Early Onset of Spontaneous Renal Preneoplastic and Neoplastic Lesions in Young Conventional Rats in Toxicity Studies

Abstract

Although occurring in aged laboratory rodents, spontaneous renal tumour are unknown in animals younger than 18 weeks. A survey on renal preneoplastic and neoplastic lesions has been performed on Sprague-Dawley rats from general toxicology studies over the period January 2004—May 2006. Data from 2249 rats necropsied and 1206 rats (688 males and 518 females) examined microscopically from 52 studies, are reported. The age range at necropsy was between 12 and 18 weeks and all the animals were obtained from the same supplier. Three cases of tubular carcinoma, 1 tubular adenoma, and 4 cases of atypical tubular hyperplasia were observed in 5 females and 3 males from both control and treated groups from 6 studies with unrelated test compounds. In treated rats, the lesions were considered spontaneous in nature, rather than related to treatment, because of: (1) their sporadic incidence, (2) the short duration of the studies, and (3) the absence of similar lesions in other rats given the same test compound. These lesions are considered a recently occurring spontaneous finding, and the similarities with the familial renal cancer models, namely the Eker and the Nihon models, strongly suggest genetic factors as responsible for the lesions.

Introduction

It has been generally observed that renal tubule tumours in laboratory rodents, whenever occurring spontaneously or being induced by chemicals, are slow-growing neoplasms. Accordingly, virtually all of the spontaneous cases recorded in literature have been encountered in safety evaluation studies at the end of 2-year carcinogenicity bioassays (Hard et al., 1994); furthermore spontaneous occurrence of rat renal neoplasia are of low incidence in F344 rats (Thurman et al., 1995; Hiasa and Ito, 1987) and uncommon or absent in Crl:CD (SD) IGS BR (Sprague Dawley) rats from long term studies (Chandra et al., 1993, Baldrick, 2005). Indeed those lesions are even not reported in Sprague–Dawley rats younger than 50 weeks of age (Son and Gopinath, 2004). The occurrence of renal tubule tumours in early toxicity studies is an extraordinary finding and, to our knowledge, unprecedented apart from toxicity studies performed in Japan in 1998, which led to the discovery and characterisation of a model of familial renal cancer, named “Nihon” (Okimoto et al., 2000). This model represents an example of a single gene mutation in the Birt-Hogg-Dubè (BHD) gene leading to the development of preneoplastic lesions from as early as 3 weeks of age and subsequent adenocarcinomas by 6 months of age (Okimoto et al., 2004).

Another rat model of familial renal cancer, the Eker rat model, was described by Eker and Mossage in the 1950s (Eker and Mossage, 1961). This is another example caused by a mutation in a single tumour suppressor gene (tsc2) encoding tuberin (Yeung et al., 1994), and 100% of heterozygous rats will develop renal neoplasms by 12 months of age (McDorman and Wolf, 2002). Both models have a Mendelian dominantly inherited predisposition for development of renal tumours, affecting different genes (Hino et al., 2006).

This paper represents a survey of preneoplastic and neoplastic renal pathology that has been emerging in the last two years in our facilities, using conventional rats younger than 18 weeks of age, from a single supplier.

The knowledge of the sporadic occurrence of spontaneous renal tumor in young Sprague–Dawley rats can be extremely useful in interpreting data from studies.

Materials and Methods

This overview refers to 52 studies run over the period January 2004 – May 2006. The studies were of differing lengths, from acute to 4-week studies and one instance of a 6-week study, for a total of 2249 Crl:CD (SD) IGS BR (Sprague–Dawley) rats (1349 males and 900 females), obtained from the same supplier (Charles River Italia, Calco, Italy).

In all studies, the animals were allocated to groups (2 rats/sex/cage), and animal room temperature and relative humidity controls were generally maintained at 21 ± 1°C and 55% ± 10, respectively. The animals were given “ad libitum” diet (Gamma Irradiated Rat and Mouse No 1 Expanded Diet, Special Diet Service Ltd–England) and free access to filtered tap water.

The rats were approximately 10-week aged on arrival and after 1 week of acclimatization, entered the study and received vehicle (e.g., saline, HPMC/TWEEN80), or test compound (belonging to different chemical or pharmacological classes), by oral gavage. All the studies were carried out in accordance with Italian regulation (governing animal welfare and protection), the European Directive 86/609/EEC, and according to the internal GlaxoSmithKline Committee on Animal Research & Ethics (CARE) review, with related codes of practice.

The age of the animals at necropsy ranged between 12 and 18 weeks, depending on the duration of the study. All animals (including premature decedent) underwent full necropsy and standard tissue sampling was performed, followed by fixation in 10% neutral buffered formalin, embedding in paraffin wax, sectioning (nominally at 4–5 μm) and staining with Hematoxylin and Eosin. Tissues were examined microscopically in accordance with the study protocol. All procedures were performed according to Good Laboratory Practice (GLP).

Microscopic evaluation of the kidneys comprised the examination of both organs by transversal section (incorporating the papilla). Initially this examination was performed on control and high dose animals; lower dose groups were examined only if macroscopic abnormalities were present, or if findings were observed in the high-dose group. In one 4-week study, a step sectioning approach for the kidneys of all animals was applied according to Eustis et al. (1994). The morphological characterisation of the lesions followed the guidelines from the Societies of Toxicologic Pathologist (Hard et al., 1995) and the classification of the WHO/IARC nomenclature (Alden et al., 1992).

Results

At macroscopic examination, whitish nodules were observed on the surface of kidneys in three animals from two different 4-week studies. The nodules were multifocal, bilateral in 1/10 female rats from the high-dose group of 1 study (Figure 1), or focal, unilateral in 2 male rats (1 from high- and 1 from intermediate-dose groups) of another study. These nodules were apparent on the surface of the kidney, single or multiple and confluent up to 5 mm, at the cranial pole. A solitary 2 mm dark nodule, unilateral, near the cranial pole was seen in one female rat from the intermediate group in a 7-day study. No masses in the regional lymph nodes or in other districts were seen in any of these cases.

Microscopic examination was performed on both kidneys from 1206 rats (688 males and 518 females), control and treated animals.

Proliferative lesions, diagnosed histologically as renal tubular carcinoma (RTC), renal tubular adenoma (RTA) and renal atypical tubular hyperplasia (RATH), were identified in 8 animals from 6 different studies.

Three cases of RTC were characterized by large multinodular masses composed of cords and nests of epithelial cells with solid to cystic pattern of growth. The neoplastic cells were eosinophilic to basophilic, often vacuolated, and with hypertrophic nucleoli. Presence of areas of necrosis, cellular pleomorphism and a few mitoses were seen (Figures 2–4). The affected rats were approximately 16 weeks of age, and all three lesions corresponded to the findings noted macroscopically for each animal (Table 1).

One case of RTA was a small cystic mass, composed of single to multilayered cuboidal cells lining a cystic lumen and few nodular solid proliferation of epithelial cells with an eosinophilic glassy to vacuolated cytoplasm, proliferating beyond the confines of the tubule and without evident necrosis (Figure 5). It was observed in a control female rat, of around 12 weeks of age.

Four cases of RATH were focal to multifocal, characterised by enlarged to occasionally cystic tubules, lined with large eosinophilic, cuboidal to columnar cells, rarely multilayered (Figure 6); the animals were ranging from 13 to 14 weeks of age. In one instance the microscopic finding was accountable for the macroscopic observation (Table 1).

The occurrence of these lesions was not associated with morbidity or mortality of the animals, since all of them survived until the end of the study, and no indication of functional impairment was observed. In the 4-week study with the 2 tumour-bearing rats, 1 from the high-dose and 1 from the intermediate-dose groups, there was evidence of nephrotoxicity related to the administration of the test compound in the high dose group. However, in these 2 animals, the RTC were not associated with other preneoplastic or neoplastic changes. The absence of any other preneoplastic or neoplastic lesions was further confirmed in all the other animals with the step sectioning approach that allows an extensive and rigorous evaluation of the whole organ (e.g., 8–10 sections from each kidney).

Altogether these 8 cases were recognized from the 1206 rats investigated macroscopically and microscopically, representing 0.66% of cases over a 2-year period. In particular, 3 out of 688 (0.44%) were male, and 5 out of 518 (0.97%) were female. These lesions appeared distributed between the 2nd quarter of 2004 and the 1st quarter of 2006. In details, one case of RTA was observed in the 2nd quarter of 2004, 2 cases of RTC and 1 case of RATH in the 2nd quarter of 2005, one case of RTC in the 3rd quarter of 2005 and 3 cases of RATH in the 1st quarter of 2006.

Discussion

This paper shows the early occurrence (≤18 weeks of age) of preneoplastic and neoplastic lesions in Sprague–Dawley rats, as observed in our facilities. The occurrence of renal tumours, in preclinical studies shorter than 90-day (Hard et al., 1994), is an extraordinary finding and, to our knowledge, unprecedented in safety evaluation.

Spontaneous renal tumours in the rat are found in aging animals with low incidence (Thurman et al., 1995; Chandra et al., 1993, Hiasa and Ito, 1987). The incidence of renal spontaneous neoplasms is about 1% in Fisher 344 rats (Thurman et al., 1995) and at 0.08% in Sprague–Dawley rats (Chandra et al., 1993). However a definitive indication of a firm estimate for renal tubular carcinoma is still undetermined (Hard, 1986; Greaves and Faccini, 1992; Lock and Hard, 2004).

Furthermore, in the context of carcinogenicity studies an increased incidence of renal tumours after 2 years of treatment has been observed after chronic administration of several compound, acting directly damaging the DNA, or known to cause indirect cytotoxicity with sustained tubular regeneration, or even exacerbating chronic progressive nephropathy (Lock and Hard, 2004). However, even in the case of the most potent chemical tested, the mycotoxin ochratoxin A, preneoplastic lesions had been observed as early as 13 weeks of age, but the first neoplastic lesions were present only after 9 months of treatment, or even later (Lock and Hard, 2004; Mantle et al., 2005).

In our facilities, in the time frame of 2 years, in 52 general toxicity studies including 1206 Sprague–Dawley rats examined microscopically, the overall incidence of preneoplastic and neoplastic lesions in the kidney was of 0.66%, with females appearing more affected than males (0.97% or 0.44%, respectively). Neoplastic lesions (RTA and the RTCs) were seen in the 0.33% of the animals.

In one 4-week oral toxicity study, where findings consistent with nephrotoxicity (e.g., tubular degeneration and regeneration) were present at the high dose, 2 RTC were observed in 1/10 high dose rats and in 1/10 intermediate dose rats. To exclude a relationship with the test compound the following considerations were made: 1) the absence of preneoplastic or neoplastic lesions in other animals of the same groups; indeed, preneoplastic lesions are always expected to be present in several rats treated with renal carcinogens (Lock and Hard, 2004); 2) in the tumor-bearing animals the tumour was not associated with any preneoplastic or other neoplastic changes in the organ; 3) the absence of nephrotoxicity in one of the two tumour-bearing animals; 4) the short length of the study; as already stated before, the most potent renal carcinogens, the ochratoxin A, can produce first neoplastic lesions only after 9 months of treatment, or even later (Lock and Hard, 2004; Mantle et al., 2005); 5) negative results in the genotoxicity assays (high-throughput fluctuation test, Mouse lymphoma test, AMES test, in vivo micronucleus test).

Confirming these interpretations, 1 rat from a concurrent 4-week study with an unrelated test-article showed RTC, and concurrently the sporadic occurrence of similar finding (i.e., renal tubular adenoma) was reported in a study from other laboratories using CD rats from the same supplier (Savard et al., 2005).

To add evidence to exclude a relationship with the treatment, multiple sections of the kidney of all the animals were examined, according to the indication of Eustis et al. (1994). RATH is, indeed, considered a precursor lesion in chemically induced renal tubular neoplasia (Hard et al.,1994; Khan and Alden, 2002; Hino, et al. 2006), which requires extensive sampling in order to be assessed adequately (Lock and Hard, 2004).

To support the spontaneous origin of these tumors in this strain of rats, there was the observation single instance of RTC, RTA, or RATH in different studies in our laboratory. These lesions were variably distributed throughout vehicle or test-article treated groups. Moreover animals were treated with chemically and pharmacologically unrelated test articles.

All the animals were from the same supplier (Charles River Italia, Calco, Italy). All 3 RTC-bearing animals, were examined between the second and third quarter of the 2005, were 15–16 weeks old and came from the same supplier breeding room. The close spatial and temporal distribution of these cases suggested a familiarity among these animals, and the hypothesis of an inherited genetic basis has been taken into account.

Hereditary tumors occurring spontaneously in the Eker rat and the Nihon rat models are well known and already characterized morphologically and genetically (Hino et al., 2006). The Eker rat model is an example of a dominantly inherited predisposition for RTC and is caused by a mutation in a tumour suppressor gene (tsc2) encoding tuberin (Yeung et al., 1994). 100% of heterozygous rats will develop renal neoplasms by 12 months of age while homozygosity of mutant allele causes embryonic lethality by 13 days of gestation (McDorman and Wolf, 2002).

Renal lesions in this model are characterised as simple cysts, papillary cystadenomas, solid eosinophilic adenomas or solid basophilic adenomas (McDorman and Wolf, 2002). The cells are large, pleomorphic, most often vacuolated and eosinophilc, with prominent nucleoli (Hard, 1994). The Nihon rat model is another example of a Mendelian dominantly inheritated RTC found in a Sprague–Dawley rat strain in Japan (Okimoto et al., 2000). It is a model of a dominantly inherited single gene mutation in the Birt-Hogg-Dubè (BHD) gene.

Preneoplastic lesions occur as early as 3 weeks of age, developing into adenocarcinomas by 6 months (Okimoto et al., 2004). The morphological appearance of the neoplasms in these animals, is composed of 2 populations of tumor cells: one was composed of large clear (PAS positive) cells with hyperchromatic nuclei and abundant clear cytoplasm variably admixed with acidophilic cells arranged in tubular or solid pattern; the other population characterised by basophilic cells with pale basophilic nuclei and a papillary pattern of growth (Okimoto et al., 2004).

In our cases, the morphological features of the RTC strikingly resembled hereditary tumours occurring spontaneously in the Eker and Nihon rat models. In conclusion, to the best of our knowledge, this is the first report of renal tubular carcinomas in rats younger than 18 weeks. The occurrence of preneoplastic and neoplastic lesions in the kidneys of rats younger than 18 weeks is a highly unusual finding. Moreover, due to the low number of animals used in short term toxicity studies, sporadic incidence of these lesions can mimic dose-response relationships, which in turn renders the findings difficult to be assessed properly.

The absence of foci of atypical tubule hyperplasia in affected animals, as well as in any other rats from the same treatment groups, should definitively help in determining the spontaneous nature of the lesions. The need for tumour incidence data from short-term rodent toxicity studies is highlighted since there is very little information available on tumor incidence in young rats. The development of background data on this issue would definitively facilitate the assessment of the spontaneous nature of these lesions in this strain of rats.

Renal tumours encountered in these studies are considered a recently occurring spontaneous finding in our rat strain. The similarities with the familial renal cancer models strongly suggest that genetic factors are responsible for these lesions, which require a further characterisation of the genetic background.