Recommendations for the Interpretation of Renal Tubule Proliferative Lesions Occurring in Rat Kidneys with Advanced Chronic Progressive Nephropathy (CPN)

Abstract

There is little guidance in the literature on the spectrum of proliferative tubule lesions in the kidneys of aging rats affected with spontaneously occurring, chronic progressive nephropathy (CPN), or their interpretation. Through accessing 2-year carcinogenicity studies in male F344 rats held in the Archives of the National Toxicology Program, NIEHS, a large number of cases of advanced CPN have been surveyed histopathologically for proliferative tubule lesions, and an attempt made to provide guidelines for discrimination of lesions common to the CPN process, from those representing precursors of neoplasia. Several proliferative lesions were identified as common in advanced CPN with no apparent evidence supporting a role in renal tubule carcinogenesis. It is recommended that these lesions be viewed generically as CPN tubule profiles, and not recorded separately from the diagnosis of CPN. Criteria were developed to distinguish these CPN-associated lesions from atypical tubule hyperplasia, a precursor of adenoma, both of which were also represented in this survey of advanced CPN.

Introduction

In the strains of laboratory rat most commonly used in 2-year studies for assessing carcinogenicity of chemicals (Fischer 344 and Sprague–Dawley rats), spontaneous, age-related, chronic progressive nephropathy (CPN) is invariably present. Particularly where conventional, high-protein diets have been provided, the severity of this disease becomes a factor with the potential to confound kidney lesion diagnosis, especially in distinguishing early stages in the development of neoplasia. Commencing around 2 months of age in some rats as single basophilic tubules associated with thickened basement membrane, the disease progresses relentlessly to involve more and more nephrons, ultimately involving the whole kidney, and leading to the onset of renal failure.

From its incipient stages, and continuing into end-stage disease, CPN is both a degenerative and regenerative disease characterized by tubule degeneration, tubule atrophy, and ongoing tubule cell proliferation. In addition to the frequency of mitotic figures, the highly proliferative nature of many CPN-affected tubules has been confirmed by autoradiographic or immunohistochemical demonstration of DNA synthesis (Sworn and Fox, 1974; Konishi and Ward, 1989; Short et al., 1989; Hard and Khan, 2004). Because of this regenerative character, a range of different phenotypes of renal tubule proliferative foci can be encountered in severe and end-stage CPN.

In carcinogenicity assessment of test compounds, there is little guidance in the published literature on how to interpret proliferative profiles of renal tubules occurring in association with advanced CPN, and even fewer attempts to understand the nature and biological behavior of these tubule profiles in the context of advanced CPN. The range of different proliferating tubule phenotypes encountered in advanced CPN include a number that have been problematic for pathologists in regard to interpretation and diagnosis. In 2-year carcinogenicity bioassays, the aim for hazard assessment related to kidney is to record those proliferative lesions that represent preneoplastic hyperplasia, as well as neoplasms. The term atypical tubule hyperplasia (ATH) has been used to describe these precursor lesions (Hard et al., 1995). In kidneys affected by severe CPN, the intent is also to recognize lesions that are common and predictable components of the progressive nephropathy process, and which are likely to have questionable, if any, biological significance in the context of preneoplasia.

In this paper, we have attempted to discriminate unusual proliferative profiles found in advanced CPN that are not necessarily on the pathway to cancer development, from pre-neoplastic lesions, and to define criteria that we hope will assist in their interpretation and diagnosis.

Materials and Methods

The kidneys from 6 different 2-year rat studies held in the Archives of the National Toxicology Program (NTP), National Institute of Environmental Health Sciences, including step sections where these had been taken, were evaluated for CPN, and the range of proliferative tubule lesions present. The emphasis in this survey was on evaluation of advanced CPN (usually equivalent to the diagnosis of marked nephropathy in NTP reports), specifically in the male rat in view of the well-known sex predisposition for this disease (Gray, 1977; Barthold, 1979). The survey also selected studies in which chemical exacerbation of CPN had been reported along with a small or marginal increase in renal tubule tumors. These bioassays included quercetin (NTP, 1992), coumarin (NTP, 1993), oxazepam (NTP, 1998a), chloroprene (NTP, 1998b), tetrahydrofuran (NTP, 1998c), and primidone (NTP, 2000). In addition, a selection of kidneys was examined from 2-year studies with 2 chemicals not associated with renal tumors but having a high control incidence and severity of CPN, namely isobutene (NTP, 1998d) and ozone (NTP, 1994). Each one of these bioassays had been conducted in the Fischer 344 rat. These re-evaluations were supported by additional studies involving serial sectioning of kidneys, special stains, and immunohistochemical staining for proliferating cell nuclear antigen (PCNA). The detailed findings from these additional studies will be reported separately.

Criteria recommended by the International Agency for Research on Cancer (IARC) and the Society of Toxicologic Pathology (STP), as reported by Alden et al. (1992) and Hard et al. (1995), respectively, were used for identifying simple tubule hyperplasia, foci of ATH (synonymous with renal tubule hyperplasia in NTP reports), and renal tubule adenomas. Additional criteria that could assist in the identification of ATH in particular were also sought during the conduct of this survey. The severity of CPN was graded according to a scheme based on the progression of the disease from minimal tubule involvement up to end-stage kidney (Hard and Khan, 2004). This system differed from that used by NTP (a 4 category grading system) in recognizing 8 grades of severity in which advanced CPN was represented by severe (grade 7) and end-stage (grade 8) CPN. These grades usually conformed to the highest grade of 4, or marked severity in the NTP system.

In the course of this survey, over 1500 kidney sections were examined, representing approximately 1100 male F344 rats. All of the rats were affected by CPN, but at least 60% of the slides surveyed involved kidneys with advanced (i.e., grades 7 or 8) CPN. The recommendations presented here are based on examining numerous profiles of CPN tubule proliferation that were not considered to be preneoplastic lesions. Criteria were developed for discriminating these CPN tubule profiles from preneoplasia based on 189 foci of ATH and 86 adenomas that were encountered in this survey. Although 21 foci of oncocytic hyperplasia or oncocytoma were also found, these present little difficulty in diagnosis and are not considered here.

Results and Recommendations

Proliferative Tubule Profiles as a Component of Advanced CPN

Based on the frequency of the proliferative tubule profiles described below, and/or the apparent absence of any morphological evidence linking them to adenomas from the many sections examined, we are recommending that these profiles not be diagnosed as preneoplastic lesions. The descriptions provided, along with accompanying illustrations, are intended to serve as criteria for identifying these profiles. Although we have provided nomenclature for each of these lesions, this has been done to facilitate their identification and description in this paper. We are not recommending that these descriptive terms be used, but if there is a need to refer to these lesions in the narrative of a study report, we suggest that a generic term such as CPN tubule profile be used. However, we also recommend that these profiles not be recorded separately from the diagnosis of CPN, either by study pathologists, or during the quality assurance process.

Tangentially Sectioned CPN Tubules

These were very frequent in severe to end-stage CPN, presenting as a multilayering or solid sheet of lining cells within a tubule. In the classic example, a partial lumen was usually present with well-defined cells showing characteristic cobblestone-like disposition of epithelial cells (Figure 1). In most examples of tangential sectioning, only the overcrowding of nuclei indicated a cross section through a CPN tubule with regenerative hyperplasia. Sometimes there was also a blurring or smudging of the associated thickened basement membrane that indicated a glancing section along the edge of a tubule.

Occasionally, tangential sections grazing the edge of a basophilic proximal tubule infiltrated by polymorphonuclear inflammatory cells (tubulitis), often a component of advanced CPN, could also be misinterpreted as ATH (Figure 2), but the neutrophil intravasation in and around the lesion identified the focus as inflammatory and not preneoplastic.

Hypertrophic Tubules

Solitary profiles of enlarged, eosinophilic tubule cells with homogeneous, abundant cytoplasm, and small bland nuclei, representing tubule hypertrophy, were commonly seen in end-stage CPN (Figure 3). On occasion, these appeared to have a lining that included papillary projections of 2 to 3 cells thickness. Rarely, tangential sectioning of a hypertrophic tubule resulted in a partially solid profile that could have been mistaken for an eosinophilic form of ATH (Figure 4).

Intratubular Cell Clusters

These profiles appeared as focal accumulations of small and compacted, basophilic cells within relatively normal proximal tubules, either as a discrete clump or dense protrusion from the tubule lining into the lumen (Figure 5). These were quite frequent in some kidneys with advanced CPN. Examination of numerous examples demonstrated that these profiles were cross sections of bends or ridge-like indentations in proximal convoluted tubules. It was characteristic that piled up, basophilic cells formed a focal multicellular (mulberry-like) aggregate at the locus of the bend or indentation, which in many sections, showed a central core of basement membrane representing the infolding of interstitium (Figure 6). The epithelial cells in the clusters were very crowded, with indistinct cytoplasm and bland nuclei. The intratubular clumps did not result in any expansion of the tubule, or bulging into the adjacent parenchyma. Intratubular cell clusters were usually associated with proximal tubules that were only partially affected by CPN, and were less frequently seen in the typical basophilic CPN tubules severely affected with basement membrane thickening.

Dilated Tubules

Two types of dilated tubule profiles in advanced CPN have proved problematic for pathologists. These tubule forms were always solitary profiles, and distinct from those containing the characteristic proteinaceous hyaline casts. One form of dilated tubule, referred to as dilated altered tubule here, was typified by an irregular outline, a single layer of deeply staining cells of low columnar shape, and no association with overtly thickened basement membrane (Figure 7). The staining was usually basophilic, but could be eosinophilic, or a combination of both. At the point of convolution of the tubule, there was usually a suggestion of multilayering, due to tangential sectioning. Dilated altered tubules represented an unusual form of simple tubule hyperplasia and were also seen occasionally in moderate grades of CPN as well as in more advanced disease.

The second form of dilated tubule had a more expanded lumen, a more rounded profile, and was surrounded by very thickened basement membrane (Figures 8 and 9). Typically these dilated tubules were between 0.3 and 0.6 mm in diameter. Although tangential sectioning could suggest multilayering in these profiles (Figure 8), the lining often appeared to be genuinely more than one-cell thick, with protruberances of several cells projecting into the lumen (Figure 9). The cells comprising the lining were cuboidal in shape and identical to those lining adjacent and distant CPN-affected tubules. There was often degeneration and detachment of rounded cells into the dilated lumen, along with some minor cell debris.

Solid Tubule Profiles

Solid tubule profiles were always set in an area of expanded basement membrane merging into collagen, and assumed a solid, plaque-like appearance that had the appearance of complex proliferation (Figures 10 and 11). Some of these have been referred to as small tubule regeneration in previous publications (Hard et al., 1997; Hard, 2002). Their size ranged from less than 0.2 up to approximately 0.7 mm at their widest dimension. The constituent cells had scant or poorly developed cytoplasm that was often rarefied, or vaguely foamy (Figure 12). These profiles were typified by an overcrowding of uniformly bland nuclei, with unremarkable, or no nucleoli. They were invariably of irregular outline, lacking the impression of circumferential expansion. Accordingly, the distribution of connective tissue cells in the surrounding, thickened basement membrane was random and did not intimately encircle the profile (Figures 10 and 11).

The profiles often showed a distinct alignment of epithelial cells peripherally, with basally polarized nuclei. Sometimes the edge appeared to consist of tubular coils of epithelium, suggesting that the lining of these profiles had multiple infoldings, perhaps representing an attempt to generate new functional epithelium (Figure 11). Some of these profiles had a tendency to a loose form of central degeneration, or some hint of a lumen. In addition there was often a small amount of brown pigment (or pigment-laden macrophages) within the profile, typical of that seen in CPN-affected tubules. They were often accompanied by smaller, but otherwise similar tubule profiles within the same island of basement membrane/collagen, which appeared to be convolutions of the same tubule.

Preneoplastic Tubule Profiles Encountered in Advanced CPN

Lesions with morphology and pathogenesis that link to adenoma development found in association with advanced CPN conformed to atypical tubule hyperplasia (ATH) as defined by STP and IARC. These focal lesions, and criteria for distinguishing them from the CPN tubule profiles listed in the previous section, are described below. Criteria for ATH in advanced CPN are also summarized in Table 1.

Basophilic Atypical Tubule Hyperplasia

The most common form of ATH appeared as a single, solid tubule profile (Figure 13), or several adjacent, solid profiles representing convolutions of the same tubule (Figure 14). Distinguishing features from the complex CPN proliferations described above (intratubular cell clusters, dilated tubules, and solid tubule profiles) included an expansile nature, cellular detail, and the tinctorial quality of basophilia. The expansile appearance characterizing ATH was manifested by a tendency to a rounded shape, bulging into the surrounding extracellular matrix, where there was an associated complete or partial margination of attenuated fibroblasts and/or small capillaries encircling the lesion (Figures 13–15). This expansile character contrasted with the typically crimped and irregular periphery of the solid (plaque-like) profiles described above, and their lack of marginated connective tissue cells. The connective tissue cells encircling ATH were usually in intimate contact with the tubule profile rather than lying exterior to any surrounding basement membrane thickening. The cells in ATH tended to be larger than those in CPN-affected tubules, with expanded cytoplasm, enhanced delineation of cell borders, and nuclei with prominent, often enlarged, nucleoli (Figure 15). There was usually a tinctorial quality about the basophilia that was different from that of the proliferative profiles commonly associated with advanced CPN. This novel quality in ATH was best described as a “glassy” or hyaline sheen. Typically, ATH lacked pigment, and a central area of degeneration was a rare event. They were usually not set in an island of expanded basement membrane.

Small Eosinophilic Cell Variant of Atypical Tubule Hyperplasia

Another variant of ATH, which was observed very rarely in advanced CPN, occurred in the inner stripe of outer medulla (zone 3). It consisted of solid profiles of epithelial cells with small nuclei and indistinct but eosinophilic cytoplasm (Figure 16), not dissimilar in morphology to the cells of Henle’s loops.

Cystic Atypical Tubule Hyperplasia

Only three cases diagnosed in this series as ATH involved an expanded, cystic tubule in which a portion of the lining was markedly thickened (Figure 17). The area of hypercellularity showed a tendency towards the glassy basophilia and cytoplasmic appearance described above for ATH. In these few cases, the basement membrane along the multicellular thickening appeared to be less prominent (Figure 18). There was no morphological evidence to suggest that these hyperplastic lesions developed from the dilated tubule structures described under CPN tubule profiles.

Neoplastic Lesions Encountered in Advanced CPN

Adenoma

As defined in the STP Guidelines, adenomas were judged as having proliferated beyond the confines of a single tubule. In this survey of 7 carcinogenicity studies, the majority of adenomas observed in male rats with advanced to end-stage CPN were borderline or incipient (Figure 19), or small (Figure 20) basophilic tumors ranging in size from 0.4 to approximately 1.0 mm at their widest diameter. Borderline or incipient tumors were considered to have too many small lobules, usually 6 or more, to be consistent with the convolutions of a single tubule. Most adenomas retained the cellular character of ATH with well-developed cellular detail, including the special quality of basophilia (Figure 21). Adenomas also exhibited either microvascular or minimal fibroproliferative ingrowth. For the most part, they had an irregular growth pattern that was not confined to circumferential expansion (Figure 20). Most were solid lesions, but a few showed tubular differentiation, and an occasional example represented a cystopapillary pattern. Some contained a few small, mineralized concretions, or a solitary area of degeneration and necrosis.

In this review of carcinogenicity studies, two small adenomas associated with advanced CPN were located at least partially in the inner stripe of outer medulla, and consisted of solid tubule like profiles of cells with small nuclei and indistinct, eosinophilic cytoplasm (Figure 22). They therefore had a similar cellular morphology to the small eosinophilic variant of ATH illustrated in Figure 16, and were found in the same medullary region, suggesting that the small hyperplastic lesions could progress to neoplasia.

Discussion

In diagnosing renal tubule hyperplasia in 2-year carcinogenicity bioassays, the aim for hazard assessment purposes is to record those hyperplastic lesions that represent a precursor of renal tubule neoplasia, that is, specifically tubule hyperplasia characterized by atypia. An increased incidence of this lesion provides additional critical evidence concerning the carcinogenic potential for the kidney of the chemical under test. The STP guidelines (Hard et al., 1995) and IARC classification (Alden et al., 1992) distinguish between simple tubule hyperplasia and atypical tubule hyperplasia (ATH) in rodent kidney.

Simple tubule hyperplasia, defined as an increase in the number of epithelial cells that extends laterally, but not inwards beyond the single cell layer of the tubule lining, is considered to be a compensatory regenerative response to continued cellular injury and loss. ATH is comprised of proliferation beyond the single layer, to usually form a solid tubule profile. This lesion is designated as renal tubule hyperplasia in the NTP Technical Reports. Like simple tubule hyperplasia, ATH is most commonly basophilic, although eosinophilic, clear, chromophobic, or oncocytic forms of ATH occur infrequently in various settings (Hard, 1990). It is now widely accepted that ATH is a preneoplastic entity, representing an early and obligatory, developmental stage in a continuum with renal tubule adenoma and carcinoma (Hard, 1990; Dietrich and Swenberg, 1991; Noguiera et al., 1993; Hard et al., 1995).

The convention often recommended for diagnosing renal tubule hyperplasia in rodents, and for discriminating this precursor from adenoma, tends to place an emphasis on lesion size (e.g., Montgomery and Seely, 1990; Eustis et al., 1994: NTP, 1998b; Mitsumori et al., 2002). Thus, some reports regard tubular hyperplasias as consisting of tubules dilated from two up to, but less than, 5 times the size of a normal tubule, while adenomas are considered to be larger, usually 5 or more tubular diameters in size. This emphasis on size may result in misinterpretation of tubule profiles common in end-stage CPN that represent lesions of doubtful relevance to cancer, as well as sometimes over-diagnosing hyperplasias as adenomas.

For example, CPN-associated tubule dilation alone can cause an increase well beyond these dimensions but does not represent a precancerous lesion. It is extremely important to have clearly defined criteria for identifying ATH. For this lesion, the set of criteria recommended by the STP and IARC places the main emphasis on intratubular proliferation to produce a solid tubule where the proliferation does not breach the integrity of the tubule outline; for early adenoma, the difference stressed is that the proliferation has gone beyond the nephron integrity and the lesion is no longer identifiable as a single tubule structure. The current study has identified additional criteria that are considered useful for distinguishing ATH from common proliferative profiles in CPN, applicable at least to the basophilic phenotype of preneoplasia that can be found in advanced CPN.

These additional criteria include the expansile nature of the lesion, cellular detail, and staining properties. The expansile character, demonstrated by a rounded profile bulging into the surrounding matrix with a partial rim of marginating connective tissue cells, is considered an especially important aid in diagnosis. The findings in this survey show that the cells in ATH and early adenoma should not be described as “normal appearing” tubule epithelium, because of cytoplasmic and nuclear features. In particular, the “glassy” quality of the basophilia, which was usually present, and very likely resulting from both the increased volume and homogeneous texture of the cytoplasm, is almost pathognomonic. Nevertheless the identification of ATH should not rest on a single criterion, but a representation of several features.

In the simple tubule hyperplasia so characteristic of most of the tubules involved in CPN, the intense crowding of cells commonly leads to a focal piling up of neighboring cells, suggesting the appearance of complex proliferation into the lumen. The probability of encountering tangentially sectioned tubules in advanced to end-stage CPN is therefore very high because of the proliferative nature of CPN tubules, especially where the lining cells are piled up or crowded together. As a consequence, tangentially sectioned CPN tubules may be easily misinterpreted and confused with ATH in pathology assessments. In addition, it is apparent from the present study that some dilated or solid tubules comprising an integral part of CPN may have a lining that is more than a 1-cell layer. Despite this extension beyond the definition of simple tubule hyperplasia, these lesions appear to be unrelated to tumor development.

It is therefore our recommendation that the lesions referred to generically as CPN tubule profiles (the lesions depicted in Figures 1–12), should not be interpreted as ATH. They can be common proliferative entities in advanced CPN, and there is no morphological evidence to indicate that they are preneoplastic lesions committed to the developmental pathway resulting in adenoma. However, it was apparent from this and previous studies (Hard et al., 1997; Hard, 2002) that at least 2 of these CPN tubule profiles, those tubules appearing as a solid or dilated form, were sometimes more frequent and/or larger in studies where CPN had been exacerbated by chemical exposure, than in control rats with equivalent stages of advanced CPN.

When directly comparing the kidneys of a very large number of rats with advanced CPN, as was done in this investigation, it becomes evident that CPN can vary markedly in subtle aspects of its appearance from rat to rat in terms of staining intensity, amount of basement membrane thickening, and nature of the epithelial cells. Sometimes this variable appearance seems to occur when kidneys are infiltrated with mononuclear cell leukemia, as is often the case with F344 rats. In particular, the cells of CPN-affected tubules in some cases can assume a somewhat unusual appearance, with more intense basophilia, perhaps enlarged nuclei, and prominent nucleoli. In these cases especially, intratubular clusters of basophilic cells are common, and often mistaken for ATH. It is therefore important to compare any proliferative lesion in question with the general CPN tubule background. ATH should stand out as very different from the CPN tubules in the particular kidney being examined. If the suspected lesion is similar to the general appearance of the CPN-affected tubules, it is probably not ATH.

In conclusion, certain lesions have been identified in this study that constitute typical components of the proliferative response in advanced CPN. Although some of these possibly represent multilayering beyond a single layer, it is our recommendation that none should be diagnosed as ATH to avoid confusion with the true precursor of neoplasia. It is inevitable, however, that pathologists will encounter an occasional renal lesion that defies accurate diagnosis. In applying the guidelines described here, there were still rare lesions in this large survey that were problematic, falling into a gray area of subjective judgment.

Nevertheless, with awareness of the range of lesions that are typical of advanced CPN, and adherence to a set of criteria for discriminating between biologically unimportant proliferative lesions and precursors of tubule neoplasia, the number of these instances should be reduced to a minimum and not influence the hazard assessment and classification of any one chemical, particularly those that are associated with an exacerbation of CPN. Finally, in a position paper on the assessment of hyperplastic lesions in rodent carcinogenicity studies published by the STP, it was cautioned that hyperplasia in any organ or tissue must always be viewed within the context of the study itself (Boorman et al., 2003). In agreement with this particularly relevant advice, the present authors stress that the recommendations presented here for lesion interpretation in rat kidney apply in the context of advanced to end-stage CPN, but may not necessarily be applicable in situations where CPN is not a confounding factor.