg Lawrence

ICF Incorporated

For:

U.S. Environmental Protection Agency

Office of Air and Radiation

Stratospheric Protection Division

May 21, 1999

I. Introduction

Trifluoroiodomethane (CF₃I) has been proposed as a candidate replacement agent for halon for use as a fire suppressant. Several studies, including acute and subchronic inhalation toxicity, cardiac sensitization, reproductive toxicity, and mutagenicity studies have been conducted with CF₃I. In addition, an exposure assessment based on a proposed use of the material has been performed and physiologically-based pharmacokinetic modeling has been performed to estimate blood levels following exposures to CF₃I. The results of these studies are discussed in the following sections to provide an assessment of the potential risks associated with CF₃I use, with consideration given to the intended uses of CF₃I and the types of exposures that may be associated with these uses. For each area of discussion, an evaluation is also performed of whether additional research is needed to support a risk assessment for CF₃I.

II. Summary of Toxicity Studies

Acute Toxicity

The acute toxicity of CF₃I has been characterized by Ledbetter (1993, 1994) and Dodd et al. (1997). Ledbetter (1993) conducted an experiment where groups of 10 (5/sex) rats were exposed to 12.7 percent CF₃I vapors via nose-only inhalation for a single 15-minute exposure. Two additional rats (1/sex) were exposed to air and served as controls. All rats were allowed a 14-day observation period post-exposure after which the animals were sacrificed and gross necropsies were conducted. Clinical signs reported immediately after exposure were salivation (all animals), rales (2/5 males) and discoloration around the mouth in one male rat. However, these effects were transient and had resolved by 1 hour post-exposure. No clinical signs were reported in the control rats. There were no treatment-related effects on body weight gains and no gross pathologic lesions were reported at necropsy.

In a follow-up experiment, Ledbetter (1994) exposed groups of 10 Sprague-Dawley rats (5/sex) to atmospheres containing approximately 24.2 percent or 28.8 percent CF₃I vapors via nose-only inhalation for a single 15 minute exposure. All female rats and 2/5 of the male rats in the high-concentration group died during exposure, while in the low-concentration group, one male rat died. The animals that survived were reported to be shaky upon removal from the exposure tubes; however, this effect was transient and the rats quickly recovered. Body weight gain during the 14 day recovery period was reported to be normal and at gross necropsy, the only lesions reported were red lungs in 2 males and 1 female in the low-concentration group. In the high-concentration group, red lungs were reported in 1 male and red foci were reported in the lungs of 2 male rats. The authors noted that the LC₅₀ was likely between the two exposure concentrations used in this study, but that little useful information would be obtained if another group of animals was exposed to CF₃I at a concentration between 24.2 percent and 28.8 percent.

Ledbetter (1994) also conducted an experiment where groups of 10 rats (5/sex) were exposed to atmospheres containing 10 percent, 12.8 percent, 20 percent or 32.2 percent CF₃I vapors for 4 hours via whole body inhalation. All rats in the high-concentration group died within 20 minutes of initiation of exposure, although the authors noted that the test article was found to be contaminated with HF. Therefore, a HF filter was used in the second exposure (20 percent CF₃I). Hydrogen fluoride was not detected in the exposure chamber during the second exposure; however, all rats died within 20 minutes. The test article was replaced with a new batch for the rats exposed to 10 percent or 12.8 percent CF₃I. No deaths were reported in these groups, although the rats became unconscious during the exposures. At gross necropsy, the lungs in the 32.2 percent exposure group were reported to be dark red and puffy. The lungs in the rats that were exposed to 20 percent CF₃I were also reported to be puffy, but were not the dark red color of the other group. The authors suggested that these differences may have been due to the HF contamination. In the rats exposed to 10 percent or 12.8 percent CF₃I, there were no effects on body weight gain over the two week recovery period. At gross necropsy, red lungs were reported in 2 of the male rats that were exposed to 12.8 percent CF₃I.

Another acute inhalation toxicity study was conducted by Kinkead et al. (1994) and published in an article by Dodd et al. (1997). In this study, groups of 30 male rats were exposed to 0 percent, 0.5 percent or 1 percent CF₃I for 4 hours via nose-only inhalation. Groups of 10 rats were sacrificed immediately or 3 or 14 days post-exposure. No deaths were reported and there were no signs of overt toxicity during exposure or the recovery periods. No treatment-related effects on body weights were reported. The authors did note that there were statistically significant differences in hematological and clinical chemistry parameters; however, none of these changes were considered to be biologically significant. There were no treatment-related gross or microscopic lesions reported.

Collectively, these data reported by Kinkead et al. (1994), Ledbetter (1993, 1994) and Dodd et al. (1997) adequately characterize the acute toxicity of CF₃I (with the exception of cardiac sensitization potential, which is discussed in the next section). Additional acute toxicity studies are unlikely to provide information that would be useful in a risk assessment for CF₃I.

Cardiac Sensitization Potential

The cardiac sensitization (CS) potential of CF₃I and C₃F₇I were assessed by Kenny et al. (1995), and published by Dodd and Vinegar (1998). The CF₃I-CS study was divided into three stages and was conducted according to the experimental procedure described by Reinhardt and company (1971). Briefly, experiments lasted for 17 minutes/dog with dogs receiving an initial challenge dose of adrenalin at 2 minutes, followed by exposure to the test substance via inhalation starting at 7 minutes, then dogs received a second challenge dose of adrenalin at 12 minutes, and lastly, discontinuation of the test substance at 17 minutes. The first stage of the study was involved in establishing the response of each dog to varying concentrations of adrenaline. A positive response in Stage 1 was defined as an increase in heart rate, followed by a decrease in heart rate and an increase in the height of the T-wave on the ECG. The doses of adrenalin to be administered to each dog in Stages 2 and 3 were determined using the results from Stage 1. The second stage of the study served as a positive control for the experimental protocol. In Stage 2, two beagle dogs were exposed to CFC-11, a known cardiac sensitizer. The positive result was defined as the appearance of a burst of multi-focal ventricular ectopic activity or ventricular fibrillation. Stage 3 of the study involved exposing the dogs to increasing concentrations of CF₃I and C₃F₇I.

Of the nine dogs tested in Stage 1, it was determined that 7 dogs could be divided into 4 groups based on responsiveness to adrenalin, a 1 µg/kg group, a 4 µg/kg group, an 8 µg/kg group, and a 12 µg/kg group. The two remaining dogs were excluded from the study because of reasons defined in the report. In Stage 2, dogs #’s 1151 and 1157 were selected for exposure to CFC-11. Dog 1151 (receiving 12 µg/kg adrenalin) showed a positive response to 2 percent CFC-11 and died of fatal ventricular fibrillation (FVF). The second dog (1157) was not exposed to CFC-11. In Stage 3, the remaining 6 dogs were exposed to CF₃I at increasing concentrations of 0.1 percent, 0.2 percent, 0.4 percent and 1.0 percent. All dogs produced negative results when exposed to concentrations of 0.1 percent and 0.2 percent CF₃I. One dog (1149) produced a positive result and died of FVF when exposed to 0.4 percent, and one dog (1147) produced a positive result and died of FVF when exposed to 1.0 percent. Exposure of subsequent dogs at concentrations that produced positive results and death were not performed based upon humane considerations.

The results observed in the CF₃I-CS study were similar to results observed in studies investigating CS for Freon 11 and Halon 1211. The LOAEL for CF₃I based on CS was established as 0.4 percent. Likewise, the LOAEL for Freon 11 based upon CS is 0.35 percent, and the LOAEL for Halon 1211 based upon CS is 1.0 percent. It is important to note that all three dogs that died in the CF₃I-CS study (dog 1151 for CFC-11, dog 1149 for 0.4 percent CF₃I, and dog 1147 for 1.0 percent CF₃I) received the highest doses of adrenalin (12 µg/kg, 8 µg/kg and 8 µg/kg respectively). At least one review article suggests that the dose of adrenalin administered may determine the sensitivity of studies investigating CS. Skaggs notes in her review that a 5 µg/kg dose of adrenalin was used in the study investigating CS for Halon 1211 and suggests “that if higher doses of epinephrine were used in the Halon study, a lower LOAEL would have resulted for Halon 1211.” Conversely, if the study investigating the CS potential for CF₃I would have used a lower dose of adrenalin, it seems likely that the LOAEL for CF₃I would have been higher.

The results of testing with CFC-113 may serve as an example that depicts the sensitivity of the dog CS test to identify agents that may induce cardiac arrhythmias (May and Blotzer 1984). In this study it was reported that the lowest concentration of CFC-113 that induced fatal cardiac arrhythmias in epinephrine-challenged dogs was 0.5 percent. However, in monkeys that were not challenged with epinephrine, fatal cardiac arrhythmias were induced only at exposure concentrations of 2.5-5 percent. Following accidental exposures in humans, based on reconstructed exposure data, fatal cardiac arrhythmias were observed at exposure concentrations on the order of 3.7-12.8 percent. Thus, the epinephrine-challenged dogs were more sensitive to the cardiotoxic effects of CFC-113, i.e., cardiotoxic effects were observed at lower exposures in the dogs than in monkeys or humans. It seems likely that the reason cardiotoxic effects were observed in dogs exposed to lower concentrations of CFC-113 was the administration of exogenous epinephrine.

In fact, the rate of epinephrine release at the typical dose of epinephrine administered in dog studies, 8 µg/kg , is about 10-fold greater than endogenous epinephrine release rates observed in humans under stress (Reinhardt et al. 1971). While fatal cardiac arrhythmias were observed in dogs that were exposed to concentrations of 0.4 percent CF₃I or greater when challenged with 8 µg/kg epinephrine (Kenny et al. 1995), there were no cardiac effects observed in dogs that were exposed to 2.5 percent CF₃I without epinephrine challenge, and only tachycardia was observed in a 5 percent exposure (ICF Kaiser/Huntingdon 1998). Therefore, it would appear that the use of the epinephrine-challenged dog as a model for human cardiac sensitization provides an inherent safety factor of as much as 10.

In conclusion, there is good evidence for a NOAEL for CS from CF₃I at 0.2 percent, with an apparent LOAEL at 0.4 percent. Based on these data a ceiling of 2,000 ppm has been proposed for CF₃I. While additional studies could refine these estimates, they would be unlikely to significantly alter the evaluation of the acceptability of CF₃I for use as a fire suppressant. As mentioned above, the CS potential of CF₃I, appears to be roughly similar to other chemicals which have had extensive use, including Freon 11 and Halon 1211.

Carcinogenicity and mutagenicity

The potential mutagenicity and genotoxicity of CF₃I has been evaluated in bone marrow micronucleus assays in mice (Mitchell 1995a) and rats (Kinkead et al. 1996; Dodd et al. 1998, the Ames reverse mutation assay (Mitchell 1995b) and mouse lymphoma forward mutation assay (Mitchell 1995c). Mitchell (1995a) conducted a bone marrow micronucleus assay where male and female mice were exposed to 0 percent, 2.5 percent, 5 percent, or 7.4 percent CF₃I (0, 25,000, 50,000 or 74,000 ppm) via nose-only inhalation, 6 hours/day for 3 consecutive days. Statistically significant concentration-related increases in micronucleus frequency (micronuclei/1000 polychromatic erythrocytes) were reported in the mid- and high-concentration male and female mice. In addition, statistically significant concentration-related decreases in the ratio of PCE/1000 erythrocytes were reported at all exposure concentrations in female mice.

A bone-marrow micronucleus induction assay was also performed as part of a 13-week nose-only inhalation study where rats were exposed to 0 percent, 2 percent, 4 percent, and 8 percent (0, 20,000, 40,000 and 80,000 ppm) CF₃I for 2 hours/day, 5 days/week for 4 or 13 weeks (Kinkead et al. 1996). After 4 weeks of exposure, concentration-related increases in micronucleus frequency were observed in the mid- and high-concentration males and females, with statistically significant positive trends reported for each sex. Concentration-related decreases in polychromatic erythrocyte/normochromatic erythrocyte (PCE/NCE) ratios, an indicator of bone marrow toxicity, were observed in all treated males and females and statistically significant trends were reported. After 90 days of exposure, concentration-related increases in micronucleus frequency and decreases in PCE/NCE ratios were observed in all treated groups, with statistically significant trends also reported for each endpoint. However, in a study by Dodd et al. (1998) where male and female rats were exposed to 0 percent, 0.2 percent, 0.7 percent, or 2.0 percent CF₃I (0, 2,000, 7,000 or 20,000 ppm) via whole-body inhalation for 7 or 14 weeks, there were no statistically significant changes in micronuclei frequency or in PCE/NCE ratios.

Historically the bone marrow micronucleus assay has been conducted in mice, and there has been greater confidence in the interpretation of data collected in that species. However, recently it has been shown that the rat is a suitable alternative to the mouse for the assessment of the potential for a chemical to induce micronuclei in bone marrow or peripheral blood (Holden et al. 1997; Wakata et al. 1998). Thus, the results from the micronucleus assays conducted in rats reported above are useful in evaluating the genotoxic effects of CF₃I. Collectively, the data from the three bone marrow micronucleus assays indicates that the induction of micronuclei by CF₃I is associated with repeated exposures to high concentrations, i.e., concentrations greater than 2.0 percent.

Other studies of genotoxicity conducted with CF₃I include an Ames assay and a forward mutation assay. The Ames assay was positive (Mitchell, 1995b), while a forward mutation assay was negative (Mitchell, 1995c). Thus, despite the fact that CF3I is slowly metabolized (Williams et al., 1994), it is clear that it has the potential to cause genotoxic effects. Nevertheless, it would be premature to assume that based on the results of these genotoxicity tests CF₃I is a carcinogen. An accurate assessment of the potential carcinogenicity of CF₃I would require data from a two-year bioassay. However, given the proposed uses of the material as a fire suppressant, exposures (apart from manufacturing) would be rare, or at least infrequent, rather than chronic. Thus even if CF₃I were found to have a relatively high carcinogenic potency, which appears unlikely given the dose-response for the micronucleus assays, such infrequent exposures would not result in a significant cumulative lifetime risk of carcinogenic effects. Therefore, a general conclusion that CF₃I should not be used as a fire suppressant based on its possible genotoxicity would be inappropriate. On the other hand, in the case of chronic exposure, such as during manufacture, the potential carcinogenicity of CF₃I could be a cause for concern.

Thyroid hormones

The potential effects of exposures to CF₃I on thyroid hormone levels have been evaluated by Dodd et al. (1998) and Kinkead et al. (1996). In a reproductive toxicity screen by Dodd et al. (1998), statistically significant increases in TSH, T₄ and rT₃ levels and decreases in T₃ levels were reported in male and female rats exposed to 0.2 percent, 0.7 percent, or 2 percent CF₃I, 6 hr/day, 5 days/week for 4 weeks prior to mating. During the 14-day mating period, gestation and lactation, the animals were exposed 6 hours/day, 7 days/week, with the exception that the dams were not exposed from gestation day 21 through day 4 of lactation to allow for parturition and early lactation. After day 21, post-partum of the last female to deliver, exposures were 6 hours/day, 5 days/week until study termination. Similar results with regard to TSH, T₄, rT₃, and T₃ levels were reported in study by Kinkead et al. (1996) where male and female rats were exposed to 2 percent, 4 percent, or 8 percent CF₃I vapors for 2 hours/day, 5 days/week for 13 weeks via nose-only inhalation. The precise mechanism by which CF₃I produced these effects is unknown; however, Dodd et al. (1998) proposed that the observed effects were likely the result of CF₃I interfering with 5'deiodinase, the enzyme responsible for the conversion of T₄ to T₃. It is obvious that the thyroid effects reported in these studies were related to CF₃I exposure; however, as discussed below, the relevance to human health is questionable, and it is uncertain if the rat is an appropriate model for the evaluation of potential thyroid effects in humans.

In the bloodstream, thyroid hormones, (T₃ and T₄) are bound to carrier proteins, such as albumin and globulins, and when bound are not subject to metabolism or degradation. Conversely, the unbound or free forms are metabolized and degraded. In rats, T₃ and T₄ are bound to albumins, while in humans T₃ and T₄ are bound with a high affinity to globulins, which are not present in rats. In the rat, due to this weaker binding, T₃ and T₄ have a shorter plasma half-life and more rapid turnover than in humans (Capen et al. 1991; Alison et al. 1994). Consequently, the demand on the rat thyroid gland to maintain homeostasis is much greater than in humans, i.e., it would be easier for humans to maintain normal physiological levels of T₃ and T₄. Therefore, rats would likely be more sensitive to thyroid effects produced by the indirect mechanism described. In particular, it seems unlikely that short-term exposures of humans to low concentrations of CF₃I would result in the thyroid effects produced in rodents, due to the differences in protein binding and plasma half-lives. Nevertheless, to protect sensitive individuals, an occupational AEL of 150 ppm as an 8 hour time-weighted average has been proposed for CF₃I based on the observed effects on thyroid hormone levels in the rat.

In rats and mice, species that are highly sensitive to chemicals that interfere with thyroid homeostasis (Capen 1991), decreased levels of T₃ and T₄ result in a compensatory increase in TSH. Continued stimulation of the thyroid gland by TSH, a mechanism that is not necessarily related to deficiencies or excesses in iodine, has been associated with proliferative lesions, adenomas, and rarely, carcinomas, in rodents. However, it is questionable if this indirect mechanism that is operative in rats is relevant in humans due to the differences in thyroid hormone protein binding and half-life times just described (Alison et al. 1994; Capen 1991). Moreover, for the proposed use of CF₃I as a fire suppressant, it seems unlikely that the relatively infrequent exposures expected would result in alterations in thyroid homeostasis in humans. Thus, the potential for CF₃I to induce thyroid cancers in humans seems extremely unlikely.

Reproductive Toxicity

Potential reproductive effects associated with exposures to CF₃I have been evaluated by Dodd et al. (1998) and to a limited extent by Kinkead et al. (1996). (Note: the results from the 90-day study of Kinkead et al. are also reported in a summary published by Dodd et al. 1997). The reproductive toxicity study conducted by Dodd et al. (1998) suggested that there were no treatment-related effects that resulted from subchronic exposure to CF₃I with regard to male reproduction parameters such as changes in mating index, fecundity index, fertility index, and gross and histological lesions.

In the study by Dodd et al., male and female rats were exposed in whole body inhalation chambers to 0.0 percent, 0.2 percent, 0.7 percent, and 2.0 percent CF₃I, 6 hours/day, 5 days/week for 30 days prior to mating. It is generally customary to expose rats 7 days/week for a minimum of 70 days prior to mating because spermatogenesis occurs over a period of 48-53 days in the rat. It could, therefore, be argued that the test protocol was inadequate for the assessment of any potential effect on the early stages of spermatogenesis, however, based upon the data it is unlikely that any effects were produced in the later stages of sperm development. Furthermore, there were no differences in testicular weights or histological observations in any of the high-dose males when compared to the control animals at the 14-week necropsy. These data suggest that under the conditions of this study, there was no evidence of a treatment-related effect on the testes.

On the other hand, the 90-day study conducted by Kinkead et al. (1996), in which rats were exposed to 0.0 percent, 2.0 percent, 4.0 percent, and 8.0 percent CF₃I in nose-only inhalation chambers, reported a 100 percent incidence of testicular atrophy in male rats receiving 4.0 percent and 8.0 percent CF₃I in the 30-day sacrifice group, and a 78 percent incidence of testicular atrophy accompanied by a decrease in spermatogonia and spermatids in high-dose males in the 90-day sacrifice group. The severity of this lesion was measured on a scale of 1-5, with grades for the lesions being 2.0 and 4.0 in the 4.0 percent and 8.0 percent groups, respectively, in the 30-day sacrifice group, and a grade of 2.2 for the high-dose group in the 90-day sacrifice group. The authors of the 90-day study noted that the effect could have been related to heat-stress associated with the nose-only exposure-chambers. Similar testicular effects were noted in rats exposed to HFC-143a by nose-only inhalation for 4-weeks. However, these testicular effects were completely absent in two subsequent studies, a 4-week study and a 90-day study, that exposed animals to HCF-143a via whole body inhalation (Malley 1993). Likewise, a mid-study change to bigger exposure tubes in the study conducted by Kinkead et al. (1996), thus diminishing the heat-stress to some degree, resulted in a substantial decrease in the occurrence and severity of testicular atrophy in high-dose males, and a complete reversal of the lesion in mid-dose males. Although Kinkead et al. (1996) noted that the testicular effects only occurred in mid-dose and high-dose treated animals and not in the low-dose or control groups, and suggested that these effects could not be dismissed totally on the basis of heat-stress associated with nose-only inhalation methods, the combined data of the reproductive study (Dodd et al., 1998) and the 90-day study (Kinkead et al., 1996) with CF₃I, as well as the studies conducted with HFC-143a, suggest that the observed testicular effects in the 90-day study were likely the result of heat stress rather than a direct effect of CF₃I on the testes.

The reproductive study conducted by Dodd et al. did appear to accurately assess female fertility and prenatal development. Starting with gestation day 0, and continuing throughout the remainder of the study, rats were exposed 6 hours/day, 7 days/week. The authors reported that there were no statistically significant differences between control and treated groups in mean number of pups/litter, pups with gross lesions, live birth index, or pup survival index for 4, 7, 14 and 21 days post-parturition. It must be noted however that the reproductive study did not expose dams from gestation day 21 through lactation day 4 to allow for parturition. Customarily dams should be exposed continuously throughout the study, however, in inhalation reproductive studies only, dams are usually not removed from their pups for the purpose of minimizing stress on the dams and the pups during this period. The authors did report a significant decrease in the sex ratio (male pups/litter) for litters from dams receiving 2.0 percent CF₃I. However, a consistent dose-response for this effect was not evident (the pup sex ratio was 0.99, 0.79, 1.07 and 0.68 for the control, low-, mid- and high-concentration groups, respectively), and this effect was not associated with a statistically significant positive trend. The toxicological significance of this effect is questionable. Additionally, Dodd et al. (1998) reported a treatment-related decrease in absolute and relative ovary weights in the 14 week sacrifice group. Kinkead et al., however, reported no significant difference in ovary weights between control and treated groups.

An overall assessment of the data in the studies by Dodd et al. (1998) and Kinkead et al. (1996) result in little concern regarding the reproductive toxicity associated with the subchronic exposure to CF₃I. It is possible that exposure to concentrations of CF₃I greater than or equal to 8 percent may result in testicular atrophy, although this statement cannot be made with certainty. Also, in rats, exposures to 2 percent CF₃I were associated with a low pup sex ratio, although the toxicological significance of this finding is questionable. At any rate, it does not appear that there are any exposure scenarios in which these effects could become the limiting toxicity.

Potential Exposure Scenarios

An outline of the possible exposure scenarios is provided in a document from Pacific Scientific (Skaggs 1996). Generally, these scenarios include repeated, low-level exposures as a result of manufacturing, transfer and filling, and accidental discharge during extinguisher maintenance, overhaul, or repair, and infrequent higher-concentration exposures resulting from firefighting, accidental discharge of extinguishers, and accidental discharge of storage cylinders. Several experiments have been conducted to estimate the ppm concentration of CF₃I that would result from different applications such as firefighting-streaming in different sized rooms, and accidental discharge in F-15 engine nacelles. In the experiments that investigated the ppm concentrations resulting from hand held fire extinguishers, maximal concentrations of CF₃I ranged from 10,000 ppm (1 percent) to 30,000 ppm (3 percent), and average concentrations for the first 30 minutes ranged from 1040 ppm (0.1 percent) to 4678 ppm (0.5 percent). Both peak and average concentrations depended on the size of the room, the height off the floor of measurement, and the amount (2.5 lb.-13 lb.) of CF₃I discharged. Likewise, the peak concentrations achieved following a 9 lb. discharge of CF₃I in an F-15 engine nacelle ranged from 9,000 ppm (0.9 percent) to 70,000 ppm (7 percent) depending on the location of measurement to the engine nacelle and the height off the floor (the highest concentration of 70,000 ppm was achieved inside the engine nacelle at head level).

Using PBPK modeling, Vinegar and coworkers (Vinegar et al. 1997) calculated the blood concentrations expected to result from accidental exposures using the concentration data collected in the F-15 engine nacelle experiments, and for a demonstration event from two salesman, wherein the two salespersons inhaled an estimated 1.25 L of pure CF₃I. The PBPK model estimated the blood concentrations resulting from accidental exposure in the F-15 engine nacelle study to range between 6 µg/ml and 40 µg/ml, and estimated blood concentrations achieved in the sale demonstration to be 2000 µg/ml. The PBPK model also estimated the expected blood concentration from a 5 minute, 4000 ppm (0.4 percent, the reported LOAEL for CF₃I) exposure to CF₃I to be 19 µg/ml. Clearly, there is evidence from both the firefighting-streaming study and the F-15 engine nacelle study that accidental human exposure to CF₃I may result in blood concentrations that exceed the 19 µg/ml estimated for the LOAEL. An experiment that measured the arterial and venous blood concentration of CFC-11 and CFC-12 in beagle dogs reported arterial concentrations of 28.6 µg/ml and 35.3 µg/ml, respectively, for the CS - LOAEL concentrations of 0.5 percent and 5.0 percent respectively. It is noteworthy that the arterial concentrations for CFC-11 and CFC-12 are similar, despite a 10 fold difference in exposure concentrations. These data suggest that the LOAEL concentrations for CFC-11 and CFC-12 may be dependent on solubility factors, and that the threshold blood concentration that leads to CS may be the same for many substances.

Regarding the PBPK analysis of the sales demonstration, these data may demonstrate differences in susceptibility to CS between humans and beagle dogs, as well as possible differences in response to CF₃I under different conditions of stress. More specifically, the sales persons inhaled a relatively large amount of CF₃I under a presumably low-stress situation, which would correlate to low levels of adrenalin, and apparently suffered no adverse consequences. The dogs on the other hand, inhaled much smaller concentrations of CF₃I, but received doses of adrenalin that were just below the threshold for causing adrenalin-induced arrhythmias, which resulted in fatal ventricular fibrillation. These two scenarios may illustrate the importance of stress and the participation of adrenalin in precipitating the adverse cardiac response. Similar differences in response were seen between two studies that investigated the CS potential of CFC-12, one in the presence of endogenous epinephrine resulting from stress (Mullin et al., 1972), and the other in the presence of 8 µg/kg exogenous epinephrine (Reinhardt et al. 1971). Based on the proposed uses of CF₃I as a fire suppressant, the potential exposures would most likely occur in more stressful situations than was seen in the sales demonstration; however, the adrenalin levels reached would not be comparable to those administered in the dog studies.

III. Conclusion

In summary, it appears that adequate toxicological data are available on the acute and subchronic effects of CF₃I to support decisions regarding safe use of the chemical. The toxicity of CF₃I is generally low, with acute and subchronic toxicity occurring only at relatively high concentrations, on the order of a percent or greater. With regard to accidental acute exposure, the critical effect appears to be cardiac sensitization, based on studies with epinephrine-challenged dogs, at concentrations as low as 0.4 percent. While this LOAEL is much lower than the threshold for cardiac sensitization for Halon 1301, the toxicity of CF₃I appears to be similar to other materials which have been used safely in the past, such as Halon 1211 and CFC-11. Moreover, as discussed previously, the interpretation of the cardiac sensitization data are difficult given that such high levels of adrenalin were administered. While the protocol used in the cardiac sensitization testing of CF₃I has been described in the literature, it is uncertain what effect, if any, that the administration of a large dose of adrenalin may have on the results. It should be noted that an investigation into this issue is currently under way. Therefore, there does not appear to be any justification for precluding the consideration of CF₃I for potential applications as a fire suppressant.

Subchronic toxicities observed with CF₃I include thyroid hormone effects, and limited evidence of genotoxicity and reproductive effects. These concerns, however, would only be relevant to long-term, regular exposure such as in manufacturing of the chemical. For these situations, personal protective equipment and other workplace controls could be applied to minimize any potential for harmful effects. In this regard, an occupational AEL for CF₃I of 150 ppm as an 8-hour Time Weighted Average with a Ceiling of 2000 ppm has been recommended.

With regard to the impact of the toxicity of CF₃I on its potential acceptability as a fire suppressant, potential exposures to CF₃I can be grouped into three categories: exposures associated with manufacture, exposures associated with intentional releases, and unexpected exposures associated with automatic or accidental releases. In the case of manufacturing, the toxicity of CF₃I is not particularly notable compared to other volatile halogenated hydrocarbons currently in production, and there is no reason to expect that good industrial hygiene practice would be unable to adequately protect the worker. In the case of intentional releases, the various potential applications of CF₃I as a hand-held streaming agent should be carefully considered with particular attention to its potential for cardiac sensitization. Fatal cardiac incidents have occurred in the past with the use of Halon 1211 as a streaming agent, even when some level of personal protective equipment was used. The trade-off between suppression effectiveness and potential for cardiac sensitization appears to be a fundamental aspect of risk management for streaming fire agents, and is not a new consideration. Different agents may be preferable in different applications. At any rate, there is no reason that CF₃I should be removed from consideration for any such applications due to its toxicity.

The third, and most problematic, category of potential exposures relates to the use of CF₃I as an automatic-release fire suppression agent. In this area, more than in any other, it is essential that possible applications be considered separately. An application in one system might be acceptable, while a similar application in another weapons system might be unacceptable. Risk assessment techniques such as fault-tree analysis are routinely practiced during the development of weapons systems to ascertain the safety of various design features. There is no clear justification for excluding a chemical such as CF₃I from any and all consideration in such analyses. It seems obvious that the decision is better made at the system level, where concrete analysis can be performed. An excellent example of one aspect of such an evaluation is provided by the preliminary assessment of exposures in an engine nacelle conducted by Vinegar and Jepson (1997). Presumably, the likelihood of unplanned exposure of personnel associated with such a release, and the probability that restricted egress could exacerbate the exposure is also undergoing analysis by the safety professionals associated with that program. The need for, and feasibility of, personal protective equipment, failsafe measures, and other controls during maintenance activities can then be evaluated. As with the case for applications as a streaming agent, the trade-offs of effectiveness and toxicity should be dealt with in the same way as they routinely are for other design questions associated with weapons system development. Arbitrary exclusion of CF₃I from consideration in such systems is unwarranted.

Finally, there are no additional acute toxicity studies that would provide additional information that would alter the hazard assessment of CF₃I for acute exposures. Similarly, with the possible exception of a two-year bioassay, there are no additional studies that would alter that hazard assessment for chronic effects. However, given the proposed uses of CF₃I, chronic lifetime exposures are unlikely; therefore, a two-year bioassay would not provide additional usable data. In fact, the most useful data that would allow for a better hazard assessment of CF₃I would not be obtained from additional toxicity testing. Rather, as discussed above, better characterization of the potential exposures associated with CF₃I based on the proposed uses of the material would allow for a more accurate and scientifically defensible risk assessment of CF₃I.

References

Alison RH, Capen CC, Prentice DE. 1994. Neoplastic Lesions of Questionable Significance to Humans. Toxicol Pathol 22:179-186.

Azar A, Trochimowicz HJ, Terrill JB, and Mullin LS. 1973. Blood levels of flourocarbon related to cardiac sensitization. American Industrial Hygiene Jounal. 34:102-109.

Capen CC, DeLellis RA, Yarrington JT. 1991. Endocrine System. In: Haschek WM and Rousseaux CG, eds. Handbook of Toxicologic Pathology. Academic Press, Inc., San Diego. pp. 705-736.

Dodd DE, Kinkead ER, Wolfe RE et al. 1997. Acute and subchronic inhalation studies on trifluoroiodomethane vapor in Fischer 344 rats. Fund and Applied Toxicol 35:64-77.

Dodd DE, Leahy HF, Feldmann ML, Vinegar A. 1998. Reproductive Toxicity Screen of Trifluoroiodomethane (CF₃I) in Sprague-Dawley Rats. Mantech-Geo-Centers Joint Venture Report for Air Force Research Laboratory, Wright-Patterson AFB. AFRL-HE-WP-TR-1998-0029.

Dodd DE, Vinegar A. 1998. Cardiac sensitization testing of the halon replacement candidates trifluoroiodomethane (CF3I) and 1,1,2,2,3,3,3‑heptafluoro‑1‑iodopropane (C3F7I). Drug Chem. Toxicol., May;21(2):137‑149.

Holden HE, Majeska JB, Studwell D. 1997. A direct comparison of mouse and rat bone marrow and blood as target tissues in the micronucleus assay. Mutat Res 391(1-2):87-89.

ICF/Huntingdon 1998. HFC 236fa, HFC 227ea, HFC125 and CF₃I. An Inhalation Study to Investigate the Blood Levels of Inhaled Halocarbons in the Beagle Dog. Huntingdon Life Sciences Ltd, England. Huntingdon Report Number IFP 001/984370. Submitted to ICF Kaiser. Washington DC.

Kenny TJ, Sheperd CK, Hardy CJ. 1995. Iodotrifluoromethane and Iodoheptafluorpropane Assessment of Cardiac Sensitisation Potential in Dogs. Huntingdon Research Centre, UK. Submitted to Armstrong Laboratories, Toxicology Division. Wright-Patterson AFB, Ohio, USA.

Kinkead ER, Salins SA, Wolfe RE, Leahy HF. 1994. Acute toxicity evaluation of Halon replacement trifluoroiodomethane (CF₃I). Mantech Environmental Technology, Inc. Dayton OH. AL/OE-TR-1994-0070.

Kinkead ER, et al. 1996. 90-Day Nose-Only Inhalation Toxicity Study of Trifluoromethane (CF₃I) in Male and Female Fischer Rats. Mantech Environmental Technologies Report for Armstrong Laboratories. AL/OE-TR-1996-0024.

Ledbetter A. 1993. Acute inhalation toxicity of iodotrifluoromethane in rats. Final report. Project number 6030-012. Mantech Environmental Technology Inc. Research Triangle Park, NC, USA.

Ledbetter A. 1994. Acute inhalation toxicity of iodotrifluoromethane in rats. Final report. Project number 1530-001. Mantech Environmental Technology Inc. Research Triangle Park, NC, USA.

Malley, LA. 1993. Subchronic Inhalation Toxicity: 90-Day Study with HFC-143a in Rats. Volume 1 of 2. E. I. Du Pont de Nemours and Company. Newark, DE, USA.

May DC and Blotzer MJ. 1984. A report of occupational deaths attributed to fluorocarbon-113. Arch Env Health 39(5):352-354.

Mitchell AD. 1995a. Genetic Toxicity Evaluation of Iodotrifluoromethane (CF₃I), Volume II of III, Results of In Vivo Mouse Bone Marrow Erythrocyte Micronucleus Testing. Genesys Study Number 94037. Genesys Research, Inc. Research Triangle Park, NC, USA.

Mitchell AD. 1995b. Genetic Toxicity Evaluation of Iodotrifluoromethane (CF₃I), Volume I of III, Results of Results of Salmonella typhimurium Histidine Reversion Assay (Ames Assay). Genesys Study Number 94035. Genesys Research, Inc. Research Triangle Park, NC, USA.

Mitchell AD. 1995c. Genetic Toxicity Evaluation of Iodotrifluoromethane (CF₃I), Volume III of III, Results of the Foreward Mutation Assay Using L5178Y Mouse Lymphoma Cells. Genesys Study Number 94036. Genesys Research, Inc., Research Triangle Park, North Carolina, USA.

Mullin LS, Azar A, Reinhardt CF, et al. 1972. Halogenated hydrocarbon-induced cardiac arrhythmias associated with release of endogenous epinephrine. American Industrial Hygiene Journal 33:389-396.

Reinhardt CR, Azar A, Maxfield ME, et al. 1971. Cardiac arrythmias and aerosol “sniffing”. Arch. Environ. Health 22:265-279.

Skaggs SR. 1996. Establishing an acceptable exposure limit for CF₃I. Pacific Scientific.

Vingar A, and Jepson GW. 1997. PBPK modeling of CF₃I releases from F15 engine nacelles. Proceedings of the Halon Options Technical Working Conference, pp 162-68.

Wakata A, Miyamae Y, Sato S, et al. 1998. Evaluation of the rat micronucleus test with bone marrow and peripheral blood: summary of the 9th collaborative study by CSGMT/JEMS. MMS. Collaborative study group for the micronucleus test. Environmental mutagen society of Japan. Mammalian mutagenicity study group. Environ Mol Mutagen 32(1):84-100.

Williams JR, et al. 1994. Gas Uptake of Bromotrifluoromethane (Halon 1301) and Its Proposed Replacement Iodotrifluromethane (CF₃I). Armstrong Laboratory, Toxicology Division, Wright-Paterson AFB, OH.