Histopathology of the Urinary Bladders of Cynomolgus Monkeys Treated with PPAR Agonists

Abstract

Peroxisome proliferator-activated receptors (PPAR) are involved in the pathogenesis of insulin resistance, diabetes, hyperlipidemias, and related complications. Consequently, a mechanistic understanding of PPAR subtypes and their activation provides promising therapeutic targets for the management of type 2 diabetes mellitus and the metabolic syndrome. Available data from rodent carcinogenicity studies, however, demonstrate that PPAR agonists can be tumorigenic in one or more species of rodents at multiple sites. Sufficient data are not yet available to explain the mode(s) of action for most of these tumor types. There has been information presented by FDA that indicates there are urothelial changes in the monkey (and possibly the dog) in addition to the rat. Outstanding questions exist regarding potency, species differences, safety margins, and other issues.

In 2005, the International Life Sciences Institute (ILSI) Health and Environmental Sciences Institute (HESI) PPAR Agonist Project Committee was established to advance research on the modes of action and potential human relevance of emerging rodent tumor data. Additionally, the HESI PPAR Agonist Project Committee authorized a Pathology Working Group (PWG) to examine the urinary bladder from cynomolgus monkeys. The focus of this PWG was to establish consistent diagnostic criteria for urothelial changes and to assess the potential relationship of these changes to treatment. Specific diagnostic criteria and nomenclature were recommended for the diagnosis of urothelial granules, vacuolation, hypertrophy, and hyperplasia in studies conducted with PPARγ and dual α/γ agonists in cynomolgus monkeys, which will assist investigators performing toxicity studies to provide data in a consistent manner between studies and laboratories. In this review of selected tissues, treatment with PPAR agonists was not associated with urothelial hypertrophy or hyperplasia, but there was an increased incidence in the size and frequency of vacuoles within the superficial urothelial and adjacent intermediate cell layers.

Introduction

Peroxisome proliferator-activated receptors (PPAR) are nuclear receptors involved in regulating lipids and the effects of insulin (Yki-Järvinen 2004). Three major receptors have been identified—alpha (α), gamma (γ), and delta (δ), which sometimes is referred to as beta (β)—with differing tissue distributions and effects. Agonists have been developed for each of these receptors, with differing pharmacologic and toxicologic effects.

PPARα receptors are present in liver, brown fat, kidney, heart, skeletal muscle, immune system, intestine, and retina. PPARα agonists have effects on maintenance of lipid levels, especially triglycerides and high-density lipoproteins (HDL) (Klaunig et al. 2003). The toxicologic effects of PPARα agonists (peroxisome proliferators) have been extensively investigated, including carcinogenic effects in rodents. They may induce liver tumors in rats and mice and frequently induce rat pancreatic acinar cell and testicular Leydig cell tumors. The modes of action for these tumors have been identified and generally been found to be not relevant to humans (at least quantitatively) (Klaunig et al. 2003).

PPARΔ receptors are ubiquitous, whereas PPARγ receptors are present in adipose tissue, endothelial cells, the immune system, and several epithelial tissues, such as urothelium and intestine (Yki-Järvinen 2004). Agonists to these receptors, either selective or with activity at more than one PPAR receptor, are being developed for clinical use in treating lipid disorders and diabetes. The effects of PPARδ agonists on carcinogenesis have not been extensively investigated but have already produced varying and conflicting results. PPARγ agonists and dual PPARγ/α agonists have been more extensively studied. Although various tumors have been produced by specific agonists, the most common types of tumors observed so far have been rat urinary bladder urothelial tumors and various mesenchymal sarcomas (especially in subcutaneous adipose tissue) in rats, mice, and hamsters. Additionally, there has been information presented by FDA that identified urothelial changes in the monkey (and possibly the dog) (El-Hage 2005). Based on these early findings, the FDA decided to require study sponsors to complete two-year rodent carcinogenicity studies and submit draft reports for agency review prior to clinical studies longer than six months in duration (El-Hage 2005).

In 2005, the HESI PPAR Agonist Project Committee was established to advance research on the modes of action and potential human relevance of emerging rodent tumor data for PPAR agonists. The most commonly observed tumor types reported were mouse hemangiosarcomas, rat fibro- and liposarcomas, and, in some cases, rat urinary bladder tumors. In 2006, the PPAR Agonist Project Committee approved a Pathology Working Group (PWG) to develop consensus on tumor diagnoses and consistency of diagnoses across multiple studies for hemangiosarcomas in mice and liposarcomas/fibrosarcomas in rats. This PWG was completed in January 2007, and the results were published (Hardisty et al. 2007).

The PPAR Agonist Project Committee also approved a PWG review of the urinary bladder from cynomolgus monkeys. The focus of this review was to establish consistent diagnostic criteria to characterize any morphologic changes that may be relevant in the urinary bladder of monkeys resulting from treatment. The PWG specifically addressed morphologic changes that may indicate toxic injury or preneoplastic proliferative responses. The intent was to provide a substantial basis upon which to design future experiments to address the mode of action associated with PPARγ and dual α/γ agonists and establish, with greater certainty, the human relevance.

Materials and Methods

Pathology Working Group

The PWG was organized and conducted by Experimental Pathology Laboratories, Inc. (EPL), Research Triangle Park, NC. All materials reviewed during the PWG were provided by the sponsoring companies. For each of the monkey cases submitted, two unstained slides were requested to be submitted to EPL.

To achieve the necessary degree of confidentiality, EPL provided glassware to be used to submit the unstained tissue sections. All glass slides were purchased from the same vendor, were of the same style and color, and had preprinted numbers. The slides were initially numbered consecutively, and then were rearranged in random order according to a computer-generated list of random numbers.

The randomized slides were distributed to the participating companies by HESI. When submitting the material, the company also provided a log containing the random number, animal number, sex, location of the bladder section, how sections were taken, whether tissue was fixed immediately or routinely collected at necropsy, treatment (control or treated), and if the monkey was sacrificed or died spontaneously. However, the exact length of time from death until fixation of the bladder for those submitted as immediately fixed was not provided. All submissions were returned to HESI, repackaged, and shipped to EPL without any accompanying information concerning the origin of the slides.

When received at EPL, the slides submitted with each log were randomized and assigned ascending consecutive identification numbers. EPL stained one slide from each case submitted with hematoxylin and eosin (H&E) to be coded for examination by the PWG. Additionally, the second slide from selected control male and female monkeys was similarly stained to be used as reference controls during the slide review.

There were a total of 197 cases submitted. This number included twenty cases routinely collected at necropsy and twenty-eight cases immediately fixed with Log A. All cases submitted with Log B (twelve cases), Log E (sixty-two cases), Log F (twenty-seven cases) and Log G (sixteen cases) were immediately fixed. Sections submitted with Log H (thirty-two cases) were routinely collected at necropsy. Two slides were submitted for each case, except for Log G with only a single slide submitted with each case. Tissues that were prepared from Logs C and D were lost in transit by a commercial carrier, and the companies did not resubmit them.

The meeting of the PWG, which consisted of independent consulting pathologists with expertise in toxicologic pathology and/or primate pathology, with specific emphasis on studies conducted in cynomolgus monkeys, was held in the Research Triangle Park, North Carolina, June 26–27, 2007.

The PWG chairperson (JFH) organized the meeting and was responsible for conducting the meeting, leading the discussion, and preparing a report of the PWG’s findings. The PWG participants included board-certified medical (SML), veterinary (DCA, JMC, FFH, LJL) and experimental (SB) pathologists (see list of authors for full names/affiliations). Several interested scientists from the HESI PPAR Agonist Project Committee or from sponsoring member companies were also in attendance as observers. At the beginning of the PWG meeting, the PWG chairperson provided the panel with a review of the issues to be addressed and an outline of the approach to be followed. The PWG examined coded slides without knowledge of treatment with a specific test article or nontreatment control, or the origin of the case selected for their review. The PWG examined the slides without knowledge of treatment. They were provided information regarding collection technique and fixation protocol. The slides were examined from each Log separately to facilitate the interpretation of differences owing to treatment from those that may be the result of the necropsy collection technique or histologic preparation of the tissue sections.

Participants recorded their diagnoses and comments on worksheets that were prepared by the PWG chairperson. The PWG participants recorded either a morphologic diagnosis of a urothelial lesion or made an observation of “Not Remarkable” if no significant changes were identified. The PWG worksheets also included an area to record other comments that might be valuable for discussion. The diagnoses for each case examined were discussed by the group, reexamined if necessary, and the final opinions were recorded by the chairperson.

Immunohistochemistry

Based on the PWG review and the further recommendations of the panel, immunohistochemical staining was performed at EPL by a board-certified veterinary pathologist (HK-R) to further characterize the affected cell type in cases where superficial urothelial vacuolation was observed. Method development for the immunohistochemistry stains was performed using normal monkey bladder samples. Remaining unstained slides from animals diagnosed with vacuolation at the PWG meeting were assigned to each IHC stain and stained according to the optimized protocols. Stained slides were reviewed and digital images of representative findings were obtained.

Briefly, slides were deparaffinized, rehydrated with distilled water, and soaked in 1X Tris buffer. Samples were quenched of endogenous peroxidase and rinsed. An additional incubation step with avidin and biotin was performed (cytokeratin only) before a protein block was applied. Slides were incubated with the appropriate primary antibody (cytokeratin, rabbit anti-cow, Dako, Carpinteria, CA, USA; CD68, monoclonal mouse anti-human, Dako, Carpinteria, CA, USA; CD45/common leukocyte antigen, monoclonal mouse anti-human, Dako, Carpinteria, CA, USA), rinsed, and incubated with secondary antibody. After treatment with streptavidin, slides were developed for a color reaction using DAB substrate. Slides were counterstained with hematoxylin, dehydrated, and coverslipped.

Results

A wide range in the number of mononuclear inflammatory cells normally present was observed in the urothelium during this slide review. The cells consisted of a mixture of macrophages and lymphocytes. The cellular infiltrate as discernible by H&E staining was most often located in the submucosa but occasionally involved the urothelium, and the distribution varied from focal to multifocal to diffuse and was often perivascular. In some instances, the mononuclear cells formed follicular structures resembling lymphoid follicles (Figure 1). Although varying degrees of cellular infiltrates were present, they were recorded during the PWG only when considered to be excessive. Differences in the incidence and degree of inflammatory cellular infiltrates between control and treated animals were not evident.

Eosinophilic granules (keratohyaline cytoplasmic inclusions) were present in the urothelium of all monkeys examined. The size and number of granules per epithelial cell varied. The granules were round to oval in shape and granular to hyaline in appearance. The cytoplasmic inclusions were typically observed in the intermediate and superficial layers of urothelium and were not associated with inflammatory or degenerative changes. The inclusions stained brightly eosinophilic in H&E stained tissue sections. Since they were observed in all sections of bladder, the granules were considered a spontaneous background finding and were not recorded during the PWG review (Figure 2).

“Basilar” vacuolation of the urothelium was a finding that was commonly observed in the basal and intermediate cell layers of control and treated animals. They consisted of small intracellular vacuoles and were considered to represent an artifact resulting from shrinkage during fixation of the tissue. Since they were considered to be an artifact and were present in most of the sections examined, they were not recorded as a morphologic diagnosis during the PWG review (Figure 3).

Vacuolation in the intermediate or superficial cell layers or between the umbrella cells and the upper intermediate cell layer were observed in control and treated animals. These vacuoles were larger and more frequent in the bladders from treated animals when compared to the control animals (Figure 4). It was difficult to determine the exact nature of the vacuoles based only on examination of H&E-stained tissue sections. It was often difficult to determine whether the vacuoles were intracytoplamic or extracellular. The larger vacuoles produced distortion of the luminal surface of the superficial cell layer. The vacuoles were generally round to oval in shape and were either clear or contained a small amount of central accumulation of lightly basophilic granular material and cellular debris. In some areas, the vacuoles appeared to coalesce and exhibit bullous formation with small accumulations of cellular debris. Pyknotic cells were often present within the vacuoles, but the degenerating cell type was not evident.

After the PWG and the identification of superficial urothelial vacuoles related to treatment, immunohistochemistry was performed on remaining unstained tissue to further characterize the identity of the degenerate cells and specific location of the vacuoles. In general, slide and tissue section quality was good. Although slides stained for cytokeratin exhibited a robust urothelial positive reaction, positive staining for CD45 (lymphocytes) and CD68 (macrophages) was cytoplasmic and produced a more subtle color reaction.

Small numbers of CD45+ intra-urothelial lymphocytes were scattered throughout the sections in the basal and more superficial epithelial layers. In sections from some treated animals, there was a slight increase in the number of lymphocytes in the superficial urothelial layers, and there was a small number of lymphocytes that were associated with or within the large superficial vacuoles found in four of thirteen treated animals (Figure 5). However, most degenerative cells and debris within vacuoles were not CD45+.

Only one slide from a control animal demonstrated CD68-positive staining within the few macrophages scattered throughout the submucosa, as was seen during method development and in the control sample of normal monkey bladder. These submucosal cells were considered the internal positive control for each slide, and because of the almost uniform absence of these cells from stained sections, it was concluded that the stain failed in the slides available from the PPAR project.

With the cytokeratin stain, positive staining occurred more intensely around the prominent, large, superficial urothelial vacuoles from treated animals. Many larger, circular vacuoles were associated with umbrella cells or the adjacent deeper urothelial layer (Figure 6). The inner surface of some vacuoles appeared to have an inner lining of coalesced, densely staining, cytokeratin-positive material (Figure 7). In some sections, vacuoles also appear to have arisen from progressive perinuclear clearing of cytoplasm and swelling of the entire cell (Figure 6). In four sections, there was separation of superficial urothelial cells owing to a presumptive degradation of intracellular junctions, which was associated with areas described as “bullous formation” (Figure 8). Cell borders on either side of these gaps were distinct and stained positively for cytokeratin. Although most of the pyknotic cells within vacuoles had minimal to no positive staining, a few cells in some sections were cytokeratin positive.

During the PWG review of the coded, H&E-stained slides, neither urothelial hypertrophy nor urothelial hyperplasia was observed. The PWG participants noted that the rounded appearance of the urothelial cells is normally present in the intermediate cell layers in contracted urinary bladders. Although some variation in the size of the urothelial cells was observed during the PWG review, none was considered to be abnormal and no consistent changes were present that would fit the criteria for the diagnosis of urothelial cell hypertrophy.

The issue of hyperplasia was much more difficult. The difficulty arose in trying to define the normal range for the monkey urothelium as compared to truly hyperplastic. Like humans, and in contrast to rodents, the number of cell layers in urinary bladder urothelium is usually greater than three and can be as many as twelve, even in sections from control monkeys. Furthermore, when tangential sectioning and artifacts of processing occur, the number of cell layers of normal urothelium can appear to be increased.Figure 9illustrates the difficulty in determination of true hyperplasia from an apparent increase in thickness of the urothelium resulting from contraction of the bladder wall and tangential sectioning in a treated male monkey. Morphologic features commonly associated with hyperplasia were not observed (marked focal, multifocal, or diffuse multicellular thickening of the urothelium). None of the slides examined was diagnosed as urothelial hyperplasia.

Following the review of slides, the PWG recommended the following nomenclature and diagnostic criteria be recorded for the changes observed in the sections of urinary bladder included in this slide review.

Recommended Nomenclature and Diagnostic Criteria

Cellular Infiltrate, Inflammatory

• Primarily mononuclear cells (macrophages and lymphocytes)

• Present most often in the submucosa but may also involve urothelium

• May be focal, multifocal, or diffuse in distribution

• Occasionally prominent perivascular component

• Sometimes form nodules of lymphoid tissue resembling lymphoid follicles

Urothelial Granules (Cytoplasmic Inclusions)

• Typically occur in the intermediate and superficial layer of urothelium

• Round to oval; granular to hyaline

• Stain brightly eosinophilic in H&E stained sections

• Not associated with inflammatory or degenerative changes

• Common spontaneous finding

Urothelial Vacuolation

• Occurs in the intermediate or superficial layer of urothelium or between the umbrella cells and upper intermediate cell layer

• Often clear, but may contain small central accumulations of light basophilic granular material or small, dark staining nuclei

• May coalesce and exhibit bullous formation with small accumulations of pyknotic cells and/or cellular debris

• Must distinguish from small intracellular vacuoles commonly observed in the basal and intermediate cell layers

Urothelial Hypertrophy

• Increased size of individual urothelial cells in the intermediate cell layer

• Increased amount of pale, granular cytoplasm

• May be focal or diffuse

• Not associated with inflammatory cellular infiltrates

• Must distinguish from rounded appearance of cells in the intermediate cell layer in contracted bladder

Urothelial Hyperplasia

• Focal, multifocal, or diffuse

• Simple, papillary, or nodular

• Multicellular thickening of the urothelial cell lining

• With or without cellular atypia

• Atypia involves cell and nuclear pleomorphism, cytoplasmic basophila, and disorganized arrangement or increased mitotic activity

Simple Hyperplasia

• Uniform increase in thickness of urothelial cell lining

• May be focal or diffuse

• Mitotic figures are uncommon

• No cellular atypia or dysplasia

• Must distinguish from contracted bladder or rotation of section

Papillary Hyperplasia

• Urothelial surface irregular owing to delicate exophytic outgrowths

• Fibrovascular supporting stroma

• Covered by urothelial cell layers of variable thickness

• May be focal or multifocal

• Diffuse papillary hyperplasia is sometimes termed “papillomatosis”

• Cellular atypia and increased mitosis may be present

Nodular Hyperplasia

• Solid round to oval islands or cords of urothelial cells

• Mainly exophytic, occasionally endophytic

• May be focal, multifocal, or diffuse

• Acute and/or chronic inflammation may accompany the hyperplasia

• May occur in conjunction with either simple or papillary hyperplasia

Discussion

The histology of the urinary bladder in nonhuman primates is very similar to that found in humans (Roberts et al. 1998). The urinary bladder is lined by epithelium that has unique morphologic and physiologic features. In the experience of the PWG participants, the thickness of the urothelium, also called transitional epithelium, varies according to the degree of distention and anatomic location. In the contracted bladder, the average thickness is usually seven to ten cells (Koss 1974), but it may be as thick as ten to twelve layers, as observed by the PWG, and still be considered normal. No urothelial hypertrophy or hyperplasia was identified during this review, and no feature of any bladder was remotely suggestive of a preneoplastic change.

Regardless of the degree of distention, normal urothelium is characterized by the presence of large superficial cells, each of which, in umbrella-like fashion, covers several smaller cells of the immediately underlying layer. Three regions can be identified: the superficial cells (which are in contact with the urinary space), the intermediate cells, and the basal cells (which lie on a basement membrane). The presence of superficial cells may be considered important evidence of normalcy of the urothelium (Koss 1974). Unfortunately, superficial cells are loosely attached to the underlying cell layer and may become detached during manipulation of the tissue at necropsy. In the intermediate cell layer, the cells are oriented with the long axis perpendicular to the basement membrane. The nuclei are oval and have finely stippled chromatin with absent or minute nucleoli. There is ample cytoplasm, which may be vacuolated. In the distended state, this layer may be inconspicuous or only one or two cells thick and flattened.

The basal layer is composed of cuboidal cells that lie on a thin but continuous basement membrane. In the distended bladder, the urothelium becomes thin and flattened along the long axis horizontal to the basement membrane. In practice, the thickness of the urothelium is dependent not only on the degree of distention, but also on the plane on which the tissue is cut. If the cut is tangential to the basement membrane, it is possible to generate an artificially thick mucosa.

The urinary bladder in the monkey has two very distinctive features that are not seen in rodents, dogs, or humans. The first of these is the presence of eosinophilic granules (cytoplasmic inclusions). These features were well delineated several years ago as keratohyaline granules (Burek et al. 1972; Lucas et al. 1972). They are present as a normal, spontaneous finding, and changes were not observed with exposure to PPAR agonists. Second, the monkey bladder has a wide range in the number of mononuclear inflammatory cells normally present in the submucosa and urothelium. It is unknown whether these inflammatory cells in the monkey are the result of an underlying infection in these animals that has not been identified or whether it is some other normal manifestation of the monkey.

Vacuolation of the urothelium appeared to begin as intercellular or intracellular vacuolization between the superficial cell layer and the intermediate cell layer. Small, dark nuclei and pyknotic cells were often present within the vacuoles. The vacuoles tended to be larger and more frequent in treated animals compared to controls. With immunohistochemical staining, large, discrete, circular, superficial vacuoles were lined by dense, cytokeratin-positive material. Progressive clearing of perinuclear cytoplasm and disruption of the cytoskeleton appeared to be a major contributing factor to the development of vacuoles containing cell debris and orphaned nuclei. The large, discrete circular vacuoles proved to be intracellular, particularly within umbrella cells and the adjacent intermediate cell layer. Irregular areas with cellular separation and bullous formation appeared to be a result of loss of cell junctions between the umbrella cells and the adjacent deeper layer. A few small cells and pyknotic nuclei stained positively for CD45 (lymphocytes), whereas a slightly higher number was positive for cytokeratin. As the attempts to stain representative samples for CD68 were unsuccessful, the contribution of macrophages to the remaining unstained cells was not discerned. However, macrophages appeared not to occur naturally in the urothelium, unlike the intraepithelial lymphocyte population.

It has been suggested based on unpublished observations in a recent investigation that the vacuoles may represent an autolytic change. The bladders from two monkeys (not treated with PPAR agonists or edematous) were collected. From each of these monkeys, the urinary bladder was immediately collected upon the death of the animal and cut into several strips, immersing one strip from each of the bladders immediately into fixative, another strip from each of the bladders at thirty minutes, and another strip at sixty minutes. The bladders at zero minutes did not have vacuoles, and each of the bladders at sixty minutes had distinctive vacuoles. Vacuoles were also present in the bladder tissue fixed at thirty minutes, but they were very subtle and difficult to find (Cohen 2005).

Normal monkey and human urothelium has been defined as urothelium that is three to seven cell layers thick, but in the slides reviewed by the PWG the upper limit considered to be normal was ten to twelve cell layers thick. This finding is in contrast to the rodent, where three or four cells is considered normal. When tangential sectioning and artifacts of processing occur, the number of cell layers can actually seem much greater. However, in the monkey and in the human, there is marked variation in the number of cells in the urothelium depending on the history of the individual, the site in the urinary bladder where the specimen is taken, and the amount of urine that has been produced by the individual prior to the obtainment of the urothelial sample for histology. Although it has not been carefully evaluated, it is generally accepted among urologic pathologists, both veterinary and human, that a major influence on the number of cell layers in the bladder urothelium is the amount of urine that is processed (Murphy et al. 2004). This is somehow related to the amount the bladder has been distended, possibly owing to generation of growth factors that can occur with stretching (Zhou et al. 2005). More importantly, there has been the suggestion, albeit never tested, that the number of cell layers is a reflection of urinary volume and/or frequency. Since the PPARγ and combined agonists produce fluid changes in monkeys like in rodents and humans, the urinary volume of the monkeys is likely to be greater as the dose of the agonist being administered is increased. This increase would be associated with increased urinary output and likely would be associated with an increase in the number of cell layers seen in the bladder urothelium, but would be diagnostically considered within the range of normal, not hyperplastic.

There has been a move away from the more traditional method of diagnosing urothelial hyperplasia (counting cell layers) in the classification of human urothelial lesions, because the assessment has not been shown to be reproducible. Consequently, a “marked” thickening is required to make the diagnosis for flat or simple hyperplasia in humans (Epstein et al. 1998). This approach is applicable to the evaluation of hyperplasia in the monkey urinary bladder as well.

It is recommended that the urinary bladders be inflated to a similar degree in each animal and fixed immediately. Three tissue sections should be examined from each animal: one taken from the dome, the second from the wall, and the third from the trigone. Care must be taken during tissue processing and embedding to avoid oblique sections.

In conclusion, the results of this review indicated that there are many difficulties that must be considered when conducting a morphologic examination of the urinary bladder. There are several technical procedures involved in the necropsy, collection, fixation, processing, and tissue staining that may affect the histologic appearance of the urothelium of the urinary bladder. These variables should be carefully monitored and eliminated when possible to provide the best material for evaluation and to ensure comparable results that can be accurately interpreted.