Developmental Toxicity Study of Pure trans -Capsaicin in Rats and Rabbits

Abstract

Human environmental and dietary exposure to trans-capsaicin—the pungent ingredient in chili peppers—is ubiquitous. Moreover, based on the highly selective agonism of trans-capsaicin for TRPV1 receptors, drug products containing high concentrations of trans-capsaicin are under development as analgesics. For instance, a high-concentration (8% w/w) pure trans-capsaicin dermal patch (designated NGX-4010) is in advanced clinical evaluation for the management of neuropathic pain of peripheral origin. Our objective was to investigate effects of trans-capsaicin on embryo/fetal development, consequent to maternal exposure, from implantation to closure of the hard palate. trans-Capsaicin was delivered systemically by means of either a patch [NGX-4010 (25, 37.5, or 50 cm²)] to pregnant Sprague-Dawley rats on days of presumed gestation (DGs) 7 through 17, or via a 10% w/v capsaicin liquid formulation (CLF), at dosages of 3, 6.5 or 13 μl/cm² applied to a 200-cm² area on the back on DGs 7 though 19 to timed-mated New Zealand white rabbits. In rats, the maternal no-observable-effect level (NOEL) was less than 25 cm² but no cesarean-sectioning or litter parameters were affected by application of NGX-4010 at patch sizes as high as 50 cm². The only test article–related observations were delays in skeletal ossification, evident as significant reductions in the average number of metatarsals and ossified hindlimb and forelimb phalanges that occurred in the 50 cm² NGX-4010 dose group. Although the values for ossified metatarsals were outside the historical control range, ossified hindlimb and forelimb phalanges were within historical control ranges. No other gross external, soft tissue, or skeletal fetal alterations (malformations or variations) were caused by application of the NGX-4010. In rabbits, the maternal NOEL was less than 3 μl/cm² CLF (or 0.3 mg/cm² trans-capsaicin) per 200 cm², but no cesarean-sectioning or litter parameters were affected. No fetal alterations (malformations or variations) were caused by dosages of CLF as high as 13 μl/cm² (or 1.3 mg/cm² trans-capsaicin). Taken together, these data suggest that trans-capsaicin should not be considered a developmental toxicant.

Overall, herbal medicines or natural products are generally regarded as safe, even though there is often no scientific basis for that belief. These natural therapies generally contain a mixture of active ingredients that may be similar to a single purified or synthesized component, and yet the safety profile for a mixture of these ingredients can be very different from that single active ingredient. Therefore, any herbal medication or natural product-based food additive should be carefully evaluated for its potential effect on fetal development before consumption during pregnancy can be recommended.

The trans-geometric isomer of capsaicin (or trans-8-methyl-N-vanillyl-6-nonenamide) is the most abundant pungent molecule in chili peppers and thus represents a primary ingredient in spicy foods. Although there are two geometric isomers of capsaicin, only trans-capsaicin occurs naturally (Cordell and Araujo 1993), and will henceforth be termed simply capsaicin. The capsaicin content of chili peppers ranges from 0.1% to 1% w/w (Govindarajan and Sathyanarayana 1991). In addition to its role as a food additive throughout the world, there is also substantial human exposure to capsaicin in the form of nonprescription (in the United States) or prescription (in the European Union) topical analgesics, and self-defense products (e.g., pepper spray). Moreover, capsaicin is used extensively in human experimental models of pain (Drewes et al. 2003) and cough (Dicpinigaitis and Alva 2005).

Capsaicin is a highly selective agonist for the transient receptor potential vanilloid receptor 1 (TRPV1; also known as VR1 according to older nomenclature) (Caterina et al. 1997). TRPV1 is a ligand-gated, nonselective, cation channel preferentially expressed in small-diameter, primary afferent neurons (C-fibers and Aδ-fibers), especially nociceptive sensory nerves. TRPV1 responds to noxious stimuli including capsaicin, heat, and extracellular acidification, and integrates simultaneous exposures to these stimuli (Tominaga et al. 1998). Based on the highly selective agonistic property of capsaicin towards TRPV1 receptors, drug products containing pure synthetic trans-capsaicin are under evaluation as topical and injectable analgesic therapies (Bley 2004).

Pharmacological studies of capsaicin with toxicological implications began when De Lille and Ramirez (1935) reported that administration of a capsaicin extract into dogs produced a fall in blood pressure accompanied by a variable effect on respiration, an increase in salivary secretion, and a relatively small increase in gastric secretion. A similar effect of capsaicin extract was found in cats (Toh, Lee, and Kiang 1955). Since then many pharmacology studies have tested capsaicin at seemingly toxic dose levels; some of these have involved administration to pregnant animals. For instance, the effects of prenatal capsaicin treatment days of gestation ([DGs] 16 and 17 only) in rats on spontaneous activity, opiate receptor binding and acid phosphatase in the spinal cord were evaluated (Kirby, Gale, and Mattio 1982).

To date, there have been no published or well-controlled toxicology studies regarding the potential of capsaicin to produce developmental toxicity—including teratogenic effects—when exposure occurs during the period of organogenesis. Moreover, the capsaicin materials tested in all of the studies cited above were either natural extracts or impure mixtures, and may not exhibit the same toxicological effects as pure trans-capsaicin (Chanda et al. 2004, 2005). Although the exact content and nature of impurities in the test articles used in these and other studies with capsaicin are often not explicitly stated, a typical capsaicin extract is a mixture of trans-capsaicin (cis-capsaicin does not occur naturally [Cordell and Araujo 1993]) and other capsaicinoids (including norhydrocapsaicin, dihydrocapsaicin, homocapsaicin, homodihydrocapsaicin). The actual percentage of trans-capsaicin and other capsaicinoids will vary depending on the peppers used as the source and the method of extraction. In fact, the United States Pharmacopoeia defines capsaicin as a product which contains ≥55% capsaicin and the combination of capsaicin and dihydrocapsaicin to be ≥75%; total capsaicinoid content may be as little as 90% ( United States Pharmacopeia 2005). Additionally, pepper extracts are expected to contain chemical entities other than vanilloid compounds.

A high-concentration (8% w/w) pure trans-capsaicin patch (designated as NGX-4010) is undergoing advanced clinical evaluation for treatment of neuropathic pain of peripheral origin. NGX-4010 is designed to deliver trans-capsaicin into the skin without significant systemic exposure to humans. As a part of the regulatory toxicology package needed for NGX-4010, the studies reported here investigated the effects of pure trans-capsaicin on fetal development (consequent to maternal exposure) from implantation to closure of the hard palate. trans-Capsaicin was delivered transdermally by means of either NGX-4010 patches applied to pregnant Sprague-Dawley rats or a 10% w/v capsaicin liquid formulation (CLF) to timed-mated New Zealand white rabbits. As the permeation of capsaicin through rat and rabbit skin is much higher than for human skin (Fang et al. 1995), consistent systemic exposure was achieved.

The authors presented a preliminary summary of these results as a poster at the 45th Annual Meeting (June 2005) of the Teratology Society at St. Pete Beach, Florida.

MATERIALS AND METHODS

All animal usage was conducted according to ethical guidelines established under the Animal Welfare Act, 7 U.S.C. 2131 et seq. Before the initiation of studies the Institutional Animal Care and Use Committee (IACUC) of the Charles River Laboratories Preclinical Services (Horsham, PA) approved all protocols.

Rat Study

Before conducting the definitive study, pilot studies were conducted in female Crl:CD (SD)IGS BR VAF/Plus rats using similar doses of NGX-4010 that were to be used in the definitive developmental toxicology study, in order to demonstrate that plasma levels of capsaicin were achieved. Serial blood samples (approximately 1 ml each) were drawn from the tail vein of the rats after single application of NGX-4010 at different time points up to 24 h post dose. Rats were treated and housed in the same manner as they were for the definitive developmental toxicology study. Details of materials and methods used in these pilot studies are not provided here.

Crl:CD (SD)IGS BR VAF/Plus rats from Charles River Laboratories (Raleigh, NC) were used for the definitive study. Male rats were used only for the purposes of breeding. Upon arrival, rats were assigned to individual housing on the basis of computer-generated random units. After acclimation for 7 days, 140 virgin female rats (approximate age upon arrival was 62 days) were placed into cohabitation with 140 breeder male rats, 1 male rat per female rat. The cohabitation period consisted of four days. On each of the 4 days of cohabitation, female rats were checked for spermatozoa in a smear of vaginal contents and for a copulatory plug. Female rats with spermatozoa observed in a smear of the vaginal contents and/or a copulatory plug observed in situ were considered to be at DG 0 and assigned to individual housing.

Healthy, mated female rats were assigned to four dosage groups (groups I through IV), by using a computer-generated randomization procedure based on the body weights recorded on DG 0. Each dosage group consisted of 25 rats. All female rats were acclimated to Elizabethan collars for 1, 2, and 4 h on 3 consecutive days during acclimation. The vehicle (group I) and NGX-4010 (groups II through IV) patches were applied to the backs of the rats once daily on DGs 7 through 17, for at least 3 hours each day. Rats in group I received a vehicle patch of approximately 50 cm² area. Rats in groups II, III, and IV received NGX-4010 patches of approximately 25, 37.5, and 50 cm², respectively. These topical dosages translate into 16, 24, and 32 mg/rat, respectively, for the 25, 37.5, and 50 cm² dosage groups. These calculations are based on the fact that NGX-4010 contains 640 μg/cm² of trans-capsaicin. Dosages were selected on the basis of a dose range-finding developmental toxicity study and other studies conducted with NGX-4010, which indicates that systemic (transdermal) exposure to capsaicin occurs in rodents.

At least 24 h before dosage application, Oster Golden A5 electrical clippers (Sunbeam Products, Boca Raton, Florida) with a size 40 blade were used to remove hair from the backs of rats. A ‘map’ of the clipped area was made for any rat with areas of apparent hair tufts, skin defects, or natural differences in skin colorization before administration. The application sites were demarcated with one mark tattooed at each corner of the site (AIMS Animal Identification and Marking System; AIMS, Piscataway, NJ; with AIMS Black Pigment No. 242). The rats were fitted with Elizabethan collars, and an anesthetic cream (Lidocaine 4%; L.M.X.4; Ferndale Laboratories, Ferndale, MI) was topically applied to the application site on all rats at approximately 32 μl/cm² for at least 1 h. Care was taken to avoid excess administration; just enough cream was applied to cover the patch application area only (plus 0.5 cm beyond the boundaries) and any excess was removed. Approximately 60 min later, the anesthetic cream was wiped off with dry gauze. The test article and vehicle patches were then applied to the backs of rats. The patches were secured using an elastic wrapping material (3M Vetrap Bandaging Tape and 3M Micropore Surgical Tape; 3M Animal Care Products and 3M Medical Division, St. Paul, MN) to overlay the patch. After at least 3 h of exposure (range was 3 to 3.5 h), the patches were removed from the backs of the rats, and the application sites were gently wiped with the cleansing gel (approximately 3 ml) and dry gauze. Cleansing gel consists of 89% polyethylene glycol and was prepared for NeurogesX, by a manufacturing facility according to current Good Manufacturing Practice (cGMP) specifications.

Rats were observed for viability at least twice each day of the study and for clinical observations and general appearance weekly during the acclimation period and on DG 0. The rats were also examined for clinical observations, abortions, premature deliveries and deaths before dosage administration, at the time of the 3-hour post-exposure cleansing, and once daily during the post-dosage period. Before the first daily dosage application, and at 24-h intervals each day thereafter, each skin site was observed for signs of skin irritation and graded using the end points described by Draize (Draize, Woodard, and Calvery 1944) and the National Research Council (National Research Council 1977). Skin sites were also observed for signs of skin irritation and graded once daily during the postdosage period until precluded by hair growth. Body weights were recorded weekly during the acclimation period, on DG 0, and daily during the dosage and postdosage periods. Feed consumption values were recorded on DGs 0, 7, 10 12, 15, 18, and 21.

All female rats were sacrificed by carbon dioxide asphyxiation on DG 21, cesarean-sectioned, and a gross necropsy of the thoracic, abdominal, and pelvic viscera was performed. Uteri of apparently nonpregnant rats were examined while being pressed between glass plates, in order to confirm the absence of implantation sites. The number and distribution of corpora lutea were recorded. The uterus of each rat was excised and examined for pregnancy, number and distribution of implantation sites, live and dead fetuses, and early and late resorptions. An early resorption was defined as one in which organogenesis was not grossly evident. A late resorption was defined as one in which the occurrence of organogenesis was grossly evident. Dead fetuses and late resorptions were differentiated by the degree of autolysis present; marked to extreme autolysis indicated that the fetus was a late resorption. Placentae were examined for size, color, and shape.

Each fetus was removed from the uterus, placed in an individual container, and individually identified with a tag noting the study number, litter number, uterine distribution, and fixative. Each fetus was subsequently evaluated following normally accepted procedures for visceral (Staples 1974) and skeletal (Staples and Schnell 1964) examinations.

All data were tabulated, summarized, and statistically analyzed using the Argus Automated Data Collection and Management System and Quattro Pro 8. Clinical observations and other proportional data were analyzed using the variance test for homogeneity of the binomial distribution (Snedecor and Cochran 1967). Continuous data (e.g., maternal body weights, body weight changes, feed consumption values, and litter averages for percent male fetuses, percent resorbed conceptuses, fetal body weights, and fetal anomaly data) were analyzed using Bartlett’s test of homogeneity of variances (Sokal and Rohlf 1969) and the analysis of variance (Snedecor and Cochran 1967), when appropriate (i.e., Bartlett’s test was not significant [p > .001]). If the analysis of variance was significant (p ≤ .05), then Dunnett’s test (Dunnett 1955) was used to identify the statistical significance of the individual groups. If the analysis of variance was not appropriate (i.e., Bartlett’s test was significant at p ≤ .001), then the Kruskal-Wallis test (Sokal and Rohlf 1969) was used, when less than or equal to 75% ties were present. In cases in which the Kruskal-Wallis test was statistically significant (p ≤ .05), Dunn’s method of multiple comparisons (Dunn 1964) was used to identify the statistical significance of the individual groups. If there were greater than 75% ties, Fisher’s exact test (Siegel 1956) was used to analyze the data. Count data obtained at cesarean-sectioning of the dams were evaluated using the procedures described above for the Kruskal-Wallis test.

Rabbit Study

Before the definitive study was initiated in rabbits, a pilot study compared plasma capsaicin levels in female rabbits after application of either NGX-4010 or CLF (10% w/v capsaicin in diethylene glycol monoethyl ether, which is the vehicle used to dissolve capsaicin for manufacture of NGX-4010) at 15 μl/cm² over 200 cm². There was no quantifiable capsaicin in the plasma of rabbits in the group receiving the NGX-4010 at any of the time points. In contrast, for the group receiving CLF, capsaicin plasma levels were first detectable between 5 and 15 min post application, were highest at approximately 1 h post application, and were generally still detectable at 3 h post application. Based on these data, administration of CLF was selected for the developmental toxicity study in rabbits. A dose range-finding developmental toxicity study was then conducted in rabbits with the CLF. Dose levels of 3, 10, and 30 μl/cm² over 200 cm² application area were used in the study. Blood samples (approximately 3 ml each) were drawn from the medial auricular artery at different time points to demonstrate exposure to capsaicin. On the basis of the information gathered for maternal toxicity from this study, a dosage of 13 μl/cm² of CLF was chosen as the highest dose level for the definitive study. Details of materials and methods used in these pilot studies are not provided here, as the protocol was essentially the same as for the definitive study.

Consequently, to test the potential teratogenicity of systemic trans-capsaicin in rabbits, dosages were selected on the basis of the dose range-identifying study in which 10% w/v capsaicin dissolved in diethylene glycol monoethyl ether was applied once daily for 3 h in duration on DGs 7 through 19 at dosages of 3, 10, and 30 μl/cm² over 200 cm²area on the back. Two of the eight does in the 30 μl/cm² dosage group were found dead. Increased incidences of adverse skin reactions occurred in all treated groups, but the occurrences of these skin observations were not necessarily dosage dependent. Body weight gains were reduced for the entire dosage period in the 10 and 30 μl/cm² dosage groups and for the entire gestation period in the 30 μl/cm² dosage group. Absolute and relative feed consumption values were reduced in the 30 μl/cm² dosage group for the entire dosage period. However, no cesarean-sectioning or litter parameters were affected by administration of 10% w/v capsaicin even at 30 μl/cm², and no fetal gross alterations occurred. Based on these results, and the appearance of plasma concentrations at both 10 and 30 μl/cm², dosage levels of 3, 6.5, and 13 μl/cm² CLF over a 200-cm² area were selected for the full study.

Eighty timed-mated New Zealand white [Hra:(NZW)SPF] rabbits (Covance Research Products, Denver, PA), approximately 6 months of age upon arrival, were used in the definitive study. Upon arrival, rabbits were assigned to four dosage groups (groups I through IV), 20 rabbits per dosage group, using a computer-generated randomization procedure which was based on body weights recorded by and at the Supplier on DG 0. The test article formulation, 10% w/v capsaicin (trans-capsaicin powder with ≥99% purity, provided by NeurogesX), or the vehicle (diethylene glycol monoethyl ether, 99%; Sigma-Aldrich, St. Louis, MO) was administered topically once daily for 3 h to timed-mated rabbits on DGs 7 through 19. The dose groups were 13 μl/cm² (group I; vehicle only), 3 μl/cm² (group II; 0.3 mg/cm² capsaicin), 6.5 μl/cm² (group III; 0.65 mg/cm² capsaicin), and 13 μl/cm² (group IV; 1.3 mg/cm² capsaicin), and dose administration took place at approximately the same time each day over a 10 × 20-cm area on the backs of the rabbits.

Formulations of capsaicin in diethylene glycol monoethyl ether were prepared at the testing facility twice and aliquotted for daily use on the day of preparation prior to storage. Prepared formulations were stored at room temperature, protected from light in amber glass containers. Concentration analyses indicated that all samples for the 10% w/v CLF were within ±10% of the target (data not shown). All female rabbits were acclimated to the Elizabethan collars for 1, 2, and 4 h on 3 respective days during acclimation. At least 24 h before dosing, Oster Golden A5 electrical clippers with size 10 and 30 blades were used to remove hair from the backs of the rabbits. Before administration, a ‘map’ of the clipped area was made for each rabbit, with areas of apparent hair tufts, skin defects, or natural differences in skin colorization noted. The administration site (approximately 10 × 20 cm) was demarcated with one mark tattooed at each corner of the site (AIMS [Animal Identification and Marking System], AIMS Black Pigment No. 242). The rabbits were fitted with Elizabethan collars, and the anesthetic cream (L.M.X.4) was topically applied to the application site on all rabbits at an approximate volume of 40 μl/cm² for at least 1 h. Care was taken to avoid excess administration; just enough cream was applied to cover the application area only (plus 0.5 cm beyond the boundaries), and any excess was removed. Approximately 60 min later, the anesthetic cream was wiped off with dry gauze. The test article CLF (3, 6.5, and 13 μl/cm²) or the vehicle alone (13 μl/cm²) was then topically administered to the rabbits. No overwrap material was used. After at least 3 h of exposure, the administration skin sites were gently wiped with the cleansing gel (approximately 10 ml) and dry gauze, and the Elizabethan collars were removed. The female rabbits were naturally bred at the supplier by breeder male rabbits of the same source and strain before shipment to the testing facility. The day of mating was considered to be DG 0. The rabbits were mated on 4 consecutive days and shipped to the testing facility to arrive on DGs 1, 2, 3, or 4. All rabbits were observed for viability at least twice each day of the study and for general appearance at least once during the predosage period. The rabbits were also examined for any effects of the test article, including adverse clinical observations, abortions, premature deliveries, or deaths before dosage administration, at the time of the 3-h postexposure cleansing, and once daily during the postdosage period. Before the first daily dosage application, and at 24-h intervals each day thereafter, each skin site was observed for signs of skin irritation and graded using the end points as described under the rat study in this paper. Skin sites were also observed for signs of skin irritation and graded once daily during the postdosage period until precluded by hair growth. Body weights were recorded on DG 0, the day of arrival at the testing facility, and daily during the dosage and postdosage period. Feed consumption values were recorded daily throughout the study.

All surviving rabbits were sacrificed on DG 29 by intravenous administration of Beuthanasia-D Special euthanasia solution (Schering-Plough Animal Health, Union, NJ) and cesarean-sectioned. A gross necropsy of the thoracic, abdominal and pelvic viscera was performed as described in detail for the rat study in this paper. Each fetus was evaluated following normally accepted procedures for visceral (Staples 1974) and skeletal (Staples and Schnell 1964) examinations. Rabbits that were found dead or were sacrificed because of moribund condition or abortion were examined for the cause of death or moribund condition on the day the observation was made. Pregnancy status and uterine contents were recorded. Aborted fetuses and/or conceptuses in utero were examined to the extent possible, using the same methods described for term fetuses.

All data were tabulated, summarized, and analyzed statistically using the same methods described for the rat study.

RESULTS

Rat Study

Systemic Exposure

The average maximum plasma concentrations of capsaicin at the 25 and 50 cm² dose levels were approximately 32 and 135 ng/ml, respectively, at the time of patch removal. Plasma levels were generally highest at the time of patch removal and decreased gradually thereafter.

Maternal Effects

Grade 1 erythema occurred in significantly increased (p ≤ .01) numbers of rats in the test article-treated groups; however, the incidences were generally comparable among the three treatment groups. Lacrimation occurred in significantly increased (p ≤ .05 or p ≤ .01) numbers of rats in the 37.5 and 50 cm² test article–treated groups. A significantly increased (p ≤ .01) number of rats in the 50 cm² test article-treated group were cold to touch. Soft or liquid feces and a peri-vaginal substance (red, brown, or black) occurred in slightly increased numbers of rats in the 50 cm² test article-treated group; however, the incidences were not statistically significant. Observations of urine-stained abdominal fur and ungroomed coat occurred in increased or significantly increased (p ≤ .01) numbers of rats in the three treated groups and were considered possibly related to an aversion to normal grooming behavior due to residual capsaicin on the skin. A summary of skin reactions and clinical observations are presented in Table 1. All other clinical observations were considered unrelated to NGX-4010 exposure because (1) the incidences were not dosage dependent; and/or (2) the observations occurred in only one or two rats per group. These observations included chromodacryorrhea, chromorhinorrhea, areas of red fur, ptosis, yellow fore- and/or hindpaws, light brown feces, localized alopecia (limbs), yellow perinasal substance, red substance in the cage pan, abrasion on the back, excess salivation, grade 1 flaking, and scabs or red skin at the application site. The numbers of rats with yellow fore- and/or hindpaws was significantly increased (p ≤ .01) in the 25 and 37.5 cm² test article–treated groups, a nondosage dependent observation considered unrelated to the test article. Persistent adverse clinical observations were confirmed at necropsy. No additional gross lesions were identified.

Absolute and relative feed consumption values (data not shown) were increased or significantly increased (p ≤ .05 or p ≤ .01) in all three treated groups on DGs 7 to 10 and for the entire dosage period (calculated as DGs 7 to 18). Absolute and relative feed consumption values were also significantly increased (p ≤ .01) on DGs 10 to 12 in the 37.5 and 50 cm² test article-treated groups. Absolute and relative feed consumption values were significantly increased (p ≤ .01) in the 50 cm² test article-treated group for the entire gestation period after the initiation of dosage (DGs 7 to 21) and the entire gestation period (DGs 0 to 21). Body weight gains tended to be increased throughout the study in the treated groups as compared to the vehicle-control group; however, no statistically significant differences occurred (data not shown). Body weight gains for the entire dosage period (calculated as DGs 7 to 18) were 117%, 108%, and 108%, of the vehicle-control group value in the three test article-treated groups. Body weights were comparable among the four dosage groups.

Cesarean Sectioning and Litter Observations

Cesarean-sectioning observations were based on 24 (96.0%), 24 (96.0%), 25 (100%), and 23 (92.0%) pregnant rats with one or more live fetuses in groups I through IV, respectively. No cesarean-sectioning or litter parameters were affected by application of NGX-4010 at areas as high as 50 cm². The litter averages for corpora lutea, implantations, litter sizes, live fetuses, early and late resorptions, fetal body weights, percentage of resorbed conceptuses, and percent live male fetuses were comparable among the four dose groups and did not significantly differ. No dam had a litter consisting of only resorbed conceptuses, and there were no dead fetuses. All placentae appeared normal. A summary of cesarean-sectioning and litter observations are presented in Table 2.

Fetal Alterations

Fetal alterations were defined as (1) malformations (irreversible changes that occur at low incidences in this species and strain); or (2) variations (common findings in this species and strain and reversible delays or accelerations in development). Litter averages were calculated for specific fetal ossification sites as part of the evaluation of the degree of fetal ossification. Fetal evaluations were based on 355, 329, 374, and 340 live, DG 21 cesarean-delivered fetuses in 24, 24, 25, and 23 litters in the 50 (vehicle), 25, 37.5, and 50 cm² groups, respectively. Each of these fetuses was examined for gross external alterations. Of these respective fetuses, 170, 159, 182, and 164 fetuses were examined for soft tissue alterations, and 185, 170, 192, and 176 fetuses were examined for skeletal alterations and fetal ossification site averages.

Summaries of fetal skeletal alterations and fetal skeletal ossification sites are presented in Tables 3 and 4, respectively. Delays in skeletal ossification occurred in the 50 cm² test article–treated group. An incompletely ossified 1st sternebra occurred in 6, 0, 8, and 12 fetuses from 6, 0, 7, and 9 litters in the four respective groups. The fetal incidence of incompletely ossified 1st sternebra was significantly increased (p ≤ .05) in the 50 cm² test article–treated group. The fetal and litter incidences of this observation were significantly reduced (p ≤ .05 or p ≤ .01) in the 25 cm² test article-treated group, reflecting the increased incidence in the vehicle-control group and considered unrelated to the test article. The fetal and litter incidences of an incompletely ossified 1st sternebra in the 50 (vehicle), 37.5, and 50 cm² groups were all outside the ranges observed historically at the Testing Facility. The average number of ossified hindlimb and forelimb phalanges and metatarsals were significantly reduced (p ≤ .01) in the 50 cm² test article-treated group. However, the values for ossified fore- and hindlimb phalanges were within the ranges observed historically at the testing facility and probably were of no biological significance. In contrast, the values for ossified metatarsals were outside the historical control range in the 50 cm²dose group.

Rabbit Study

Plasma Levels

The first rabbit pilot study which measured plasma levels of capsaicin after a single application of either NGX-4010 or CLF (15 μl/cm²) revealed no measurable values produced by NGX-4010, but an average peak plasma level of approximately 26 ng/ml (at 120 min post application) following CLF treatment. Based on this study, CLF was chosen for further investigation in rabbits.

In the second pilot study, the 3 μl/cm² dose group showed minimal systemic exposure on the first day of dosing (DG 7). However, the 10 and 30 μl/cm² groups displayed significant systemic capsaicin exposure, with average maximum plasma levels of 13 ng/mL and 54 ng/ml, respectively, at 3 h post application. Plasma levels remained almost unchanged at 6 h post application for both the groups. On the last day of dosing (DG 19), plasma levels of capsaicin were detected in all the dose groups at predose, indicating circulating levels of capsaicin remained from previous applications; predose levels were 57, 190, and 286 ng/ml in the 3, 10, and 30 μl/cm² groups, respectively. Postdose capsaicin levels remained around 100 ng/ml for the 3 μl/cm² group, starting 5 min post dose until the last time point measured (6 h post dose), with a general trend towards a slight decrease after the 3-h time point. For the 10 and 30 μl/cm² groups, average plasma levels increased to about 350 and 700 ng/ml, respectively, by 15 min post dose and generally remained around that level until 6 h post dose.

Maternal Effects

There were no test article-related deaths or abortions. Erythema (grades 1 and 2), flaking (grades 2 and 3), wrinkling and lesions at the application site occurred at statistically significant incidences (p ≤ .05 or p ≤ .01) in the 3, 6.5, and 13 μl/cm² dosage groups. However, these observations tended to occur at similar incidences in the three treated groups, with no apparent dosage-dependent differences. Additionally, the number of affected vehicle control group does with grade 1 erythema and grade 2 flaking were similar to the numbers of affected does in the treated groups (i.e., 16 in the vehicle group compared to 19 or 20 in the treated groups). A summary of skin reactions and clinical observations are presented in Table 5. All other skin reactions and all clinical observations were considered unrelated to capsaicin because (1) the incidences were not dosage dependent; (2) the observations occurred in only one or two rabbits in a group; and/or (3) the observations were related to an injury. These clinical observations included discolored application site, decreased motor activity, spastic paralysis, impaired or lost proprioceptive positioning, localized alopecia on the limbs, soft or liquid feces, vocalization, scant feces, no feces in the cage pan, ungroomed coat, emaciation, hypotension of the hindlimbs, enophthalmos, cold to touch, pale mucous membranes, hunched posture, dehydration, grade 3 erythema, grades 1 and 2 edema, and grade 1 flaking.

There were a total of four rabbits that were either found dead or sacrificed in moribund condition in the vehicle-control group. Out these four rabbits one rabbit (sacrificed moribund on DG 11) had a confirmed back injury, which probably occurred during dosing. The litter of this rabbit consisted of six embryos. One rabbit aborted and was sacrificed on DG 20 in this group. This rabbit generally lost weight and feed consumption values were reduced after DG 11. The stomach contained a trichobezoar; all other tissues examined appeared normal at necropsy. The litter consisted of four fetuses and eight late resorptions. All fetuses appeared normal at gross external and soft tissue evaluations. One rabbit was sacrificed due to moribund condition on DG 20. This rabbit lost weight after DG 9 and feed consumption values were reduced after DG 7. The stomach contained a trichobezoar; all other tissues examined appeared normal at necropsy. The litter consisted of 12 fetuses. All fetuses appeared normal at gross external and soft tissue evaluations. The fourth rabbit in the vehicle control group was sacrificed in moribund condition on DG 18. This rabbit lost weight and feed consumption values were reduced after DG 11. All tissues examined appeared normal at necropsy. The litter consisted of nine embryos that appeared normal at gross external examination. One rabbit was found dead on DG 21 in the 3 μl/cm² group. This rabbit had generally lost body weight and feed consumption values were reduced after DG 8. Necropsy revealed black areas on the mucosal surface of the stomach (ranging in size from pinpoint to 0.1 × 0.4 cm). All other tissues appeared normal. The litter consisted of seven late resoprtions and four early resorptions. One rabbit in the 13 μl/cm² group was sacrificed on DG 11 before administration of the 5th daily dosage because of a suspected back injury. This rabbit had decreased motor activity, spastic paralysis and impaired proprioceptive positioning on DG 11. This rabbit lost weight on DG 11 and feed consumption values were slightly reduced on DGs 7 to 8 and 9 to 10. All tissues examined appeared normal at necropsy; no fractures in either the spinal column or the hindlimbs were detected. The litter consisted of seven embryos.

Body weights and body weight gains were unaffected by dosages of capsaicin as high as 13 μl/cm² (data not shown). Values were generally comparable among the four dosage groups throughout the dosage and postdosage periods. A statistically significant increase (p ≤ .05) in body weight gain in the 6.5 μl/cm² dosage group on DGs 10 to 13 was considered unrelated to the test article because the increase was not dosage dependent and did not persist.

Absolute and relative feed consumption values were unaffected by dosages of capsaicin as high as 13 μl/cm² (data not shown). Statistically significant increases (p ≤ .01) in absolute and relative feed consumption values in the 6.5 μl/cm² dosage group on DGs 10 to 13 and 13 to 16 were considered unrelated to the test article because the increases were not dosage dependent.

Cesarean Sectioning and Litter Observations

Pregnancy occurred in 19 or 20 does in each dosage group. As a result of some unscheduled deaths as described under Maternal Effects, cesarean-sectioning observations were based on 15, 19, 19, and 18 pregnant does with one or more live fetuses in groups I through IV, respectively. No cesarean-sectioning or litter parameters were affected by dosages of capsaicin as high as 13 μl/cm². The litter averages for corpora lutea, implantations, litter sizes, live fetuses, early and late resorptions, fetal body weights, percent resorbed conceptuses, and percent live male fetuses were comparable among the four dosage groups and did not differ significantly. No doe had a litter consisting of only resorbed conceptuses, and there were no dead fetuses. All placentae appeared normal. A summary of cesarean-sectioning and litter observations are presented in Table 6.

Fetal Alterations

Fetal alterations were defined as previously described. Fetal evaluations were based on 133, 175, 147, and 159 live, DG 29 cesarean-delivered fetuses in 15, 19, 19, and 18 litters in the 13 (vehicle), 3, 6.5, and 13 μl/cm² dosage groups, respectively. Based on examination of each of these fetuses, no gross external, soft tissue, or skeletal fetal alterations (malformations or variations) were caused by dosages of capsaicin as high as 13 μl/cm². There were no dosage-dependent or biologically important differences in the litter or fetal incidences of any gross external, soft tissue or skeletal alterations. Summaries of fetal skeletal alterations and fetal skeletal ossification sites are presented in Tables 7 and 8, respectively.

DISCUSSION

The data presented here constitute the first reported evaluation of trans-capsaicin in standard developmental toxicity models. Several previous studies (e.g., Kirby, Gale, and Mattio 1982; Atkinson and Chaggar 1983) have focused on the prenatal administration of very high dose levels of capsaicin (e.g., 50 mg/kg) for short exposure periods, in which wide spread neurotoxic effects can be produced. In contrast, the data presented here are much more relevant to the systemic trans-capsaicin exposures, which may occur following typical environmental exposures to capsaicin-containing foods or medical and self-defense products. The consumption of capsaicin-containing peppers was reported to be about 2.5 g/person/day in India, 5 g/person/day in Thailand (Monsereenusorn 1983), and 20 g/person/day in Mexico (Lopez-Carrillo, Hernandez Avila, and Dubrow 1994). The estimated maximum daily intake of capsaicin based on consumption of mild chilies and paprika is about 1.5 mg/person/day in the United States and Europe (European Commission 2002). However, considering that trans-capsaicin (the only isomer that occurs naturally) content of pepper plants can vary widely, and that no data are available regarding plasma capsaicin levels following food or environmental exposures, the calculation of safety factor predictions for developmental toxicities based on the data presented here is not possible.

The dosages administered in the current study represent capsaicin dose levels several-fold higher than what will be administered clinically by NGX-4010, in part because systemic absorption of capsaicin through rat or rabbit skin is much higher than for human skin (Fang et al. 1995). In addition, it is expected that NGX-4010 will be applied no more than once every 12 weeks for a period of 1 h; this is a much lower frequency and duration than have been used in the current teratology studies (which was once per day for 3 h on DG 7 through 17 in rats and DG 7 through 19 in rabbits). Still, even at these very high capsaicin dose levels, there was no evidence for teratogenic effects following dermal applications of either NGX-4010 to rats or CLF to rabbits, even though dose range–finding studies have confirmed that plasma capsaicin concentrations are consistently produced.

Based on the results of the rat study, the maternal no-observable-effect level (NOEL) of NGX-4010 is less than a patch size of 25 cm². Erythema (grade 1), urine-stained abdominal fur and ungroomed coat occurred in all test article–treated groups at generally comparable incidences, lacrimation occurred in increased numbers of rats in the 37.5 and 50 cm² test article–treated groups, and an increased number of rats in the 50 cm² test article–treated group were cold to touch. The developmental NOEL for the NGX-4010 is a patch size of 37.5 cm². Delays in skeletal ossification (reductions in the average number of ossified hindlimb and forelimb phalanges and metatarsals) occurred in the 50 cm² test article-treated group. However, the values for ossified fore- and hindlimb phalanges were within the ranges observed historically at the testing facility and probably were of no biological significance. The values for ossified metatarsals were outside the historical control range in the 50 cm² group.

Based on the rabbit study results, the maternal NOEL of capsaicin is less than 3 μl/cm² or 0.3 mg/cm² trans-capsaicin. Erythema (grades 1 and 2), flaking (grades 2 and 3), wrinkling, and lesions at the application site occurred at all dose levels tested, with no dosage-dependent differences among the dosage groups. The developmental NOEL is greater than 13 μl/cm², or 1.3 mg/cm² trans-capsaicin, as no effects were observed at the highest dose level tested. There were no adverse effects on embryo-fetal development as evaluated in this study.

Though the current studies suggest little or no developmental toxicity potential for pure trans-capsaicin, products that contain trans-capsaicin along with other capsaicinoids or chemicals could exhibit a different profile. Indeed, based on our own experience with capsaicin, we have observed outcomes in toxicology studies with pure synthetic trans-capsaicin that differ from literature-reported results with capsaicin extracts (Chanda et al. 2004, 2005).

In conclusion, data collected from two species suggest that pure trans-capsaicin should not be identified as a developmental toxicant. No findings have been identified which would raise safety concerns for either capsaicin-containing food supplements or drug products.