Abstract
Hibernomas are rare neoplasms originating in brown adipose tissue of humans and other animal species, including laboratory animals. Background incidence values for these tumors in all common strains of laboratory rats are generally accepted as being <0.1%. Between April 2000 and April 2007, however, sixty-two hibernomas (an overall prevalence of 3.52%) were observed in a total of 1760 Sprague-Dawley rats assigned to three carcinogenesis bioassays at two separate research laboratories. All rats were obtained from Charles River’s breeding facilities in either Portage, Michigan, or Raleigh, North Carolina. Tumors (twenty-nine benign and thirty-three malignant) were randomly distributed among test article–treated and control groups and were considered to be spontaneous. Most tumors originated in the thoracic cavity, and they were usually described as soft, mottled to tan masses with nodular to lobulated profiles. Immunohistochemical procedures for uncoupling protein 1 (UCP1) confirmed brown adipose tissue as the site of origin rather than white fat. The marked increase in hibernomas in our studies suggests that greater numbers of spontaneous hibernomas may be sporadically encountered in future carcinogenesis studies with Sprague-Dawley rats. The increased potential for hibernomas to arise as spontaneous neoplasms has important implications in studies involving peroxisome proliferators–activated receptor (PPAR) drugs, lipophilic environmental chemicals (e.g., polychlorinated biphenyls), and other molecules or physiologic processes (e.g., β-adrenergic stimulation) that may target brown fat adipocytes.
Introduction
Hibernomas are rare neoplasms that originate in brown adipose tissue (BAT) of both humans and animals. Brown adipose tissue is a specialized adipose tissue whose primary function is to provide nonshivering thermogenesis during periods of cold-induced stress (Cannon and Nedergaard 2004; Iatropoulos and Williams 2004). Brown adipose tissue is especially critical for thermal maintenance in neonates, where it is abundantly present in deposits within the abdominal and thoracic cavities as well as in subcutaneous tissue of the interscapular regions (Figure 1A). In rodents, BAT deposits normally persist in multiple locations throughout life.
During routine necropsy procedures associated with three Good Laboratory Practices–compliant, proprietary, chronic toxicity/carcinogenesis studies conducted at two separate research laboratories, twenty-nine benign and thirty-three malignant hibernomas were observed in a total population of 1760 Sprague-Dawley rats (combined incidence of 3.5%). This rate of prevalence markedly exceeded historical incidence values for the conducting laboratories, and for values reported in the open literature and in compilations reported by government agencies and other commercial laboratories. Most important, however, was the observation that hibernomas in all three studies were randomly distributed among all test article–treated and control groups, suggesting that there was no causal relationship with the test articles. Precise causative factors for the increase in hibernomas were not identified. The marked increase over historical control values, however, indicated that greater numbers of hibernomas may be expected in future carcinogenesis bioassays with Sprague-Dawley rats. The aim of this article is to inform pathologists, toxicologists, and allied biomedical scientists that, for undetermined reasons, hibernomas have apparently become rather common, spontaneous neoplasms in aging Sprague-Dawley rats.
Materials and Methods
All three studies reported in this article were proprietary research investigations, and the sponsors for each study requested that the precise nature of the respective test article(s) remain confidential. Two studies (hereafter identified as Studies A and B) were conducted at WIL Research Laboratories, LLC, Ashland, Ohio, and the third study (Study C) was conducted at Covance Laboratories, Madison, Wisconsin. All studies were conducted in facilities accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care (AAALAC) and were in compliance with the U. S. Food and Drug Administration’s Good Laboratory Practice regulations, the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and the Animal Welfare Act. Environmental conditions, including food and water intake, were not considered to be influential in hibernoma development, and all other forms of neoplasia and reactive or degenerative diseases observed in each study were within normal background limits for Sprague-Dawley rats.
Study A was a twenty-four-month dietary study of two potential food additives, either alone or in combination. Other than dietary ingredients, no vehicle was required. This study included 380 male and female rats (760 total) assigned to six dietary test groups plus a control group that was fed a standard ration. Rats included in this study were obtained from Charles River’s breeding facility in Raleigh, North Carolina.
Study B was a twenty-four-month oral carcinogenicity study including 240 male and female rats (480 total) assigned to three test article–treated groups and a control group. The vehicle for the test article was deionized water. Rats in this study were obtained from Charles River’s breeding facility in Raleigh, North Carolina.
Study C was a scheduled 104-week oral gavage carcinogenicity and toxicokinetic study involving 260 male and female (520 total) rats assigned to three test article–treated groups and one control group. The vehicle for the test article was 2% (w/v) Tween 80 in 1% (w/v) aqueous carboxymethylcellulose in reverse-osmosis water. Males were terminated at ninety-eight weeks when survivors in the high-dose group reached twenty animals. Rats in this study were obtained from Charles River’s breeding facility in Portage, Michigan.
Necropsy procedures for all studies included the collection, processing, and microscopic examination of a complete set of organs (approximately fifty-five tissues per animal). Tumors that were subsequently determined to be hibernomas were usually collected as solitary, soft tissue masses with varying external profiles (nodular to lobulated), tinctorial qualities (tan to red-brown), and texture (soft to variably fibrous). The precise tissue of origin for these soft-tissue masses usually could not be determined at necropsy. Collected specimens were immersed in10%neutral-buffered formalin, processed into hematoxylin and eosin–stained microslides, and examined his-tologically along with other protocol-required tissues (full trim of all major organs). Although histologic features were often highly compatible with tumors of brown fat origin, select tumors were stained immunohistochemically for the presence of uncoupling protein 1 (UCP1), a unique protein that is constitutionally expressed in brown fat adipocytes (Figure 1B). Uncoupling protein 1 expression distinguishes brown fat tumors from other poorly differentiated mesenchymal tumors that may have vacuolated cytoplasm (e.g., liposarcomas). The primary antibody was a rabbit polyclonal to UCP1 obtained from Abcam (Cambridge, MA). The secondary antibody was a biotin-SP–conjugated AffiniPure F(ab’)2 fragment goat anti- rabbit IgG (H+L) purchased from Jackson Immuno (West Grove, PA). The negative control was an IgG protein–matched rabbit negative control purchased from Zymed (Carlsbad, CA). The ABC Vector Rabbit Elite kit (Vector Labs, Burlingame, CA) was incubated on the tissue for thirty minutes prior to staining with DAB. This kit was obtained from Dako (Carpinteria, CA).
Results
Histologically, collected masses were determined to be either benign or malignant hibernomas originating from brown fat adipocytes. Benign tumors were characterized by expanding populations of round to oval cells that were encapsulated and subdivided into irregular clusters by fibrous strands of varying thickness. Proliferating cells were well differentiated and displayed modest amounts of eosinophilic, foamy to multivacuolated cytoplasm along with central nuclei that exhibited low mitotic activity (Figures 1C and 1D). Malignant tumors often were locally invasive and displayed more pleomorphic cell populations with hyperchromatic nuclei that exhibited increased mitotic activity (Figure 1E). Vascular invasion was sometimes observed, with embolic clusters of tumor cells reaching pulmonary blood vessels (Figure 1F).
In Study A, hibernomas were observed in 10/380 males (2.6%) and in 4/380 females (1.1%). All hibernomas were observed grossly as masses originating in the thoracic cavity, where they often circumscribed the aorta and/or compressed the lungs or mediastinal structures. Twelve tumors were considered to be benign; however, two (one male and one female) displayed malignant features. Metastases to distant sites were not observed. Of the ten males with hibernomas, eight were either found dead or euthanized in extremis between study days 378 and 732, and the cause of death was uniformly attributed to expanding hibernoma in the thoracic cavity. Of the four females with hibernomas, all were found dead or euthanized in extremis between study days 593 and 722; however, hibernoma was considered to be the cause of death in only one animal. Treatment-based incidence values for hibernomas observed in Study A are presented in Table 1.
In Study B, hibernomas were observed in 15/240 males (6.3%) and 10/240 females (4.2%). Most hibernomas were observed grossly as masses originating in the thoracic cavity; however, in two males the abdominal cavity was determined to be the primary site, and in one male, the hibernoma appeared to develop in the subcutaneous fat near the mammary gland. In the females, all hibernomas appeared to originate in the thoracic cavity. Sixteen tumors were considered to be benign (eight males and eight females), and nine were malignant (seven males and two females). Metastases to the lungs (intravascular emboli) were observed in one male and one female, and metastasis to the adrenal gland was noted with one male. Of the fifteen males with hibernomas, twelve were either found dead or euthanized in extremis between study days 269 and 727. Hibernomas were considered to be the cause of death for eleven of the males and six of the females. Treatment-based incidence values for hibernomas observed in Study B are presented in Table 2.
In Study C, hibernomas were observed in 13/260 males (5.0%) and 10/260 females (3.8%). Although brown fat of the interscapular region was a protocol-required tissue, all hiberno-mas were collected grossly as masses in the thoracic or abdominal cavities or in the mammary gland region. In relative contrast to studies A and B, twenty-two of twenty-three hibernomas in Study C were regarded as malignant. Eighteen tumors (seventeen malignant and one benign) were present in the thoracic cavity, whereas five additional malignant tumors were observed in the abdominal cavity (one) or in the mammary gland region (four). Metastatic lesions were identified for six primary malignant tumors. Malignant hibernoma was regarded as the cause of death for seventeen rats that were either found dead or euthanized in extremis between study day 264 and the terminal necropsy (day 684 for males and days 728–734 for females).
In all three studies, hibernomas were slightly more prevalent in males, and most tumors originated in the thoracic cavity. When compared with Studies A and B, malignant hibernomas were strikingly more common in Study C. Reasons for this trend were not apparent; however, Study C was completed approximately two years following the completion of Studies A and B, and rats assigned to Study C were obtained from Charles River’s breeding facility in Portage, Michigan, rather than the facility in Raleigh, North Carolina. Essentially all hibernoma diagnoses were based on grossly apparent masses that were collected at necropsy from the thoracic or abdominal cavities or the mammary gland region. In this regard, it is possible that some small hibernomas escaped gross detection and were not processed for microscopic examination. Microscopic findings in other tissues collected at necropsy from all three studies were generally consistent with common, spontaneous changes observed in aging Sprague-Dawley rats, and no neoplasms were observed that were considered to be test article–related. All studies (A, B, and C) were accompanied by companion, long-term mouse studies with the same test article(s) using similar exposure regimens. Hibernomas were not observed in the mouse studies.
Discussion
Incidence rates for hibernomas in these three studies greatly exceeded historical values for the two conducting laboratories and values reported by the National Toxicology Program (NTP), NIEHS, and other commercial laboratories. Search of the NTP tumor database identified a total of nine hibernomas in rats and two in mice assigned, respectively, to 213 rat studies (92,296 rats) and 201 mouse studies (91,366 mice). For both species, the incidence was < 0.001%. Although most of the NTP studies involved Fischer-344 (F-344) rats and B6C3F1 mice, six of the nine hibernomas were reported in Osborne-Mendel rats assigned to two different studies that were completed during the early phases of the NTP carcinogenesis bioassay program.
Hibernomas were not reported in a compilation of spontaneous neoplasms reported by Charles River Laboratories in March 2004. This survey included nontreated control Crl:CD(SD) rats representing thirty-one studies (104 weeks) conducted in eight different testing facilities in the United States, Europe, Canada, and Japan between 1989 and 2002. In twelve of these studies, rats were confirmed to be from Crl:CD(SD) colonies produced under the International Genetic Standard (IGS) breeding system that was designed to stabilize the degree of genetic diversity among colonies of Crl:CD(SD) rats on a worldwide basis.
In a similar compilation of spontaneous neoplasms in control Crl:CD BR rats assigned to thirty-six different studies that were conducted prior to February 1989, Charles River reported one hibernoma in the thoracic cavity of a male. In addition to the Charles River compilations, Al Zubaidy and Finn (1983), Coleman (1980), and Stefanski (1987) reported small numbers of hibernomas as incidental findings in the thoracic cavity of rats assigned to three unrelated studies. Two possible exceptions were noted in reports by Poulet et al. (2004) and by Brees et al. (2008). In Poulet’s report, eleven hibernomas were observed in a population of 520 Sprague-Dawley [Crl:CD BR] rats (260/sex) assigned to a two-year carcinogenesis study of phentolamine mesylate (an α-adrenergic antagonist) administered at daily doses of 10, 50, and 150 mg/kg. In this study, three males and one female given 10 mg/kg, four males given 50 mg/kg, and three males given 150 mg/kg displayed hiber-nomas in either the thoracic cavity or in retroperitoneal (abdominal) locations. In the report by Brees, hibernomas were observed in the mediastinum of three male rats administered high doses of varenicline (a partial nicotinic agonist) during a two-year carcinogenesis study (unpublished data). Although there was no distinct dose relationship for the hibernomas observed in these two studies, phentolamine mesylate and var-enicline could not be entirely ruled out as carcinogens because hibernomas were restricted to treated animals only and were considered to represent a rare form of neoplasia. In both studies, other forms of neoplasia were within historical control values for the performing laboratory. Furthermore, there were no significant histopathological findings in a reported twenty-six-week
In humans, hibernomas have been described as rare neoplasms that occur most commonly in adults, with a mean age of 38.0 years and an age range of 2–75 years. In a review of 170 cases from the Armed Forces Institute of Pathology, ninety-nine tumors were observed in men and seventy-one in women. The most common anatomic locations were thigh (n = 50), shoulder (n = 20), back (n = 17), neck (n = 16), chest (n = 11), arm (n = 11), and retroperitoneum (n = 10) (Furlong et al. 2001). Additional sites where hibernomas have been reported include the forehead, axilla, and interscapular regions (Baskurt et al. 2004; Chen et al. 1998; Chitoku et al. 1998; Lay et al. 2000; Wilhelm et al. 1993). No association with drugs or xenobiotic chemicals has been reported for hibernomas in humans; however, in a report by Lele et al. (2002), two hibernomas in human females were associated with an overexpression of
Although brown fat appears to be relatively unimportant in adult humans, it is critical in laboratory rodents that have higher basal metabolic rates and substantially greater surface-to-body-mass ratios. In rodents, brown fat deposits in the thoracic and abdominal cavities and in the subcutaneous tissue are richly supplied with adrenergic stimulation from the sympathetic trunks. During periods of cold-induced stress, hypothalamic signaling via sympathetic nerves releases norepinephrine, which binds to receptors on brown adipocytes, resulting in a cascade of intracellular signals that activate lipoprotein lipase with the subsequent liberation of free fatty acids. In brown adipocytes, the metabolism of fatty acids is largely regulated by a unique protein UCP1 that is present along the inner mitochondrial membrane. This protein (also termed
Despite the importance of BAT in thermoregulation and metabolic homeostasis of laboratory rodents, this tissue has been largely ignored during the histopathological evaluation of animals assigned to preclinical drug safety and efficacy bioassays. In part, this omission has resulted because brown fat exists as irregular, widely disseminated deposits that complicate uniform collection and organ weighing procedures. Additionally, distinct morphologic changes are seldom recognized grossly or microscopically in BAT, and finally, most metabolic pathways that are selectively targeted by xenobiotics are not functionally operative in brown adipocytes. More recently, however, several classes of new or investigational drugs have emerged that may target brown fat either directly or indirectly. These drug classes include peroxisome proliferators–activated receptor agonists (for diabetes and atherosclerosis), adrenergic agonists (norepinephrine), and adrenergic antagonists (phentolamine mesylate). In addition, fat-soluble environmental chemicals such as DDT and polychlorinated biphenyls have the potential for altering the metabolism of brown fat adipocytes. Based on increasing knowledge of the pathophysiology of brown fat and the apparent increase in hibernomas in recent studies, it would appear prudent to expand the collection and examination of brown fat deposits in future rodent carcinogenesis bioassays.
Conclusion
The overall incidence of hibernomas in Sprague-Dawley rats assigned to three carcinogenesis bioassays with unrelated, proprietary test articles greatly exceeded historical control values for the testing laboratories and background incidence values reported in the open literature. Tumors were slightly more prevalent in males, and most originated in brown fat of the thoracic cavity. In Study A, fourteen hibernomas (twelve benign and two malignant) were observed in total of 760 animals (380/sex). In Study B, twenty-five hibernomas (sixteen benign and nine malignant) were observed in 480 rats (240/ sex), and in Study C, twenty-three hibernomas (one benign and twenty-two malignant) were noted in 520 rats (260/sex). In each study, tumors were randomly distributed among test article–treated and control groups, and none was considered to be test article–related. Furthermore, test article–related carcinogenesis was not observed in other organs in any study, and there was no evidence of hibernomas in companion mouse studies. Collectively, pathologic findings in these three studies strongly suggested that hibernomas represented spontaneous neoplasms unrelated to respective test articles, and that genetic and/or environmental factors should be considered as likely causes. Results also suggest that increased numbers of hibernomas can be expected in future carcinogenesis studies with Sprague-Dawley rats, and that brown fat should receive greater attention in future safety and efficacy studies.