The Role of Pathology in Evaluation of Reproductive, Developmental, and Juvenile Toxicity

Species Selection

Although no specific nonclinical species best predicts male infertility, species selection should be justified. Mammalian species should be used and if applicable, should express the pharmacological target; this is especially important if the target is present in the human male reproductive system. Ideally, the same species and strain used in the repeat-dose toxicity studies are utilized for functional fertility studies in order to minimize animal use since pharmacokinetics and metabolism have already been well characterized and can be used for dose setting. See ADME P rinciples in S mall M olecule D rug D iscovery and D evelopment— A n I ndustrial P erspectiv e Vol 1, Chap 3 and Biotherapeutic ADME and PK/PD Principles Vol 1 , Chap 4 . In addition, use of the same species for repeat-dose toxicity and functional fertility studies allows for better correlation and integration of the overall effects on the male reproductive system. Rodents are the default test system for functional fertility studies with the rat most often used followed by the mouse. Nontraditional species for functional fertility studies include the rabbit, guinea pig, and hamster. Functional fertility studies are not recommended for nonhuman primates (NHPs) or dogs due the practical challenges with low background fertility, limited number of cycles per year (female dog), and relatively long gestation length. Table 7.2 summarizes differences in key male reproductive considerations among standard toxicology species. When there is concern for male reproductive toxicities not specific to the testis, the effects can be better characterized using surrogate fertility endpoints such as sperm assessment and hormones. In addition, sensitivity of specific cell types can be investigated in vitro and support in vivo studies of reproductive toxicology ( ; ; ), specifically the testis and, in the future, potentially more complex models capable of organ-to-organ hormonal signaling during reproduction ( ; ).

Sexual Maturity

A critical component of experimental test systems to appropriately evaluate potential effects on male fertility is that the test system should be sexually mature. The Society of Toxicologic Pathology (STP) recently endorsed Best Practices for the recording and reporting of sexual maturity status ( ). An immature or developing testis can in many ways resemble types of testicular toxicity and at best confound an otherwise adequate assessment of test item–related toxicity ( ; ; ). Based on microscopic evaluation of the male reproductive tract including sperm evaluation, the Wistar and Sprague Dawley rat are considered sexually mature by 10 weeks of age ( ). As such, because male reproductive safety in first-in-human (FIH) studies is typically supported by repeat-dose toxicity studies, it is important to include sexually mature animals in those studies. For functional fertility studies, however, it is common to mate rats between 12 and 24 weeks of age, as that seems to represent peak reproductive capacity in most rat strains. When younger or older rats are used, there is more variability as animals first mature, and then as they age, their reproductive capacity appears to slowly decline ( ; ; ; ; ). Based on microscopy of the testis, it is recommended that mice be older than 7 weeks of age at termination ( ).

Sometimes a male reproductive toxicant will not be apparent in a rodent but will present in a nonrodent species, either as part of routine safety evaluation ( ) or, as is often the case with biologics, the NHP as the only pharmacologically relevant species. Regardless, if the repeat-dose study is the only evaluation of the reproductive tract, it is important that the evaluation of potential effects on male reproduction be done using sexually mature dogs or NHPs. Mature animals are typically used in subchronic or chronic studies and not necessarily FIH studies unless there is a specific cause for concern. It is recommended that dogs be older than 9 months of age at necropsy for an appropriate evaluation of potential testicular toxicity ( ; ). Unlike rodents, and to some extent dogs, determination of sexual maturity in male NHPs based on age, body weight, and/or testicular volume is insufficient to predict sexual maturation (defined by testicular histopathology) ( ). However, the presence of sperm in a single semen sample unequivocally demonstrates capacity for complete spermatogenesis and is recommended to support selection of male NHPs when the evaluation needs to include an assessment of toxicity to the male reproductive system ( ).

Confounding Factors

There are confounding factors that can affect interpretation of potential effects on male reproduction. In addition to the maturity status previously described, nonspecific toxicity such as stress and/or effects resulting in decreased body weight and/or food consumption can complicate otherwise clear signals of male reproductive toxicity. This type of nonspecific stress introduced by feed restriction can result in suppression of the hypothalamic–pituitary–gonad (HPG) axis, gonadotrophin releasing hormone (GnRH), and downstream effects on the male rodent reproductive system such as decreased testosterone, organ weights (decreased seminal vesicles and prostate, but not testis), atrophy of accessory sex organs, and/or testicular degeneration ( ; ; ; ).

A specific analysis looking at the value of organ-to-body weight ratio or organ-to-brain weight ratios confirmed that relative testis weights are not useful determinants of toxicity in the rat ( ). Interestingly, these effects on the morphology of male reproductive tissues do not seem to negatively impact functional fertility in the rat ( ) and only variably so in the mouse ( ). In general, the reproductive system of dogs and NHPs appears less sensitive to effects of nonspecific stress compared with rodents, although dogs may show atrophy of the prostatic epithelium, and NHPs may have decreased testicular size, weight, and spermatogenesis ( ; ; ).

Blood–Testis Barrier

A unique aspect of male reproductive physiology is the blood–testis barrier (BTB) and blood–epididymis barrier (BEB). These barriers are an interaction of the structural, physiological, and immunological components ( ; ) and can be important considerations in test article–related toxicity and potential species-specific sensitivity. Although both barriers are important for sperm maturation, little is known about the BEB ( ) compared with the expanding research on the BTB ( ; ). Specifically, drug biodistribution at the BTB is influenced by transporters such as breast cancer resistance protein on endothelial cells ( ), transferrin on Sertoli cells ( ), and P-glycoprotein on several different cell types ( ; ), and these can impact reproductive function ( ).

Organ Weights

There are specific recommendations by the STP for which organ weights should be collected and in which species ( ). Specifically, testes should be weighed in all repeat-dose toxicity studies of all species, epididymides and prostate glands should be weighed in all rat studies (and case by case in nonrodent and mouse studies), and thyroid and pituitary gland weights should be collected for all species except mice. As described earlier, organ weights of male reproductive tissues are most appropriate in sexually mature animals. Seminal vesicle (with or without fluid) weights in male rodents provide similar information as prostate weights, but if done consistently or together with the prostate (commonly referred to as the accessory sex organ/glands) can provide a sensitive indication of androgen status ( ). While slight changes in male reproductive organ weights in the absence of other effects may not be the cause for concern (particularly in nonrodent species) ( ), specific changes (even unilaterally) can significantly contribute to a better understanding and interpretation of observations in other endpoints including macroscopic and microscopic changes, as well as sperm assessment and fertility indices ( ).

Macroscopic and Microscopic Evaluation

In addition to macroscopic examination, there are specific recommendations by the STP as to which tissues should be collected and microscopically evaluated for all repeat-dose toxicity studies ( ). These include testes, epididymides, prostate gland, seminal vesicles, and pituitary gland, which serve as a critical part of the evaluation for potential effects on the male reproductive system. A robust microscopic evaluation of the male reproductive tract requires consistent and appropriate trimming and fixation of these tissues, as reviewed by the STP ( ; ; ). Standardized terminology (See Nomenclature and Diagnostic Resources in Anatomic Toxicologic Pathology , Vol 1, Chap 25 ) and lesion characterization is important and has been published for the female reproductive system with the rat and mouse ( ). Just as important as a qualitative, stage-aware, evaluation of spermatogenesis is an integrated assessment of all the reproductive tissues ( ; ) and appropriate recognition of immature tissues (e.g., testis) relative to a potential lesion in nonrodent species such as NHP ( ; ).

Sperm Assessment

Sperm assessment is required for some chemical toxicity testing (e.g., extended one-generation study) but is not routinely required for toxicity screening of pharmaceuticals. Regardless, evaluation for potential effects on sperm counts, motility, and morphology can be triggered to further characterize male reproductive toxicity and is one of the few endpoints that can be evaluated in clinical trials. Figure 7.3 illustrates rat sperm on a slide that is used by computer-assisted sperm analysis to count sperm and calculate percent motility.

Sperm assessment is an important complementary endpoint for male reproductive assessment that falls between effects on male fertility (functional mating) and morphologic changes in male reproductive tissues. In rodents, sperm assessment is usually conducted from testicular tissue and/or cauda epididymis or vas deferens collected at necropsy ( ). Sperm assessment can be a more sensitive measure of effects on spermatogenesis than fertility in rodents because there is a large sperm reserve in these species such that as much as a 90% decrease in sperm count may still result in normal fertility ( ; ; ). As such, without sperm evaluation, a posttesticular effect on sperm maturation may be missed in rodents ( ).

For effects on spermatogenesis in the testis, microscopic evaluation is considered the most sensitive endpoint ( ; ; ). In rabbits, dogs, and NHP, serial semen samples can be collected before, during, and after dose administration in a toxicity study, allowing onset and recovery of effects on sperm to be characterized ( ; ; ; ; ; ; ). In all species, sperm assessment can be conducted in all types of studies as an optional endpoint, including standard repeat-dose toxicity studies and/or fertility studies.

Biomarkers

There are few practical biomarkers of male reproductive health that can be measured in blood. These include hormones and more novel markers such as sperm mRNA transcripts as indicators of testicular damage. As a general screen for detecting male reproductive toxicity in nonclinical species, effects on circulating hormones can be difficult to interpret ( ). In particular, some hormones are subject to wide variability due to pulsatility of secretion, diurnal rhythms, and stress; as such, it is critical to factor this into the design of any study including these endpoints, as statistical power to detect changes can be limited ( ; ). There have been several reviews describing changes in luteinizing hormone (LH), follicle-stimulating hormone (FSH), inhibin, testosterone, or prolactin, as well as how different “fingerprints” correlate to different microscopic changes in the male reproductive tract ( ; ). One commonly observed change in hormones associated with testicular damage is decreased testosterone, which can be secondary to decreased GnRH and/or LH, or through a direct effect on Leydig cell steroidogenesis. This type of effect can occur quickly, as demonstrated in a 7-day study with a neurokinin 1, 2, and 3 receptor antagonist ( ). After many years of anecdotally using inhibin B as a biomarker of testicular damage, a robust assay was evaluated by many companies and compounds with equivocal results in that approximately half of the exposures demonstrated inhibin B as a lagging biomarker requiring significant testicular damage ( ). In an effort to identify additional biomarkers of testicular damage, it was found that sperm mRNA transcripts are sensitive indicators of both Sertoli and germ cell damage, while sperm DNA methylation changes are not ( ). The lack of a premonitory biomarker of male reproductive toxicity limits clinical management of potential risk and thus large therapeutic margins and/or mechanistic understanding are often needed to assure safety.

Mating Trials

As mentioned earlier, the minimum amount of data needed to detect the majority of potential effects on male reproduction using animal models are organ weights and histopathology, but mating trials (fertility indices) are needed to detect other functional effects unrelated to testicular sperm production such as posttesticular sperm maturation, behavior, libido, and fecundity ( ). It is this latter endpoint of fertility that evaluates the relatively complicated process of gamete maturation and successful delivery via the ejaculate to complete the mating process (copulatory performance) leading to presumptive fertilization of the released oocyte in the female partner, and blastocyte implantation in the uterine wall confirming pregnancy. Mating trials are most often conducted in dedicated fertility studies with treated males and females, or, in the case of male fertility studies, with treated males mated to treatment-naïve females ( ; ). However, it is also possible to do mating trials with treated males on repeat-dose toxicity studies coupled with treatment-naïve females ( ). These options are described later under different study types.

The vast majority of mating trials are conducted in rats, but similar designs can be conducted in mice, rabbits ( ), or guinea pigs ( ). Mating trials are not practical in dogs or NHPs, so surrogate endpoints of fertility (e.g., sperm assessment) may be used in addition to organ weights and histopathology to fully characterize male reproductive toxicity in those species. In a rodent reproductive toxicity study, functional fertility is evaluated by pairing the treated male with a female for a set period of time, typically 2–3 weeks, until evidence of mating is confirmed by copulatory plug in the vagina or sperm present in a vaginal smear. The animals are unpaired, and the mated female is allowed to progress through gestation. The female is either subject to Cesarean section or allowed to deliver, but the primary fertility endpoint is usually to confirm pregnancy status and then to evaluate the number of viable and nonviable conceptuses or pups relative to the number of implantation sites.

In addition to confirming pregnancy, mating trials can inform the viability of the offspring which can be a sensitive endpoint of male (rodent) reproductive toxicity. The limited apical endpoints calculated in rodent mating trials include time to mating and several different types of reproductive indices such as mating index ([number males with evidence of mating (or females confirmed pregnant)/total number of males used for mating] × 100) and fertility index ([number of males siring a litter/total number of males used for mating] × 100).

As mentioned earlier, evaluation of potential effects on male reproduction ideally consists of data from repeat-dose toxicity studies as well as a functional fertility study that includes a mating trial. It is the integration of all available data across studies and species, accounting for differences in design and species, that should provide a complete evaluation of male reproductive safety of a test item.

Repeat-Dose Toxicity Studies

Microscopic evaluation of male reproductive tissues is a critical component in all repeat-dose toxicity studies ( ), and specific chapters in this Edition will cover the specific pathologies in these tissues (see Male R eproductive T ract, Vol 4, Chap 9). While testicular toxicity can be detected in a short-term study, effects may not fully manifest acutely, or may become more severe with a longer duration of dosing. There are many different targets of toxicity within the testis including interstitial cells (Leydig cells), Sertoli cells, and perhaps most challenging are the spermatogonia, spermatocytes, and spermatids.

The sensitivity to detect the primary cell type affected by a specific toxicant can be challenging; testicular toxicity often simply manifests as seminiferous tubule degeneration and atrophy identified at necropsy and histologically. However, the severity of the damage and cell types impacted can dictate whether the lesion will recover and how long it will take to recover. It is not uncommon in short-term studies for the recovery animals to actually have more severe testicular pathology, as it can take time for affected cell types to manifest as damaged tubules. It is also possible for unilateral testicular toxicity to be secondary to efferent duct toxicity, which can be useful to identify because it appears to be a rat-specific lesion that is not relevant for humans ( ).

It is more difficult to detect male reproductive toxicity other than testicular damage in repeat-dose studies unless other endpoints are included such as sperm assessment and/or in the case of rodents, mating trials ( ; ; ; ; ; ; ). If the rodent or rabbit are not biologically or pharmacologically relevant and there is a concern for male reproductive toxicity not involving the testis, the addition of sperm assessment can be the most reasonable way to evaluate those potential effects in dogs and NHPs. In fact, ICH S6(R1) describes the scenario for when NHPs are the only pharmacologically relevant species as described earlier ( ).

Male Fertility Study

Ideally, the evaluation of testicular morphology in repeat-dose toxicity studies is complemented by a study including a mating trial typically conducted in a male fertility study in rodents. The key feature of a male fertility evaluation is a mating trial, but also can include male reproductive organ weights, microscopic evaluation of the male reproductive tract, sperm assessment, or other endpoints, if warranted.

Historically, the male fertility study included dosing for a full 8–10 weeks prior to mating to encompass the full developmental period of spermatogenesis on the impact to mating and fertility. However, current ICH S5(R3) guidance supports a minimum of 2 weeks of dosing prior to mating for detecting potential effects on male mating and fertility, provided no effects on male reproductive organs were observed in previous repeat-dose toxicity studies of at least 2 weeks duration and that testicular microscopic examination was conducted in a “stage-aware” manner ( ; ). The 2-week dosing period prior to mating was justified by a detailed review of known male reproductive toxicants demonstrating that organ weights and microscopic examination detected the majority of male reproductive toxicants ( ), and by demonstrating that most testicular toxicants could be detected microscopically after only 2–4 weeks of dosing ( ; ). Importantly, 2 weeks of dosing prior to mating in a fertility study can detect posttesticular effects on sperm, accessory male organ function, male libido, and male copulatory function. This duration is further supported by the approximate 2-week posttesticular transit time in rats as it is these posttesticular sperm that are ejaculated during the subsequent mating period.

If a functional evaluation of sperm that have been exposed during testicular spermatogenesis is needed (based on findings observed in repeat-dose toxicity studies), then the period of time prior to mating can be extended. Key examples of male reproductive toxicity in the absence of overt effects on testicular weights or histopathology include alpha-chlorohydrin and ornidazole that cause male infertility via effects on sperm maturation (motility and capacitation) in the epididymis ( ; ).

There are some other examples where lack of testicular toxicity and 2 weeks of dosing prior to mating do not detect male reproductive toxicity in the rat. One example is the dihydrotestosterone inhibitor, finasteride, where no testicular histopathology was observed, and no effect on fertility was observed after 6 or 12 weeks of dosing; effects on male rat fertility were only observed at 24 weeks or longer of dosing ( ). Another example is a nicotinic acetylcholine receptor agonist that caused profound effects on male fertility after 11 weeks of dosing, but did not cause changes in testicular histopathology, sperm counts, or sperm motility after 6 months of dosing in rats; follow-up investigations revealed that only increased preimplantation loss was observed after 14 days of dosing, with no fertility effect after 5 or 14 days of dosing with the nicotinic acetylcholine receptor agonist ( ). The nicotinic acetylcholine receptor agonist example highlights that an extended premating period may have better sensitivity to detect the magnitude of male fertility effects, particularly posttesticular effects, such as those targeting epididymal and ejaculated sperm.

Species Selection

Much like the male, specific experimental considerations for evaluating potential effects on female fertility include selection of an appropriate animal model, the stage of sexual maturity of the test system, controlling for confounding effects, and selection/use of appropriate endpoints associated with female fertility. These considerations apply to in vivo testing, including both general toxicity and functional fertility (mating) studies. Importantly, many of the considerations and endpoints of reproduction are interrelated, and therefore, data interpretation should be done within the context of the other aspects of female reproductive function. Discrete aspects of cell and organ function can also be investigated in vitro and support in vivo studies of reproductive toxicology, specifically the ovary ( ; ) and, in the future, potentially more complex models capable of organ-to-organ hormonal signaling during reproduction ( ).

Many aspects of female reproduction are conserved across species; thus, toxicity to reproductive tissues can be evaluated as part of general toxicity testing. General toxicity testing does not typically include evaluation of reproductive function, and therefore, stand-alone functional fertility studies are used to supplement repeat-dose toxicity testing. Therefore, rodents are the default test system for functional fertility studies with the rat most often used followed by the mouse. Although mating trials in NHP are impractical, the female NHP reproductive cycle (endocrinology and physiology) is very similar to human and has been very useful for understanding effects on female fertility ( ; ; ). Nontraditional species for functional fertility studies include the rabbit, guinea pig, and hamster. In most species (including NHP), surrogate fertility endpoints such as estrous cycling and hormones can be used to improve evaluation of potential effects on female fertility. Table 7.3 summarizes differences in key female reproductive milestones among standard toxicology species.

Sexual Maturity

Robust evaluation of potential effects on female fertility should be done in female animals of reproductive potential. If the animals are not yet sexually mature (still in puberty) or too old (reproductively senescent), cyclicity and fertility can be inconsistent and is likely to confound interpretation of test item–related effects. As such, documentation of sexual maturity is important ( ).

In rats, there is some variability in maturation of estrous cyclicity up to 8 weeks old with morphological maturity by 10 weeks old. However, it is common practice to conduct the mating trial in a functional fertility study using female rats at least 12 weeks of age. Importantly, female rats should cycle normally (4–5 days per cycle) between 3 and 6 months of age. Although abnormal estrous cycles can occur occasionally between 3 and 6 months of age, the number and duration of abnormal cycles start increasing further around 6 months of age, and soon thereafter, fertility indices start decreasing, including embryo loss ( ; ). Age-related reproductive senescence in rats usually progresses into a state of persistent diestrus with some percent of rats also going through a protracted estrous stage during the progression to persistent diestrus ( ). Normal reproductive senescence is generally considered a consequence of age-related changes in the HPG axis and exhaustion of oocyte stores in the ovary ( ).

In the beagle dog, first estrous occurs between 8 and 14 months old, and they cycle only 1–2 times per year ( ) making evaluation of any surrogate female fertility endpoints in the dog impractical, although microscopic evaluation of female reproductive organs is appropriate if they are sexually mature.

In the cynomolgus macaque (NHP), most females are sexually mature by 4 years of age and weighing at least 3 kg, but, as with male NHPs, age and body weight are imperfect likely due to biological variability in response to a variety of environmental factors ( ). Since female NHPs can cycle normally approximately every 28 days, it is possible to monitor cycling in NHPs in controlled conditions to confirm sexual maturity with high confidence prior to the start of a toxicity study ( ). Regardless of the animal model, understanding whether the animals in the experiment are sexually mature is critical to a valid assessment of potential effects on female fertility, whether done as part of a functional fertility study or in a general toxicity study where microscopic evaluation of the reproductive tract is the primary endpoint.

Confounding Factors

In addition to the maturity status of the animal model, there are other potentially confounding factors that can interfere with interpretation of potential effects on female fertility. These other factors include nonspecific toxicity such as stress and/or effects resulting in decreased body weight and/or food consumption, housing conditions, and a combination of these stressors. In the adult female, the most sensitive reproductive endpoint affected by stress is irregular cycling and possibly fewer corpora lutea . With increased stress such as that induced by notable effects on body weight and/or food consumption or in situations where stress has induced hormonal perturbations, several effects including persistent cycling disruption (e.g., diestrus), fewer corpora lutea, and reduced fertility have been observed ( ; ).

Although lower ovarian and uterine organ weights may occur with stress, these can be difficult to distinguish from inherent variability due to stage of the reproductive cycle and sampling of these tissues. Microscopic observations consequent to food restriction in female rodents include decreased or absent corpora lutea , increased follicular atresia in the ovary, and evidence of atrophy in the uterus and vagina ( ; ). In general, reproductive effects consequent to dietary restriction appear more pronounced in mice than rats. The effects of stress and/or effects resulting in decreased body weight and/or food consumption on reproductive parameters in dogs are not well characterized. This effect is likely due to the common use of immature dogs in general toxicity studies and the protracted duration of the estrous cycle reducing the sensitivity of detecting cycle alterations even for mature dogs ( ; ).

In addition to dietary alterations, in the NHP model, exercise and housing changes can result in changes in female reproductive parameters. Of particular importance to toxicity studies, the social status of female NHPs, such as a change of cage-mates and moving/shipping of animals ( ; ), can alter cycling. An altered social status can be a problem if introduced at the beginning of a toxicity study where measurement of cyclicity is an endpoint ( ). These changes are usually transient and can be controlled by allowing sufficient time to return to normal cycling, which is usually about 3 months ( ). A larger than expected impact on the female reproductive axis can occur when these minor stressors occur in combination as compared with individual stressors alone ( ).

Evaluation for potential effects on female fertility includes cyclicity, fertility indices (including implantation), systemic biomarkers, and standard toxicity testing endpoints of organ weights, macroscopic, and microscopic examination of tissues in the female reproductive tract. Most of these endpoints can be evaluated in a general toxicity study using nonpregnant females, but a mating trial with males is needed to detect potential effects on time to mating, tubal transport, implantation, and development of preimplantation stages of embryonic development. As in the male, many reproductive endpoints have biological relationships ( ), such that while some endpoints may be redundant, many are complementary and interpretation of potential effects on female fertility should account for those interdependencies in determining the biological importance of any findings.

Organ Weights

Female reproductive organ weights can have marked variability due to normal interanimal reproductive cycling and age, and thus, as recommended by the STP, are not included as tissues for routine weighing in repeat-dose general toxicity studies ( ). If ovary weights are collected in rodents, it is recommended that it be done in study of less than 6 months duration due to variability attributed to reproductive senescence. When NHP are the only pharmacologically relevant species, and in the absence of a reproductive cause for concern based on literature, mature female reproductive organ weights and microscopic examination are considered sufficient evaluation for potential effects on female fertility ( ).