Context.—Specimens from the prostate and bladder are commonly encountered by the general surgical pathologist. Emphasis is usually placed on neoplasms of the bladder and prostate, particularly if malignant, owing to their therapeutic consequences. A good command of benign lesions occurring in the bladder and prostate, and knowledge of their preneoplastic potential will help pathologists confidently diagnose malignancy versus its benign mimickers and guide the urologists in choosing the appropriate therapy and follow-up for the patient.
Objective.—To present a mixture of benign entities, and discuss their histologic and clinical characteristics, hoping to provide a practical review for the general surgical pathologist.
Data Sources.—An extensive review of the literature on the entities discussed was performed.
Conclusions.—A wide variety of benign entities are present in the prostate and bladder. Benign lesions in the prostate can be age related, such as prostatic atrophy and benign prostatic hyperplasia; transition zone associated, such as basal cell hyperplasia, adenosis, and sclerosing adenosis; or prostatic urethra associated. Benign lesions of the bladder encompass a wide variety of reactive changes that can occur in the urothelium, as well as hyperplastic lesions or reactive proliferations that could be misdiagnosed as malignant. The bladder responds to chronic irritation through several reactive/metaplastic lesions such as cystitis cystica/glandularis, keratinizing squamous metaplasia, or nephrogenic metaplasia. The urothelium can also give rise to hyperplastic/proliferative lesions, in particular von Brunn nest hyperplasia, papillary polypoid cystitis, and pseudocarcinomatous proliferation, which should be distinguished from malignant processes. Ectopic tissue, such as prostatic or mullerian, can also be seen.
Nonneoplastic Lesions Of The Prostate
The prostate gland is an accessory sex gland located underneath the bladder and surrounding the prostatic urethra. It is composed of glands and ducts arranged in a lobular fashion within a fibromuscular stroma. The glands and ducts are composed of secretory cells, basal cells, neuroendocrine cells, stem cells, and transit-amplifying cells.1 The surrounding stroma is composed of fibroblasts, myofibroblasts, and extracellular matrix. A variety of benign processes occur in the prostate; a selection of these processes and their salient features are discussed.
Prostatitis occurs in approximately 10% to 15% of men and is a costly, psychologically and physically debilitating disease.2 Prostatitis is classified into 4 categories: acute bacterial (National Institutes of Health [NIH] type I); chronic bacterial (NIH type II); chronic prostatitis/chronic pelvic pain syndrome (NIH type III), which is divided into inflammatory (type IIIA) and noninflammatory (type IIIB); and finally, asymptomatic (NIH type IV).3 The most common type of prostatitis encountered by the surgical pathologist is asymptomatic prostatitis or NIH type IV. Patients who undergo needle core biopsies of the prostate for elevated prostate-specific antigen (PSA) levels show evidence of prostatitis on histology in approximately 27% of the cases.4 Most have chronic inflammation (94%), followed by acute inflammation (6%), and rarely, granulomatous inflammation (0.2%). Studies describing inflammation in the prostate have subclassified it by inflammatory cell type, pattern and location of inflammation (glandular, periglandular, and stromal), and extent/grade. There is no correlation between any type, pattern, or grade of inflammation seen on needle core biopsy and bacterial cultures or PSA level.5 Any extent of acute or granulomatous inflammation that is present on the prostatic biopsy specimen should be mentioned; however, chronic inflammation is mentioned only when it is prominent. In the setting of extensive chronic inflammation, particularly if lymphocytic, a low-grade lymphoma involving the prostate should be ruled out. Chronic inflammation of the prostate has been invoked as a predisposing factor in the pathogenesis of prostatic adenocarcinoma, and future studies are necessary to further elucidate this association.6
Granulomatous inflammation can be seen in the prostate in less than 1% of needle core biopsies.7 The most common form of granulomatous prostatitis consists of “nonspecific granulomatous prostatitis” and is characterized by granulomatous inflammation arranged concentrically around prostatic ducts or glands (Figure 1), and can be accompanied by giant cells and a mixture of inflammatory cells including eosinophils. It is often accompanied by acute inflammation in the remaining prostatic tissue. Other causes of granulomatous prostatitis include infectious, postbiopsy, and systemic granulomatous prostatitis. Infectious prostatitis commonly occurs after Bacillus Calmette-Guerin (BCG) instillations for urothelial carcinoma of the bladder. It is characterized by well-formed granulomas, with or without caseating necrosis, and/or sheets of histiocytes. In the setting of a patient with known history of BCG therapy, it is not necessary to perform special staining, since treatment depends on the presence of symptoms or systemic disease.8 In addition to BCG granulomas, a whole array of infectious agents have been reported in the prostate, including other mycobacterial microorganisms,9 coccidioidomycosis,10 and blastomycosis,11 among others. Systemic granulomatous diseases are also very rare causes of granulomatous prostatitis and include asthma-related allergic conditions, sarcoidosis,12 and Wegener granulomatosis.13
Prostatic Gland Atrophy
Atrophy of the prostatic gland is subclassified as focal or diffuse. Diffuse epithelial atrophy usually refers to a global atrophic phenomenon involving the entire prostate and attributed to hormonal therapy for prostate cancer. It involves the prostatic gland globally and is characterized by prominence of the basal cell layer and a decrease in cellular size and cytoplasm of the luminal cells. Focal atrophy refers to a patchy phenomenon that is present normally within the prostate and increases with age. Although the cause of atrophy is unknown, several hypotheses have been put forward, such as ischemia, inflammation, hormones, and compressive effects of benign prostatic hyperplasia. Atrophy has been recently classified into 4 subtypes: simple atrophy, postatrophic hyperplasia, simple atrophy with cyst formation, and partial atrophy.14
Simple atrophy is easily detectable on low magnification owing to its basophilic appearance imparted by the scant cytoplasm of the low cuboidal to flattened cells populating the glands. Glands of simple atrophy lack the papillations classically seen in benign prostatic glands and often have an irregular or angulated shape. Architecturally, the spatial arrangement and glandular size are similar to normal prostatic glands. The glands can be cystically dilated but they are not distended enough to form a round shape. The surrounding stroma can be fibrotic, and it is usually associated with chronic inflammation.
Postatrophic hyperplasia is also referred to as nodular hyperplasia or lobular atrophy. Similar to simple atrophy, they all have scant cytoplasm and lack the normal papillations of benign prostatic glands. Architecturally, it is composed of closely packed small glands arranged in a lobular architecture. The glands are mostly round and lined by cuboidal cells, which can have a prominent nucleolus. Often a dilated feeder duct is seen at the center of the lobule. The surrounding stroma can be fibrotic and it is usually associated with chronic inflammation.
The hallmark of simple atrophy with cyst formation is the presence of cyst-like glands arranged in a back-to-back configuration. The glands should have a round configuration and not just be distended. In addition, attenuated flattened lining cells with scant cytoplasm are also present. The cystic glands can be small or large (>1 mm). The stroma is usually not fibrotic and no associated inflammation is present.
Partial atrophy is characterized by more cytoplasmic volume than the other atrophies. The cytoplasm is usually clear and the cytoplasmic membrane is prominent (Figure 2). Moderately prominent nucleoli can be present. The glands are usually small to medium sized and can be angulated. No particular architectural pattern is consistently present. However, in large enough foci, a transition to complete atrophy is seen. The stroma is usually not fibrotic and no associated inflammation is present.
In most patients, a mixture of the several patterns of atrophy can be seen, rather than a pure form or another. In addition, the distinction between postatrophic hyperplasia and partial atrophy can be challenging at times, pushing some authors to lump these 2 categories together.15
Proliferative inflammatory atrophy of the prostate is defined as atrophy associated with an inflammatory infiltrate and a high proliferative rate. Included in this category are simple atrophy and postatrophic hyperplasia.14 Proliferative inflammatory atrophy of the prostate has been proposed as a potential precursor of prostatic adenocarcinoma; however, studies looking at screening prostate biopsy specimens from asymptomatic men did not find an association between the presence of atrophy and the subsequent development of cancer.16 Therefore, the diagnosis of atrophy on needle core biopsy does not have to be reported, until further studies prove otherwise.
Usually, atrophy is easily diagnosed on routine hematoxylin-eosin (H&E)-stained sections. However, in some instances, foci with features worrisome for carcinoma can be seen. The most challenging distinction is with the atrophic subtype of prostatic adenocarcinoma. Thorough histologic examination, accompanied by immunohistochemical stains for basal cell markers, is usually helpful to make that distinction. An infiltrative growth pattern, presence of atypical nuclear and nucleolar features, and lack of basal cells by immunohistochemistry are features that support a carcinoma diagnosis. It is important to remember that a patchy basal cell layer can be present in atrophic glands, and sometimes entire single glands are devoid of basal cells. In addition, racemase or α-methylacyl-CoA racemase (AMACR), can be positive, potentially misleading the pathologist into making a diagnosis of atypical small acinar proliferation. In these instances, it is important to evaluate groups of similar-appearing glands as a unit, rather than as separate individual glands. When the absence of basal cells is worrisome, we have found it helpful to examine basal cell markers separately on several levels.
Benign Prostatic Hyperplasia
The normal prostate weighs up to 20.0 ± 6.0 g. The weight of the prostatic gland increases with age owing to the development of benign prostatic hyperplasia (BPH), averaging 33.0 ± 16.0 g, but reaching up to 100.0 g in a small percentage of men older than 70 years.17 The incidence of BPH increases with age and is measured at 70% at 70 years of age. Some believe that the process of BPH is initiated before the patient is 30 years old. Histologically, BPH arises predominantly in the transition zone and consists of nodules composed of a mixture of glandular and stromal elements (Figure 3), which vary in percentage to include strictly stromal nodules. Benign prostatic hyperplasia glands show histologic appearances ranging from complex glands with luminal undulations and papillary fronds to small and crowded glands. Cystic dilatation of the glands can occur. Benign prostatic hyperplasia has been reported to arise less commonly in the posterior peripheral zone adjacent to the boundary with the transition zone.18 The etiology of BPH is thought to be multifactorial and includes environmental and genetic factors. Recently, chronic inflammation of the prostate is thought to play a role in the genesis of BPH, and future studies to clarify this role and perhaps target it therapeutically are in the works.19,20 Clinically, not all patients with BPH are symptomatic or have lower–urinary tract obstructive symptoms. The International Prostate Symptom Score is used clinically to quantify the symptoms of BPH.
The diagnosis of BPH from needle core biopsies of the prostate is discouraged, since no correlation was found between the histologic findings on needle core biopsy and clinical or symptomatic BPH.21 Stromal nodules can be confidently identified on needle core biopsies; however, they are not unique to BPH, since they have been identified in small-sized prostates (15.0 g).21
Basal Cell Hyperplasia
Basal cells are a normal constituent of benign prostatic glands, located in the outermost layer, and are thought to harbor self-renewable stem cells.22 Immunohistochemically, they stain for p63 and high-molecular-weight cytokeratin (HMWCK), which is helpful diagnostically and serves to differentiate benign glands and high-grade prostatic intraepithelial neoplasia from prostatic adenocarcinoma. Basal cells show zonal variations; in the peripheral zone, they tend to be triangular and express HMWCK and p63, whereas in the transition zone, they are flattened or low cuboidal and express p63 and variable HMWCK (weak or absent).22
Basal cell hyperplasia occurs in up to 10% of prostates.23 It is divided into usual basal cell hyperplasia, atypical basal cell hyperplasia, atrophy-associated basal cell hyperplasia, and adenoid cyst-like hyperplasia.
Usual basal cell hyperplasia is defined as the presence of a basal cell proliferation composed of 2 or more cell layers of small cells with scant cytoplasm present as glands or solid nests24 (Figure 4). Since it is usually a component of benign prostatic hyperplasia,24 it is normally present in the transition zone. However, basal cell hyperplasia can be seen in the peripheral zone and is most often associated with chronic inflammation.23 Rare atypical features can be encountered, such as perineural involvement, prominent nucleoli, dense intraluminal secretions, luminal blue mucin, crystalloids,25 and individual cell necrosis.23 These features are problematic on needle core biopsies, where the lobular architecture is not always obvious. A combination of basal cell markers including p63 is helpful in ruling out carcinoma.
Atrophy-associated basal cell hyperplasia is seen in prostates post hormonal therapy, radiation, and/or cryosurgery.
Adenoid cysticlike basal cell hyperplasia is a rare form of basal cell hyperplasia.28 Most patients have concomitant usual basal cell hyperplasia. Architecturally, focal glandular anastomosis and cribriform structures with small luminal spaces, with or without collagenous spherules, are seen. It can be multifocal, but has a noninfiltrative lobular architecture. Residual luminal cells are sometimes present. Focal metaplastic change can be seen, predominantly of the squamous type. However, sebaceous and mucinous metaplasia can rarely occur. They are distinguished from adenoid cystic carcinoma of the prostate by their lack of infiltrative growth pattern, and lack of extensive cribriform architecture and glandular anastomosis. In addition, perineural and extraprostatic invasion should not be present. They usually stain for basal cell markers such as HMWCK. Of note, while adenoid cyst-like basal cell hyperplasia presents in older patients, adenoid cystic carcinoma presents in a relatively younger age group.
Clear Cell Cribriform Hyperplasia
Clear cell cribriform hyperplasia is associated with benign prostatic hyperplasia and consists of enlarged glands filled with anastomosing clear cells forming cribriform and/or papillary structures.29 The cytology is usually bland with inconspicuous nucleoli and absence of mitoses. A basal layer is present and demonstrable by immunohistochemistry.30 It can be a pitfall for high-grade cribriform prostatic adenocarcinoma; however, the bland cytology and presence of a basal cell layer are helpful in this distinction. In addition, clear cell cribriform hyperplasia usually arises in the transition zone, whereas prostatic adenocarcinoma more commonly arises in the peripheral zone.
Adenosis is also referred to as atypical adenomatous hyperplasia, atypical adenosis, small acinar atypical hyperplasia, and small gland hyperplasia. This lesion is currently considered benign and the term adenosis is recommended since several studies have shown no definite link to prostatic adenocarcinoma. Adenosis refers to a relatively well-circumscribed nodule of closely packed prostatic glands with clear cytoplasm predominantly arising in the transition zone (Figure 5). Glands range from small round to larger with papillary infoldings. Prominent nucleoli, crystalloids, and luminal blue mucin can be present. Basal cells are not always readily seen on routine H&E and can be very patchy on immunohistochemical stains for basal cells. In addition, racemase/AMACR/p504S staining is positive in this lesion. Features that distinguish adenosis from prostatic adenocarcinoma are the infiltrative pattern of adenocarcinoma, particularly at the periphery of the nodule; the amphophilic quality of the carcinoma cytoplasm; and the atypical nuclear features of adenocarcinoma.31 In addition, another helpful hint lies in the fact that the nuclear/cytologic features of adenosis are usually similar to the adjacent benign glands. However, in certain instances, accurate distinction on routine H&E from low-grade prostatic adenocarcinoma is not possible and immunohistochemical stains for basal cell markers are recommended.
Sclerosing adenosis is a rare benign lesion arising in the transition zone of the prostate, characterized by a cellular proliferation of variably sized glands and solid nests surrounded by basement membrane material in a cellular stroma, which can be myxoid at times. The contours of the lesion can be well circumscribed or infiltrative, raising concern for carcinoma. Immunohistochemically32 and ultrastructurally,33 a basal cell layer with myoepithelial differentiation is demonstrable and is positive for S100, HHF-35, and HMWCK. Sclerosing adenosis can show minimal cytologic atypia; however, rare cases labeled atypical sclerosing adenosis demonstrate nuclear atypia, hyperchromasia, pleomorphism, or prominent nucleoli.34 These lesions have a similar benign course as classical sclerosing adenosis and the distinction serves to raise the awareness of the pathologist to the cytologic atypia that can be encountered in this lesion.
Prostatic Gland Metaplasia
A variety of metaplastic changes can be seen in the prostate, including transitional cell metaplasia, mucinous metaplasia, squamous metaplasia, and eosinophilic metaplasia. Metaplasia is encountered in 0.6% of prostates; squamous and transitional metaplasias are among the most common forms encountered.35 Squamous metaplasia occurs in normal prostatic glands, as well as in a variety of reactive settings such as adjacent to prostatic infarcts36 and in hormonally treated prostates,37 including estrogen therapy.38 Although squamous metaplasia can show reactive atypia, particularly if present adjacent to prostatic infarcts, it usually lacks overt “malignant” atypia and should be distinguished from squamous differentiation of residual or recurrent prostatic adenocarcinoma, a finding that can be also encountered post radiation or hormonal therapy.
Transitional metaplasia can be a pitfall for high-grade prostatic intraepithelial neoplasia and is distinguished by the presence of nuclear elongation and grooves, as well as inconspicuous nucleoli. Similarly to squamous metaplasia, cytologic atypia is usually absent, and when present, raises the possibility of extension of urothelial carcinoma within prostatic glands and ducts. In this setting, a previous or concomitant history of urothelial carcinoma is helpful.
It is also important to recognize mucinous metaplasia, so as not to confuse it with intraluminal mucin, associated with prostatic adenocarcinoma and other benign lesions.
Verumontanum Mucosal Gland Hyperplasia
Verumontanum mucosal gland hyperplasia is a benign lesion, present in 14% to 29% of prostates, and can be seen in the urethra at the verumontanum, ejaculatory ducts, prostatic ducts, and utricle.39,40 It is a relatively well-circumscribed nodule of closely packed microacinar prostatic glands with both basal and luminal cells. Occasionally, large, more complex glands can also be present. They can contain intraluminal rust-colored concretions resembling corpora amylacea.41 Distinction of this lesion from a low-grade prostatic adenocarcinoma can be easily performed through immunohistochemical stains for basal cells (p63 and HMWCK).
Prostatic Urethral Lesions
The prostatic urethra is lined by urothelial mucosa and can be involved in any of the processes affecting the urothelial mucosa of the bladder, discussed subsequently. However 2 types of polypoid lesions are worthy of mention: the prostatic urethral polyp and the fibroepithelial polyp.
Prostatic urethral polyps are papillary/polypoid projections lined by normal prostatic epithelium, but can be a mixture of urothelial and prostatic epithelium. The epithelium is underlined by a fibrovascular stroma that can contain prostatic glands. They are benign lesions and several theories around their formation have been proposed, including a reactive process and persistence of evaginations normally occurring in embryonic development.
Fibroepithelial polyps are rare benign lesions usually presenting in childhood and are thought to be predominantly congenital.42 They have also been described to present in adulthood43 and are more common in males. They can be seen in the renal pelvicaliceal system, bladder, or urethra. They usually do not recur after transurethral resection. Three histologic patterns are recognized,43 the most common of which is a cloverleaf-shaped polypoid mass with secondary changes of cystitis cystica/glandularis. Back-to-back glands and anastomosing nests can be present; however, they are usually focal. In contrast to inverted papillomas, most nests are round. Atypical reactive stromal cells can be seen and can provide the clue to the right diagnosis. The second pattern consists of numerous, small, rounded, densely fibrotic fibrovascular cores, as opposed to the thin and loose cores present in papilloma. Areas of glandular differentiation can be present. The last pattern consists of a polypoid projection with secondary, usually broad-based, edematous, fingerlike papillary structures. This last pattern shares overlapping features with papillary/polypoid cystitis; however, in papillary polypoid cystitis, the polypoid and papillary projections arise from the urothelial lining, whereas in fibroepithelial polyps, they arise from the main polypoid mass. Cystoscopically, fibroepithelial polyp is thought to represent a polypoid mass, whereas papillary polypoid cystitis is thought to be an inflammatory lesion.
Mesonephric remnants are mesonephric duct vestiges that can be identified histologically in both the male and female genital tract.44–47 In males, mesonephric remnants have been described in the prostate, prostatic urethra, and seminal vesicle.46 They are rare benign mimickers of prostatic adenocarcinoma. Three most common architectural patterns are seen in mesonephric remnant hyperplasia and include (1) small acini arranged in a lobular appearance with characteristic colloidlike eosinophilic secretions; (2) ill-formed glands and cords of cells, which could either be lobulated or have an infiltrating pattern; and (3) acini with cystic dilatation (Figure 6). Less common patterns include papillary or micropapillary tufting and a spindle cell component. Mild cytologic atypia including prominent nucleoli can be present; however, severe cytologic atypia should raise the possibility of a malignant process. Architecturally, mesonephric remnants can be present within extraprostatic tissue or associated with nerves, raising the possibility of carcinoma.48 Once the possibility of mesonephric remnant hyperplasia is raised, the distinction can be easily made with an immunohistochemical panel that includes PSA and PAX8. They are usually positive for PAX8 and negative for PSA, in contrast to prostatic adenocarcinoma, which is positive for PSA and negative for PAX8. Mesonephric remnant hyperplasia can be positive for HMWCK and/or racemase, which is particularly treacherous if racemase results are positive and HMWCK results are negative, resembling the staining pattern of prostatic adenocarcinoma.
Benign Glands With Perineural Involvement
Benign glands can be closely associated with nerves in several patterns: indenting the nerves, tracking along the nerve, encircling the nerve, and intraneural49 (Figure 7). Distinction between benign gland perineural involvement and perineural invasion by prostatic adenocarcinoma can be resolved by performing immunohistochemical stains for basal cell markers and comparing the glands in question to the surrounding benign glands. It is important to note that in rare occasions, where benign glands with partial atrophy are involving perineural spaces, basal cells could be absent and the diagnosis rests on histologic comparison of the glands in question with the surrounding benign glands.
Nonneoplastic Lesions Of The Bladder
The bladder, renal pelvicalyceal system, ureters, and the proximal urethra are lined by urothelial mucosa, formerly called transitional mucosa. Similar processes that can involve the urothelial mucosa in the bladder can be seen in the other organs. The urothelium in the bladder is 5 to 7 cell layers thick, which could decrease to 2 to 3 cell layers when the bladder is distended. It is composed of 3 types of cells: (1) the basal cells, which are cylindrical to flat and in contact with the underlying fibroconnective tissue or lamina propria—the basal cell layer is thought to harbor urothelial stem cells; (2) the intermediate cells, which are cuboidal to columnar cells, overlie the basal cells, and usually have the same orientation toward the bladder lumen; and (3) the umbrella cells, which are larger flat eosinophilic cells, with ample pink cytoplasm and variable nuclear reactive/degenerative atypia such as binucleation, smudging, and nucleoli. A variety of lesions and metaplastic phenomena are seen in the bladder and urothelium and will be the subject of discussion of the rest of the article.
The urothelium can demonstrate reactive changes due to a variety of causes including catheterization, infection, and urinary calculi. These processes are usually accompanied by inflammation, edema, and congestion. The normal pattern of maturation is usually seen in nondenuded reactive mucosa. Mucosal polarity is for the most part preserved. Nuclei can be slightly enlarged, although nuclear pleomorphism is unusual. A relatively smooth nuclear contour and evenly distributed fine chromatin are features that are still preserved. Nucleoli can be prominent, as in most reactive conditions. Mitoses, when present, are confined to the lower layers and are not atypical. Overt cytologic atypia, including nuclear hyperchromasia and pleomorphism, should prompt exclusion of a neoplastic process. Immunohistochemical stains for CK20, CD44, and p53 show a distinct pattern of staining in normal or reactive urothelium.50 The normal urothelium shows CK20 positivity confined to the umbrella cell layer, CD44 positivity in the basal cell layer, and a weak and patchy p53 nuclear staining. On the other hand, urothelial carcinoma in situ is positive for CK20 and p53 (in a strong and diffuse fashion) and negative for CD44. Reactive urothelium usually shows positive CD44 staining, CK20 staining confined to the umbrella cell layer, and weak and patchy p53 nuclear staining. However, it is important to remember that the use of these stains should be supplemental to the histologic impression and not the determining factor.
Urothelial Atypia Of Unknown Significance
In rare cases, the urothelial atypia is such that it goes beyond what one would confidently accept for a reactive urothelium, particularly in the presence of a disproportionate inflammatory infiltrate, with respect to the level of atypia seen; however, the atypia is not severe enough to warrant a preneoplastic/neoplastic diagnosis of “dysplasia”/flat urothelial carcinoma in situ. Immunohistochemical staining for CK20, CD44, and p53 can be performed and can be helpful. A clinical history favoring reactive atypia, if present, can be informative. However, in rare cases, it is difficult to be reasonably confident either way, and we resort to using this descriptive World Health Organization (WHO)/International Society of Urological Pathology category to relay to the urologist our uncertainty and prompt adequate follow-up. The frequent use of this term is discouraged and it should not be used as a wastebasket for difficult cases. We use this term rarely, often in posttreatment patients with a history of urothelial carcinoma, with biopsy specimens showing inflammation and mucosal atypia. In this setting, distinction between severe posttreatment reactive atypia and residual in situ carcinoma can be difficult. Comparison of the current atypical urothelial lining to the patient's previously diagnosed carcinoma in situ, when available, might also be helpful in this scenario.
Cystitis can be broadly subcategorized into infectious and noninfectious. Infectious cystitis includes bacterial, viral, and rarely, fungal etiologies. Viral cystitis can be suspected by the presence of viral inclusions in an inflammatory background on H&E and confirmed with positive immunohistochemical staining for the corresponding viruses, such as cytomegalovirus (CMV) (Figure 8), adenovirus, BK virus (Figure 9), and Herpes simplex. Bacterial and fungal cystitis are characterized by a mixed inflammatory infiltrate and presence of microorganisms on culture or special stains. When microorganisms are not seen, the histologic features are not specific; rare forms of bacterial cystitis, such as encrusted cystitis and emphysematous cystitis, have identifiable histologic appearance.
Encrusted cystitis is caused by bacterial infections associated with alkaline urine in a predisposed immunocompromised patient. Corynebacterium is frequently involved.51 Histologically, it is characterized by a necrotic surface containing encrusted calcifications and underlined by necrotic tissue with inflammatory cells and bacterial colonies.
Emphysematous cystitis is a complication of urinary tract infection, characterized by the presence of air within the bladder wall, caused by gas-forming bacteria. It occurs more commonly in women, and particularly in immunocompromised patients, including those suffering from diabetes mellitus.52 It is usually suspected clinically or radiologically. Histologically, it shows a mixed inflammatory infiltrate, as well as clear air spaces in the lamina propria, which can be surrounded by a foreign body giant cell reaction.
Noninfectious cystitis includes interstitial cystitis, eosinophilic cystitis, drug-induced cystitis, and postsurgical necrobiotic granulomas.
Interstitial cystitis presents with pelvic pain and urinary symptoms with a female predilection. It is characterized cystoscopically by ulceration and linear fissures of the urothelial mucosa. In patients without ulceration or fissures, dotlike small hemorrhages, which bleed on filling of the bladder, are present.53 The mainstay for the diagnosis of interstitial cystitis rests on clinical criteria, since the histologic findings associated with this entity are not specific.54 A cystoscopic biopsy is performed primarily to rule out other possible confounding etiologies such as carcinoma in situ. The spectrum of histologic findings seen on biopsy specimens from patients with interstitial cystitis ranges from histologically normal to abnormal. Some of the histologic features seen in abnormal biopsy specimens include ulceration, granulation tissue, a mononuclear inflammatory infiltrate, and perineural inflammation. In addition, mast cells can be seen in the mucosa or muscularis propria.
Eosinophilic cystitis is a rare disease that occurs more frequently in males and can affect any age group. It can be seen associated with allergies, drug reactions, parasitic infections, and other causes. It is characterized by the presence of a dense inflammatory infiltrate composed primarily of eosinophils involving the bladder wall. The eosinophilia can be accompanied by muscle necrosis and edema.
Postsurgical granulomas, similar to their counterpart in the prostate, can also be seen. They can be linear and palisading. They are surrounded by a foreign body giant cell reaction and a mixed inflammatory background.55
The possibility that chronic inflammation could have a role in the pathogenesis of urothelial carcinoma has been suggested, especially as this association is relatively well established in the pathogenesis of squamous cell carcinoma of the bladder subsequent to schistosomiasis infection as a predisposing factor. However, its role in the development of urothelial carcinoma remains hypothetical and further studies should be performed to elucidate this relationship, if any.
Malakoplakia is thought to be an inflammatory mass-forming process, composed of sheets of granular pink histiocytes (von Hansemann cells), which contain distinctive Michaelis-Gutmann bodies. It occurs more frequently in women and commonly occurs in the genitourinary tract, particularly the bladder.56 Michaelis-Gutmann bodies are small cytoplasmic inclusions measuring approximately 5 to 10 µm in size,56 which appear targetlike and mostly contain calcium hydroxyapatite. Clinically, malakoplakia occurs in association with Escherichia coli infection or in immunocompromised states.57 Cystoscopically, it is a yellow plaquelike lesion with hyperemia around the bases. Malakoplakia of the bladder is considered a self-limiting process and the treatment is primarily conservative, in contrast to a more aggressive surgical approach in the upper genitourinary tract. Extension into the adjacent pelvic soft tissue has been rarely reported.
Urothelial hyperplasia is defined as the presence of a markedly thickened urothelium (usually more than 10 cell layers thick) with normal cytology. The normal constituents of the urothelium are present, including basal cells, intermediate cells, and umbrella cells. Cytologic atypia is absent and mitotic activity, when present, is confined to the lower layers and should not be atypical. Nucleoli are usually inconspicuous.
Hyperplasia has been divided into flat and papillary urothelial hyperplasia. In addition, flat urothelial hyperplasia can be divided into isolated flat urothelial hyperplasia and flat hyperplasia accompanying papillary neoplasms. A separate discussion of these entities is warranted.
Isolated flat urothelial hyperplasia is thought to be a variant of normal/reactive conditions and is not considered a malignant precursor. Reactive changes involving the adjacent urothelial mucosa can also accompany urothelial hyperplasia, and a slight deviation of the classical normal urothelium should not be worrisome. Tangential sectioning mimicking urothelial hyperplasia should be ruled out. In addition, close examination for fibrovascular cores, to rule out papillary neoplasms, such as papillary urothelial neoplasm of low malignant potential, is necessary. This becomes more problematic in the distinction with papillary urothelial hyperplasia.
Papillary urothelial hyperplasia is considered a precursor of urothelial neoplasms and is defined as undulating urothelium with narrow papillary folds.58 Similar to other forms of hyperplasia, cytologic atypia should be absent and preservation of organization should be maintained. It is not a term without controversy and some prefer to label this lesion as neoplastic in lieu of hyperplastic, since a study on papillary hyperplasia has found genetic alterations, predominantly on chromosome 9, in approximately half of the cases.59 We have made this diagnosis rarely and mostly in patients with previous history of papillary neoplasms with follow-up biopsy specimens exhibiting a thickened, normal-looking urothelium with slim minute fibrovascular protuberances, raising the question of a papillary neoplasm, but falling short of this diagnosis owing to lack of branching or secondary budding. Deeper levels on such lesions are helpful in ruling out the possibility of papillary neoplasms. If such lesions were to be encountered in patients being worked up for hematuria and who lack a history of papillary neoplasms, they should warrant close follow-up.
Flat urothelial hyperplasia associated with “low-grade” papillary neoplasms has been recognized and molecular studies performed on these have found genetic alterations involving chromosome 9,60,61 raising the possibility of its preneoplastic potential. In our practice, we do not report urothelial hyperplasia adjacent to papillary neoplasms, since the prognosis and follow-up of the patient is dictated by the papillary neoplasm. In addition, we believe that papillary neoplasms can extend and involve the adjacent flat urothelium, also known as the “shoulder” of the papillary neoplasm. In the setting of papillary urothelial neoplasms of low malignant potential and even the lower spectrum of low-grade papillary urothelial carcinoma, the shoulder of these neoplasms can be histologically similar to urothelial hyperplasia. In addition, the same genetic changes seen in hyperplasia associated with papillary neoplasms were also seen in the normal adjacent urothelium in these cases,61 raising the possibility that these genetic alterations are part of the genetic disturbances leading to the neoplastic process rather than changes that are particular to this entity. In light of this, we believe more studies are warranted to further elucidate this entity.
Von Brunn Nest Hyperplasia
von Brunn nests are present in the normal urothelium and are thought to arise from invaginations of the urothelial mucosa into the underlying lamina propria. The prevalence of von Brunn nests was reported as 89% in a postmortem examination of 100 normal bladders.62 They usually lack cytologic atypia; however, all processes occurring in the normal urothelium can involve the underlying von Brunn nests, including reactive atypia, hyperplasia, dysplasia, and carcinoma in situ. They are well recognized among pathologists and most often do not cause concern. However, they can be problematic when they are hyperplastic, composed of predominantly solid nests, or are deep seated in the lamina propria. When hyperplastic, they could form a mass-like lesion on cystoscopic examination. However, keeping in mind a few of their histologic characteristics can help in proper identification and distinction from the nested variant of urothelial carcinoma, which is a high-grade carcinoma with a deceptively bland cytologic appearance. Hyperplastic von Brunn nests show well-rounded contours and variability in size, including the presence of larger nests.63 They often show regular spacing between nests and can show changes of cystitis cystica/glandularis, defined as central lumen formation with possible cystic dilatation and apical columnar cells overlying the urothelium (Figure 10). When cytologic atypia is present, it is within the range of reactive atypia, unless there is a superimposed dysplastic/in situ neoplastic process. Architecturally, the nests lack an infiltrative growth pattern, maintain a uniform depth of involvement, and often have a symmetrical shape. They should not involve the muscularis propria when it is sampled in the biopsy specimens. The presence of an accompanying tangential capillary network, hugging the basement membrane, similar to that seen underneath normal urothelial mucosa, is an additional clue in their distinction from the nested variant of invasive urothelial carcinoma. Immunohistochemical stains in the differential diagnosis with nested variant of urothelial carcinoma, including MIB1, CK20, p27, and p53, are not useful and the diagnosis relies on routine histology.63
Cystitis Cystica/cystitis Glandularis
Cystitis cystica is a benign urothelial lesion characterized by the presence of cystically dilated von Brunn nests. The prevalence of cystitis cystica has been reported at 60% in normal bladders on autopsy.62 When these nests have an overlying glandular (cuboidal or columnar) lining, they are referred to as cystitis cystica et glandularis or cystitis glandularis. Cystitis cystica, cystitis glandularis, and von Brunn nest hyperplasia often occur together and are thought to represent a reactive process. Their presence in specimens from patients with bladder carcinoma is thought to be coincidental as long-term follow-up of cystitis cystica and glandularis, without associated carcinoma, did not portend an increased risk of developing carcinoma.64 They occur in male and female adults but can also occur in childhood.65 Cystitis glandularis can have intestinal metaplasia, with presence of columnar cells and interspersed goblet cells that resemble intestinal mucosa; it is then referred to as cystitis glandularis of intestinal type. Paneth cells can be rarely seen and the intestinal metaplasia stains similarly to colonic-type mucosa (CK20+, CDX2+). We generally use the term intestinal metaplasia instead of cystitis glandularis of intestinal type, to refer to lesions with extensive intestinal metaplasia with a minimal cystitis glandularis component or exclusive intestinal metaplasia.
These lesions are rarely a diagnostic challenge in their classical forms; however, cystitis glandularis of intestinal type/intestinal metaplasia can become worrisome in the setting of mucin extravasation, raising the possibility of colloid adenocarcinoma (Figure 11). The distinction can be tricky, and it is good to keep in mind that lack of cellular atypia, absence of epithelial nests floating in the mucin, and lack of involvement of the muscularis propria support benignity.
The possibility that intestinal metaplasia is a putative precursor lesion to adenocarcinoma has been raised owing to the finding of significant telomere shortening in intestinal metaplasia with or without associated cystitis glandularis.66 However, clinicopathologic correlation studies do not show an increased risk of developing cancer in patients with long-term follow-up of intestinal metaplasia.64 Further studies to clarify the role of intestinal metaplasia in the development of carcinoma are warranted.
Nonkeratinizing squamous metaplasia, “vaginal subtype,” is considered a normal variant of the urothelial mucosa and is predominantly seen in women in the trigone and bladder neck. It has been reported in a smaller percentage in men (7%).62 Keratinizing squamous metaplasia, however, is rare, occurs in both males and females, and is thought to arise from chronic irritation such as indwelling catheters or recurrent infections.67 It can be limited (<50% of mucosal surface involvement), or extensive (>50%). It may be associated with concurrent squamous cell carcinoma or subsequent development of squamous cell carcinoma,68,69 such that complete conservative removal and close follow-up are necessary. The distinction between nonkeratinizing squamous metaplasia, a normal variant of the urothelial mucosa, and keratinizing squamous metaplasia is therefore important on biopsy material, since it changes the course of follow-up and management. Similarly, the verrucous form of squamous metaplasia or verrucous squamous hyperplasia (Figure 12) is thought to be associated with increased risk of concurrent squamous cell carcinoma, necessitating proper follow-up.67
Papillary Polypoid Cystitis
Papillary polypoid cystitis is a reactive benign urothelial lesion with polypoid to papillary architecture that is often confused with low-grade papillary neoplasms.70 It is usually a mixture of simple large edematous polypoid fronds and interspersed thin papillae. With time, the edematous bulbous fronds can become densely fibrotic. The polypoid/papillary structures are lined by normal, hyperplastic, or reactive urothelium and do not usually show secondary branching. The bases of the fronds can be narrow or broad but are often a mixture of both (Figure 13). Mild inflammation, as well as other reactive conditions, such as cystitis cystica, can accompany the lesion. When in doubt, a clinical history can usually illustrate an irritative cause and the cystoscopic impression is usually that of an inflammatory lesion, rather than a neoplasm.
Nephrogenic metaplasia/adenoma is a benign lesion, which occurs in adults and children with a slight male predilection. A history of previous instrumentation, surgery, or chronic insult, such as calculi and cystitis, is usually present. In addition, it is relatively common in renal transplant patients. Cystoscopically, it presents as a polypoid mass or a flat lesion, which can be single or multifocal.71 Histologically, several patterns have been described, the most common of which is a tubular architecture.72 The tubules are lined by cuboidal to low columnar cells, occasionally hobnailing, with little cytoplasm (Figure 14). The cytoplasm is pink and occasionally clear. Marked cytologic atypia and necrosis are absent. Mitoses, when present, are rare. The tubules are mostly simple and separate, with an occasional back-to-back architecture. They involve the lamina propria in a haphazard manner and rarely involve the superficial layers of the muscularis propria. A small percentage of the cases show a prominent basement membrane material surrounding the tubules. The second pattern includes cysts and cystic dilatation of the tubules, accompanied by eosinophilic or basophilic secretions. The third pattern shows a polypoid to papillary architecture, usually lined by a single layer of cuboidal to low columnar cells. The papillary architecture is predominantly simple with minor secondary branching. Other rare patterns include solid nests, cords of cells, and spindle cells.73 Recurrence occurs infrequently after transurethral resection.
A few entities come into the differential diagnoses of nephrogenic metaplasia. The most difficult is thankfully a rare entity: nephrogenic adenoma-like clear cell carcinoma. Architecturally, nephrogenic adenoma-like clear cell carcinoma shows tubular, cystic, and papillary architecture similar to nephrogenic metaplasia. However, features that favor nephrogenic adenoma-like clear cell carcinoma include abundant clear cytoplasm and hobnailing, nuclear pleomorphism and hyperchromasia, solid nests of cells, necrosis, widespread infiltration including of muscularis propria, lymphovascular invasion, a high Ki-67 mitotic index, and positive p53 nuclear immunostaining.74 Other differential diagnoses include prostatic adenocarcinoma; however, an immunohistochemical panel including PAX2, PAX8 (positive in nephrogenic metaplasia), and PSA (positive in prostatic adenocarcinoma) is helpful. Of note, racemase or AMACR staining can be positive in both and would not be helpful in this differential diagnosis.
Perhaps the most controversial aspect of nephrogenic adenoma is its derivation and whether it represents a metaplasia or a benign tumor. It is currently accepted that nephrogenic metaplasia/adenoma is a metaplastic lesion. In an elegant study, Mazal et al75 showed that nephrogenic adenoma arising in renal transplant patients with opposite sex donor kidneys were derived from the donor kidneys, suggesting that some nephrogenic adenomas are indeed derived from renal tubule seeding, rather than a metaplastic phenomenon.
The preneoplastic potential of nephrogenic adenoma is also hotly contested. Hartmann et al76 reported on a patient with recurrent nephrogenic adenoma, who subsequently developed clear cell carcinoma, and found that the initial nephrogenic metaplasia, the recurrence, and the clear cell carcinoma had similar genetic alterations, suggesting a clonal evolution. In addition, Pycha et al77 found aberrations of chromosome 7 and 9 in nephrogenic metaplasia, also raising the possibility of its preneoplastic potential. Additional studies are necessary to further elucidate the preneoplastic potential of nephrogenic metaplasia.
Pseudocarcinomatous proliferation refers to a rare reactive proliferation of the bladder mucosa with cytologically and architecturally worrisome features, seen post radiation, chemotherapy, or ischemic/irritative insults.
In the setting of radiation or chemotherapy, patients commonly present with hematuria, on average 2 years after treatment but can present anytime from 0 to more than 7 years78 post treatment. Pseudocarcinomatous proliferation shows cytologic and architectural features that are worrisome for carcinoma. It is composed of small irregular nests of cells that can have an infiltrative pattern, but are always confined to the lamina propria (Figure 15). Nuclear pleomorphism can be mild to moderate. Mitotic figures are commonly absent but can be present and brisk (up to 8 per 10 high-power fields).78 The background of ulceration, inflammation, hemorrhage, vascular congestion, fibrin deposition, and edema should offer a clue to the diagnosis. In addition, helpful hints include an accurate clinical history of treatment, squamous differentiation,79 infiltrating nests wrapping around blood vessels with fibrin deposition,78 and lack of in situ carcinoma away from the lesion. However, one should keep in mind that radiation and chemotherapy, particularly cyclophosphamide, are risk factors for the development of carcinoma, and careful histologic examination should be performed to exclude posttreatment carcinoma. In addition, patients with previous history of urothelial carcinoma should be critically evaluated.
Pseudocarcinomatous proliferation can also rarely occur in patients without a history of radiation or chemotherapy, but who do have predisposing ischemic or irritative etiologies such as chronic catheterization.80 Similar to the proliferations that occur in the setting of radiation or chemotherapy, the surrounding background of inflammation, edema, hemorrhage, ectatic blood vessels, and fibrin and hemosiderin deposition provides the clue to the correct diagnosis. Interestingly, in this setting the cells tend to show a more eosinophilic cytoplasm, which could be confused with paradoxical differentiation or squamous differentiation of invasive urothelial carcinoma.80 Despite their infiltrative-appearing architecture, they do not involve the muscularis propria or show marked cytologic atypia. Whether these lesions represent a preneoplastic lesion still needs to be studied; however, owing to the rarity of this lesion, we might never get a definitive answer.
Müllerianosis of the bladder is uncommon and usually refers to 3 related benign glandular entities: endometriosis, endosalpingiosis, and endocervicosis. Endosalpingiosis often occurs with either endometriosis or endocervicosis, although any combination of these 3 can be seen, warranting the term müllerianosis, when substantial amounts of each are present. While endometriosis usually presents in women of reproductive age, endocervicosis presents in perimenopausal women. A history of previous pelvic surgery is present in a large proportion of women with endometriosis or endocervicosis, raising the possibility of implantation due to surgery. Alternative possibilities could include a metaplastic phenomenon, due to the coexistence of these entities with other metaplastic phenomena such as intestinal metaplasia81 and nephrogenic metaplasia82 in the neighboring mucosa. Histologically, these processes involve the posterior wall or dome of the bladder and can be haphazardly present in any layer of the wall: mucosa, lamina propria, muscularis propria, or perivesical fat.
Endometriosis is characterized by the presence of endometrial glands surrounded by endometrial stroma often with evidence of hemorrhage (Figure 16). Endosalpingiosis refers to the presence of glands lined by ciliated columnar epithelium and peg cells, resembling fallopian tube mucosa. Endocervicosis refers to glands lined by mucinous columnar cells resembling endocervical mucosa. The presence of goblet-like or even true goblet cells has been reported in these müllerian lesions.83 In all 3 entities, marked cytologic atypia is not seen, and architectural features of malignancy, such as single infiltrating cells and back-to-back glands, are not present. Mitoses are absent to rare. In addition, a desmoplastic stromal response is usually absent. If present, these features should raise the possibility of an invasive adenocarcinoma.
The distinction between endocervicosis and low-grade cervical adenocarcinoma can be challenging. Immunohistochemically, features that favor endocervicosis are a diffuse pattern of estrogen and progesterone receptors, a low proliferative index on MIB1/Ki-67, and lack of gastric mucin.84
Ectopic Prostatic Tissue
Prostatic-type epithelium has been described in the bladder and urethra of male patients and can present as a polypoid mass. Histologic examination reveals prostatic-type glands lined with 2 cell types: the basal cells and luminal cells, located in the lamina propria of the bladder, underlying the urothelial surface. It is commonly present in the bladder neck and trigone region, where prostatic evaginations during embryonic development have been reported.85 They are thought to arise from persistent or migrating embryonic tissue, although the possibility of a metaplastic origin has also been raised. They stain like prostatic epithelium with PSA and prostatic-specific acid phosphatase and are benign.86 Occurrence of prostatic tissue in women is very rare but has been reported in the urethra,86 vagina, and cervix.87
Bladder diverticula are defined as outpouchings of the bladder mucosa and lamina propria, with or without a muscular wall.88 Currently, bladder diverticula are divided into congenital, which usually affect a younger patient population,89 and acquired, which present in an older patient population. Congenital diverticula are attributed to an inherent weakness or malformation of the bladder wall,88 whereas acquired bladder diverticula are thought to be caused by bladder outlet obstruction.90 Some studies suggest that diverticula are caused by embryologic defects with subsequent superimposed obstruction. Whether bladder diverticula have an increased risk of cancer development is controversial. Some studies suggest that urine stagnation, which occurs in diverticula, may cause chronic irritation, which in turn promotes urothelial neoplasia.89 Others believe that diverticular urothelium is subject to the same field effect as the rest of the bladder, since some patients with diverticular tumors present with extradiverticular disease upon follow-up.91 Perhaps a combination of both settings is at play in the formation of bladder diverticular neoplasia.
A wide variety of interesting lesions is seen in the prostate and bladder and is more commonly encountered by the general surgical pathologist. Knowledge of the histologic characteristics of these lesions, their preneoplastic potential, and the possible pitfalls can help the pathologist serve the patient better.
From the Department of Pathology and Cell Biology, Columbia University Medical Center and the New York Presbyterian Hospital, New York, New York.
The authors have no relevant financial interest in the products or companies described in this article.