Revisiting the testicular toxicity of cyanide: Assessment of the new and existing literature

Based on new testing, we re‐assess U.S. EPA and California OEHHA conclusions regarding male reproductive toxicity associated with cyanide exposure.


| INTRODUCTION
Cyanides are chemicals that contain a terminal -CN group. Under physiological conditions within the mammalian body, both simple cyanides (hydrogen, sodium, and potassium) and acetone cyanohydrin dissociate to form the cyanide anion (CNˉ), which is the common toxic species (ATSDR, 2006;ECETOC, 2007;US EPA, 2010a). Cyanides are known acute respiratory and oral poisons that interfere with the action of mitochondrial cytochrome enzymes (Keilin, 1929;Way, 1984) and can produce severe life-threatening effects. As such, most cyanides are classified as highly acutely toxic. Nevertheless, cyanides are ubiquitous in the environment due to natural and human-generated sources, and people are exposed to cyanides routinely at low air concentrations throughout the environment. Cyanide is released during volcanic eruptions and as a result of the normal biological processes of some higher plants, bacteria, and fungi (IPCS, 2004). Cyanide and cyanogenic substances are also present in many foods, including almonds, spinach, lima beans, millet sprouts, and cassava tubers. In particular, cyanide can be bound to sugars to form cyanogenic glycosides such as amygdalin that are found in the pits of various stone fruits (e.g., cherries, peaches, apricots) and in apple seeds (IPCS, 2004).
Cyanides are also present in many anthropogenic activities, including the industrial processes of metallurgy, electroplating, mining, ore extraction, and manufacturing (ATSDR, 2006;IPCS, 2004;US EPA, 2010a). Due to their asphyxiant properties, cyanides are used in some circumstances as a fumigant and a rodenticide (US EPA, 2010a). They are also components of tobacco smoke. On a global scale, however, the greatest source of cyanide occurs as the result of the burning of biomass (IPCS, 2004).
In order to protect the body from cyanide-induced inhibition of cellular respiration, cyanide is quickly conjugated to thiosulfate and eliminated from the body in the urine as thiocyanate. Although thiocyanate is less toxic than cyanide, it is a competitive inhibitor of iodine uptake by the thyroid gland and can impair the function of the thyroid gland, especially in the presence of an iodine-deficient diet (ATSDR, 2006;ECETOC, 2007).
As a consequence of their broad distribution and exposures to humans, the potential toxicity due to repeated low-dose exposures to cyanides has been investigated. Some information in humans is also available from inadvertent or accidental cyanide exposures and from epidemiology studies, most of which are focused on the neurological consequences of exposure. Additionally, the possible indirect impact of cyanide exposure on thyroid function via thiocyanate formation has also been evaluated.
Based on findings in a 90-day drinking water study conducted by the National Toxicology Program (NTP, 1993), the US Environmental Protection Agency (US EPA, 2010a) and the Office of Environmental Health Hazard Assessment of the State of California (OEHHA, 2013) have determined that cyanides should be considered male reproductive toxicants. No related data are available in humans because epidemiologic studies evaluating potential reproductive toxicity due to cyanide exposures have not been reported. The male reproductive system changes reported in the NTP study of sodium cyanide have been also used as the basis for calculating the current reference dose (RfD) for cyanide (US EPA, 2010a). The RfD is an estimated daily chronic oral exposure to humans that is likely to be without adverse effects (US EPA, 1993). It should be noted, however, that the Agency's confidence in the RfD value derived for sodium cyanide is low to medium (US EPA, 2010b). Furthermore, when judged against contemporary study designs that are generally used to generate data for risk assessment, the NTP 90-day drinking water study exhibits some limitations (including group caging; lack of individual water consumption measurements; absence of a post-treatment recovery group; fixation of the testes and epididymides in formalin; no determination of thyroid hormone status) as well as observations that deserve follow-up. Because of this, the US EPA's confidence in the study (which serves as the basis for the RfD) was only medium (US EPA, 2010b). One deficiency is that the NTP study reported only absolute organ weights, and not organ weights relative to body weight. The US EPA, taking these limitations into consideration in its IRIS assessment (US EPA, 2010a), calculated the relative reproductive organ weights for the NTP study before drawing conclusions for the purpose of deriving an RfD. However, the rats in the higher dose groups also consistently drank less water than those in the control and low dose groups and exhibited lower body weights, which could have affected the absolute and relative weights of the reproductive organs, as has been reported for hamsters as well as several strains of mice (Janský et al., 1986;Nelson, 1988;Nelson et al., 1995;Nelson & Desjardins, 1987).
Recently, a modern 90-day drinking water study of sodium cyanide was completed using the same rat strain and cyanide concentrations as in the NTP study (Tyner & Greeley, 2022). This new study incorporated additional design features to address limitations of the NTP study, including individual caging of animals, the addition of a "paired water" control group, the measurement of thyroid hormone levels, and the incorporation of a posttreatment recovery group. Tyner and Greeley (2022) found a small decrease in water consumption (consistent with reduced palatability) and increased thyroid/ parathyroid weights at the highest dose tested, but no treatment-related changes in male reproductive measures, thyroid hormone levels, or thyroid organ histopathology. The purpose of the present report is to consider the new results as part of an updated assessment of the potential male reproductive toxicity associated with cyanide exposure.

| Literature search
A search of the published scientific literature was conducted for the identification of human or animal studies that reported cyanide exposures and assessment of male reproductive outcomes. This search was completed on initially completed on May 13, 2020, and updated again on August 5, 2022. The search was conducted using the PubMed database, available through the National Center for Biotechnology Information (NCBI) at the U.S. National Library of Medicine (NLM) (https://www. ncbi.nlm.nih.gov/pubmed) employing the following PubMed query: ["cyanide"] AND ["male" AND "reproduction"] OR ["male" AND "genitalia"] OR ["sperm"].
Search results were initially limited to 2006 or later. Cyanide-containing chemicals are commonly used as mechanistic tools to inhibit mitochondrial respiration in cellular investigations. Therefore, the search results were culled to a body of potentially relevant in vivo studies conducted in mammalian species through a review of titles and abstracts. These were then compared against the studies that had been reviewed in the September 2010 US EPA Toxicological Review of Cyanide for IRIS (US EPA, 2010a), the 2006 Toxicological Profile for cyanide from the Agency for Toxic Substances and Disease Registry (ATSDR, 2006), and the 2007 European Centre for Ecotoxicology and Toxicology of Chemicals' review on cyanides (ECETOC, 2007) to ensure that no relevant studies previously reviewed by regulatory authorities and other scientific expert organizations were excluded or missed. To these studies, the newest 90-day study drinking water study of sodium cyanide (Tyner & Greeley, 2022) was added.

| Data quality assessment
No epidemiologic or other human studies related to the potential male reproductive toxicity of cyanide exposure were identified as a result of the literature search. The animal studies relevant to male reproductive toxicity were assessed for data quality using the Toxicological data Reliability Assessment Tool (ToxRTool) (Schneider et al., 2009; https://ec.europa.eu/jrc/en/scientific-tool/ toxrtool-toxicological-data-reliability-assessment-tool), which evaluates study quality based on Klimisch criteria (Klimisch et al., 1997). Studies were scored 1 through 3 (1-reliable without restrictions; 2-reliable with restrictions; or 3-not reliable) based on the reported study methods and level of data documentation. The studies identified in the literature search and via the scientific expert review reports (ATSDR, 2006;ECETOC, 2007;US EPA, 2010a) were evaluated independently by two authors using the ToxRTool. Any discrepancies in scoring were discussed such that the authors came to an agreement on the overall score assigned to each study.

| Data analysis and synthesis
The present assessment focuses primarily on the results from those studies that, through use of the ToxRTool, were assigned a Klimisch reliability score of 1; studies with Klimisch scores of 2 or 3 are considered in a cursory manner. Because the male reproductive changes reported in the 1993 NTP 90-day drinking water study of sodium cyanide (NTP, 1993) serve as the basis for multiple regulatory actions, the analysis focuses first on describing this study, its strengths and weaknesses, and the influence of potential confounding factors on the study results (for example, the impact of greatly reduced water consumption in the high-exposure group). The most recently conducted 90-day drinking water study of sodium cyanide (Tyner & Greeley, 2022) is then reviewed, its strengths and weaknesses are identified, and the findings of this study are compared to those of the earlier NTP investigation. Because all other studies assigned a Klimisch score of 1 were conducted prior to the 1993 NTP study, these studies were considered lastly as additional supportive information. Similar to the other studies, their strengths and weaknesses are also addressed herein.

| Literature search and data quality assessment
The search of the literature for studies of cyanide exposures and male reproductive endpoints was conducted in mid-May 2020 and updated again in August 2022. The initial search identified 193 articles published in 2006 or later (after the ATSDR toxicological profile was published). A review of titles and abstracts, however, showed only one potentially relevant animal study of male reproductive endpoints. The search was then expanded to include articles published prior to 2006, which identified an additional 56 articles. In both searches, no human studies that evaluated male reproductive parameters after cyanide exposure were identified. A review of titles and abstracts from the expanded literature search showed that all relevant animal studies had been identified in at least one of the three scientific expert reports being utilized as resources for the identification of relevant studies that evaluated male reproductive parameters after cyanide exposures (ATSDR, 2006;ECETOC, 2007;US EPA, 2010a). An additional two studies were added as a result of a subsequent search of the literature conducted in August 2022. To these, the study conducted at Charles River Laboratories (Tyner & Greeley, 2022) was added. As shown in Table 1, 12 in vivo animal studies examining the influence of cyanide exposure on male reproductive parameters were identified: 11 through the literature searches and review of expert reports and 1 being newly completed. It should be noted that 2 unpublished 90-day rat studies (of copper cyanide; Gerhart, 1986 and potassium silver cyanide; Gerhart, 1987) were discussed in some of the expert reports, but these reports could not be located for the purposes of this analysis and therefore are not included. According to ATSDR (2006), the Gerhart studies both reported increased male gonadal weights with treatment; however, the doses at which these findings were observed were associated with increased mortality, significant body weight reductions, as well as other adverse effects. Furthermore, the results are considered confounded by co-exposure to the metals copper and silver. Therefore, it is unlikely that the results of these studies would alter the conclusions drawn herein.
The animal studies listed in Table 1 were scored for data quality using the ToxRTool as described above. Studies assigned a score of 1 form the basis of the present evaluation. Studies assigned scores of 2 or 3 are briefly summarized in Appendix S1 to this report and are considered only as additional information in the analysis presented herein. Additionally, the study by Seiguer and Mancini (1971) was not assessed because it was considered a mechanistic evaluation due to the route of exposure (injection) and the extremely short duration between dosing and examination (3 min to 3 hr). Those studies assigned a ToxRTool score of 1 were considered in detail in this evaluation. Although the NTP (1993) study could have been assigned a ToxRTool quality score of 2 due to the omission of an appropriate waterrestricted control group, it was upgraded to 1 because it included measurement of water intake, which allows for some conclusions regarding the influence of water consumption on the reported data. All other quality 1 studies included a paired drinking water control group or were conducted via inhalation and thus, water consumption was not a confounder.
Because the results of the NTP (1993) study of sodium cyanide serve as the basis for multiple regulatory actions in the United States, this study was evaluated first. The newest subchronic study of sodium cyanide (Tyner & Greeley, 2022), which is the only quality score 1 study conducted since the NTP study was completed, is then examined in detail. The results of the three other quality score 1 studies, which were all conducted prior to the NTP study, are considered thereafter.

| Methods
The NTP conducted good laboratory practices (GLP)compliant 90-day (13-week) drinking water studies of sodium cyanide (NaCN; 99.9% purity) in F344 rats and B6C3F1 mice (10/sex/group for each species). The rats were housed 5/cage/sex; the mice were housed individually. The drinking water concentrations administered in both studies were 0, 3, 10, 30, 100, and 300 ppm. These drinking water concentrations were selected based on the results of 2-week drinking water studies in which animals exposed to concentrations greater than 300 ppm T A B L E 1 Animal studies of cyanide evaluating male reproductive endpoints.

Study
ToxRTool score Because assessments were conducted at 3 min to 3 hr post-dosing, this study was considered a mechanistic evaluation and was not included in the overall evaluation. exhibited significantly depressed body weight gains related to reduced water consumption (at 1000 ppm, 50%-86% lower body weight gains than controls; at 3000 ppm, body weight losses). In both the rat and mouse studies, animals were observed twice daily; body weights and water consumption were measured weekly. From these data, compound intakes were calculated. Hematologic and clinical chemistry evaluations were conducted on blood samples taken throughout the study in rats (Days 5,25,45,and 92)  Urinalyses were conducted in rats and included evaluation of thiocyanate concentrations as a biomarker of exposure (Days 8,22,43,and 88). Sperm motility and vaginal cytology evaluations were conducted on mice and rats in the 0, 30, 100, and 300 ppm groups only at the end of the study; male reproductive organ weights and spermatogenic stages as well as female estrous cyclicity and stage lengths were recorded at this time.
At necropsy, the weights of select organs (heart, right kidney, liver, lungs, right testis, and thymus) were recorded. Organs listed in Table 2 were fixed in 10% formalin and processed for microscopic evaluation, with full evaluations and pathology peer-review being conducted primarily on specimens from the control and high-dose groups.

| Results-rat study
All rats survived until the end of study. Male mean body weight gains were slightly lower than control in the 10 ppm (À10%) and 300 ppm (À8%) groups; final mean body weights were 4% and 5% less than control at these doses, respectively. Mean body weights/body weight gains were unaffected in females and at 30 ppm and 100 ppm in males. Likely due to palatability issues, water intake at the top two concentrations of 100 ppm and 300 ppm were approximately 10% and 18% less than control, respectively, in males and approximately 16% and 26% less than control, respectively, in females (Table 3).
Some statistically significant alterations in hematologic and clinical chemistry parameters were noted in rats (particularly in the 100 ppm and 300 ppm groups), but these changes were minor, did not show consistency throughout the study, and were considered unrelated to treatment. Due to the reduced water intake noted at 300 ppm, urinary volume and specific gravity were both statistically significantly altered from control values in male rats (data for female rats not presented) at this drinking water concentration. Urine thiocyanate concentrations increased in a dose-related manner, indicating exposure. Some statistically significant differences from control organ weights were reported at 300 ppm, but these appeared to be sporadic and unrelated to treatment. For example, absolute (but not relative) heart and lung weights were statistically reduced in males, but not in females. Additionally, the absolute kidney weight, as well as absolute and relative liver weights, were increased in females, but not males. Of relevance to this analysis, the right testis weight was slightly ($5%), but not statistically, reduced at 300 ppm compared to control; relative weights were unaffected (Table 4).
Neither gross nor histologic indications of pathology in the organs were increased with treatment of either sex of animals (data not shown). The NTP report also noted that no differences in thyroid follicle size, colloid staining or follicular epithelium were observed; however, thyroid weights were not reported.
In the reproductive tissue examination, the absolute weights of the left testis and left epididymis were statistically reduced at the highest drinking water concentration of 300 ppm (Table 4); the left cauda epididymal weights were statistically reduced at 30, 100, and 300 ppm. Importantly, when adjusted for body weight, as calculated by US EPA (2010a), the relative weights of the testes and epididymides were not statistically different from control values (Table 4). However, relative cauda epididymal weights remain statistically significantly different from control at the top three exposure concentrations, although the dose-response appears essentially flat.
The sperm analyses (Table 5) showed no statistically significant differences in testicular sperm counts per gram of tissue. Total count per testis and testicular suspension concentration were significantly reduced at 300 ppm, but both the concentration per gram of epididymal tissue and total count per cauda epididymis were unaffected by treatment. Sperm motility was significantly lower than control at 30, 100, and 300 ppm. It is not known if these changes could be a reflection of the hydration status of the animals; however, these data did not show a dose-response, were within expected control ranges, and, thus, were not considered to be treatmentrelated.

| Results-mouse study
The doses administered to mice on a mg/kg body weight basis were considerably higher than those administered to rats-in most cases, by approximately two-fold. Two female mice (1 control, 1 at 30 ppm) died on study; all males survived until the end of study. Male mean body weight gains were greater than control by 10%-20% in all dose groups except 300 ppm; consequently, final male mean body weights were similar to (or slightly greater than) control in all exposure groups. Female mean body weight gains were greater than control by 16%-20% at 3, 10, and 100 ppm. In contrast, female mean body weight gain and final mean body weight at 300 ppm were 8% and 7% less than control, respectively. Likely due to palatability issues, water intake at the top concentrations of 300 ppm was approximately 16% and 23% less than control in males and females, respectively (Table 6).
A few sporadic statistically significant alterations in hematologic and clinical chemistry parameters were noted in mice, but these were minor and generally unrelated to dose; thus, they were considered not biologically significant. No urinalyses were conducted.
Some statistically significant differences from control in organ weights were reported. Relative kidney and liver weights were statistically increased in both sexes at 300 ppm; in females, relative kidney weights were also increased at 100 ppm and relative liver weights were increased at ≥10 ppm. Of relevance to this analysis, the absolute and relative right testis weights were not affected by treatment (Table 7). Neither treatment-related gross nor histologic indications of pathology were observed in either sex (data not shown).
In the reproductive tissue examination, the absolute weights of the left epididymis and cauda epididymis were statistically reduced at the highest drinking water concentration of 300 ppm (Table 7). However, the relative (to body weight) reproductive organ weights calculated by US EPA (2010a) found no significant change in the epididymis, although the left cauda epididymal relative weights were significantly lower at both 100 ppm and 300 ppm.
In contrast to the data for the rat, neither male reproductive organ weights nor sperm analyses in the mouse showed changes associated with drinking water exposure to sodium cyanide (sperm parameters shown in Table 8).

| Tyner & Greeley, 2022
Tyner and Greeley (2022) evaluated the potential toxicity of NaCN (97.65% pure) administered via drinking water to CDF (F344/DuCrl) male rats for at least 90 consecutive T A B L E 4 Male rat absolute and relative reproductive organ weights (mean ± standard error) in the control and top three NaCN exposure groups (data from NTP, 1993 and US EPA, 2010a). Absolute weights in grams; relative weights in mg/g body weight.

Exposure
c Calculated in US EPA, 2010a.

Note:
Values that were statistically signficantly different from control were bolded.
days and investigated the recovery, persistence, or progression of any alterations following a minimum of a 70-day recovery period. The study was conducted under GLP and generally followed guidance for subchronic toxicity studies as set forth in OECD 408 (OECD, 1998). The study built upon the design of the previously described NTP (1993) drinking water study of NaCN but included several additional design features to address some of the limitations of the NTP study. These enhancements included the addition of a water-restricted control group (paired to the high-dose group mean water consumption); individual housing of animals in order to permit measurement of individual water and feed consumption data; fixation of the testes and epididymides in modified Davidson's solution; assessment of thyroid hormone concentrations and thyroid histology, evaluating for tissue pathology as well as using a five-point semi-quantitative grading scheme based on thyroid follicular height and colloid area (Stump et al., 2014), consistent with that recommended in US EPA guidance for histologic assessment of the thyroid (US EPA, 2009b; US EPA, 2009a). The exposure design duplicated the NaCN concentrations used in the NTP (1993) study (Table 9). Animals were observed daily for mortality/morbidity. Detailed clinical evaluations were performed at dosing initiation and then every 2 weeks until the end of the dosing period. Body weights and water consumptions were measured twice per week (except for the waterrestricted group, for which measurements were made daily); food consumption was measured weekly. The animals were subjected to ophthalmologic observations approximately 6 days prior to the start of dosing, at the end of dosing, and at recovery euthanasia. Clinical pathology parameters (hematology, serum chemistry, and urinalysis) were measured at the time of terminal euthanasia and at the end of the recovery period. Blood was collected via jugular venipuncture for thyroid hormone and plasma thiocyanate measurements on study days 28, 56, 90, 118, and 160. Both the terminal and recovery cohorts were subjected to gross necropsies. As shown in Table 10, the organs from all groups were weighed and/or subjected to histopathologic examination.
Average compound consumption during the treatment period ranged between 0.23 mg/kg/day (3 ppm) and 21.66 mg/kg/day (300 ppm) ( Table 11). Analysis of thiocyanate in plasma showed a dose-dependent increase during the treatment period (day 1-day 91); by study day 118, thiocyanate values in plasma were similar across all groups, showing recovery. All test substance-treated animals survived the scheduled necropsies. A single male in the restricted water control group was found dead on Day 54. The cause of death was undetermined but was not test substance-related given that this animal was not administered test substance.
At the end of the treatment period, slightly lower (À11%) water consumption was noted in the 100 and 300 ppm groups (Table 11); the differences were most substantial at the initiation of dosing and were reported to frequently reach statistical significance. It should be noted that the volume administered to the restricted water group was adjusted for the residual water volume retained in the bottle cap. Beginning on Day 45, when this amount was reduced from 9 ml to 5.5 ml, the magnitude of the difference between water consumption in the restricted water control group and the 300-ppm group was reduced. There were no test substance-related clinical observations or changes in body weight, food consumption, water utilization, hematology, serum chemistry, or urinalysis measures when compared to T A B L E 5 Rat sperm parameters (mean ± standard error) in control and top three NaCN exposure groups (data from NTP, 1993 andUS EPA, 2010a both the normal control and the water-restricted control; nor were there test substance-related ophthalmic or macroscopic findings. Except for a nondose-related reduction in the T4 mean serum concentration on Day 90 in the 300-ppm group (Table 12), no alterations in other thyroid hormones (T3 and TSH) were seen throughout the study and no change in T4 were seen at earlier time points in the study. Furthermore, no thyroid hormone changes were detected during the recovery period (Days 118 and 160). The laboratory additionally conducted a histomorphologic evaluation of the thyroid glands based on follicular cell height and colloid area (similar to the type of assessment currently required in thyroid assessments to address potential endocrine disruption) but found no differences across groups. The authors, therefore, attributed the single T4 finding to biological variability, noting that the individual values were within the range of those observed in the water-restricted control group; thus, the change was not considered to be treatment-related. Test substance-related organ weight changes at the terminal euthanasia were limited to higher absolute and relative thyroid/parathyroid weights in the 30, 100, and 300 ppm groups; these changes were statistically significant at 300 ppm only and occurred in the absence of thyroid hormone changes and histopathologic correlates, which is suggestive of an adaptive response. Although minimally higher liver weights were noted in the 300-ppm group, these changes were not statistically significant, nor of a degree to suggest a definitive relation to treatment (2.4% increase in absolute weight, 6.3% change relative to body weight) and were without histologically correlated observations. Importantly, there were no significant alterations in male reproductive absolute or relative organ weights (Table 13); nor were there T A B L E 6 Mean water consumption and NaCN intake in mice (data from NTP, 1993 cyanide-related histopathological findings in the male reproductive organs. Sperm concentration in both the testis and cauda epididymis; sperm morphology; and sperm motility, were evaluated in both the terminal and recovery cohorts (Table 14). Compared to negative controls, no significant treatment-related alterations were detected in sperm parameters of rats that consumed water formulated with NaCN. Although the sperm concentration per gram of testis in the high-dose group was lower than, but not statistically different from, controls at study termination, the testicular sperm concentration of the high-dose group exceeded that of controls at the end of the recovery period, thus indicating that, if treatment-related, the change was minor and transient.
Based on the results of this study, oral administration of NaCN to rats at dosage levels of 3, 10, 30, 100, and 300 ppm for a minimum of 90 days was well tolerated at all dosages with no mortality or adverse findings. Importantly, no observations were noted in the reproductive organs of either the terminal or recovery cohort. Nonadverse test-substance-related findings were limited to lower water consumption and higher thyroid/parathyroid weights in the 30, 100, and 300 ppm groups. There were no accompanying treatment-related alterations in thyroid hormone levels, no histopathological alterations associated with the increased weights, and the weights were similar to control values after the recovery phase. The study authors, therefore, interpreted the thyroid weight changes as being nonadverse and considered the NOAEL to be 300 ppm, equivalent to 21.66 mg/kg/day. Based on the lower mean water consumption and higher thyroid/ parathyroid and liver weights at 300 ppm, the study NOEL was considered to be 100 ppm, equivalent to 7.46 mg/kg/day.

| Blank & Thake, 1984
Groups of 15 male and 15 female Sprague Dawley rats were exposed via whole-body inhalation (6 hr per day, 5 days per week) to acetone cyanohydrin (98.5% purity) at target concentrations of 0, 10, 30, and 60 ppm for a period of approximately 14 weeks (69 total exposure days). The animals were observed daily for clinical signs and survival; body weights were assessed weekly. Samples were collected for urinalysis the day prior to necropsy and for hematologic and clinical chemistry evaluation (including thyroid hormone [T3 and T4]) at the time of necropsy. The animals were sacrificed over a period of 3 days (five/sex per group on each day). Organs T A B L E 8 Mouse sperm parameters (mean ± standard error) in control and top three NaCN exposure groups (data from NTP, 1993 and US EPA, 2010a).
Exposure ( were weighed and tissues, including the gonads (testes with epididymides and prostate), were collected for histopathologic examination (tissues fixed in 10% neutral formalin), with evaluations being conducted on the control and high-dose animals. Mean analytical concentrations were found to be 0, 10.1, 28.6, and 57.7 ppm. No mortality or treatmentrelated signs of toxicity were observed. Except on exposure Day 7, when body weights of the high-dose males were significantly lower than those of controls, no statistical differences in body weights were observed across dose groups over the course of the study. Some hematological differences from control were seen in the low and mid-dose groups, but these were considered to be within the range of normal biological variation. Importantly, no differences in T3 and T4 were observed across treatment groups. At necropsy, no significant differences from control in absolute or relative organ weights, including those of the male reproductive system (Table 15), were observed across the treatment groups, and no treatmentrelated histopathologic lesions were detected in the highdose group.
3.3.2 | Kier et al., 1985 Groups of 15 male Sprague Dawley rats were exposed via whole-body inhalation (6 hr per day, 5 days per week) to acetone cyanohydrin (98.5% purity) at target concentrations of 0, 10, 30, and 60 ppm for 69 days (48 total exposure days), after which they were mated to untreated females (three/male) while also continuing exposure. The duration of exposure prior to mating was selected to ensure exposure over a complete spermatogenic cycle in the rat. Mated females were sacrificed mid-gestation (approximately gestation days [GDs] 13-15) for analysis of pregnancy status and litter parameters. Males were observed daily; clinical signs and body weights were assessed weekly. They were sacrificed approximately 3 weeks after the last exposure. Although the male reproductive organs were collected at the end of study, it does not appear that these were weighed or examined histopathologically.
Mean analytical concentrations were found to be 0, 10.0, 28.5, and 57.2 ppm. All males survived to the end of study and no treatment-related alterations in clinical signs or statistically significant body weight differences from control were observed. The pregnancy rates of females mated to the treated males were 85%, 94%, 97%, and 95% in the 0, 10, 30, and 60 ppm groups, respectively. Additionally, the numbers of live implantations and resorptions (and total implantations) were similar across groups. Treatment-related gross findings were not observed in the testes; whether other male reproductive organs were examined grossly is not clear.

| Leuschner, 1989
A 90-day (13-week) drinking water study of potassium cyanide (KCN; >99% purity) was conducted in male Sprague Dawley rats under GLP. Singly housed rats were T A B L E 9 Exposures to NaCN in a 90-day drinking water study of male F344 rats (Tyner & Greeley, 2022 (Table 16); a number of control groups were also included in the study to address issues related to palatability and restricted water intake. These drinking water concentrations were selected based on the results of a 34-day drinking water study in which animals exposed to doses of ≥41.2 mg/kg showed reductions in water consumption; lower doses were not evaluated. In the definitive study, the paired drinking control was offered the "minimum amount of drinking water (in a test tube) which had been consumed on average by any one group in the previous test week" (Leuschner, 1989). With exception of the paired drinking control for which drinking solutions were changed out daily, all other drinking solutions were replaced twice weekly with fresh solutions. Additionally, in week 12, the high-dose of 160 mg/kg/day was decreased to 140 mg/kg/day due to excessive toxicity. For the purposes of this evaluation, results for the +10% alcohol groups (Groups 3 and 7) will not be discussed. The rats in study were observed daily; body weights and food consumption were measured weekly, and water consumption was recorded twice per week. From these data, compound intakes were calculated. Hematologic and clinical chemistry evaluations were conducted in weeks 6 and 13 (five animals per group); urinalyses were conducted on all rats at these same study times. At necropsy, organ weights (including those of the gonads) were recorded. All organs were then fixed in 10% formalin, but only select organs (including the testes) underwent microscopic evaluation. All treatment groups were compared statistically to the normal control; the highest exposure group (160/140 mg/kg/day) was also statistically compared to the paired drinking control.
One (1) normal control, 1 paired drinking control and 13 high-dose animals (4 on Day 7) died or were euthanized prematurely. Rats in the high-exposure group and the paired drinking control were reported to be severely emaciated and slightly sedated in the first study week but behaved normally thereafter. Water consumption values in all KCN treatment groups were significantly lower than those of control in a dose-related manner. These differences were most substantial in the first week of dosing, with water intake at 160/140 mg/kg/day being only 29% that of control. Thereafter, water intake improved, but remained low (in the high-dose group, 61%-78% of the normal control). Body weights were significantly reduced in the 80 and 160/140 mg/kg/day treatment groups compared to the normal control; in contrast, body weights of the high-dose group animals were similar to those of paired drinking control. At the mid and high doses, body weights were 12%-15% and 40%-46% lower than the normal control, respectively; furthermore, these reductions were clearly evident by the end of the first week of treatment and maintained throughout the study. While food consumption values at the low and mid-dose were similar to those of the normal control, food consumption in the high-dose group was significantly higher (rather than lower), but generally similar to that of the paired drinking control. These data indicate that the body weight reductions were not related to reduced food intake, but rather because of lower water consumption.
No treatment-related differences in hematologic or clinical chemistry parameters were reported; rather, upon comparison to the paired drinking control, any differences noted in the KCN-treated groups versus the normal control were attributed to reduced water intake. At necropsy, absolute testis weights were slightly reduced at the mid-dose, and even more so at the high dose, compared to the normal control (Table 17); relative weights were increased at the high dose versus the normal control. However, the differences seen in the treated groups were due primarily to reduced water consumption, as both absolute and relative testis weights at the high dose were similar to those of the paired drinking control. No T A B L E 1 1 Mean water consumption and NaCN intake in male F344 rats a (Tyner & Greeley, 2022 T A B L E 1 2 Thyroid hormone levels (mean ± standard deviation) in male rats in the control and NaCN exposure groups (data from Tyner & Greeley, 2022).
Exposure ( T A B L E 1 3 Terminal absolute and relative male rat reproductive organ weights (mean ± standard deviation) in the control and NaCN exposure groups (data from Tyner & Greeley, 2022

Note:
No statistically significant differences between treatment groups and normal control or water-restricted control.
a Percent of negative control/% of water restricted control.
b Absolute weights in grams; relative weights in mg/g body weight.
T A B L E 1 4 Sperm parameters (mean ± standard deviation) in male rats at termination and recovery in the control and NaCN exposure groups (data from Tyner & Greeley, 2022 treatment-related histopathologic alterations were observed in the testes. However, erosive alterations were observed in the stomachs of animals from the high-dose group as well as the paired drinking group.

| Results analysis
Based on data from the 13-week drinking water studies of NaCN, the NTP (1993) concluded that cyanide exposure in the drinking water may be associated with subtle yet significant alterations in male reproduction. Although the statistically significant changes in rat sperm motility were judged not biologically important, the reduced rat epididymal weights and testicular sperm counts at the highest exposure concentration in the drinking water (300 ppm) were deemed likely adverse. NTP further noted that, while the changes were probably not of a degree to adversely affect rat reproductive function, humans are considered generally more sensitive. Therefore, the potential for adverse effects of cyanide exposure on human reproductive function could not be discounted. The US EPA, in its 2010 IRIS assessment (US EPA, 2010a), used these data to set a chronic oral reference dose for cyanide. US EPA considered 30 ppm to be the lowest observed adverse effect level (LOAEL) in the rat based on significantly reduced absolute and relative cauda epididymal weights. They similarly called the LOAEL in the mouse based on reduced relative cauda epididymal weights at 100 ppm. Ultimately, US EPA selected decreased cauda epididymal weights as the critical effect of cyanide exposure based on results of the NTP study. Using benchmark dosing and application of 3,000-fold uncertain factors (10Â for interspecies extrapolation, 10Â for human variability, 10Â for subchronic-tochronic extrapolation, and 3Â for database deficiencies), the Agency derived an oral RfD value of 6.3 Â 10 À4 mg/kg/day. However, as noted previously, the Agency's confidence in the RfD was low to medium (US EPA, 2010b).
In setting the IRIS RfD for cyanide, the Agency noted that "the observed reproductive effects following exposure to cyanide may be mediated through decreases in thyroid hormones mediated through the cyanide metabolite thiocyanate" (US EPA, 2010a). While the NTP study showed increased urinary thiocyanate concentrations due to cyanide exposure-a finding corroborated by the increased plasma thiocyanate concentrations reported in the study by Tyner and Greeley (2022), EPA's conclusion was reached in the absence of concurrent thyroid organ weight or thyroid hormone data and despite the reported lack of histopathologic changes to the thyroid of either the rat or mouse. The NTP further noted no observed differences in thyroid follicle size, colloid staining, or follicular epithelium in rats after NaCN drinking water exposure (these parameters were not evaluated in the mouse). The newest NaCN drinking water study (Tyner & Greeley, 2022) found no treatment-related alterations in serum thyroid concentrations with NaCN drinking water exposure and also reported no alterations in thyroid follicular cell height or colloid volume.
Sodium iodide symporter (NIS) inhibitors such as thiocyanate, have the potential to affect thyroid activity by restricting the rate-limited transport of iodide into thyrocytes (Bizhanova & Kopp, 2009), thereby altering T3 and T4 biosynthesis. Hallinger et al. (2017) developed an in vitro screening approach for evaluating for the T A B L E 1 5 Absolute and relative rat testes plus epididymides (combined) weights (mean ± standard error) in the control and acetone cyanohydrin exposure groups (data from Blank & Thake, 1984 Standard error values, although reported in the study, were missing for relative testes weights for some exposure groups, and thus, are not included here. T A B L E 1 6 Dose groups and numbers of male rats per group in Leuschner (1989). potential of a compound to interact at the NIS. Perchlorate anion (ClO 4 À ), a well-known NIS inhibitor, was found by Hallinger et al. (2017) to have a greater affinity for NIS than thiocyanate (NIS selectivity score of 4.00 for perchlorate versus 2.26 for thiocyanate). Despite having a greater NIS inhibition potential than thiocyanate, perchlorate has not been shown to cause adverse effects on the male reproductive system. For example, Siglin et al. (2000) conducted a 13-week drinking water study of ammonium perchlorate in rats at doses of up to 10 mg/kg/day. In this study, perchlorate exposure resulted in significant changes in T3 and T4 levels at 90 days at all doses administered (≥0.01 mg/kg/day); TSH concentrations were affected in males at ≥0.20 mg/ kg/day. As expected, absolute thyroid weights were increased, and thyroid histopathologic changes noted at 10 mg/kg/day. Despite the substantial thyroid-related changes mediated through NIS inhibition, perchlorate exposure did not affect testes weights, histopathology of the male reproductive organs (testes, epididymides, prostate, and seminal vesicles), or on sperm parameters. York et al. (2001) conducted a two-generation reproductive study of ammonium perchlorate in rats at drinking water doses up to 30 mg/kg/day. The time of initiation of exposure prior to mating was not reported. Similar to the results reported by Siglin et al. (2000), thyroid weights were significantly increased in P-generation animals at 30 mg/kg/day and in F1-generation adults at ≥3 mg/kg/ day; thyroid histopathologic changes were seen at ≥3 mg/kg/day in both generations. Inconsistent thyroid hormone alterations were reported across doses and generations. Despite obvious treatment-related thyroid alterations, no adverse effects of treatment were observed on the weights of the male reproductive organs (testes, epididymides, prostate and seminal vesicles), sperm parameters, or fertility; histopathologic examination of the male reproductive organs was not reported. In sum, the lack of male reproductive system findings with perchlorate, a strong NIS inhibitor, argues against the hypothesis that the NTP (1993) NaCN study findings resulted from thiocyanate-mediated alterations in thyroid function. Because obvious restrictions in water intake were observed in the NTP (1993) study, the other possibility exists that the observed male reproductive organ weight changes were secondary to reduced water consumption. The impacts of reduced water consumption in rodents are, however, complex. Various regimens of restricted water intake have been associated with disparate effects on the male reproductive systems of rodents. Although information specific to B6C3F1 mice (the strain used in the NTP studies) is lacking, Nelson (1988) restricted daily water intake to 50% of the ad lib amount for 5 weeks and then further reduced the daily amount to 25% of ad lib in two strains of male mice. One strain was the fourth generation of wild-caught mice. The wild-type mice experienced a 12.6% reduction in body weight and substantially reduced reproductive organ weights. Spermatogenesis was reported to be significantly reduced. Inbred CF-1 mice, in contrast, were surprisingly resilient. CF-1 mice experienced a 25.5% reduction in body weights, but no reduction in testicular or epididymal weights, although seminal vesicle weight was reduced; spermatogenesis was not significantly affected. In a second experiment, Nelson et al. (1995) provided California mice with daily water amounts that were 50% of ad lib amounts over a 10-week period, resulting in a 10% decrease in body weights at study termination. Reproductive organ weights were reduced, including those of the epididymis (decreased 72%) and seminal vesicles (reduced 56%). Mean testicular sperm count was 76% of the control value. Thus, reduced water intake in mice tended to adversely affect the male reproductive organs, but the impact was both variable and strain-specific. In contrast, Amario et al. (1983) restricted water intake in adult male Wistar rats by providing water for only 1 hr per day during the morning for 28 days. Over the course of the experiment, the waterrestricted rats experienced a 30% decrease in body weight gain, but showed increases in the weights of the testes, seminal vesicles, liver, and brain relative to body weights; the authors ascribed the increased relative organ weights to decreased body weights because the absolute organ weights were minimally affected by water restriction. Consequently, a potential exists for restricted water intake to affect male reproductive organ weights in rodents, but the direction of the change and the specific impacts are not predictable.
To reduce this uncertainty and account for potential confounding, Tyner and Greeley (2022) added a paireddrinking water control group. Overall, reduced water intake due to NaCN in the drinking water was minimal in the Tyner and Greeley (2022) study. This finding appears to be due to the fact that, through the first half of the study, the volume of water administered to the paired drinking water control group was adjusted by 9 ml to account for residual water retained in the bottle cap. On Day 45, however, the adjustment volume was corrected to 5.5 ml, which improved the pairing, as shown by the similar water intakes seen in the paired drinking water control group and the 300-ppm group after Day 45. Other adjustments were made to correct for water bottle malfunction and animal behavior (i.e., playing with the sipper tubes), which also affected the water consumption values. Thus, the water consumption data from both the NTP (1993) and the Tyner and Greeley (2022) studies may be imprecise. Nevertheless, at 100 ppm, the reduction in water consumption from control levels was similar across the two studies: 10%-11%. At 300 ppm, however, NTP (1993) reported that, on average, male rats drank 18.3% less water per day than controls, while Tyner and Greeley (2022) only observed an 11% difference in water intake between the highest exposure group and the normal controls. In fact, water intake in the Tyner and Greeley (2022) study was generally lower than that reported in the earlier NTP (1993) study across the board. For example, in Tyner and Greeley (2022), rats in the normal control group drank approximately 19 ml/day versus approximately 25 ml/day reported to have been consumed by controls in the NTP (1993) study. Even when these values are converted to ml/kg/day based on body weights, water consumption among all groups in the Tyner and Greeley (2022) study was considerably lower than that in the NTP (1993) study. While the reasons for this difference are not known, they may be related to the group housing conditions in the NTP study versus individual caging in Tyner and Greeley (2022). Nevertheless, the fact that changes on the male reproductive organs were seen only in the study that also showed substantial restrictions in water consumption raises questions regarding their relation to treatment.
The lack of a treatment-related change in the cauda epididymal weights of the NaCN groups from the Tyner and Greeley (2022) study further raises the question of whether or not the changes seen in the NTP (1993) study (which EPA considered the most sensitive effect of cyanide exposure) were related to treatment. Table 18 below shows the absolute and relative (to body weights) cauda epididymal weights observed in the NTP study control groups and the range of values reported in the 30-300 ppm groups for both rats and mice. These values are compared to historical control data from fifty 13-week studies conducted in F344 rats and B6C3F1 mice at NTP during the same relative time period or slightly before the NTP NaCN study (Morissey et al., 1988). This comparison shows that, based on values obtained from 239 F344 male rats from 24 studies, both the absolute and relative cauda epididymal weights reported in the NaCN treatment groups of the NTP study, in addition to being affected in a nondose-related manner, were within expected historical control ranges. In contrast, the absolute and relative weights of these organs for the NTP control group were both higher than the reported historical control ranges. This observation, in combination with the fact that the changes observed with NaCN exposure were relatively minor, suggests that the findings were not treatment related; rather, the statistical significance appears to be a function of the abnormally high cauda epididymal weights reported for the NTP study control group.
In mice, the absolute weights of male reproductive organs in all groups from the NaCN study (including T A B L E 1 7 Absolute and relative male rat testis weights (mean ± standard deviation) in the control and KCN exposure groups (data from Leuschner, 1989).
Exposure ( Absolute weights in grams; relative weights in mg/g body weight.
c Standard deviation values were not reported, and statistical significance was not assessed for relative organ weights in Leuschner, 1989.

Note:
Values that were statistically signficantly different from control were bolded.
control) were within the reported historical control range (Morissey et al., 1988). While the relative weights of the study control group were within the expected range, the relative values reported in the 30-300 ppm groups were below the range of values typically reported in control mice. However, as shown in the studies of Nelson (1988), Nelson et al. (1995) and Amario et al. (1983), the impact of water restriction on male reproductive organ weights varies depending on the specific strain of mouse. The only other statistically significant changes observed with NaCN exposure in the NTP (1993) study in rats were a decrease in the number of spermatid heads per testis and per ml of homogenate at 300 ppm. Although the exact methods used were not described, it is likely that the whole testis (after removal of the tunica albuginea) was homogenized in a fixed volume of solution. Thus, these two parameters are basically different measures of the same endpoint and the divergence from control in each exposure group was the same for both parameters. Testicular size can vary among animals and can be affected by temporary conditions, such as inflammation or edema (Meistrich, 1989). Consequently, the number of sperm normalized to a gram of tissue (e.g., number of sperm heads per gram of testis) can be a more informative index of potential reproductive toxicity (Meistrich, 1989). In the NTP (1993) study, the numbers of spermatid heads normalized per gram of testis in rats showed no impact of cyanide exposure on sperm production (Table 8). This finding was confirmed in the Tyner and Greeley (2022) study, which found no statistically significant decrease in the number of sperm per gram of testis at study termination (Table 14). Importantly, the numbers of sperm per gram of epididymis did not differ significantly between the control and NaCN-exposed rats in the NTP (1993) study, nor at both the terminal and recovery euthanasias in Tyner and Greeley (2022), further demonstrating that sperm production is not affected adversely by exposure to 300 ppm cyanide in the drinking water. Additionally, sperm parameters were unaffected by NaCN exposure in the mouse (NTP, 1993). Thus, the differences seen in the rat sperm parameters at 300 ppm are unlikely to be due to cyanide treatment. Table 19 shows the cyanide doses for the studies discussed herein. Despite the administration of a diverse array of cyanide compounds by two different routes of exposure (drinking water and inhalation), the cyanide doses delivered in these studies-with the exception of those applied in Leuschner (1989)-are all of a similar magnitude or range. The doses for Leuschner (1989) are reported in the study both as 40, 80, and 160/140 mg KCN/100 ml/day and as 40, 80, and 160/140 mg/kg/day. If one assumes an average daily water intake of 25 ml and a mean rat body weight of 250 g, these doses roughly translate as reported; however, it is unlikely that these T A B L E 1 8 Rat and mouse absolute and relative cauda epididymal weights from the NTP NaCN drinking water study and contemporaneous historical control data (reported in Morissey et al., 1988 Doses calculated using an average rat body weight of 350 g in the study of Blank and Thake (1984) and a rat minute volume calculated based on information from OEHHA (2018).
water intake and body weight values apply across all dose groups. Further, in a two-week dose range-finding study reported by NTP (1993), the rats that received 1,000 ppm NaCN weighed 50% to 86% less than controls and those that received 3,000 ppm actually lost weight compared to their initial weights. These data suggest that it would be difficult to sustain similar doses for a 13-week period, as reported in Leuschner (1989). In our data quality review, the Leuschner (1989) study was a priori assigned a ToxR-Tool score of 1 because of its use of appropriate control groups, large numbers of animals and extensive data reporting. In retrospect, this study should be downgraded due to uncertainties regarding the calculation of the doses administered. Furthermore, the study provides limited information of value to our assessment: only reduced testis weights (without confirmatory histopathological evaluation) that were shown to be related to water restriction. Although conducted via inhalation rather than oral drinking water exposure, the studies of Blank and Thake (1984) and Kier et al. (1985) appear to have administered cyanide doses that were similar to those reported in both the NTP (1993) rat study and the Tyner and Greeley (2022) study. These studies also benefit from the fact that water restriction due to unpalatability was not an issue. Combined, the two inhalation studies showed no impact of cyanide exposure on testicular weights, adnexal male reproductive organ weights, or male fertility and mating parameters. Additionally, T3 and T4 levels were unaffected by treatment. While these data support the absence of alterations in the male reproductive system at these cyanide doses, they also have limitations. For one, neither study reported any adverse or otherwise treatmentrelated effects, which suggests that doses were not of sufficient magnitude for a toxicology safety study. Additionally, neither study reported cauda epididymal weights, the endpoint that EPA selected as most sensitive to cyanide exposure.
In establishing the RfD for cyanide, EPA cited data from a study in dogs (Kamalu, 1993) as additional support for the male reproductive changes reported in the NTP (1993) study; this study is discussed briefly in the Appendix S1. In Kamalu (1993), dogs were fed control diets, cassava or NaCN-formulated feed for 14 weeks. Dogs that were fed the cassava showed some abnormal germ cells and other testicular histopathologic changes; those fed diets supplemented with NaCN were reported to have reduced stage eight testicular tubules and substantial testicular histopathology. In our data quality assessment, the Kamalu (1993) study only received a ToxRTool score of 3 due to numerous design flaws and deficiencies, many of which are outlined in ECETOC (2007) and that affect the study's utility for risk assessment. For one, the control animal data show evidence that the control diets, despite having been supplemented with iodine, were still iodine deficient. Consequently, there may have been dietary effects on thyroid function irrespective of CN exposure. Most concerning, the dogs used in the study were mongrels bought at a village market and had to undergo 4 months of treatment for parasites before the study could begin; also, the ages, sizes, and number of dogs per group are unknown. Because different breeds can exhibit substantial differences in thyroid hormone turnover (ECETOC, 2007) and other biological functions, the use of dogs of undefined origin may have confounded the study results. Concerns have also been raised regarding the statistical analyses conducted (ECETOC, 2007). Furthermore, no information was reported regarding the source and purity of the NaCN used in the study. Thus, these data-which are incongruent with the data reported in all the studies that received ToxRTool scores of 1 (including the NTP study)-are given little to no weight in our evaluation.

| SUMMARY AND CONCLUSIONS
This report is an updated assessment of the conclusions from U.S. EPA and California OEHHA on the potential for male reproductive system toxicity due to cyanide exposure. This assessment focuses on the recently completed NaCN drinking water study by Tyner and Greeley (2022) as well as the results of previously considered studies. The 1993 NTP 13-week drinking water study of NaCN in rats and mice reported significantly decreased drinking water intake, especially in the high-dose group (300 ppm; 18% decrease compared to control), and concurrently reduced mean absolute and relative weights of the cauda epididymis at the top doses (≥ 30 ppm) in both species compared to controls (NTP, 1993); some testicular sperm parameters also differed from concurrent control values in rats (but not mice) at 300 ppm. These data were considered indicative of an adverse effect on the male reproductive system and used by the US EPA to derive an RfD for cyanide as well as by California OEHHA to judge cyanide to be a male reproductive toxicant.
An updated 13-week drinking water study was recently completed by Tyner and Greeley (2022) that utilized the same strain of rats and NaCN drinking water concentrations as in the NTP study. The Tyner and Greeley (2022) study incorporated enhanced design features, including a paired drinking water control group, fixation of reproductive organs in modified Davidson's solution, assessment of thyroid parameters (organ weights, histopathology, and hormone levels), and a posttreatment recovery period. Tyner and Greeley (2022) found a lesser effect of NaCN on mean water consumption than in the NTP study (at 300 ppm, an 11% decrease in water consumption compared to control versus an 18% decrease in the NTP study). NaCN exposure was not associated with changes in sperm parameters or male reproductive organ weights, including that of the cauda epididymis. Additionally, no changes in the thyroid hormone levels were seen over the course of the study, thus negating the hypothesis posed by US EPA that adverse effects of NaCN on the male gonadal system may be mediated through the interactions of thiocyanate with the thyroid hormonal axis.
Evaluation of the NTP study results in context of contemporary historical control data (Morissey et al., 1988) reveals that the values for the absolute and relative cauda epididymal weights of the NaCN-treated rats were within the expected range; in contrast, the values for the control rats were higher than the upper range of the historical control data, suggesting that the nondose related statistical difference reported in the NTP study was due to divergence of the control value from the expected rather than to an effect of treatment on the experimental groups. For mice, the control cauda epididymal weights were within the reported historical control range (Morissey et al., 1988), and other data support a potential influence of water restriction on this measure, which likely explains the lower cauda epididymal weights with NaCN treatment (relative weights were below the historical control range at ≥100 ppm).
With regard to rat sperm parameters, the least confounded measures reported in the NTP study were unaffected by NaCN exposure and the absence of any change in sperm was confirmed by Tyner and Greeley (2022) as well as in mice in the NTP study. Finally, none of the other high-quality studies of cyanide reported adverse effects of treatment on the male reproductive system. However, it is also recognized that these other studies did not evaluate the same sensitive endpoints (i.e., cauda epididymal weight, sperm parameters) measured in the NTP and Tyner and Greeley studies.
In light of the lack of male reproductive toxicity findings in the most recent 13-week drinking water study of NaCN, and due to limitations of the earlier NTP (1993) study, re-evaluation of the human health assessment of NaCN on the male reproductive system is needed.