ToxSci Advance Access originally published online on April 10, 2007
Toxicological Sciences 2007 98(1):258-270; doi:10.1093/toxsci/kfm083
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Rodent Carcinogenicity Profile of the Antidiabetic Dual PPAR 
and 
Agonist Muraglitazar
* Bristol-Myers Squibb Pharmaceutical Research Institute, Department of Drug Safety Evaluation Mt. Vernon, Indiana 47721
Covance Laboratories, Vienna, Virginia 22182
Department of Pathology and Microbiology, University of Nebraska Medical Center, Omaha, Nebraska 68198
1 To whom all correspondence should be addressed at Department of Drug Safety Evaluation, Bristol-Myers Squibb Pharmaceutical Research Institute, 2400 West Lloyd Expressway, P3, Evansville, IN 47721-0001. Fax: (812) 429-8469. E-mail: sarah.tannehill-gregg{at}bms.com.
Received November 30, 2006; accepted March 23, 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
The carcinogenic potential of muraglitazar, a dual human peroxisome proliferator–activated receptor
/
agonist, was evaluated in 2-year studies in mice (1, 5, 20, and 40 mg/kg) and rats (1, 5, 30, and 50 mg/kg). Benign gallbladder adenomas occurred at low incidences in male mice at 20 and 40 mg/kg (area under the curve [AUC] exposures 
62 times human exposure at 5 mg/day) and were considered drug related due to an increased incidence of gallbladder mucosal hyperplasia at these doses. There was a dose-related increased incidence of transitional cell papilloma and carcinoma of the urinary bladder in male rats at 5, 30, and 50 mg/kg (AUC exposures 8 times human exposure at 5 mg/day). At 30 and 50 mg/kg, the urinary bladder tumors were accompanied by evidence of increased urine solids. Subsequent investigative studies established that the urinary bladder carcinogenic effect was mediated by urolithiasis rather than a direct pharmacologic effect on urothelium. Incidences of subcutaneous liposarcoma in male rats and subcutaneous lipoma in female rats were increased at 50 mg/kg (AUC exposures 48 times human exposure at 5 mg/day) and attributed, in part, to persistent pharmacologic stimulation of preadipocytes. Toxicologically relevant nonneoplastic changes in target tissues included thinning of cortical bone in mice and hyperplastic and metaplastic adipocyte changes in mice and rats. Considering that muraglitazar is nongenotoxic, the observed tumorigenic effects in mice and rats have no established clinical relevance since they occurred at either clinically nonrelevant exposures (gallbladder and adipose tumors) or by a species-specific mechanism (urinary bladder tumors). Key Words: muraglitazar; rodent carcinogenicity; peroxisome proliferator–activated receptor (PPAR) agonist; gallbladder adenoma; urinary bladder carcinogenesis; urolithiasis; liposarcoma; lipoma.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Peroxisome proliferator–activated receptors (PPARs) are ligand-activated transcription factors which are members of the nuclear hormone receptor superfamily (Rosen et al., 2000
). Three receptor types belonging to this family have been identified including alpha (), gamma (), and delta (
)—also called beta. PPARs are potent insulin sensitizers; PPAR (expressed mainly in the liver, heart, and skeletal muscle) is involved in the metabolism of lipoproteins and fatty acids, while PPAR (expressed mainly in adipose tissue) is involved in the differentiation of fat cells, as well as the metabolism of fatty acids and glucose (Yki-Järvinen, 2004).
Many agonists of these receptors have been developed for use as therapeutic agents in the treatment of type 2 diabetes, including the marketed PPAR agonists rosiglitazone and pioglitazone. Diabetes is a chronic, debilitating disease that affects at least 4% of the adult population worldwide (most of whom have type 2 diabetes) (King et al., 1998). The incidence of type 2 diabetes is increasing rapidly in industrialized nations, and it is estimated that there will be 221 million diabetics worldwide by the year 2010 (King et al., 1998). Moreover, since the percentage of type 2 diabetic patients that achieve glycemic control with oral antidiabetic treatments is estimated to be less than 50%, there remains a significant unmet medical need for better control of hyperglycemia in diabetic patients (Koro et al., 2004). Control of the dyslipidemia which is commonly associated with diabetes (characterized by elevated levels of triglyceride and decreased levels of high-density lipoprotein cholesterol) is another attractive medical target for treatment in diabetics (Cox, 2005). Muraglitazar is a nonthiazolidinedione, oxybenzylglycine dual PPAR / agonist which was previously in development for the treatment of type 2 diabetes and associated dyslipidemia. As a dual PPAR agonist, muraglitazar combines the insulin sensitizing and glucose lowering effects of PPAR agonism with the antidyslipidemic effects of PPAR agonism.
PPARs have a wide tissue distribution, and 2-year mouse and rat carcinogenicity bioassays for 12 PPAR agonists (six and six dual / agonists, including muraglitazar) reviewed by the United States Food and Drug Administration (U.S. FDA) have demonstrated that the tumors types which develop in mice and rats are consistent with this distribution (El Hage, 2005). The most commonly occurring tumor types induced by PPAR dual / agonists include urinary bladder and adipose (lipoma and/or liposarcoma) tumors in rats and vascular tumors (hemangiosarcoma) in mice. PPAR agonists most commonly induced hemangiosarcomas in mice and adipose tissue (lipoma and/or liposarcoma) tumors in rats. It is interesting that those tumors typically associated with the fibrate PPAR agonists (liver, pancreatic, and/or testicular tumors) (Klaunig et al., 2003) have not generally been observed with the dual / agonists.
Interestingly, PPAR dual / agonist–related urinary bladder tumors occur with a somewhat greater frequency in male than in female rats and not in mice. Urinary bladder tumors have previously been described with the marketed PPAR agonist, pioglitazone, and a host of chemicals including sodium saccharin and other sodium salts when administered at high doses (Cohen, 1995; PDR Electronic Library, 2006a). With sodium saccharin, the rat appears to be the only sensitive species—mice, hamsters, guinea pigs, and monkeys fail to demonstrate evidence of a proliferative response (Cohen, 1995). Considerable evidence suggests an indirect mechanism of bladder tumorigenesis in male rats treated with drugs or chemicals that increase formation of urinary solids (Cohen, 2005), which may cause mucosal injury and consequent regenerative hyperplasia, eventually leading to tumor formation.
Since muraglitazar was intended for chronic administration to human diabetics, it was evaluated in conventional 2-year carcinogenicity studies in mice and rats. The outcome of those studies and a preclinical assessment of the carcinogenic potential of muraglitazar are presented here.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Test substance.
Muraglitazar is described chemically as N-[(4-methoxyphenoxy)carbonyl]-N-[[4-[2-(5-methyl-2-phenyl-4-oxazolyl)ethoxy]phenyl]methyl]glycine (Fig. 1). Over the concentration range used, muraglitazar was stable in the vehicle, 96% polyethylene glycol 400 containing 4% 1M sodium hydroxide in aqueous solution (alkaline PEG-400 solution), for 14 days when stored at 2°C–8°C, protected from light. Analysis of dosing preparations at 13-week intervals during the studies confirmed that intended concentrations of the test article were achieved, with dosing formulations ranging from 90% to 108.1% of target. Bulk drug integrity and stability were maintained over the period of use.
|
Animals.
Random-bred, barrier-raised Crl:CD-1®(ICR)BR mice and Hsd:Sprague Dawley®SD® rats were obtained from Charles River Laboratories, Raleigh, NC and Harlan Laboratories, Frederick, MD, respectively. Animals were acclimated for a 1- to 3-week period prior to study initiation. Mice were 8 weeks old and rats were 5–8 weeks old at study initiation. Animals were housed individually in suspended, stainless-steel cages in environmentally controlled rooms with a 12-h light/12-h dark cycle, a humidity range of 30–70%, and a temperature range of 18°C–26°C. Tap water and certified rodent diet (#8728C, Harlan Teklad) were available ad libitum.
Experimental design.
Studies were conducted at Covance Laboratories Inc., Vienna, VA, and were compliant with U.S. FDA Good Laboratory Practice regulations. Further, animal use was in accordance with the Guide for the Care and Use of Laboratory Animals.
Muraglitazar was administered to 60 mice per sex per group by oral gavage once daily for up to 105 weeks. There were two studies in mice—doses of 1, 5, or 20 mg/kg were evaluated in the first (initial) and 40 mg/kg in the second (supplemental). The first study contained two vehicle control groups, whereas one vehicle control group was included in the second study. All control groups contained 60 mice per sex per group and were administered alkaline PEG-400 solution. In addition, 19 mice per sex per group were designated for toxicokinetic analysis for each study and received muraglitazar by oral gavage once daily at 1, 5, 20, or 40 mg/kg for 26 weeks. In the carcinogenicity study in rats, 65 rats per sex per group were given muraglitazar by oral gavage once daily at 1, 5, 30, or 50 mg/kg for up to 105 weeks. Two control groups (65 rats per sex per group) were given the same vehicle used in mice, alkaline PEG-400 solution. The dose volume for all studies was 5 ml/kg. In the rat carcinogenicity study, composite toxicokinetic analyses were conducted on blood samples collected from study animals rather than from separate satellite groups.
Animals were observed daily for clinical signs of toxicity. Individual body weights and food consumption were recorded weekly for weeks 1–13 or 14, monthly thereafter, and at study termination.
In all studies, plasma concentrations of muraglitazar were determined from samples collected during week 26 from three animals per sex per group each at 0.5-, 2-, 4-, 6-, 8- and 24-h postdose. In the initial mouse study, miscalculated collection times resulted in samples being collected from two rather than three mice during the 4-, 8-, and 24-h timepoints. All samples were from nonfasted animals. Dipotassium ethylenediaminetetraaceticacid was used as the anticoagulant, and samples were stored frozen at –10°C to –30°C prior to analysis.
All animals received a complete post-mortem examination, and protocol-specified tissues were collected and processed for microscopic evaluation. Due to increased incidences of urinary bladder tumors in decedent male rats at 30 and 50 mg/kg, the dorsal and ventral surfaces of the urinary bladder in males were marked for reference before removal beginning in week 92. Additionally, the urinary bladders were opened and examined for the presence of calculi and masses, and the location of any masses was described (dorsal vs. ventral, neck vs. apex) when possible.
Urinalyses.
Because of the increased incidences of hematuria and urinary bladder tumors in decedent male rats at 30 and 50 mg/kg, urinalyses were conducted in rats late in the study. Overnight pan-sample collections were performed in week 86 (18 h) and week 87 (8 h) in the first 25 surviving rats in the first control group and in all 30 and 50 mg/kg males. Samples were collected on wet ice from food-fasted animals, and 0.1 ml of 1% sodium azide was added to each tube prior to collection. Tests included appearance/color, volume, specific gravity, pH, total protein, urobilinogen, bilirubin, blood, glucose, ketones, and microscopic examination of sediment. During week 87, urine sodium and calcium levels were also measured. During weeks 86 and 87, urine sediment smears were prepared and stained with Wright–Giemsa and a cytological evaluation was performed.
Fresh-void urine samples were collected from 10 male and 10 female nonfasted rats in the first control group and all treated groups during week 90. Samples were collected from approximately 0700–0900 h. Urinary pH was immediately measured using a microelectrode. After centrifuging the samples and collecting the sediment on a vacuum-dried filter, the supernatant was evaluated for creatinine, calcium (nonacidified), magnesium, phosphorus, total protein, urea nitrogen, sodium, potassium, and chloride. Urine sediment smears collected from males in the control and 30 and 50 mg/kg groups were stained with Alcian Blue to qualitatively evaluate mucopolysaccharide-containing material. Additionally, filtered sediment was examined by scanning electron microscopy (SEM) and energy dispersive x-ray probe microanalysis for crystals, calculi, and/or aggregates.
Histopathology.
Tissues from all animals were evaluated microscopically by study pathologists at Covance Laboratories using established diagnostic criteria. All drug-related nonneoplastic findings and tumors were peer reviewed.
Subcutaneous neoplasms from six rats were stained with Oil-Red-O (a stain for neutral lipids) and examined microscopically to help define that adipocytes (fat cells) were the cell of origin. In addition, immunostaining for S-100 protein was performed on subcutaneous mesenchymal neoplasms from 26 rats, including the liposarcomas from 18 control and treated animals, to determine whether some poorly differentiated neoplasms were derived from adipocytes (S-100 positive) or fibroblasts (S-100 negative) (Cocchia et al., 1983; Hashimoto et al., 1984). Using a Ventana Benchmark automated immunostainer (Ventana Medical Systems Inc., Tucson, AZ), 4-µm sections were stained with primary (S-100) or negative control (normal rabbit IgG) antibodies purchased prepacked/prediluted from Ventana in a ready-to-use staining dispenser. The secondary antibody was biotinylated goat antirabbit IgG antibody (Vector Laboratories, Burlingame, CA) diluted 1:200, and the immunoreaction was performed by using activated 3–3' diaminobenzadine-tetrahydrochloride dihydrate. At least one positive control slide (rat peripheral nerve) and one negative control slide (rat peripheral nerve stained with normal rabbit IgG antibodies) were stained per run. Immunostained sections were evaluated by light microscopy.
Statistical analyses.
The two-sided Dunnett's t-test was used for pairwise comparisons for body weights, body weight changes, and food consumption. Levene's test was performed to test for homogeneity of variance. When heterogeneity of variance was p 
0.05, then log transformations were used to stabilize the variance. Comparison tests took variance heterogeneity into consideration.
Mortality data were evaluated for a dose-related trend using the Cox–Tarone test (two sided). For all tests, differences with p 0.05 were considered significant. Survival data were analyzed using the National Cancer Institute (NCI) Life Table Package (Thomas et al., 1977). The methods consisted of Kaplan–Meier product limit estimates and Cox–Tarone binary regression life tables. Week 105 was treated as the end of the study in the NCI package for both males and females. Two-sided tail probabilities for trend and group comparisons were evaluated at the 5.0% significance level. When two vehicle control groups were present, each vehicle control group was compared to each other, and the treated groups were compared to the combined vehicle control groups.
Neoplastic lesions were chosen for statistical analyses if the incidence in at least one treated group for each sex was increased by at least two occurrences over that of the control group(s). The occult tumors (incidental alone or incidental and fatal combined) were analyzed by a modified asymptomatic fixed interval-based prevalence test (Peto et al., 1980). The cut-off points for the interval-based test were weeks 0–52, 53–78, 79–92, 93 to before scheduled terminal sacrifice, and scheduled terminal sacrifice (week 105). Both the incidental and the fatal tumors of the same types were analyzed over the fixed intervals. At each of the fixed intervals, the number of animals at risk for each group for fatal tumors was taken to be the number of animals surviving at the end of the previous interval. The test was implemented using PROC MULTTEST in the SAS system (release 8.0; SAS Institute, Inc., Cary, NC). The palpable tumors were analyzed by life table techniques (Thomas et al., 1977) using the first palpation time as onset time, and assuming that the number of animals at risk was the number of animals examined macroscopically. Benign and malignant lesions were evaluated individually and combined, when appropriate. Combined evaluation followed the recommendations of McConnell et al. (1986). Using nominal dose levels, all groups were compared to the group 1 and group 2 controls combined, when two control groups were present. One-sided positive trend was evaluated at 0.005 for a common tumor (incidence rate 1%) or 0.025 for a rare tumor (incidence rate < 1%) based on data from recent historical controls, as indicated in the FDA Draft Guidance for Industry (2001). In cases where the first analysis showed a significant positive trend under the common or rare denomination, subsequent analyses were conducted by deleting the data from the highest dose group(s) and then repeating the one-sided test for trend on the reduced data set. The procedure was continued until the remaining groups did not show any positive trend.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Mouse Carcinogenicity Studies
Clinical evaluation.
There were no direct drug-related clinical signs in either sex at any dose level. Throughout most of the initial study, mean body weights of the 20 mg/kg females were slightly increased (frequently statistically significantly) compared with those of control group 1, an effect considered to be drug related. At week 105, the mean body weight of the 20 mg/kg females was 105% of the control group 1 mean. There were no significant effects on body weight in males at 1, 5, or 20 mg/kg. Statistically significant, drug-related increases in mean body weights for males and females were noted throughout the supplemental study at 40 mg/kg. Mean body weights were approximately 111% (males) and 108% (females) that of respective controls at week 105. At 40 mg/kg, mean food consumption was generally higher (often statistically significantly) for both males and females when compared to controls. Mean total food consumption values for weeks 1–53 were 104% (males) and 105% (females) that of respective controls, and for weeks 1–104 were 101% (males) and 106% (females) that of respective controls. There were no significant effects on mean food consumption at 1, 5, and 20 mg/kg (body weight and food consumption data not shown).
Toxicokinetics.
Muraglitazar AUC(0–24 h) exposures on day 177 (week 26) were generally dose proportional with slightly greater exposures in females compared to males at all dose levels. The ratios of mean AUC values for muraglitazar (male/female) were 0.70, 0.66, 0.71, and 0.92 at 1, 5, 20, and 40 mg/kg, respectively. The doses evaluated were associated with muraglitazar plasma AUC(0–24 h) values of approximately 3–154 times the mean human AUC at 5 mg/day (Fig. 2).
|
Survival.
Mortality rates were statistically significantly increased in males at 40 mg/kg (mainly during the final 6 months of the study), and in females at 20 mg/kg (mainly during the final year of the study) (Supplementary Data 1). In the initial study, overall survival at 105 weeks was 42%, 35%, 38%, 39%, and 36% in males and 43%, 35%, 36%, 40%, and 18% in females at 0 (control 1), 0 (control 2), 1, 5, and 20 mg/kg, respectively. In the supplemental study, overall survival at 105 weeks was 48% and 14% in males, and 30% and 25% in females in the control and 40 mg/kg groups, respectively. The cause of increased mortality in the 20 mg/kg females was not determined, whereas a treatment-related increased incidence of atrial thrombosis was the cause of death in 10 males and two females at 40 mg/kg. The most common nonneoplastic processes causing death were ovarian hematocyst, suppurative inflammation/abscess, dermal ulceration, and amyloidosis, which were noted without relationship to treatment. The most commonly observed neoplasms as the cause of death were lymphoma, fibrosarcoma, histiocytic sarcoma, bronchiolar–alveolar carcinoma, and hemangiosarcoma. There was no evidence of an increased incidence or an earlier onset of these commonly occurring neoplasms in treated mice when compared to control mice.
Nonneoplastic pathology.
Drug-related macroscopic findings were noted at 40 mg/kg and included increased incidences of enlarged heart and enlarged and mottled atria in males and enlarged spleens in females. The macroscopic cardiac findings in the 40 mg/kg males correlated with degenerative cardiomyopathy, which microscopically was present with increased incidence and severity in both males and females (Table 1). The enlarged and mottled atria seen macroscopically in the 40 mg/kg males correlated histologically to an increased incidence of atrial dilatation and thrombosis (Table 1). The increased incidences of both degenerative cardiomyopathy and atrial thrombosis were likely an exacerbation of these common spontaneous age-related changes due to drug-related plasma volume expansion. The enlarged spleens noted macroscopically in 40 mg/kg females correlated to an increased incidence of extramedullary hematopoiesis in this group. This likely represented slight exacerbation of a common background finding in normal female mice (Suttie, 2006); however, the basis for this exacerbation was unclear. An increased incidence of dark mesenteric lymph nodes was noted macroscopically in females at 1, 5, and 20 mg/kg; however, there was no histopathologic correlate for this finding. Additionally, the incidence of this change in the 1 and 5 mg/kg female groups was comparable to that seen in the control males.
|
Drug-related nonneoplastic microscopic findings included minimal to severe, focal hyperplasia of the gallbladder mucosal epithelium in males at all doses and in females at 5, 20, and 40 mg/kg. In the liver, there was minimal to moderate, centrilobular to midzonal (20 mg/kg) or diffuse (40 mg/kg) hepatocellular hypertrophy in both sexes, with an associated minimal to moderate amount of greenish-brown pigment (consistent with lipofuscin) in centrilobular to midzonal hepatocytes at 40 mg/kg only. The hepatocellular hypertrophy was considered to be an adaptive change associated with induction of hepatic drug metabolizing enzyme activity and likely peroxisomal proliferation. The increased hepatic lipofuscin pigment was considered reflective of increased cell membrane/organelle turnover secondary to lipid peroxidation, and was likely an exacerbation of a spontaneous age-related liver change since it was also seen to a lesser extent in control mice (Irisarri and Hollander, 1994
; Kumar et al., 2004).
Dose-related, pharmacologically mediated alterations in adipose tissue included increased incidences of minimal to moderately severe macrovesiculation of brown adipose tissue in both sexes at all doses, minimal to severe microvesiculation of white adipocytes in both sexes at 5 mg/kg, increased white adipocytes in the bone marrow of the femur and sternum in males at all doses and females at 5, 20, and 40 mg/kg (Table 1). Cortical bone defects [thinning] were often associated with the increased marrow adipocytes, notably in females at 20 mg/kg and in both males and females at 40 mg/kg. When severe, the microvesiculation of white adipose tissue at 40 mg/kg resulted in the presence of very small residual cells with little resemblance to normal white adipose tissue. Other microscopic findings occurred at similar incidences in treated and control groups and were considered to be typical of spontaneous lesions in this strain of mouse and in a study of this duration.
Tumor analysis.
An overall summary of tumor incidence in mice is presented in Tables 2 and 3 (see Supplementary Data 2 and 3 for a complete list of neoplastic lesions in mice). There were no statistically significant positive trends (p 0.005 for a common neoplasm, or p 0.025 for a rare neoplasm) in the incidences of neoplasms in muraglitazar-treated males and females as compared to the two concurrent control groups combined. However, single benign gallbladder adenomas observed in one male mouse at 20 mg/kg and two male mice at 40 mg/kg were considered to be drug related because of the concurrent increased incidence in focal gallbladder mucosal hyperplasia.
|
|
Other neoplasms observed in this study were those commonly observed in CD-1 mice, and there was no evidence of an earlier onset of these neoplasms in treated mice. The incidence of hemangiosarcoma (multiple organs) appeared to be slightly higher in female mice administered 20 mg/kg (5/60; 8.3%) and 40 mg/kg (3/60; 5.0%), and the combined incidences of hemangioma and hemangiosarcoma in multiple organs in female mice appeared to be minimally increased at 5 and 20 mg/kg (5/60; 8.3% in each group), when compared to concurrent controls. However, these increases were not statistically significant and were not accompanied by preneoplastic changes or increased hemangiomas.
The incidences of hepatocellular carcinoma in male mice administered 20 mg/kg (4/60 animals; 6.7%) and 40 mg/kg (3/60 animals; 5.0%) were slightly higher than those of concurrent control(s); however, this was considered to be incidental to treatment with muraglitazar, as there was no associated statistically significant positive trend tests and the incidences were well within the historical incidences in control male CD-1 mice at Covance Laboratories, Inc., Vienna, Virginia (38/430; 8.84%). The incidence of malignant lymphoma was slightly higher in male mice administered 40 mg/kg (6/60; 10.0%) compared to the concurrent control group; however, this was considered unrelated to treatment as the incidence was less than that seen in one of the control groups from the initial study (8/60; 13.3%) which was conducted at approximately the same time and in the same testing facility. The incidence of mammary gland carcinoma in 40 mg/kg females (4/59, 6.78%) was higher than that in control females; however, the difference was not statistically significant at the p 0.005 level for a common neoplasm and the incidence fell within the historical range (0.78–8.33%) for this tumor type in CD-1 mice (Charles River Laboratories; Raleigh, NC).
Rat Carcinogenicity Study
Clinical evaluation.
Drug-related clinical signs were limited to an increased incidence of red genital discharge in males at 30 and 50 mg/kg (likely related to urinary tract lesions detailed below), swollen ventral abdominal area in females at 30 and 50 mg/kg, and swollen axillary region in males at 50 mg/kg. Throughout most of the study, mean body weights of the 30 and 50 mg/kg males were slightly increased (frequently statistically significant) compared with those of control group 1, with mean body weights at these doses 111% and 106%, respectively, at week 105. Treatment did not significantly effect body weight gain in females; however, mean food consumption was slightly increased in both males and females at 30 and 50 mg/kg during the first 6 months of treatment. At week 21, mean food consumption of the 30 and 50 mg/kg groups was 109% and 113%, respectively, of control group 1 in males, and 109% and 109%, respectively, of control group 1 in females.
Toxicokinetics.
Systemic AUC exposures to muraglitazar in rats on day 177 were dose related and generally increased dose proportionally up to 30 mg/kg, and less than proportional to the dose increment at 50 mg/kg. The doses evaluated were associated with muraglitazar plasma AUC(0–24 h) values of approximately 1–59 times the mean human AUC at 5 mg/day (Fig. 3). There were no substantial gender-related differences. The ratio of mean AUC values for muraglitazar (male/female) was 0.70, 1.13, 0.81, and 0.81 for 1, 5, 30, and 50 mg/kg, respectively.
|
Survival.
Survival of the males ranged from 77% to 88% after 1 year (week 53). An increase in mortality for the 50 mg/kg males was noted during the second year of treatment, with survival of the males at this dose level reduced to 8% at study termination. Although not statistically significant, mortality was slightly increased in the 30 mg/kg males (16% survival at study termination). In contrast, the 5 mg/kg group showed a decreased mortality (52% survival at study termination) compared to the combined controls. The two male control groups were approximately equivalent in survival rates (27% and 23%, respectively). Survival of the females ranged from 86% to 94% after 1 year and from 26% to 41% at study termination with no significant positive trend or increase in mortality in the treated female groups. In addition, the two female control groups were similar in their survival rates (Supplementary Data 4).
The increased mortality in the 30 and 50 mg/kg males was attributed to an increased incidence of malignant epithelial neoplasms of the urinary bladder. The deaths of 14 of 55 (25.4%) and 23 of 60 (38.3%) males at 30 and 50 mg/kg, respectively, were attributed to neoplasms of the urinary bladder. The cause of death in male rats in other groups, when it could be determined, was varied and appeared unrelated to treatment. Nephropathy was the most common nonneoplastic cause of death in males, with the highest incidence in control and 1 mg/kg groups, as these survived for a longer period than the males given higher doses of muraglitazar. Mammary neoplasia was the most common cause of moribundity/death in females and occurred with the greatest frequency in the control, 1 and 5 mg/kg groups. Nonneoplastic uterine pathology (hyperplasia and squamous metaplasia of the uterine mucosa, inflammation, hemorrhage, thrombosis, dilatation, and necrosis), which occurred without relationship to treatment, was the second most common cause of death in females, followed by deaths associated with uterine neoplasms.
Urinalysis.
Relevant findings in overnight pan-collected urine from 30 and 50 mg/kg males at weeks 86 and/or 87 included increases in the incidence and/or severity of red and white blood cells (findings consistent with hematuria and/or mild inflammation of the urinary tract), as well as in the number and/or incidence of urine crystals (triple phosphate [magnesium ammonium phosphate] and birefringent amorphous [calcium phosphate], most commonly occurring as single crystals). Transitional cells present in the urine sediment did not appear atypical or dysplastic for any animal, but cell preservation was not optimum. Minor increases in urine pH (week 86 only) and urine sodium concentration, and decreases in urine soluble calcium were also noted (data not shown).
Drug-related changes in freshly voided urine collected at week 90 from a cohort of rats (10 rats per group) at all doses included a trend of increased pH in males and females at 5 mg/kg (with pH slightly higher in males than females) and dose-dependent decreases in protein in males at all doses with mean protein/creatinine ratios ranging from 79% of the control mean at 1 mg/kg to 14% of the control mean at 50 mg/kg (data not shown). Since there was only a limited number of animals evaluated with sufficient urine volume for electrolyte analyses, no firm conclusions could be drawn. Analysis of urine sediment by SEM and energy dispersive x-ray probe microanalysis revealed increased incidences and larger aggregates of magnesium ammonium phosphate (triple phosphate/struvite) crystals and the presence of jagged calcium phosphate-containing crystals in males at 30 and 50 mg/kg, calcium phosphate-containing calculi in 3 of 10 males at 50 mg/kg, and a single jagged calcium phosphate-containing crystal in 2 of 10 females at 50 mg/kg (Fig. 4). Additionally, urine sediment smears showed no evidence of a treatment-related effect on the amount of mucopolysaccharide-containing material.
|
Nonneoplastic pathology.
Treatment-related nonneoplastic changes were observed in the urinary bladder, glandular stomach, brown and white adipose tissues, bone marrow, and adrenal cortex (Table 4). Increased incidences of focal or diffuse hyperplasia of the urinary bladder epithelium were noted in males at 5, 30, and 50 mg/kg and females at 30 and 50 mg/kg. Generally dose-related adipocyte alterations in male and female rats that were considered pharmacologically mediated included increased white adipocytes in the muscular tunic of the glandular stomach at all doses and increased macrovesiculation of brown adipose tissue, microvesiculation of white adipose tissue, and increased white adipocytes in sternal bone marrow at 5, 30, and 50 mg/kg.
|
In the adrenal cortex, a dose-dependent increased incidence of vacuolization of cells in the zona glomerulosa was observed in males at 5, 30, and 50 mg/kg and in females at 30 and 50 mg/kg. In addition, slight hypertrophy of the zona glomerulosa was evident in a few males at 5, 30, and 50 mg/kg. These adrenal gland findings were considered a muraglitazar-related exacerbation of spontaneous age-related changes.
Tumor analysis.
An overall summary of tumor incidence in rats is presented in Tables 5 and 6 (see Supplementary Data 5 for a complete list of neoplastic lesions in rats). An increased incidence of masses in the urinary bladder was evident in males given 5, 30, or 50 mg/kg of muraglitazar. These masses were noted to be principally located in the mucosa of the ventral bladder near the apex, and correlated with a dose-dependent increased incidence of transitional cell carcinoma of the urinary bladder observed microscopically at these doses (Table 5). Additionally, benign transitional cell papilloma was observed at a low incidence in muraglitazar-treated males. The incidences of transitional cell carcinoma and/or transitional cell papilloma and carcinoma in treated male rats also showed a statistically significant positive trend at p 0.025 for a rare tumor (incidence 1%) at 5, 30, and 50 mg/kg. Transitional cell neoplasms of the urinary bladder were not observed in female control or treated rats.
|
|
An increased incidence of subcutaneous malignant liposarcoma (confirmed by S-100 immunostaining—Cocchia et al., 1983
; Hashimoto et al., 1984—with select tumors additionally confirmed by Oil-Red-O staining) was observed in the 50 mg/kg males (Table 5). In addition, an increased incidence of subcutaneous benign lipoma was observed in the 50 mg/kg females (Table 6). The Peto trend tests for these adipocyte neoplasms were positive at p 0.025 for a rare tumor (incidence 1%) only when the 50 mg/kg groups were included in the analysis. The combined incidence of adipocyte neoplasms (lipoma, fibrolipoma, and liposarcoma) in male rats at 50 mg/kg also showed a statistically significant positive trend (p 0.025); however, this effect was attributable exclusively to the increased incidence of liposarcomas. There was a higher incidence of uterine squamous cell carcinomas (ranging from 0% to 4.6% across groups) and combined incidences of uterine adenoma and carcinoma (ranging from 3.1% to 10.8% across groups) and uterine squamous cell papilloma and carcinoma (ranging from 1.5% to 6.2% across groups) in treated females; however, there were no clear relationships to dose and the trend was not statistically significant.
There were no other statistically significant positive trends in the incidences of neoplastic lesions in other tissues in muraglitazar-treated male or female rats when compared to the two combined control groups. All other neoplastic findings in this study occurred sporadically with no relationship to treatment, and unbalanced distributions of specific lesions among groups were attributed to differences in survival or reflected normal biologic variation.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
The overall carcinogenicity profile of muraglitazar included a low incidence of benign gallbladder adenoma in male mice at high doses and exposures, a dose-dependent increase in transitional cell papilloma and carcinoma in the urinary bladder of male rats, and an increased incidence of subcutaneous adipocyte tumors in rats at the highest dose tested. This profile differs from most PPAR dual agonists due to the absence of an increased incidence of vascular tumors in mice, the presence of fatty tumors only at high exposure multiples in rats, and the presence of urinary bladder tumors in male rats only (El Hage, 2005
). These rodent tumorigenic responses were considered to have likely been pharmacologically mediated since muraglitazar demonstrated no evidence of cytotoxicity in affected tissues. Moreover, muraglitazar was not mutagenic at concentrations up to 500 (Salmonella) or 5000 (E. coli) µg per plate in the bacterial reverse-mutation assay and was not clastogenic at concentrations up to 150 µg/ml in an in vitro cytogenetics study in primary human lymphocytes. In addition, muraglitazar was not clastogenic in an in vivo micronucleus test up to a maximum tolerated dose (1200 mg/kg/day) in female rats or the limit dose (2000 mg/kg/day) in male rats.
Although there were no statistically significant positive trends in the incidences of neoplasms in muraglitazar-treated mice, the low incidence of benign gallbladder adenomas in male mice at clinically nonrelevant exposures (x62 human AUC exposure) was considered drug related given the background of increased focal gallbladder mucosal hyperplasia at all doses. Gallbladder adenomas were not present in female mice, although there was an increase in the incidence of focal gallbladder mucosal hyperplasia at 5 mg/kg. The mode of action for the development of proliferative gallbladder changes in mice is unknown.
Although hemangiosarcoma is a tumor type commonly increased in mice treated with PPAR dual agonists (El Hage, 2005), the incidences of hemangiosarcoma and combined hemangioma and hemangiosarcoma in mice treated with muraglitazar were not statistically significantly increased. The apparent increase in hemangiosarcoma in female mice given 20 and 40 mg/kg was considered reflective of normal biologic variability given the lack of dose dependency for this effect when data from both carcinogenicity studies were considered collectively. The incidences of hemangiosarcoma in females at these dose levels (8% at 20 mg/kg; 5% at 40 mg/kg) were comparable to that documented in the historical control database at the testing facility (3.95%; range 1.67–9.23%), in the male control groups (5.0–6.67%), and in one of three female control groups (5%) in these carcinogenicity studies. A potential pharmacological basis for the apparent marginal increase in hemangiosarcoma in muraglitazar-treated female mice at 20 or 40 mg/kg was considered unlikely given that survival and AUC exposures in the 40 mg/kg females exceeded that of the 20 mg/kg females where the highest incidence of hemangiosarcoma was observed. Moreover, there was no evidence of an increased incidence of hemangiosarcoma in males at 40 mg/kg, where exposure also exceeded that in the 20 mg/kg females (x1.62).
In male rats, muraglitazar treatment at doses 5 mg/kg was associated with a significant increase in transitional cell papilloma and carcinoma in the urinary bladder and a low incidence of focal to diffuse urothelial hyperplasia. Although transitional cell tumors of the urinary bladder were not present in female rats, there was a minimal increase in the incidence of focal to diffuse urothelial hyperplasia at 30 mg/kg. Urothelial tumors of the urinary bladder have been described as common neoplasms in rats, particularly male rats, treated with other PPAR / agonists suggesting a pharmacologically mediated mode of action. The demonstration of increased urinary solids in 30 and 50 mg/kg male rats, and the ventral disposition of the bladder tumors at the end-of-dose necropsies, provided preliminary evidence that the male rat-specific urinary bladder tumorigenic response was potentially a consequence of muraglitazar-related alterations in urine composition. In a subsequent investigative study, chronic mucosal injury and proliferation secondary to muraglitazar-induced urolithiasis was shown to be the nongenotoxic mechanism for the male rat-specific urinary bladder tumorigenic response to treatment with muraglitazar (Dominick et al., 2006).
The marked decrease in the normally high urine protein levels of aged male rats (Hard, 1995) to levels approaching those seen in females noted in male rats treated with 50 mg/kg of muraglitazar may provide additional support for this mechanism. While the basis for this decrease in urine protein is unknown, pharmacodynamic effects of muraglitazar on glomerular and/or tubular function may have contributed to this phenomenon. In support of this notion, the PPAR agonist rosiglitazone was shown to delay the development of, and retard the progression of, nephropathy in Zucker fatty rats (Buckingham et al., 1998), while the PPAR agonist pioglitazone increased renal tubular uptake of albumin in an in vitro model (Zafiriou et al., 2004). Additionally, treatment with the PPAR agonist troglitazone reduced albuminuria in humans with type II diabetes mellitus (Imano et al., 1998). The increased urine solids noted in treated male rats in this study may have contributed to decreases in urine protein as protein may self aggregate in a hypocitraturic environment (Hess et al., 1993) and thereby act as a nidus for the formation of urinary crystals and calculi. Interestingly, there was no clear evidence for a decreased severity of spontaneous nephropathy in association with the reduced urinary protein levels in males rats treated with muraglitazar.
The increased incidences of subcutaneous adipocyte tumors in rats were considered potentially due to pharmacologic stimulation of preadipocytes. PPAR receptors are highly expressed in adipose tissue and are involved in transcriptional regulation of adipocyte proliferation and differentiation (Gurnell, 2005). Recent FDA overviews of the carcinogenic potential of PPAR agonists have shown increased lipomas and/or liposarcomas in rats, but not mice, for three of six gamma () and two of five dual (/) agonists (El Hage, 2005). The increased incidences of subcutaneous adipocyte tumors in rats treated with muraglitazar occurred only at the highest dose tested (50 mg/kg) where exposures were clearly nonrelevant (x48–59 the mean human AUC at 5 mg).
Although uterine squamous cell carcinoma was not observed in control females in the oral carcinogenicity study in rats, its occurrence in treated female rats was considered incidental to treatment. This interpretation was supported by the lack of a clear dose response and because the incidences of uterine squamous cell carcinoma in individual treatment groups in the rat study were well within the historical control range (0–8.3%) for Bristol-Myers Squibb carcinogenicity studies with Harlan Sprague–Dawley rats. Moreover, based on the historical control data, nine uterine squamous cell carcinomas would be predicted to occur spontaneously in the 390 female rats assigned to this study, whereas seven tumors were observed. If muraglitazar had induced uterine squamous cell carcinoma, the number of tumors observed would be expected to exceed the number predicted based on background incidence rates. Therefore, the occurrence of uterine squamous cell carcinoma only in treated females in the present study was considered a random occurrence. Importantly, relative to other rat strains, female Harlan Sprague–Dawley rats are predisposed to development of extensive nonneoplastic degenerative/inflammatory changes in the uterus and to a higher incidence of uterine neoplasms.
The nonneoplastic adipocyte (fat cells) changes noted in mice and rats, including macrovesiculation of brown adipose tissue, microvesiculation of white adipose tissue, increased bone marrow adipocytes, and white adipocytes in the muscular tunic of the stomach, reflect the expected effects of PPAR agonism on the differentiation and metabolism of fat cells. Thiazolidinedione PPAR agonists such as rosiglitazone and troglitazone have been shown to induce similar changes in brown and/or white adipocytes including hypertrophy, hyperplasia, and a phenotypic change so that the two types of fat resemble one another (Brieder et al., 1999; Toseland et al., 2001), and fatty proliferation and fat deposition in various organs, as well as fatty infiltration of the bone marrow, are known pharmacologic effects of PPAR agonists (El Hage, 2005).
The increased incidence of cortical bone thinning noted in mice treated with muraglitazar may reflect direct pharmacologic effects of PPAR agonism on adipocytes and osteoblasts. Treatment of rodents with the thiazolidinedione PPAR activator rosiglitazone has been shown to result in bone loss, although treatment with the thiazolidinedione PPAR activator troglitazone had no such effect (Rzonca et al., 2004; Tornvig et al., 2001). The reduced bone formation after drug-induced PPAR activation is postulated to be a result of the increase in bone marrow adipocytes, as well as decreases in osteoblastogenesis (Schwartz et al., 2006).
In conclusion, the carcinogenic profile of muraglitazar was fully characterized in 2-year studies in mice and rats. Drug-related tumors were limited to a low incidence of benign gallbladder adenomas in male mice at 20 and 40 mg/kg, increased transitional cell papilloma and carcinoma of the urinary bladder in male rats at 5 mg/kg, and increased adipocyte tumors in male and female rats at 50 mg/kg. These neoplasms occurred by nongenotoxic mechanisms and, with the exception of urinary bladder tumors, at clinically nonrelevant systemic AUC exposures. And finally, investigative studies conducted by our laboratory have defined pharmacologically mediated changes in urine composition that predispose to increased urinary solids as the inciting event in the mode of action for urinary bladder tumors in male rats. This mechanism has no established clinical relevance as crystalluria has not been identified as a risk factor for bladder cancer in humans (Burin et al., 1995; Chappel, 1992).
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Supplementary data 1–5 are available online at http://toxsci.oxfordjournals.org/.
|
NOTES |
|---|
Portions of this report were presented at the 45th annual meeting of the Society of Toxicology, San Diego, CA, March 2006 (Abstract # 2095).
|
ACKNOWLEDGMENTS |
|---|
The authors would like to acknowledge the contributions of the technical and administrative staffs of Covance Laboratories, Inc. (Vienna, VA) in the conduct of this study. In addition, the authors would like to acknowledge Stan Hansen and Shannon Boring for technical assistance in the preparation of figures and tables.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Breider MA, Gough AW, Haskins JR, Sobocinski G, De La Inglesia FA. Troglitazone-induced heart and adipose tissue proliferation in mice. Toxicol. Pathol. (1999) 27:545–552.
Buckingham RE, Al-Barazanji KA, Toseland CDN, Slaughter M, Connor SC, West A, Bond B, Turner NC, Clapham JC. Peroxisome proliferator-activated receptor- agonist, rosiglitazone, protects against nephropathy and pancreatic islet abnormalities in Zucker fatty rats. Diabetes (1998) 47:1326–1334.[Abstract]
Burin GJ, Gibb HJ, Hill RN. Human bladder cancer: Evidence for a potential irritation-induced mechanism. Food Chem. Toxicol. (1995) 33:785–795.[CrossRef][ISI][Medline]
Chappel CI. A review and biological risk assessment of sodium saccharin. Regul. Toxicol. Pharmacol. (1992) 15:253–270.[CrossRef][ISI][Medline]
Cocchia D, Lauriola L, Stolfi VM, Tallini G, Michetti F. S-100 antigen labels neoplastic cells in liposarcoma and cartilaginous tumours. Virchows Arch. A Pathol. Anat. Histopathol. (1983) 402:139–145.[CrossRef][ISI][Medline]
Cohen SM. Cell proliferation in the bladder and implications for cancer risk assessment. Toxicology (1995) 102:149–159.[CrossRef][ISI][Medline]
Cohen SM. Effects of PPAR gamma and combined agonists on the urinary tract of rats and other species. Toxicol. Sci. (2005) 87:322–327.
Cox SL. Muraglitazar: An agent for the treatment of type 2 diabetes and associated dyslipidemia. Drugs Today (Barc) (2005) 41:579–587.[CrossRef][Medline]
Dominick MA, White MR, Sanderson TP, Van Vleet TR, Cohen SE, Arnold LE, Tannehill-Gregg SH, Moehlenkamp JD, Waites CR, Schilling BE. Urothelial carcinogenesis in the urinary bladder of male rats treated with muraglitazar, a PPAR / agonist: Evidence for urolithiasis as the inciting event in the mode of action. Toxicol. Pathol. (2006) 34:903–920.[CrossRef][ISI][Medline]
El Hage J. (2005) http://www.fda.gov/ohrms/dockets/ac/05/slides/2005-4169S2_02_02-FDA-ElHage.ppt. Accessed November 2006.
FDA Draft Guidance for Industry. Statistical Aspects of the Design, Analysis, and Interpretation of Chronic Rodent Carcinogenicity Studies of Pharmaceuticals (2001) Rockville, MD: U.S. Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research (CDER). Available at: http://www.fda.gov/cder/guidance/index.htm. Accessed November 2006.
Gurnell M. Peroxisome proliferator-activated receptor and the regulation of adipocyte function: Lessons from human genetic studies. Best Pract. Res. Clin. Endocrinol. Metab. (2005) 19:501–523.[CrossRef][Medline]
Hard GC. Species comparison of the content and composition of urinary proteins. Food Chem. Toxicol. (1995) 33:731–746.[CrossRef][ISI][Medline]
Hashimoto H, Daimaru Y, Enjoji M. S-100 protein distribution in liposarcoma. An immunoperoxidase study with special reference to the distinction of liposarcoma from myxoid malignant fibrous histiocytoma. Virchows Arch. A Pathol. Anat. Histopathol. (1984) 405:1–10.[CrossRef][ISI][Medline]
Hess B, Zipperle L, Jaeger P. Citrate and calcium effects on Tamm-Horsfall glycoprotein as a modifier of calcium oxalate crystal aggregation. Am. J. Physiol. (1993) 265:F784–F791.[ISI][Medline]
Imano E, Kanda T, Nakatani Y, Nishida T, Arai K, Motomura M, Kajimoto Y, Yamasaki Y, Hori M. Effect of troglitazone on microalbuminuria in patients with incipient diabetic nephropathy. Diabetes Care (1998) 21:2135–2139.[Abstract]
Irisarri E, Hollander CF. Aging of the liver. In: Pathobiology of the Aging Rat—Mohr U, Dungworth DL, Capen CC, eds. (1994) Vol. 2, 1st ed. Washington, D.C.: ILSI Press. 342–349.
King H, Auert RE, Herman WH. Global burden of diabetes, 1995–2025. Diabetes Care (1998) 21:1414–1431.[Abstract]
Klaunig JE, Babich MA, Baetche KP, Cook JC, Corton JC, David RM, Deluca JG, Lai DY, McKee RH, Peters JM, et al. PPAR agonist-induced rodent tumors: Modes of action and human relevance. Crit. Rev. Toxicol. (2003) 33:655–780.[ISI][Medline]
Koro CE, Bowlin SJ, Bourgeois N, Fedder DO. Glycemic control from 1988 to 2000 among U.S. adults diagnosed with type 2 diabetes. Diabetes Care (2004) 27:17–20.
Kumar V, Fausto N, Abbas AK, eds. Cellular adaptations, cell injury, and cell death. In: Robbins & Cotran Pathologic Basis of Disease (2004) 7th ed. Philadelphia: W. B. Saunders Company. 3–46.
McConnell EE, Solleveld HA, Swenberg JA, Boorman GA. Guidelines for combining neoplasms for evaluation of rodent carcinogenicity studies. J. Natl. Cancer Inst. (1986) 76:283–289.[ISI][Medline]
PDR Electronic Library. (2006a) Actos (Takeda) (Pioglitazone Hydrochloride) Tablets. Physicians' Desk Reference (online). Retrieved September 21, 2006, from http://www.thomsonhc.com/pdrel/librarian/PFPUI. Thomson Micromedex, Greenwood Village, CO.
PDR Electronic Library. (2006b) Fibrate Compounds (Various). Physicians' Desk Reference (online). http://www.thomsonhc.com/pdrel/librarian/PFPUI. Thomson Micromedex, Greenwood Village, CO.
Peto R, Pike MC, Day NE, Gray RG, Lee PN, Parish S, Peto J, Richards S, Wahrendorf J. Guidelines for simple, sensitive significance tests for carcinogenic effects in long-term animal experiments. In: IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, Suppl. 2, Long-term and Short-term Screening Assays for Carcinogens: A Critical Appraisal (1980) Lyon, France: International Agency for Cancer Research [IARC]. 311–426.
Rosen ED, Walkey CJ, Puigserver P, Spiegelman BM. Transcriptional regulation of adipogenesis. Genes Dev. (2000) 14:1293–1307.
Rzonca SO, Suva LJ, Gaddy D, Montague DC, Lecka-Czernik B. Bone is a target for the antidiabetic compound rosiglitazone. Endocrinology (2004) 145:401–406.
Schwartz AV, Sellmeyer DE, Vittinghoff E, Palermo L, Lecka-Czernik B, Feingold KR, Strotmeyer ES, Resnick HE, Carbone L, Beamer BA, et al. Thiazolidinedione use and bone loss in older diabetic adults. J. Clin. Endocrinol. Metab. (2006) 91:3349–3354.
Suttie AW. Histopathology of the spleen. Toxicol. Pathol. (2006) 34:466–503.[CrossRef][ISI][Medline]
Thomas DG, Breslow N, Gart JJ. Trend and homogeneity analyses of proportions and life table data. Comput. Biomed. Res. (1977) 10:373–381.[CrossRef][ISI][Medline]
Tornvig L, Mosekilde LI, Justesen J, Falk E, Kassem M. Troglitazone treatment increases bone marrow adipose tissue volume but does not affect trabecular bone volume in mice. Calcif. Tissue Int. (2001) 69:46–50.[CrossRef][ISI][Medline]
Toseland CDN, Campbell S, Francis I, Bugelski PJ, Mehdi N. Comparison of adipose tissue changes following administration of rosiglitazone in the dog and rat. Diabetes Obes. Metab. (2001) 3:163–170.[CrossRef][ISI][Medline]
Yki-Järvinen H. Thiazolidinediones. N. Engl. J. Med. (2004) 351:1106–1118.
Zafiriou X, Stanners SR, Polhill TS, Poronnik P, Pollock CA. Pioglitazone increases renal tubular cell albumin uptake but limits proinflammatory and fibrotic responses. Kidney Int. (2004) 65:1647–1653.[CrossRef][ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. Chopra, J. Hinley, M. B. Oleksiewicz, and J. Southgate Trans-Species Comparison of PPAR and RXR Expression by Rat and Human Urothelial Tissues Toxicol Pathol, April 1, 2008; 36(3): 485 - 495. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 |