* Drug Safety Evaluation, Pfizer Global Research & Development, 2800 Plymouth Road, Ann Arbor, Michigan 48105; and
Biostatistics, Discovery & Early Development Projects, Pfizer Global Research & Development, 2800 Plymouth Road, Ann Arbor, Michigan 48105
Received November 1, 2000; accepted February 2, 2001
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key Words: sperm analysis; male beagle dogs; semen analysis; HMG-CoA reductase inhibitor; atorvastatin; sperm motility..
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Atorvastatin is a potent inhibitor of HMG-CoA reductase, the rate-limiting enzyme in the biosynthesis of cholesterol. Atorvastatin is highly effective in lowering LDL cholesterol and increasing HDL cholesterol in hyperlipidemic patients, producing 3341% reductions in LDL cholesterol in patients with primary hypertriglyceridemia (Bakker-Arkema et al., 1996). It has been postulated that severe reductions in endogenous cholesterol biosynthesis might alter steroidogenesis in experimental animals or humans. However, there is very little evidence to support this theory. Testicular degeneration was observed sporadically and in varying degrees in dogs following treatment with other HMG-CoA reductase inhibitors such as simvastatin (Gerson et al., 1989
) and lovastatin (MacDonald et al., 1988
). Human studies with pravastatin (Dobs et al., 1993
) and simvastatin (Rossato et al., 1993
) suggested alterations in steroidogenesis, but neither study demonstrated effects on circulating testosterone, LH, or FSH. Numerous studies in humans have since shown no effects of HMG-CoA reductase inhibitors on gonadal function and/or sex steroid hormones (Bernini et al., 1998
; Dobs et al., 2000
; Travia et al., 1995
). In addition, a male fertility study with atorvastatin in rats (Dostal et al., 1996
) demonstrated no effects on reproduction or fertility. In that study, there were no effects on testes, epididymides, or accessory organ weights or histopathology, epididymal sperm counts, sperm motility, or sperm morphology. In response to a regulatory request received during the development of atorvastatin, repeated semen analyses, as well as terminal epididymal sperm counts, were incorporated into an ongoing chronic toxicity study of atorvastatin in dogs. The results of these analyses presented in this paper demonstrate the lack of an effect of atorvastatin on male reproductive parameters in a non-rodent species that shows marked decreases in cholesterol in response to the agent.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Atorvastatin was administered to groups of 10 male dogs at 0, 10, 40, or 120 mg/kg for up to 104 weeks. Throughout the study, clinical signs, body weights, clinical pathology, and plasma drug concentrations were evaluated, and complete gross and histopathological evaluations were conducted at necropsy (Walsh and Rothwell, 1999). Results of the serum cholesterol levels, reproductive organ evaluations, and semen analysis are reported here. Three dogs/group were euthanized after 52 weeks of dosing and 5/group were euthanized at Week 104. Two dogs/group were withdrawn from treatment after 52 weeks and were euthanized after a 12-week reversal period at Week 64.
Serum cholesterol.
Blood was collected and serum prepared for analysis of cholesterol pretest and during Weeks 13, 26, 39, 52, 64, 78, 91, and 104. Animals were fasted overnight prior to blood collection. Total cholesterol was analyzed on a Johnson & Johnson Vitros® Automated Chemistry Analyzer (Raritan, NJ). The system uses a dry chemistry technique in which cholesterol esters and cholesterol are converted to a colored dye via cholesterol oxidase and peroxidase.
Semen collection.
Semen samples were collected during Weeks 48 and 49 to familiarize the dogs with the collection procedure, and semen was discarded. Since the study was ongoing at the time of the regulatory request, it was not possible to conduct pretest evaluation of semen. The dogs were approximately 2 years old and fully mature at the time semen analyses were incorporated, and semen samples were obtained from all of the dogs during the study. In our experience, semen collection is more successful in beagle dogs that are 1 year old. Optimally, semen analyses can be incorporated into a study prior to its initiation, and pretest semen analyses would be useful to familiarize dogs with the collection procedure and to exclude animals with consistently poor semen quality or uncooperative behavior.
Semen analysis was conducted on all surviving dogs once each week at Weeks 50, 51, 52, 64, 78, 91, and 104. The analyses at Weeks 5052 were conducted weekly to determine the variability within each animal and to optimize the evaluations at a time when all animals were still on study. The subsequent time points corresponded to the end of the reversal period (Week 64), termination of the study (Week 104), and 2 intervening periods (Weeks 78 and 91) when other toxicity endpoints (e.g., clinical pathology) were already scheduled.
Semen samples were collected by a slight modification of the method described by Seager and Platz (1977). Briefly, semen was collected manually using an artificial vagina (AV) (Nasco Farm and Ranch, Cat. No. C6157N) connected to a graduated 15-mL conical tube. On the specified days, individual males were taken to a separate room where a female dog, preferably in estrus, was held until the male became aroused. The AV, with a small amount of lubricating jelly (Vaseline® or KY® jelly) on the top cuff, was then slipped over the penis and semen was collected. In dogs, the first fraction of semen is a small pre-sperm fraction that is usually relatively clear. A sperm-rich fraction, usually cloudy, and then a prostatic fluid fraction, which can be very large in volume and contains negligible sperm, immediately follow. In this study, the pre-sperm fraction was collected in the same tube with the sperm-rich fraction to avoid sample loss during the time of active thrusting of the dog. When the sperm-rich fraction was obtained (determined by visual determination and the stepping-over behavior of the male), the sample tube was changed. Often the sperm-rich fraction continued into the second collection tube (based on cloudiness of the semen), necessitating changing collection tubes multiple times. The last fraction of prostatic fluid was collected in the tube until the fluid appeared clear and 23 ml was collected; then the AV was removed. The semen tubes were placed immediately in a 37°C water bath until motility was assessed, and sperm morphology slides were prepared, generally within 10 min. Semen color was determined in each tube, and semen volume was calculated as the sum of the volumes (to the nearest 0.5 ml) of all sperm-rich fractions.
Allowing the pre-sperm fraction to mix with the sperm-rich fraction during the collection did not appear to affect sperm motility when motility analyses were conducted immediately after each sample collection. However, we have found in subsequent method-development experiments that more complex media, such as DMEM/F12 (Gibco) are needed to maintain sperm motility for longer than 510 min, or when samples must be transported from animal rooms to other laboratories.
Semen analysis.
Immediately (< 12 min) after obtaining the sperm-rich fraction of semen, an aliquot was diluted in 37°C Dulbecco's phosphate buffered saline, pH 7.4 (D-PBS; Gibco) to an appropriate concentration, and loaded immediately onto a 20 µm µCell chamber (Fertility Technologies) on a 37°C warming plate. The chamber was placed on a 37°C stage warmer on an Olympus CH-2 microscope. Live images of sperm were videotaped using a 10x objective, a video camera, a Panasonic WJ-810 time-date generator, and a Panasonic Model AG-1970 videotape recorder. At least 10 unique fields containing approximately 525 sperm each were recorded. Later the videotapes were played back by observers blinded to the animal and group numbers. Sperm motility and progressiveness in approximately 710 fields were rated visually on a scale of 05. The motility scale was: 0, no moving sperm; 1, < 20% motile sperm; 2, > 20% and < 40% motile sperm; 3, > 40% and < 60% motile sperm; 4, > 60% and < 80% motile sperm; 5, > 80% motile sperm. The progressiveness scale was 0, no motility; 1, slight side-to-side movement, no forward progression; 2, rapid side-to-side movement, no forward progression; 3, rapid side-to-side movement, occasional forward progression; 4, slow, steady forward progression; 5, rapid steady forward progression (adapted from Christiansen, 1984).
Sperm morphology slides were prepared by placing a drop of eosin/nigrosin sperm morphology stain on a glass slide. A drop of semen from the sperm-rich sample was placed on the slide near the drop of stain. The semen and stain were mixed on the slide using the long edge of a glass pipette, the mixture was spread in a thin film across the slide and allowed to air dry. A minimum of 200 sperm were evaluated at 100x by technicians blinded to the group and animal numbers. Sperm were classified as normal, abnormal head, abnormal midpiece, abnormal tail, detached head, proximal cytoplasmic droplet, distal cytoplasmic droplet, and bent tail using criteria described previously (Ball et al., 1983; Oettle and Soley, 1988
). Analyses were multiparametric: each sperm was categorized for each defect separately.
Since semen volume is the total volume of the sperm-rich fraction (i.e., excluding the prostatic fluid fraction), semen sperm counts were determined in each sample tube, then the total volume of the sperm-rich fraction was calculated. Semen sperm counts for each tube were determined in 4 replicate counts on a hemocytometer following dilution of semen fractions with D-PBS. Samples were diluted so that the sperm count for each chamber ranged from 15 to 150 sperm whenever possible. The 4 counts were averaged and the total sperm per ejaculate was calculated by adding the sperm counts from all sperm-rich fractions. Sample tubes containing less than 5 million sperm were considered prostatic fluid fractions, and were not included in the semen volume calculation.
Epididymal sperm counts and histopathology.
At necropsy, the left cauda epididymis was weighed and frozen (except at Week 52 when caput rather than cauda epididymis was inadvertently saved). When thawed, the tissues were minced in a dilute solution of Triton X-100 in 0.9% saline, then homogenized with a Virtishear tissue homogenizer for 1 min. The homogenates were sonicated for 3 min and diluted in 0.9% saline to 100-ml final volume. Sperm counts were determined on a hemocytometer in 4 replicate counts. Total sperm per cauda and sperm/gram cauda were calculated. Paired testes and epididymides were weighed, epididymal samples taken for sperm counts, and the remaining tissue fixed in Bouin's; prostates were weighed and fixed in 10% neutral buffered formalin. All tissues were embedded in paraffin, stained with hematoxylin and eosin, and examined microscopically for abnormalities. All pathology data were subjected to peer review and consensus reports were generated.
Statistics.
To control for the multiplicity of statistical comparisons (i.e., to reduce the likelihood of false-positive conclusions), the parameters were divided into distinct classes of related parameters (e.g., sperm counts in one, sperm motion parameters in another, etc.). The classwise significance level was then allocated to each parameter proportionally by the inverse of the square root of the number of parameters in a class (Tukey et al., 1985). For each animal, a combined mean of semen samples collected over Weeks 50 through 52 were calculated and used for statistical comparisons of group means. Group mean comparisons were conducted for Weeks 5052, Week-64 treated animals, and Weeks 78, 91, and 104 by Tukey's sequential trend test, using the rank dose scale and rank-transformed data, 1- or 2-tailed, at the 5% classwise significance level (Park, 1985
; Tukey et al., 1985
). A trend reversal test was performed at the 1% classwise significance level.
To assure that an adequate number of dogs were used to detect real treatment effects in the presence of the experimental variation, statistical power calculations were conducted. Statistical power is a measure of the chance of detecting a true treatment effect for a given sample size in the presence of the methodological variation. As statistical power is a measure of chance, it is expressed as a probability between 0 and 100 percent. The higher the statistical power of a test, the more likely it is that the parameter or measurement will be able to identify a real treatment effect. Retrospective power calculations were performed on semen-analysis and sperm-count parameters to assess the operating characteristics of the sequential linear-trend test; i.e., to evaluate the probability of detecting a biologically relevant difference from control at a 5% classwise significance level with the sequential linear-trend test. The lower bound for the power of the sequential linear trend test was computed with SAS JMP® software, Version 3.1. Acceptable statistical power (at least 80%) was considered evidence that sample sizes were sufficient to demonstrate significant differences in the data. These power calculations apply specifically to this study and would vary with different experimental conditions (i.e., different experimental variation). This retrospective statistical power calculation is different from a prospective calculation in which the variability of the parameters in the treated groups is assumed to be the same as controls. In the retrospective calculations, the variance is not assumed but estimated from the actual results. A prospective power calculation indicates the number of animals per group needed to detect a certain magnitude of treatment effect at a given level of significance, using a variance assumed to be a reasonable estimate. These variance estimates are derived from previous studies, and it is assumed that the experimental conditions of previous studies will be the same as future studies (with the exception of choice of treatment).
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
|
Epididymal sperm counts.
Epididymal sperm counts were comparable in all groups at Weeks 52 and 64 (not shown), and 104 (Table 3). The statistically significant increase in sperm per gram cauda epididymis in 120 mg/kg animals at Week 104 was not considered biologically significant, due to the magnitude of the change and the concomitant increase in total semen sperm count (Table 2
). The power calculations demonstrate good statistical power (> 90%) for mean comparisons of epididymal sperm count at Week 104.
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
It has been hypothesized that severe reductions in plasma cholesterol might lead to reductions in steroidogenesis by reducing the supply of circulating cholesterol for endocrine organ uptake or by inhibiting in situ cholesterol synthesis (Dobs et al., 1993). Although testicular degeneration has been reported in dogs with simvastatin (Gerson et al., 1989
) and lovastatin (MacDonald et al., 1988
), the present study showed no evidence of testicular effects or adverse effects on semen parameters, even at doses which severely reduced serum cholesterol levels. This is the first study to examine the effect of an HMG-CoA reductase inhibitor at multiple time points in dogs. However, several studies of the potential effects of other HMG-CoA reductase inhibitors on endocrine parameters, semen quality, and steroidogenesis have been reported in humans. In one study, sperm motility was reduced by lovastatin in men with Type-II hyperlipoproteinaemia (Farnsworth et al., 1987
), while all other indications of testicular function (testis size, sperm count, morphology, and serum testosterone, LH, FSH, or prolactin concentrations) were unaffected. Human studies with simvastatin and pravastatin (Bernini et al., 1998
; Dobs et al., 2000
; Purvis et al., 1992
; Travia et al., 1995
) also demonstrated no effects on testicular-hypothalamic-pituitary hormones and semen parameters, including sperm motility, at doses which produced marked reductions in serum total or LDL-cholesterol. Therefore, the majority of studies conducted in humans and dogs with HMG-CoA reductase inhibitors indicate no significant effect of reduced plasma cholesterol on male reproductive function. The present study confirms the lack of effects of low cholesterol on testicular histopathology and male reproductive parameters in dogs.
Because of the testicular changes observed in dogs with other HMG-CoA reductase inhibitors and a regulatory request, repeated analysis of semen quality was incorporated into this 2-year toxicity study in beagle dogs. Since the study was ongoing at the time of incorporation of these analyses, it was not possible to develop and validate a computer-assisted sperm analysis (CASA) method for assessing sperm count, motility, and progressiveness in dogs. Therefore, semen analysis procedures were used that are commonly employed in theriogenology and veterinary medicine (Christiansen, 1984; Johnston, 1991
; Seager and Platz, 1977
). These semen parameters, including semiquantitative measures of sperm motion, are useful for predicting fertility in dogs (Nothling et al., 1997
; Oettle, 1993
). In most cases, the data from the present study were similar to those reported previously (Johnston, 1991
; Oettle and Soley, 1988
; Olar et al., 1983
; Purswell et al., 1992
; Seager and Platz, 1977
), with the possible exception of sperm-tail abnormalities discussed below. It was apparent from the videotapes that a large amount of debris was present in the semen of some dogs that appeared to have the same size and shape as sperm heads using normal illumination. This indicates that techniques such as specific DNA staining would be required to accurately identify and track dog sperm for CASA. Currently, semen analysis methods for dog sperm are being developed in our laboratory using the Hamilton-Thorne IVOS® instrument utilizing live IDENT staining.
The power calculations performed on the various sperm parameters demonstrate very good statistical power for the statistical comparisons of semen sperm counts and normal morphology and for epididymal sperm counts when group sizes were at least 35. Acceptable power for sperm motility and progressiveness rating was achieved when group sizes were 810. These data also demonstrate the advantages of repeated evaluations of parameters such as semen since the statistical power can vary within the same group of animals during a study. In addition, treatment-related trends over time could be observed, while spurious statistically significant differences (i.e., not biologically relevant) could be discounted if the values fall within the range of previous values within the same study.
The power values calculated for other parameters were lower. The statistical power for the progressiveness rating (Scale 05) was 63% and for semen volume only 1% for Weeks 5052 versus values greater than 80% for other parameters. However, the cause of these lower power values was not the poor performance of the parameters during the execution of the study, but the inherent properties of the parameters themselves. For example, the progressiveness rating parameter in Table 1 has very small variation (as indicated by the small standard error). The difference in these group means was very slight relative to the variation: the difference in progressiveness rating means between the vehicle controls and the treated groups was at most 0.4 units, while the largest standard error for a group mean was 0.25 units. Given this variability one might expect it to be very difficult to detect a statistically significant difference unless the treatment produced a very large difference in means (e.g., in the case of progressiveness, an effect of greater than 1). This is confirmed in the statistical power analysis. Parameters such as the progressiveness rating might have a smaller chance of producing statistically significant findings for sample sizes less than 10, even with exposure to other agents, due to the small variation inherent in the measurement.
Sperm morphology abnormalities in the dog are generally classified as primary and secondary (Ball et al., 1983; Christiansen, 1984
; Johnston, 1991
). In the classification system used in this study, the primary abnormalities included abnormal heads, midpieces, and tails, and the secondary abnormalities included proximal and distal cytoplasmic droplets and bent tails. The primary abnormalities are believed to originate from spermatogenesis, while secondary abnormalities arise during epididymal maturation, transit, or ejaculation (Christiansen, 1984
). Secondary abnormalities have been shown to occur with a frequency of 2.488.0%. In the present study, bent sperm tails occurred with relatively high frequency, causing the percent-normal sperm morphology to be very low in some groups. This high frequency may be due to the use of strict scoring criteria, or to sexual abstinence which was associated with an increase in sperm abnormalities in some studies (Christensen, 1984) but not in others (Taha et al., 1983
). Numerous factors have been shown to affect semen parameters in dogs; large variability can occur due to sample handling and collection procedures, temperature, inflammation, and hormones (Freshman et al., 1988
).
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
NOTES |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Ball, L., Ott, R. S., and Mortimer, R. G. (1983). Manual for breeding soundness examination of bulls. J. Soc. Theriogenol. 12, 3854.
Bernini, G. P., Brogi, G., Argenio, G. F., Moretti, A., and Salvetti, A. (1998). Effects of long-term pravastatin treatment on spermatogenesis and on adrenal and testicular steroidogenesis in male hypercholesterolemic patients. J. Endocrinol. Invest. 21, 310317.[ISI][Medline]
Christiansen, I. J. (1984). Part 1. Andrology of the normal male. In Reproduction in the Dog and Cat, pp. 80109. Bailliere Tindall, London.
Dobs, A. S., Miller, S., Neri, G., Weiss, S., Tate, A. C., Shapiro, D. R., and Musliner, T. A. (2000). Effects of simvastatin and pravastatin on gonadal function in male hypercholesterolemic patients. Metabolism 49, 115121.[ISI][Medline]
Dobs, A. S., Sarma, P. S., and Schteingart, D. (1993). Long-term endocrine function in hypercholesterolemic patients treated with pravastatin, a new 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor. Metabolism 42, 11461152.[ISI][Medline]
Dostal, L. A., Whitfield, L. R., and Anderson, J. A. (1996). Fertility and general reproduction studies in rats with the HMG-CoA reductase inhibitor, atorvastatin. Fundam. Appl. Toxicol. 32, 285292.[ISI][Medline]
Farnsworth, W. H., Hoeg, J. M., Maher, M., Brittain E. H., Sherins, R. J., and Brewer, H. B., Jr. (1987). Testicular function in Type-II hyperlipoproteinemic patients treated with lovastatin (mevinolin) or neomycin. J. Clin. Endocrinol. Metab. 65, 546550.[Abstract]
Freshman, J. L., Amann, R. P., Soderberg, S. F., and Olson, P. N. (1988). Clinical evaluation of infertility in dogs. Compendium on Continuing Education for the Practising Veterinarian 10, 443460.
Gerson, R. J., MacDonald, J. S., Alberts, A. W., Kornbrust, D. J., Majka, J. A., Stubbs, R. J., and Bokelman, D. L. (1989). Animal safety and toxicology of simvastatin and related hydroxy-methylglutaryl-coenzyme A reductase inhibitors. Am. J. Med. 87 (Suppl. 4A), 2838S.[ISI][Medline]
Johnston, S. D. (1991). Performing a complete canine semen evaluation in a small animal hospital. Verterinary Clinics of North America: Small Animal Practice 21, 545551.[Medline]
MacDonald, J. S., Gerson, R. J., Kornbrust, D. J., Kloss, M. W., Prahalada, S., Berry, P. H., and Bokelman, D. L. (1988). Preclinical evlauation of lovastatin. Am. J. Cardiol. 62, 1627J.
Nothling, J. O., Gerstenberg, C., and Volkmann, D. H. (1997). Semen quality after thawing: Correlation with fertility and fresh semen quality in dogs. J. Reprod. Fertil. Suppl. 31, 109116.
Oettle, E. E. (1993). Sperm morphology and fertility in the dog. J. Reprod. Fertil. Suppl. 47, 257260.[Medline]
Oettle, E. E., and Soley, J. T. (1988). Sperm abnormalities in the dog: A light and electron microscopic study. Vet. Med. Rev. 59, 2870.
Olar, T. T., Amann, R. P., and Pickett, B. W. (1983). Relationships among testicular size, daily production, and output of spermatozoa, and extragonadal spermatozoal reserves of the dog. Biol. Reprod. 29, 11141120.[Abstract]
Park, Y. C. (1985). Nonparametric sequential trend test. Proc. Tenth Annual SAS Users Group International Conference, 809813.
Purswell, B. J., Althouse, G. C., and Root, M. V. (1992). Guidelines for using the canine breeding soundness evaluation form. Proc. Ann. Meeting Soc. Theriogen., 174182.
Purvis, K., Tollefsrud, A., Rui, H., Haug, E., Norseth, J., Viksmoen, L., Ose, L., and Lund, H. (1992). Short-term effects of treatment with simvastatin on testicular function in patients with heterozygous familial hypercholesterolaemia. Eur. J. Clin. Pharmacol. 42, 6164.[ISI][Medline]
Rossato, M., Guarneri, G., Lavagnini, T., Padovan, D., and Foresta, C. (1993). Simvastatin influences testicular steroidogenesis in human. Horm. Metab. Res. 25, 503505.[ISI][Medline]
Seager, S. W., and Platz, C. C. (1977). Collection and evaluation of canine semen. Vet. Clin. North Am. 7, 765773.[ISI][Medline]
Taha, M. B., Noakes, D. E., and Allen, W. E. (1983). The effect of the frequency of ejaculation on seminal characteristics and libido in the Beagle dog. J. Small Anim. Pract. 24, 309315.[ISI]
Travia, D., Tosi, F., Negri, C., Faccini, G., Moghetti, P., and Muggeo, M. (1995). Sustained therapy with 3-hydroxy-3-methylglutaryl-coenzyme-A reductase inhibitors does not impair steroidogenesis by adrenals and gonads. J. Clin. Endocrinol. Metab. 80, 836840.[Abstract]
Tukey, J. W., Ciminera, J. L., and Heyse, J. F. (1985). Testing the statistical certainty of a response to increasing doses of a drug. Biometrics 42, 295301.
Walsh, K. M., and C. E. Rothwell. (1999). Hepatic effects in beagle dogs administered atorvastatin, a 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor, for 2 years. Toxicol. Pathol. 27, 395401.[ISI][Medline]