Standard Morphologic Evaluation of the Heart in the Laboratory Dog and Monkey

Abstract

The nonrodent species most commonly utilized in preclinical safety studies are the purpose-bred beagle dog and cynomolgus macaque (Macaca fascicularis). Potential effects of a new chemical entity (NCE) on the heart pose serious concerns; consequently in vivo testing is focused on detection of functional alterations as well as morphological changes. Macroscopic and microscopic evaluation of the heart is based on a standard survey of key structures to properly assess presence of spontaneous and potential drug-induced lesions. Evaluation of historical controls to determine type and frequency of background change is valuable, as studies with non-rodent species generally have a small sample size. Archived control dog and monkey data were retrospectively reviewed, including terminal body weight (BW), heart weight (HW), and archival glass slides of heart. Control dogs had minimal background changes that included myxomatous or cartilagenous change in the cardiac skeleton and a variable degree of vacuolation in Purkinje fibers. Control monkey hearts commonly contained inflammatory cell infiltrates, myocyte anisokaryosis, and handling artifacts, while myocyte degeneration, squamous plaques, pigment, and intimal plaques were occasionally observed. These findings highlight the utility of consistently recorded and readily accessible archived control data when attempting to discern background spontaneous changes and artifacts from test-article induced changes.

Introduction

Registration of NCE or biologics requires a sponsoring company to demonstrate adequate safety through in vitro and in vivo evaluations. In vivo toxicology testing is performed in both rodent and nonrodent species and the nonrodent species most commonly utilized are the purpose-bred beagle dog and the cynomolgus macaque (Macaca fascicularis). The latter species may be wild-caught or selected from breeding colonies at isolated locations. Potential effects of an NCE on heart are a serious concern and in vivo testing is focused on detection of functional alterations via safety pharmacology studies and evaluation of morphological changes identified macroscopically or microscopically in repeat dose toxicology studies. As with other organ systems, the key to consistent morphologic evaluation and interpretation of histologic data is based on a standardized approach to collection, dissection, and sampling. Studies with nonrodent species have a small number of animals per group (3 or 4/sex) and understanding the biological relevance of morphologic observations in toxicology studies is dependent on knowledge of spontaneous background changes to better differentiate these from test article-induced pathology. A review of control data at one of the GlaxoSmithKline laboratories was completed to assess the range (type and incidence) of spontaneous changes in heart of dogs and monkeys.

Materials and Methods

Animals

All procedures and care of animals were in accordance with the principles for humane treatment outlined by the Institute of Laboratory Animal Resources Guide for the Care and Use of Laboratory Animals of the National Institutes of Health, USDA Guidelines Animal Welfare Regulations, and were reviewed and approved by the GlaxoSmithKline Institutional Animal Care and Use protocols Committee.

Beagle dogs (103 males and 98 females) aged 9–20 months from studies with durations of 1 to 6 months were used for this review. These dogs received vehicle only by the oral or intravenous routes of administration. Dogs were obtained from Marshall BioResources (formerly known as Marshall Farms USA, Inc.), North Rose, NY. Dogs were singly housed in climate controlled rooms at 64 to 84°F with a relative humidity of 30 to 70%. The rooms had approximately 10–15 air changes per hour and a photoperiod cycle of 12 hours light/12 hours dark. Dogs were fed with a commercially available complete, pelleted diet and generally received 300 to 400 grams/day. Water was provided ad libitum.

Cynomolgus monkeys (64 males and 56 females) aged 2–7 years from studies with durations of 10 days to 1 year were used for this review. These monkeys received vehicle only by either the oral, intravenous, or subcutaneous routes of administration. The monkeys were obtained from Covance/Hazelton, Primate Products (Miami, FL) and Charles River (Houston, TX). Monkeys originated from either China, Indonesia or Mauritius; all were captive bred and raised in groups outdoors in their country of origin before shipment to the United States. Monkeys were singly housed in climate controlled rooms at 64 to 84°F, with a relative humidity of 30 to 70%. The rooms had approximately 10–15 air changes per hour and a photoperiod cycle of 12 hours light/12 hours dark. Monkeys were fed with a commercially available complete diet and generally received 8 (male) or 6 (female) biscuits/day; a daily allotment of fresh fruit was also provided. Water was provided ad libitum.

Prior to necropsy, each dog was pretreated with an intramuscular or subcutaneous injection of acepromazine maleate (approximately 0.5 mg/kg) and then anesthetized with an intravenous injection of sodium pentobarbital (starting at approximately 30 mg/kg). The dogs were killed by exsanguination.

Prior to necropsy, each monkey was pre-treated with an intramuscular or subcutaneous injection of ketamine HCl (approximately 10.0 mg/kg) and then anesthetized with an intravenous injection of sodium pentobarbital (starting at approximately 30 mg/kg). The monkeys were killed by exsanguination.

The heart was carefully removed from the thoracic cavity, rinsed with saline, weighed and all 4 chambers were opened to allow for visual inspection of valves and endocardial surface. The heart was then immersed in 10% neutral buffered formalin and fixed for 24 hours at room temperature. The time from exsanguination to immersion fixation was generally less than 30 minutes.

Procedures

Historical control data included BW, HW, and glass slides containing heart and surrounding tissues. A 95% prediction interval was generated that represents the boundaries within which future observed values (HW) are expected to fall with a 95% predication, based upon the fit of the present experiment’s data. The figures were generated by SAS/GRAPH version 8.2. Assessment of histologic changes for each species was performed by a single pathologist (dog, C.M.K; monkey J.D.V.). All slides were evaluated without knowledge of previously recorded diagnoses and a peer review was conducted to ensure that microscopic changes were recorded consistently. Sections of right and left atria, right and left ventricles, papillary muscle, atrio-ventricular (AV) valves, aortic valve, and AV node were evaluated for each animal.

Results—Dogs

One hundred and sixty dogs had both BW and HW (79 males and 81 females) values available. A summary of weights is presented in Table 1 and a correlation of HW to BW is provided in Figure 1. Collection, handling, and fixation artifacts were occasionally observed in the dog and were similar to some observed in the monkey; these are described next.

Microscopic changes are summarized in Table 2. The most common change was the presence of myxomatous tissue and/or cartilage at the base of the heart, typically seen in the section which included aortic valve and aorta; this was observed in 17 of 103 males (17%) and in 15 of 98 females (15%). This is a normal structural component of the “cardiac skeleton” or fibrous base of the heart and may become more evident with age (Figure 2).

Prominent vacuolation of Purkinje fibers was present in 15 of 103 males (15%) and in 12 of 98 females (12%). The vacuolation varied in prominence but was more evident in fibers following vascular adventitia into the ventricles (Figure 3). Inflammatory infiltrate, predominantly mononuclear cells, was uncommon and tended to be minimal with one focus per section and in variable locations. This finding was present in 5 of 103 males (5%) and in 2 of 98 females (2%). Coronary artery medial hypertrophy was characterized by a thickened vessel wall and a reduced lumen (Figure 4). This was typically seen only in intramural arteries and was present in 12 of 103 males (12%) and in 4 of 98 females (4%). Mineralization of aorta (Figure 5) was noted along the outflow tract present in heart sections and was seen in 2 of 103 males (2%) and in 2 of 98 females (2%).

Some observations that occurred infrequently included: variable prominence of adipose tissue in atria or ventricles (right side more frequently than left in 3 of 103 males and 1 of 98 females) (Figure 6); dilated lymphatics in AV valve in 1 female, mineralization of papillary muscle in 1 male, single myofiber degeneration with inflammatory infiltrate in 1 male, inflammation of AV valve in 1 female, minimal myxomatous change in AV valve in 1 female, epicardial inflammation in 1 male, arteritis of extramural coronary artery in 1 female and intimal hyperplasia of a coronary artery in 1 male.

A normal morphological feature was prominent endocardial and subendocardial connective tissue at the insertion of chordae tendinae (Figure 7). Depending on section orientation, this presented with variable thickness and disrupted adjacent myocardial fibers.

Results—Monkeys

Seventy-eight monkeys had both BW and HW (44 males and 34 females). A summary of the weights is presented in Table 3 and correlation of HW to BW is provided in Figure 8. Collection, handling, and fixation artifacts were commonly observed in monkey heart. These included linear aggregates of hypereosinophilic myocytes adjacent to a cut edge (Figure 9), mechanical separation of myofibers by clear space, focal distortion and hypereosinophilia representing crush effects (linear or circular patterns), linear hypereosinophilia mimicking contraction band necrosis (Figure 10) and foci of individual and/or clusters of hypereosinophilic myocytes representing differential myocyte contraction (Figure 11).

Microscopic observations are summarized in Table 4. Individual myofiber degeneration/necrosis was characterized by a fragmented, hypereosinophilic myofiber surrounded by an inflammatory infiltrate (Figures 12–13). Generally these were found as a single focus with variable distribution and noted in 3 of 64 males (5%) and 2 of 56 females (4%).

Squamous plaques, characterized by a focus of epicardial stratified squamous epithelium, was noted in 3 of 64 males (5%) and 1 of 56 females (2%) (Figure 14). Intracytoplasmic granular, light brown pigment was most commonly seen in the left ventricle and noted in 5 of 64 males (8%) and 4 of 56 females (7%) (Figure 15).

Inflammatory infiltrates were observed as single or multiple aggregates of mixed cell types consisting of predominantly lymphocytes, with occasional histiocytes, multinucleated giant cells, and/or eosinophils. These foci were scattered randomly throughout all portions of heart and pericardial fat and noted in 39 of 64 males (61%) and 36 of 56 females (64%). Of the 75 total observations, the severity scores were as follows: 59 minimal (Figure 16), 15 mild (Figure 17) and 1 moderate (Figure 18).

Anisokaryosis and karyomegaly were most commonly noted in left ventricle and interventricular septum (Figure 19). This was observed in 34 of 64 males (53%) and 27 of 56 females (48%). Of the 61 total observations, 41 were minimal, 16 mild, and 4 moderate.

Intimal plaques within coronary arteries were characterized by focal expansion of loose connective tissue and scattered mononuclear cells (Figure 20). This occurred in 2 of 64 males (3%) and 3 of 56 females (5%). One female monkey had epicarditis with mixed inflammatory infiltrate and necrotic debris.

Discussion

To provide for the best assessments of cardiac changes in toxicology studies it is important to standardize tissue dissection, trimming and the specific areas for histologic evaluation. Consistent collection provides the best foundation for comparison within and across studies of both spontaneous changes as well as compound-related effects. Although there may be a variety of mechanisms for drug-induced cardiac change, there are certain key anatomical regions that are important to survey, such as papillary muscle, subendocardium, valves and coronary arteries (Greaves, 1998). Our laboratory routinely evaluates both atria, ventricles (with papillary muscle), AV valves, interventricular septum and aortic valve; additionally sinoatrial node and AV node may be present. A detailed review of methods to evaluate the cardiac conduction system has been reported (Palate et al., 1995).

Handling and processing of fresh contractile tissues such as cardiac muscle may introduce a number of artifacts (Winters and Schoen, 2001). The heart must be carefully removed and handled by the prosector with gentle use of forceps to minimize crush artifacts. Crush artifacts cause focal distortions of cardiac myocytes leading to indistinct cell borders, irregular nuclei, and hypereosinophilic cytoplasm. In addition to crush artifacts, handling of the heart can cause separation of myofibers, which may be mistaken for edema (Winters and Schoen, 2001). Incising the fresh heart causes focal myofiber contraction and myocytes along the cut edge are often irregular and hypereosinophilic (Baroldi, 2001). Both handling and immersion fixation can lead to differential myocyte contraction, which appears microscopically as individual and/or irregular groups of hypereosinophilic myocytes scattered throughout the myocardium. Handling artifacts causing post mortem myocyte contraction and irregular cytoplasmic eosinophilia can have a similar histologic appearance to contraction band necrosis (Winters and Schoen, 2001; Baroldi, 2001). Therefore it is critical to handle the heart gently during dissection in order to minimize the number of artifacts encountered. Immediate immersion of the heart into chilled normal saline after removal from the thorax but prior to dissection to open chambers may help minimize artifacts.

A good knowledge of the species is important to be able to discern spontaneous changes and artifacts from test-article induced changes. The species utilized have varying types and incidence of spontaneous changes. Factors that can affect type and incidence of change include origin of the animal, environmental conditions, provisions for routine health care, type of vehicle used, and the route of administration. The threshold for assessment of spontaneous change can be quite variable, with some pathologists capturing minimal, discrete changes while others may opt to diagnose only the more extensive mild to moderate changes. Either approach is suitable as long as there is consistency within a study. With the advent of routine peer review in many laboratories, agreement on threshold for spontaneous changes is becoming standard and this will improve consistency between studies.

Generally, young beagle dogs utilized in toxicology studies have minimal spontaneous cardiac changes. Most reviews of canine cardiac pathology are based on multiple breeds from a clinical perspective (Luginbuhl and Detweiler, 1965; Sheridan, 1967). There are a few reports summarizing spontaneous changes in purpose-bred beagles and the minor lesions referenced include focal myocarditis, focal fibrosis, focal calcification, granuloma and medial degeneration of the aorta (Hottendorf and Hirth, 1974; Oghiso et al., 1982). While certain reports have indicated a high prevalence of spontaneous arteritis (Hartman, 1989; Hayes et al., 1989; Son, 2004), this finding was not common in our laboratory, as it was only noted in an extramural coronary artery of one female dog. The coronary artery hypertrophy observed in this survey of control dogs may be more a consequence of hyper-contraction than a true pathologic change. The hypertrophy was only noted in intramural arteries and generally tended to be near cut edges, suggesting the possibility of a recoil effect. The minimal, sporadic observations in young beagles are in contrast to observations noted in older beagles with common changes including endocardiosis (myxomatous degeneration) of AV valves and to a lesser extent endocarditis, myocardial degeneration, myocardial hypertrophy, myocardial calcification, and myocardial fibrosis (Van Vleet, 2001).

Cynomolgus macaques were obtained from China, Indonesia, or Mauritius and although the source companies do screen for several spontaneous primate diseases, many monkeys have a variety of parasitic, or other clinical conditions (Lowenstine, 2003). Despite common use of the Cynomolgus monkey as a model for atherosclerosis (Williams and Suparto, 2004), there are no reports summarizing commonly observed spontaneous cardiac changes in this species. In our survey, the most frequently encountered finding was inflammatory cell infiltrates. Inflammatory infiltrates are commonly observed in a variety of organs in macaques (Lowenstine, 2003) and, in this review, infiltrates were randomly distributed in all parts of the heart and surrounding tissues. The majority of these infiltrates were single or multiple foci of small numbers of lymphocytes that only minimally separated myofibers. A number of these minimal foci likely fall into a category of changes not typically recorded by many toxicologic pathologists and as a result, the high incidence of inflammatory infiltrates in this study may be a function of applying a low threshold for diagnosing this finding.

Anisokaryosis and karyomegaly of myocyte nuclei were commonly observed in this survey. As with inflammatory infiltrates, the majority of these nuclear changes were minimal and likely of little functional significance. This common change has been reported in nonhuman primates (Lowenstine, 2003) and in humans (Veinot et al., 2001) and may represent polyploidy.

Individual myocyte degeneration/necrosis with a surrounding inflammatory infiltrate was identified in 4% of the study population. This has been observed as a spontaneous change in nonhuman primates (Cowan et al., 1983) and humans (Lewis and Silver, 2001) and has been produced experimentally in monkeys by administration of catecholamines (Khullar et al., 1989). Elevations of endogenous catecholamines secondary to routine manipulations during general toxicology study may contribute to the generation of this lesion and must be differentiated from drug-induced cardiac pathology.

Additional findings of low incidence included squamous plaques, intracytoplasmic pigment and intimal arterial plaques. Squamous plaques have been reported in monkeys and humans and are thought to be displaced foregut epithelium (Kaspareit et al., 2003). Several monkeys had a light brown, granular pigment within myotcyte cytoplasm. This pigment is histologically similar to lipofuscin, which is commonly reported in humans and the incidence increases with age (Lester, 2001). However, an association with age in this survey could not be evaluated due to the small number of affected monkeys and the unavailability of precise breeding history/age in study monkeys. Intimal plaques within coronary arteries have been identified previously in cynomolgus macaques (Clarkson et al., 1994).

In summary, standardization of approach provides the best foundation for comparison within and across studies to assure identification of compound-related effects. Our experience suggests minimal changes occur in hearts of young beagle dogs, the most frequent being myxomatous or cartilagenous change in cardiac skeleton and secondly vacuolation of Purkinje fibers. For the monkey, inflammatory cell infiltrates and myocyte anisokaryosis were the most common findings.