Diagnosis, evaluation and treatment of carcinoma in situ of the urinary bladder: The state of the art

1.1. Definition 

As with CIS in other organ systems, urothelial CIS refers to replacement of part or all of the normal epithelium by cells which have microscopic – and now molecular – features of carcinoma, yet are confined to the epithelium. The number of cell layers in CIS may be increased, normal, or decreased. Other architectural changes, such as those of non-invasive papillary urothelial carcinoma, are by definition excluded and are classified separately. CIS is therefore a “flat” lesion, designated Tis for TNM pathologic staging. For clinical staging, the “is” suffix may be added to any primary tumor to designate the presence of associated CIS. In contrast, non-invasive papillary urothelial carcinomas are considered Ta lesions [6]. In many cases, an in situ carcinoma component is adjacent to or associated with invasive carcinoma. In such cases the CIS is referred to as secondary. In the less common clinical scenario, invasive carcinoma is absent and CIS is associated with a lower grade intraurothelial lesion (dysplasia) [7]. These cases are referred to as primary CIS [8].

1.2. History 

Urothelial carcinoma and its in situ counterpart were traditionally referred to as transitional cell carcinoma (TCC) and TCC in situ, based on the premise that urothelium has features intermediate or transitional between stratified squamous epithelium and pseudostratified columnar epithelium. In recent years, the term “urothelium” has been increasingly used to describe this specialized epithelium. In 1952, Melicow recognized the significance of examining the grossly normal bladder mucosa between exophytic tumors, in an attempt to explain the high recurrence rate of bladder cancer, even after extensive organ-preserving surgical treatment [9]. Later that year, Melicow and Hollowell described CIS of the urinary system, referring to lesions “…whose gross features are inconspicuous and seemingly benign but whose microscopic picture is that of malignancy [3].” This concise description remains accurate today.

Nomenclature for urothelial CIS has historically been somewhat confusing. In Melicow's original papers, the terms “Bowen's disease” and “intraurothelial cancer” are used somewhat interchangeably with CIS. Over the years, precancerous bladder lesions have been referred to by a variety of names, including dysplasia (mild, moderate, or severe), intraepithelial neoplasia (low grade, high grade), intraurothelial neoplasia (grade I, grade II), atypical hyperplasia, and marked atypia [7].

The nomenclature for preinvasive urothelial lesions has gained some uniformity in recent years, influenced by a classification system proposed by the World Health Organization and International Society of Urologic Pathology (WHO/ISUP) in 1998 and another derived from the Ancona Consultation on the Diagnosis of Non-Invasive Urothelial Neoplasms in 2001 [10]. These two classification systems both separate putative premalignant lesions (dysplasia and CIS) from purely benign lesions or lesions with unclear premalignant potential.

1.3. Primary versus secondary CIS 

Definitions of primary and secondary CIS vary somewhat in the pathologic and urologic literature. Takenaka and colleagues separate lesions into the following categories: (1) primary CIS, occurring without associated previous urothelial tumor, (2) concomitant CIS, occurring in conjunction with a newly diagnosed bladder tumor, and (3) secondary CIS, diagnosed during follow-up of a known bladder tumor, with or without a concomitant tumor at the time of CIS diagnosis. This convention is used by some other authors, and may be gaining in usage [11]. In many cases, however, the terminology is less clearly defined, so primary and secondary refer presumptively to the presence or absence of any current or past tumor. Admittedly, this is perhaps the most important distinction, since it appears to make little clinical difference whether CIS is related to a synchronous or metachronous primary tumor. Other synonyms are sometimes used, including “isolated”, “solitary”, and “concurrent” CIS, but the meaning of these is even less clearly defined than the terms above.

1.4. Epidemiology 

Urothelial carcinoma of the bladder is the second most common malignancy of the genitourinary system, after prostate cancer. Urothelial carcinoma is by far the most common histologic type of bladder cancer and accounts for over 90% of bladder malignancies [1]. Bladder carcinoma ranks among the most frequent of male cancers, having its highest incidence in the developed countries of North America and Western Europe [12], [13]. In the United States, Canada, and the European Union, the urinary bladder is estimated to be the fourth leading site of new cancer diagnoses in men, following prostatic, pulmonary, and colorectal cancers. Bladder cancer accounts for 3% of US male cancer deaths [1]. CIS in particular seems to maintain the same male predilection as invasive bladder cancer, having a male to female ratio ranging from 4:1 to 7:1 and a mean age of 65–73 years in studies which focus on CIS [4], [11], [14], [15], [16]. In women, bladder cancer is diagnosed less frequently, although rates have been increasing slightly each year [1].

Of bladder cancers, tumors that are superficial or non-muscle-invasive at the time of diagnosis make up the majority of cases (up to 80%). CIS as a primary entity, however, makes up only a small component of all bladder cancer. It is most frequently seen in conjunction with invasive urothelial carcinoma, accompanying 45–65% of such tumors and often designated “secondary CIS” in that setting. Primary CIS accounts for only about 1–3% of all urothelial neoplasms [8], [10]. Shariat et al. found that 46% of patients undergoing cystectomy for bladder cancer had concomitant CIS at the time of resection [17], and conversely, up to 33% of patients who underwent radical cystectomy for a clinical diagnosis of refractory CIS were found to have invasive disease at the time of resection [18]. Thus it is clear that urothelial CIS and invasive urothelial carcinoma are intimately related entities.

1.5. Etiology 

Risk factors specific to CIS are infrequently discussed apart from risk factors for all bladder cancer, since primary CIS is uncommon and any CIS is frequently associated with invasive cancer. Risk factors for CIS appear largely to parallel those for bladder cancer in general. For this reason, risk factors common to all types of urothelial neoplasia are discussed here.

Cigarette smoking is a well documented risk factor for bladder carcinoma, and cumulative risk parallels the degree of tobacco consumption. Current smokers have up to a threefold higher risk than nonsmokers, with greater number of years smoking and cigarettes per day correlating with higher risk. Older age at first exposure is associated inversely with bladder cancer risk [13]. Smoking cessation is followed by a notably brisk reduction in risk, up to 50% within 3 years [19]. Patients with superficial urothelial carcinoma (including CIS) who continue to smoke are more likely to experience adverse outcomes, as evidenced by the later age at presentation of CIS in ex-smokers, and the more rapid time to disease recurrence in continuing smokers. The latter have a diminished survival free of adverse events compared to ex-smokers and nonsmokers [20]. The association of bladder carcinoma with cigar and pipe smoking is less clear and less extensively reported, and may be weaker than the risk from cigarette smoking. Likewise, studies regarding secondary exposure to tobacco smoke have found a small, but convincingly increased risk. The mechanisms whereby smoking induces bladder carcinogenesis are not well understood, however. It is unclear which of the chemicals present in cigarette smoke, whether aromatic amines, such as 4-aminobiphenyl, or aldehydes, such as acrolein, is the likely cause of increased risk [13].

Exposure to aromatic amines from other sources has been unequivocally implicated in bladder carcinogenesis, particularly in the setting of occupational exposure. With occupational exposure there may some overlap with the same chemicals present in cigarette smoke, but without the complex mixture of cofactors present in cigarette smoke. In occupational exposure, either inhalation through the lungs or absorption through the skin may represent the mechanism of absorption. Polycyclic aromatic hydrocarbons (PAHs), diesel exhaust, and paint substances have also recently been implicated as urothelial carcinogens. These agents, unlike other chemical carcinogens, remain less strictly monitored [21]. Occupational exposure is most common and now most closely monitored in the chemical, dye, rubber and textile industries. Industrial agents implicated in bladder carcinogenesis include 2-naphthylamine, 4-aminobiphenyl, benzidine (and benzidine-derived azo-dyes), 4,4-methylene bis 2-chloroaniline (MBOCA), o-toluidine, and methylene dianiline. In some settings where there appears to be a clearly increased risk for bladder cancer, specific agents are unknown or may be present in combination with other agents. Although many carcinogenic compounds are now prohibited or strictly controlled, patients with remote exposure may still have an increased risk. Occupational exposure is thought to account for up to one fourth of urothelial carcinomas in heavily industrialized countries [12].

In a meta-analysis of 30 epidemiologic studies, Zeegers et al. found that alcohol consumption in men, when adjusted for cigarette smoking, confers a small increased risk compared to no alcohol consumption. This result, however, was not statistically significant. The same relationship in women was not demonstrated. Similarly, coffee consumption is heavily confounded by a frequent association with tobacco smoking and yet has been shown by some investigators to confer a slightly increased risk [13].

Analgesic use has also been implicated in bladder carcinogenesis, particularly with regard to phenacetin-containing compounds. This observation has raised concern over the use of acetaminophen (its metabolite) and non-steroidal anti-inflammatory drugs (NSAIDs) [22]. Conversely, some studies have found that NSAIDs, including acetaminophen, may be associated with decreased risk [22], [23].

A variety of infectious conditions, including urinary tract infection, gonorrhea, syphilis, other bacteria, human papilloma virus (HPV), human immunodeficiency virus (HIV), herpes simplex virus (HSV), and BK virus have been studied as potential risk factors for bladder carcinoma. However, only in schistosomiasis has a strong association between infection and carcinoma been found; in a setting of chronic schistosomal infection of the bladder, there is a demonstrably high risk for development of carcinomas, of which squamous cell carcinoma is the most common [24].

Other putative risk factors include urinary tract lithiasis, radiation, and exposure to chemotherapeutic agents, particularly cyclophosphamide.

2.1. Clinical differential diagnosis 

2.2. Location 

CIS is frequently multifocal. Two or more separate locations are involved by CIS in 50% of cases, with a predilection for the trigone, lateral wall, and dome of the bladder [8], [27]. Additional sites of multifocal involvement include distal ureters, prostatic urethra (20–67% of cases), prostatic ducts or acini (up to 40%), renal pelvis and proximal ureters [8].

2.3. Cystoscopic findings 

Cystoscopically, detection and accurate assessment of flat lesions is difficult. The mucosa may range from unremarkable to erythematous, edematous, or eroded and lesions may be found both near an invasive tumor and at a distance from it [8]. Characteristic lesions are sometimes described as a “red, velvety patch,” although this finding is nonspecific. Swinn et al. found that 12% of “red patch” biopsies yielded a finding of malignancy (78.3% of which were CIS). Thus, performing biopsies on “red patch” lesions is warranted, particularly on follow-up for known carcinoma or in patients over 60 years of age. Below age 60 no examples of malignancy were found [28]. Fernando et al. found a similar age range (60–70 years) for positive biopsies and found that 8% of “red patch” biopsies were positive for CIS (4 of 25 patients) [15]. Prudent cystoscopic examination, therefore, should include sampling of areas with erythema, elevation or erosion, as well as extensive sampling of grossly normal bladder mucosa in a high-suspicion patient.

One technique for improving in vivo detection of flat lesions is photodynamic diagnosis (PDD) or fluorescence cystoscopy. This approach takes advantage of concentration differences in fluorescent molecules to distinguish diseased from normal tissue. Common fluorescent agents for this purpose include hypericin, 5-aminolaevulinic acid (5-ALA), and hexaminolevulinate (HAL) [29]. The sensitivity of this technique in some studies is greater than white light cystoscopy (WLC) alone, particularly for detecting CIS. Using this modality, photoactive porphyrins accumulate preferentially in neoplastic tissue and under the excitation of blue light emit lower energy and longer wavelength light, which appears red [30]. Colombo and colleagues studied patients with CIS only undergoing intravesical therapy (although some patients had previous papillary tumors), and found that PDD findings correlated strongly with CIS. As no cases were missed using the combination of WLC and PDD techniques, the authors suggest that in the absence of findings by both modalities, bladder “mapping” or extensive biopsy of random sites may be reasonably avoided. Although PDD carries an increased false-positive rate compared to WLC, they found the overall side effects of PDD to be negligible [31].

Other promising techniques are being investigated and may provide better predictors of the biopsy showing CIS. Raman spectroscopy, for example, relies on a principle known as the Raman effect. When incident light interacts with tissue molecules, the vibrational energy changes, resulting in the release of new photons. By measuring the spectrum of photons released, it is possible to generate a “picture” of the tissue composition and potentially differentiate between normal, inflammatory, and malignant processes, such as CIS. Primarily, however, this technique has been tested on tissue samples from the bladder, rather than in vivo. In the future, cystoscopic application of this technique may prove promising for the resolution of the CIS differential diagnosis. Similarly, optical coherence tomography measures the amplitude of scattered light, resulting in an image similar to histopathology. Interpretation of the generated image, however, does depend on the skill level of the observer. These techniques, however, require a specific target lesion and are less suitable for screening of the entire bladder [29].

3.1. Morphologic classification 

It is useful to recognize the histopathologic variants of urothelial CIS, such as large cell CIS, Pagetoid CIS and small cell CIS, in order to avoid misdiagnosis. Nevertheless, these variants connote no known prognostic differences. Morphologic grading has historically been applied to CIS; but the modern WHO/ISUP classification system recognizes only two grades of atypia, defined as dysplasia and CIS. Each of these grades has been shown to have clinical prognostic importance.

3.2. Microinvasion 

Microinvasive CIS (CISmic) was originally defined as lamina propria invasion to a depth 5mm or less from the basement membrane. Currently, a cancer exhibiting more than 20 cells measured from the stromal–epithelial interface should be regarded as fully invasive. The microinvasive cells are frequently arranged in cords, single cells and clusters of cells, surrounded by retraction artifact. A desmoplastic stromal response is a helpful diagnostic feature, although it is not always present [10], [34]. Detection of microinvasive or fully invasive carcinoma can be facilitated by keratin immunostaining, particularly when obscuring inflammation conceals the neoplastic cells [10].

3.3. Adenocarcinoma in situ 

In situ adenocarcinoma (AIS) is a rare entity infrequently discussed in the literature. Chan and Epstein describe three predominant architectural patterns of AIS, papillary, cribriform and flat. These non-invasive glandular lesions demonstrate atypical columnar epithelium with apical cytoplasm. The authors note that such lesions are more strongly associated with invasive carcinoma than is typical urothelial CIS. They also suggest that AIS may occur more frequently than would be inferred from the existing literature. In their series AIS made up 0.14% of their institutional bladder biopsy cases. The most likely candidate for confusion of AIS with urothelial CIS is the flat pattern of in situ adenocarcinoma. This and other variants of AIS, including pleomorphic and non-papillary types, may be differentiated from traditional CIS by the presence of definitive glandular structures. This architectural feature is in contrast to the small, mucin-containing spaces without columnar epithelium, so-called “gland-like lumina,” which are present in some cases of urothelial carcinoma [35].

3.4. Pathologic differential diagnosis 

Flat urothelial hyperplasia (FUH) is characterized by an increase in number of cell layers (usually greater than 10) without the hyperchromasia, nuclear membrane irregularities or architectural disarray of carcinoma or CIS. There may be slight nuclear enlargement, but the nuclei of hyperplastic epithelium are of approximately uniform size. Hyperplasia as an isolated lesion is not a documented premalignant condition. FUH is frequently seen in conjunction with low-grade papillary urothelial tumors, as well as inflammatory disorders, and lithiasis. Chromosomal aberrations similar to those of papillary tumors, such as chromosome 9 deletions, may be present in FUH, as well as in histologically normal urothelium. The observation of genetic lesions in FUH similar to those of papillary tumors, suggests that FUH may represent an early step in the development of papillary bladder tumors [10], [33].

Urothelial dysplasia, or low-grade intraurothelial neoplasia, includes a spectrum of cytologic and architectural abnormalities that do appear preneoplastic but are insufficient for the diagnosis of CIS. There may be nuclear rounding, crowding, some cytologic atypia, and architectural disturbance including loss of polarity, but altogether to a lesser degree than those of unequivocal CIS. Lesions classified as urothelial dysplasia are frequently observed in a context of known cancer diagnosis (secondary dysplasia) or in conjunction with a non-invasive papillary neoplasm. Nevertheless, dysplasia is also sometimes identified de novo (primary dysplasia) and has the potential to progress to neoplasia. Unfortunately, data regarding primary dysplasia is limited [8], [10]. In a study of 36 patients with primary dysplasia, 19% progressed to CIS, papillary carcinoma, or death from bladder cancer in a mean time period of 2.5 years (6 months to 8 years) [36].

In reactive atypia, nuclear abnormalities are less than those of dysplasia or CIS and these changes occur in conjunction with mucosal inflammation. Thus, the nuclear aberrations of reactive atypia can reasonably be attributed to urothelial repair or regeneration. These cells have enlarged vesicular nuclei, prominent nucleoli, and abundant cytoplasm, and their presence can often be correlated with a clinical setting of previous manipulation or irritation [8], [10].

Atypia of unknown or unclear significance has been proposed as a diagnosis to encompass lesions with nuclear abnormalities similar to those of reactive changes (less than those of dysplasia), but out of proportion to the degree of inflammation [10]. These lesions, however, are not associated with adverse outcomes and may reasonably be combined diagnostically with reactive changes [25].

Therapy-related atypia: Many of the therapeutic modalities currently used for bladder carcinoma and drugs used for other cancers can cause therapy-related changes that confound the diagnosis of flat urothelial lesions like CIS. Such cases may have markedly abnormal cytologic and histologic features in non-neoplastic urothelium. Thiotepa (triethylenethiophosphoramide) and mitomycin-C, for example, may induce cell exfoliation and mucosal denudation, similar to denuding cystitis. After intravesical administration of these agents, umbrella cells become large, vacuolated and multinucleated; prominent small nucleoli may be apparent within these enlarged nuclei. These findings may persist for weeks or months after discontinuation of treatment. Similarly, systemic cyclophosphamide treatment may produce large binucleate or multinucleate cells with enlarged bizarre nuclei. There is both cellular and nuclear enlargement, because cyclophosphamide causes arrest of the cell cycle. Nucleoli in these cells may be single or double, with angulated edges.

Radiation therapy likewise causes urothelial cell enlargement, multinucleation, and vacuolization, with a normal nuclear–cytoplasmic ratio. Hemorrhage, fibrin deposition, and multinucleated stromal cells may create a tumor-like appearance in longstanding cases. There may even be nodules of squamoid epithelium pushing into the lamina propria without true infiltrative growth [37].

Awareness of the therapy-induced changes noted above may assist the pathologist in avoiding a misdiagnosis of CIS. Unfortunately, the needed information is often not provided by the clinician when the diagnostic specimens are submitted for pathologic evaluation.

3.5. Diagnostic immunohistochemistry 

Immunohistochemical staining (IHC) may be useful in differentiating CIS from reactive changes in difficult biopsy cases. Normal urothelium typically exhibits CK20 staining in only the umbrella cell layer and p53 staining primarily in the basal layer. A pattern of CK20 and p53 positivity throughout the neoplastic urothelium combined with negative staining for CD44 favors a diagnosis of CIS. McKenney et al. concluded that this immunohistochemical pattern may reasonably be used to confirm cases of strongly favored CIS, primary CIS, or the less common patterns of undermining and Pagetoid CIS [38].

In the reactive setting, the urothelium is often strongly positive for CD44 and lacks extensive staining for CK20 and p53. The authors recommend that a panel of all three antibodies be used in correlation with hematoxylin and eosin morphology [38]. Mallofre and colleagues reported similar findings utilizing CK20 and p53 immunostaining, and proposed Ki67 as a third “positive marker” in the neoplastic urothelium of CIS [39].

3.6. Selected prognostic markers 

3.7. Molecular features 

Non-invasive papillary tumors and CIS or urothelial dysplasia represent distinct entities with different frequencies of progression to invasive cancer. It is, therefore, fitting that they differ at the genetic level. In general, loss of heterozygosity (LOH) of chromosome 9 is reported more frequently in papillary tumors than in CIS, while the reverse is true for p53 gene mutation [48], [49]. However, Hartmann et al. found that many cases of CIS (86%) had significant chromosome 9 abnormalities, calling this relationship into question. These investigators found that cases of moderate dysplasia were slightly less likely to exhibit chromosome 9 deletions (75%) than CIS or high-grade dysplasia, supporting the notion that dysplasia is a precursor to CIS. In addition, they found by LOH studies and fluorescent in situ hybridization (FISH) that deletions of 17p13.1 at the p53 locus were present in 84% of CIS cases and 53% of dysplasia cases, reinforcing the association between p53 abnormalities and high-grade lesions [50].

Hopman et al. propose that molecular differences exist between primary CIS (termed “isolated”) and secondary CIS (associated with a papillary tumor). In their study, chromosome 9 deletions were not present in cases of primary CIS and were frequently present in secondary CIS [51]. This finding supports the hypothesis that p53 mutations precede chromosome 9 aberrations in the progression from CIS to invasive cancer. In contrast, the Hartmann study did not identify differences between primary and secondary CIS, perhaps due to the limited total number of cases, a majority of which included concurrent papillary tumors [50].

Zieger et al. have recently further elucidated this issue. They demonstrated gains of chromosome 5p only in CIS, whereas FGFR3 mutations were identified only in papillary tumors [52]. These findings suggest that non-invasive papillary-type tumors and CIS-type tumors originate along different pathways, merging in some cases with disease progression, as CIS eventually develops in some patients with papillary tumors. This proposed mechanistic pathway is also compatible with the findings of overlap between chromosome 9 and p53 abnormalities in CIS by some investigators.

3.8. Gene expression profiling 

Dyrskjot and colleagues identified a multi-gene molecular classifier for CIS using microarray analysis, comparing CIS lesions to normal urothelium, papillary tumors with and without associated CIS, and invasive tumors [53]. Interestingly, the group found that in CIS cases with associated papillary tumors, the CIS lesions demonstrated similar expression patterns to both the concurrent papillary tumors and the adjacent normal urothelium. This pattern of similar molecular findings even in adjacent normal urothelium would support the “field effect” theory of bladder carcinogenesis. In papillary lesions without concurrent CIS, the expression profile was notably different between these primary papillary tumors and the usual CIS profile. Contrary to the common hypothesis that invasive tumors typically arise from CIS-type lesions, however; the group found a distinct characteristic profile for invasive carcinoma [53]. In a large-scale validation, a 68-gene signature had 80% sensitivity (36 of 45 samples) and 68% specificity (71 of 105) for CIS [54]. Together, these findings suggest that such molecular techniques may provide an accurate prediction of CIS presence even when CIS itself is not seen in the tissue biopsy.

3.9. Proteomics 

Proteomic-based methodologies for identification of tumor markers are promising, but not immediately useful diagnostically. High-resolution two-dimensional electrophoresis, mass spectrometry and surface-enhanced laser desorption/ionization (SELDI) mass spectrometry have identified markers of interest in some cancers [55], [56]. Munro et al. utilized the SELDI technique, which has been studied in cancers of other sites, to discriminate urothelial carcinoma from benign lesions at a sensitivity (78.3%) and specificity (65.0%) comparable to traditional methods. Only 2 cases, however, relating to CIS were examined [57]. Sanchez-Carbayo examined serum protein profiles by antibody array, successfully discriminating bladder cancer patients from controls at a rate of 93.7% (n=95). Of note, 22 of 37 patients with bladder cancer had CIS [58]. Currently, there are no published data related to application of proteomic techniques in assessing primary CIS; this is potentially a fruitful area of research in screening for early bladder cancer.

3.10. Urine markers 

4.1. Prognosis 

4.2. Upper urinary tract CIS 

4.3. Prostatic ducts and prostatic urethra involvement 

Patel and colleagues in 2009 examined 308 patients who underwent radical cystoprostatectomy with whole-mount processing of the prostate component [91]. Of the 121 (39.3%) patients with prostatic urothelial carcinoma, 59 (48.8%) had dysplasia/CIS of the prostatic urethra and 20 (16.5%) had involvement of the prostatic ducts. The remaining patients had involvement by invasive urothelial carcinoma. Risk factors for overall prostatic involvement included a bladder tumor location in the trigone and associated bladder CIS by both univariate and multivariate analysis. Presence of bladder CIS was the sole risk factor associated specifically with non-stromal involvement (CIS/dysplasia of ducts/urethra) [91]. Similar to the tumor cell mobility seen in the bladder, the malignant cells may spread in a Pagetoid fashion through the prostatic epithelium or undermine the epithelium of the prostatic ducts/acini, resulting in a layer of malignant cells sandwiched between the basal cell layer and epithelial layer.

In the newly revised 2010 TNM staging system [6], subepithelial invasion of prostatic urethra will not constitute T4a staging status. T4a tumor is defined by prostatic stromal invasion directly from the bladder cancer.

4.4. Urethral involvement 

Tobisu et al. in 1997 examined 52 patients who underwent radical cystoprostatectomy with simultaneous en bloc urethrectomy for bladder cancer [92]. This included 21 patients with diffuse primary CIS (with or without microscopic invasion). Of them, 4 (19%) had abnormalities of the anterior urethra, 3 with CIS in the bulbar urethra extending from the prostatic and membranous urethra and 1 with severe dysplasia. Of the 10 patients with diffuse CIS in addition to nodular/papillary tumor, 1 had invasive urothelial carcinoma involving the corpus spongiosum of the penile urethra. These authors conclude that diffuse bladder CIS extending to the prostatic urethra is a risk factor for synchronous anterior urethra involvement [92].