Secondary Logo

Journal Logo

The diagnostic utility of antibodies to smoothelin and desmin in staging of urothelial carcinoma

Kamel, Nora N.a; Helmy, Noha A.H.a; Hassanein, Amina A.a; Sharaf, Walid M.a; Moneer, Bahaa I.b; Salem, Mostafa M.b

Egyptian Journal of Pathology: July 2015 - Volume 35 - Issue 1 - p 14–23
doi: 10.1097/01.XEJ.0000462730.11654.7fss
ORIGINAL ARTICLES
Free

Background and objectives Accurate recognition of urinary bladder muscularis propria (MP) invasion by urothelial carcinoma is crucial as it is the critical crossroad between conservative and aggressive management for the patient. It is now widely known that an inconsistent layer of muscularis mucosae (MM) muscle exists in the lamina propria, which can mimic the MP muscle, particularly when hyperplastic, making staging extremely challenging in some limited, unoriented, or highly cauterized specimens. Smoothelin is a novel smooth muscle-specific contractile protein expressed only by fully differentiated smooth muscle cells, and not by proliferative or noncontractile smooth muscle cells and myofibroblasts.

Materials and methods We performed immunohistochemical staining in the bladder for smoothelin to evaluate its expression in MM and MP muscle in 60 cystectomy specimens and by comparing the staining pattern with desmin.

Results In contrast to desmin, which equitably stained both types of muscle fibers, smoothelin showed striking differential immunoreactivity between MM and MP muscle. Smoothelin was expressed in different intensities and showed different immunoreactivity between MM and MP muscles. The MM showed absent (44/60, 73%) or weak staining (15/60, 25%), only one showed moderately positive staining (1/60, 1.6%), and none showed strong staining (0/60, 0%). In contrast to the MM, the MP predominantly showed strong staining (36/60, 60%) and moderate staining (24/60, 40%), with none showing weak or negative staining (0/60, 0%). There was always a significant and striking disparity in the staining intensity between the MM and the MP in every section (in cases where smoothelin was weak in the MM, it was always stronger in the MP).

Conclusion The relatively distinct immunohistochemical staining pattern of smoothelin between MP and MM (including its hyperplastic forms) makes it a robust and attractive marker to be incorporated into the contemporary diagnostic armamentarium for the sometimes difficult area of staging bladder urothelial carcinoma. It also confirmed the use of desmin to differentiate between smooth muscle cells (whether MM or MP) and proliferating myofibroblasts.

aPathology Department, National Research Center (NRC)

bPathology Department, Cairo University, Cairo, Egypt

Correspondence to Noha A.H. Helmy, MD, Pathology Department, National Research Center (NRC), Cairo, Egypt E-mail: nohahelmy@gmail.com

Received October 1, 2014

Accepted October 12, 2014

Back to Top | Article Outline

Introduction

Urothelial carcinoma is the fourth most common malignancy in men and the 10th most common malignancy in women (Gofrit et al., 2009).

The single most important prognostic factor in urothelial carcinoma is pathological stage, which includes the anatomic depth of invasion on the basis of which therapeutic decisions are made. Accordingly, identification of invasion, and determining its depth if present, is an essential component in the evaluation of bladder biopsies and transurethral resection of bladder tumor (TURBT) specimens that harbor carcinoma (Council and Hameed, 2009).

There are three potential pitfalls in this assessment. In the lamina propria, there is an inconsistent layer of muscularis mucosae (MM) muscle that can occasionally be hyperplastic and mimic the muscularis propria (MP) muscle, making staging extremely difficult sometimes in limited or unoriented specimens such as biopsies or TURBT specimens (Paner et al., 2007).

Another compounding factor is the presence of a desmoplastic stromal response in which prominent myofibroblasts may mimic muscle bundles (Jimenez et al., 2000). In addition to tumor surrounding MP, tumor may infiltrate the muscle bundles, fracturing their round contours and splaying the muscle fibers (De Peralta-Venturina et al., 2002). In these situations, distinction from MM or desmoplastic stromal response may be problematic.

Back to Top | Article Outline

Aim of this work

  • Assess the immunohistochemical expression of smoothelin in the bladder wall smooth muscle cells of the MM and MP, and in myofibroblasts.
  • Assess the immunohistochemical expression of desmin in the bladder wall smooth muscle cells of the MM and MP, and in myofibroblasts.
  • Compare the immunohistochemical expression of smoothelin and desmin in the bladder wall to assess their potential use as a discriminatory stain between MM and MP muscle.
Back to Top | Article Outline

Materials and methods

Case selection

A total of 60 specimens of bladder urothelial carcinoma were examined in this study. Specimens were obtained from the Department of Pathology, Cairo University, and from other private laboratories in the period from May 2009 up to August 2012. All the specimens were radical cystectomy specimens: as in contrast to some transurethral resection specimens, the depth of invasion and type of muscle can be determined on the basis of hematoxylin and eosin (H&E)-stained sections, and this morphologically determined depth of invasion and type of muscle was considered to be the gold standard. For all specimens, some clinical data were available on the computer files including age of the patients, clinical history, and history of previous resection with recurrence.

Back to Top | Article Outline

Histological review

All specimens were formalin fixed, routinely processed, and embedded in paraffin. Three (5-μm-thick) sections were prepared from each tissue block, one of them stained by H&E for re-evaluation; the other two were mounted on poly-L-lysine-coated slides (superfrost slides) and subjected to the two immunohistochemical markers mentioned later.

Back to Top | Article Outline

Immunohistochemical staining

Immunohistochemistry for smoothelin and desmin

Superfrost slides were placed in the BenchMark IHC/ISH staining module (Ventana Medical Systems Inc.,Tuscon, Arizona, USA) and the following steps were performed:

  • Deparaffinization.
  • Cell conditioning.
  • Mild CC1 application.
  • Standard CC1 application.
  • Application of one drop of Monoclonal Mouse Anti-Human Smoothelin (cell marque, R4A) and Monoclonal Mouse Anti-Human Desmin (cell marque, D33).
  • Application of a cover slip and incubation for 32 min.
  • Application of one drop of hematoxylin II (counterstain).
  • Application of a cover slip and incubation for 8 min.
  • Application of Bluing Reagent.
  • Application of a cover slip and incubation for 4 min.
Back to Top | Article Outline

Evaluation of smoothelin expression

Interpretation of immunohistochemical stains in each case was performed semiquantitatively by analyzing the intensity of staining separately in each compartment (myofibroblasts, smooth muscle of the MM, and smooth muscle of the MP). Vascular smooth muscle was used as an internal control.

For MP, the inner muscle layer or muscle bundles bordering the lamina propria was selected to evaluate staining of MP bundles of the inner half of the bladder wall. These muscle bundles were most likely to be included in TURBT specimens.

Furthermore, we could confidently evaluate these with respect to the overlying urothelium in all available specimens.

Staining intensity was evaluated as weak, moderate, and strong.

Back to Top | Article Outline

Evaluation of desmin expression

Diffuse cytoplasmic staining within the cell was considered positive.

Staining was evaluated separately in each compartment (myofibroblasts, smooth muscle of the MM, and smooth muscle of the MP). Vascular smooth muscle was used as an internal control.

Back to Top | Article Outline

Statistical analysis

The sensitivity and the specificity of both smoothelin and desmin were calculated using the following equations:

Back to Top | Article Outline

Results

Immunohistochemistry

Desmin

Desmin was expressed with similar intensity and pattern in MM (60/60, 100%) and MP (60/60, 100%), including the cases showing hyperplastic MM (Figs 1c and 2c, Table 1).

Fig. 1

Fig. 1

Fig. 2

Fig. 2

Table 1

Table 1

When evaluating the myofibroblastic/desmoplastic reaction surrounding the tumors, Desmin was negative in all 53 cases (100%) (Fig. 5c, Table 1).

Back to Top | Article Outline

Smoothelin

Smoothelin was expressed in different intensities and showed different immunoreactivity between MM and MP muscles. The MM showed absent (44/60, 73%) (Fig. 1b) or weak staining (15/60, 25%), only one showed moderately positive staining (1/60, 1.6%), and none showed strong staining (0/60, 0%) (Fig. 3, Table 2).

Fig. 3

Fig. 3

Table 2

Table 2

In contrast to the MM, the MP predominantly showed strong staining (36/60, 60%) (Fig. 4) and moderate staining (24/60, 40%), with none showing weak or negative staining (0/60, 0%). There was always a significant and striking disparity in the staining intensity between the MM and the MP in every section (in cases where smoothelin was weak in the MM, it was always stronger in the MP); smoothelin accentuated individual muscle fibers within groups (Fig. 2b, Table 3).

Fig. 4

Fig. 4

Table 3

Table 3

When evaluating the myofibroblastic reaction surrounding the tumor, most cases were negative (52/53, 98.2%), with only one case showing weak positive staining (1/53, 1.8%), (Table 3, Fig. 5b).

Fig. 5

Fig. 5

When all smoothelin staining irrespective of intensity and pattern (weak or moderate or strong) was considered positive, the specificity for smoothelin in MP (vs. MM) was 73.33% and the sensitivity was 100%.

If we only considered strong and moderate intensity as positive staining, the specificity improved to 100% (Tables 4 and 5).

Table 4

Table 4

Table 5

Table 5

Back to Top | Article Outline

Comparison between desmin and smoothelin expressions in differentiating muscularis mucosae and muscularis propria

Desmin immunostaining was equally effective in highlighting smooth muscle cells of both compartments (MM and MP), with a sensitivity of 100%. Smoothelin showed 100% sensitivity for labeling smooth muscle cells of the MP, but its specificity was variable according to the intensity of staining (Tables 6 and 7).

Table 6

Table 6

Table 7

Table 7

The sensitivity and specificity of smoothelin in highlighting the MP are shown in Table 8.

Table 8

Table 8

Back to Top | Article Outline

Discussion

The single most important prognostic factor in urothelial carcinoma is the pathological stage of the tumor. This includes the anatomic depth of invasion on which major therapeutic decisions are made. Similar to noninvasive tumors, a carcinoma confined to the lamina propria is usually treated conservatively, whereas one extending into the MP almost always dictates a more radical surgical management (Hall et al., 2007).

Accordingly, identification of invasion, and determining its depth if present, is an essential component in the evaluation of bladder biopsies and TURBT specimens that harbor carcinoma. Thus, it is imperative for pathologists to document not only the presence (and involvement by tumor) of MP but also the absence of this key histologic landmark in the initial biopsy or TURBT specimens. This, however, is not always straightforward and both understaging and overstaging have been reported to occur in a significant proportion of cases (Lopez-Beltran et al., 2008).

There are three potential pitfalls in the assessment of invasion of urothelial carcinoma into MP: first, the presence of the MM layer within the lamina propria of the urinary bladder. The presence of this layer was first reported by Dixon and Gosling, 1983. It is now known that the majority of cystectomy specimens contain MM within the lamina propria. It usually consists of one o three thin, wispy, and often-discontinuous smooth muscle fibers that are closely associated with the blood vessels of the lamina propria. This layer can readily be distinguished from the thick smooth muscle bundles of the MP in cystectomy specimens and well-oriented bladder biopsies (Paner et al., 2007). This is not always the case in TURBT specimens. The fragmentation, tangential sectioning, and thermal artifacts inherent to these specimens as well as the inflammatory and desmoplastic responses often elicited by invasive tumors can lead to poor orientation and difficulty in distinguishing the type of smooth muscle invaded by carcinoma (MM vs. MP). Also, it has been emphasized in many recent studies that the thickened hyperplastic appearance of MM may resemble the thick bundles of the MP. Thus, in TURBT, it may cause difficulty in staging of cancers (Vakar-Lopez et al., 2007).

The second potential problem in staging bladder carcinomas is that, similar to invasive carcinomas elsewhere, these tumors are often associated with a variably cellular, spindle-cell myofibroblastic stromal response. For the most part, these myofibroblasts can be readily distinguished from the smooth muscle cells of the MM or MP. However, in some cases, the proliferating myofibroblasts may mimic smooth muscle fibers of the MM or MP and may lead to misdiagnoses of invasive disease solely on the basis of the presence a cellular myofibroblastic response surrounding the tumor (McKenney, 2005).

Finally, in MP invasive urothelial carcinoma, in addition to the tumor surrounding MP. The tumor may infiltrate the muscle bundles, fracturing their round contours, and splaying the muscle fibers, causing difficulty in identifying the presence of MP and difficulty in distinction between it and the thin fibers of MM or the desmoplastic reaction (De Peralta-Venturina et al., 2002)

This raises several important questions: can immunohistochemistry help to distinguish between reactive myofibroblasts and smooth muscle cells of the bladder, and given the importance of the issue, between MM and MP? If so, can immunohistochemistry be used as an adjunct in staging carcinomas of the bladder, especially under the circumstances discussed above? (McKenney and Amin, 2005).

In the present study, the focus was on the potential use of smoothelin, which is a new marker that is used to discriminate MP from MM in diagnostic surgical pathology; we also evaluated the role of desmin in distinguishing between myofibroblastic reaction and the smooth muscle cells of the urinary bladder, whether it is MM or MP.

Maake et al. (2006) were the first to detect smoothelin expression in the urinary bladder; they found that the gene and protein (detected by RT-PCR and immunofluorescence, respectively) were expressed by the smooth muscle cells of the MP in eight normal and 13 overactive bladders. Subsequently, Kuijpers et al. (2007) found that protein expression was limited to the MP and was not identified in the MM in eight cystectomy specimens.

In this study, smoothelin stained all the MP of all 60 radical cystectomy specimens with strong or moderate intensity, showing a sensitivity of 100% and a specificity of 98.33% (when only strong and moderate staining were considered positive). When any staining intensity was considered positive, the specificity decreased to 73.33%. The previous findings were similar to the those of a study carried out in Council and Hameed (2009), in which they evaluated the expression of different immunohistochemical markers in the staining of MM and MP in radical cystectomy specimens and found that if any staining intensity of smoothelin was considered positive, the specificity for smooth muscle cells of the MP was 36%, and when the staining intensity cut-off increased to ‘moderate’, the specificity increased to 91%, whereas with using only ‘strong’ expression as a cut-off, the specificity increased to 100%.

This was also confirmed in a larger study by Paner et al. (2009), who found that the MP of all 42 evaluated sections of MP showed moderate or strong smoothelin expression. Absent or focal staining with Smoothelin was not observed in the MP in any of the cases. They also highlighted the striking disparity in the staining intensity between MM and MP muscle in every section.

Evaluating the MM, in the present study, smoothelin showed negative or weak staining in 59 out of the 60 cases, with only one case showing moderate staining. This also included the cases that showed hyperplastic MM, which mostly showed negative staining, 0%. The cases that showed weakly positive staining showed significant difference in the intensity and pattern of staining between MM and MP. Those findings were similar to those of Council and Hameed (2009), with all MM of their cases showing negative or weak expression and a single case showing moderate positivity, but a striking difference between the staining intensity of MM and MP. Those findings were also similar to the those of a study carried out by Paner et al. (2009) in which smoothelin showed predominantly absent staining in 45% of cases and weak staining in 43% of cases, with a only few (12%) showing moderate staining, with none showing strong staining. They also showed that the cystectomy specimens that had hyperplastic MM had predominantly absent or weak staining, with none showing moderate or strong staining.

The previous findings highlighted the importance of the staining intensity of smoothelin in its potential use in differentiating MM and MP. In all the studies presented, including ours, the expression of smoothelin was never weak in MP, and if it was expressed in MM, the staining pattern was almost always focal and weak, thus showing a striking difference between the MM and the MP. Therefore, the staining intensity of smoothelin was considered an important factor to differentiate between MM and MP.

Several studies also evaluated the use of smoothelin in TURBT specimens. Paner et al. (2009) evaluated 10 TUR specimens, in which MP was morphologically easily identifiable by H&E sections (with no problematic cases). Smoothelin showed strong or moderately positive staining and could identify the muscle layer proper in all cases. Later on Paner et al. (2010) also carried out a study on TUR specimens, but this time, they used problematic cases in which the identification of MP was not straightforward. All their cases were evaluated by H&E and assessment of invasion was performed in all cases; evaluation of immunohistochemistry to smoothelin was carried out afterwards and a concordance rate of 97% was achieved between H&E microscopy and smoothelin immunohistochemistry evaluations. Smoothelin staining was strong or moderate in 98% of the TURBT sections with morphologically confirmed MP. There were two discrepant cases: one in which MP showed weak staining; in the other case, smoothelin was moderately positive in a thin muscle fragment that was completely indistinguishable in the corresponding H&E slide. Thus, smoothelin showed the ability to identify MP in all cases with a sensitivity of 83% and a specificity of 95%. Furthermore, the degree of cautery or thermal artifactual distortion in routine TURBT specimens had no effect on smoothelin interpretation.

In a study carried out by Miyamoto et al. (2010), they also used diagnostically difficult TUR specimens in which the pathologists were uncertain of the pathological stage. Smoothelin showed moderate or strong expression in the eight cases showing uninvolved MP, and moderate or strong expression in MP involving the tumor in the rest of the cases. There were no differences in smoothelin expression in the muscle bundles, whether involving the tumor or not. However, in 25% of their cases, there was an overlapping staining pattern between MM and MP, but typically, Smoothelin immunoreactivity was stronger in MP than MM compared with that observed in smooth muscle of blood vessels in the same case, which is very similar to this study.

Evaluating the myofibroblastic profile, our study showed negative staining in 99% of the cases and weakly positive staining in only one case. This finding confirms the role of smoothelin as a specific marker for MP. Smoothelin could not differentiate between MM and myofibroblastic reaction as it was negative in both.

Thus, on the basis of the findings in the present study, a strong or moderate smoothelin staining on suspected MP on H&E may be interpreted as true MP; however, this requires caution and greater reliance must be placed on morphology as the final arbitrator of the interpretation.

Several findings in this and earlier studies (Paner et al., 2009, 2010) may warrant caution in the use of smoothelin in TURBT samples. First, hyperplastic MM muscles may sometimes stain for smoothelin with moderate intensity and classic MP muscle may sometimes stain weakly positive with smoothelin, as found in 2% of cases in the study carried out by Paner et al. (2010). Such staining differences can potentially be enhanced by systematic staining variations or suboptimal staining quality that may produce either too dark (false positive) or too weak (false negative) staining. Second, although the lamina propria blood vessels are a good internal positive control, they may lead to false-positive interpretation when the vascular structure is not fully appreciable on H&E. Finally, the thin trigonal superficial muscle bundles may stain strongly positive with smoothelin.

For optimal smoothelin immunohistochemistry interpretation, these approaches were recommended to contend with the potential interpretation pitfalls mentioned above. First, there should be a concomitant bladder control section on the same slide as the test section, thereby showing differential staining of MP strong immunoreactivity versus MM weak or negative immunoreactivity and validating the smoothelin staining to be optimal. Second, immunohistochemistry staining for MP in a test section should be similar to or greater in intensity than the MP in the concomitant control section. This is to minimize variability in interpretation owing to systematic or random staining variations. Third, the smoothelin immunohistochemistry staining outline or silhouette in the muscle bundles should be compatible with or in the appropriate architectural context of MP to rule out the occasional strong staining in superficial thin muscle bundles common in the trigone.

One possible approach of limiting a false-positive reaction for smoothelin in MM is to increase the smoothelin antibody dilution. The study by Khayyata et al. (2009) used a 1 : 500 dilution for smoothelin and reported that none of the 10 normal bladder sections stained with more than a weakly positive intensity in the MM. Although increased dilution may compromise the sensitivity owing to a possible weaker MP staining, this is not a concern as the main use of smoothelin to confirm the presence of suspected MP muscle. Until further optimization and enhanced criteria are fulfilled, we recommend that smoothelin should not be used to confirm the absence of MP. This is because of the few MP muscle bundles that may stain weakly with smoothelin.

It is imperative that smoothelin immunohistochemistry interpretation should always complement, and should not precede, H&E morphologic assessment. Helpful clues to the recognition of hyperplastic MM muscle bundles by morphology are haphazard intertwining of individual fibers with irregular outlines that typically occur as isolated muscle bundles in the superficial lamina propria and/or in close association with the lamina propria vascular plexus. In contrast, MP muscle bundles are frequently present in groups of compact muscle bundles with a smooth regular outline often arranged in several layers (Paner et al., 2007).

Finally, it must be emphasized that not all strong and diffuse smoothelin staining represents classic bladder MP, and this may also be present in superficial muscle bundles at the trigone and superficial smaller caliber MP of the ureter (Paner et al., 2007).

In this study, desmin was consistently positive in smooth muscle cells of the urinary bladder; it showed positive staining of MP and MM of all 60 cases whereas it showed negative staining of the myofibroblastic reaction in all 53 cases. Desmin showed 100% sensitivity and 100% specificity in staining of smooth muscle cells of the urinary bladder (whether MM or MP). These findings were similar to those of Council and Hameed (2009), who reported that Desmin stained both MM and MP in all cases with a sensitivity of 100% and a specificity 100% and was negative in the myofibroblastic reaction in all cases. In their study, they used several smooth muscle markers, and it was found that desmin, in contrast to the other markers, consistently produced strong staining of smooth muscle with negative staining of myofibroblasts. This indicates the role of desmin as a useful marker in the assessment of invasion (whether it is MM or MP invasion) and differentiation from myofibroblasts (Saha et al., 2014).

The findings in this study clearly suggest that there is a potential role for immunohistochemistry in staging of bladder carcinoma, which is somehow impacted by relying on different staining intensities in distinguishing between MM and MP. The use of vascular smooth muscle as an internal reference for comparison would help in minimizing potential variations in intensity because of different fixation methods and different antibody dilutions. We suggest the use of desmin first in diagnostically difficult TURBT cases. Desmin positivity determines the presence of invasion of smooth muscle fibers, but we cannot confirm whether it is MM or MP; thus, we use smoothelin to document the MP invasion. If desmin is negative, then there is no evidence of invasion, and the proliferating spindle cells are just myofibroblasts and not smooth muscle.

In conclusion, the data presented in this study confirm the use of smoothelin in the accurate recognition of MP from either MM or myofibroblastic reactions, thereby facilitating appropriate pathologic stage designation in often challenging TURBT specimens. It also confirmed the use of desmin to differentiate between smooth muscle cells (whether MM or MP) and proliferating myofibroblasts. Knowledge of morphology and variations of MM and MP muscle by H&E evaluation remains paramount and must always take precedence when evaluating the depth of urothelial carcinoma invasion.

Back to Top | Article Outline

Acknowledgements

Conflicts of interest

There are no conflicts of interest.

Table

Table

Table

Table

Back to Top | Article Outline

References

Council LN, Hameed O (2009). Myofibroblasts vs. smooth muscle cells and muscularis propria: a potential role for immunohistochemistry in staging bladder carcinoma. Mod Pathol 22:639–650.
De Peralta-Venturina MN, Cohen C, Amin MB (2002). Potential utility of smooth muscle myosin-heavy chain antibody in the assessment of level of invasion in urothelial carcinoma in transurethral resection of bladder tumor specimens. Mod Pathol 15:160A.
Dixon JS, Gosling JA (1983). Histology and muscle structure of the muscularis mucosa of the human urinary bladder. J Anat 136:265–271.
    Gofrit ON, Pode D, Pizov G, Zorn KC, Katz R, Duvdevani M, Shapiro A (2009). The natural history of bladder carcinoma in situ after initial response to bacillus Calmette-Guerin immunotherapy. Urol Oncol 27:258–262.
    Hall MC, Chang SS, Dalbagni G, Pruthi RS, Seigne JD, Skinner EC, et al.. (2007). Guideline for the management of nonmuscle invasive bladder cancer (stages Ta, T1, and Tis): 2007 update. J Urol 178:2314–2330.
    Jimenez RE, Keane TE, Hardy HT, Amin MB (2000). pT1 urothelial carcinoma of the bladder: criteria for diagnosis, pitfalls, and clinical implications. Adv Anat Pathol 7:13–25.
    Khayyata S, Dudas M, Rohan SM (2009). Distribution of smoothelin expression in the musculature of the genitourinary tract. Mod Pathol 22:175A.
    Kuijpers KA, Heesakkers JP, Jansen CF, Schalken JA (2007). Cadherin-11 is expressed in detrusor smooth muscle cells and myofibroblasts of normal human bladder. Eur Urol 52:1213–1221.
    Lopez-Beltran A, Requena MJ, Cheng L, Montironi R (2008). Pathological variants of invasive bladder cancer according to their suggested clinical significance. BJU Int 101:275–281.
    Maake C, Landman M, Wang X, Schmid DM, Ziegler U, John H (2006). Expression of smoothelin in the normal and the over reactive bladder. Urology 175:1152–1157.
    McKenney JK (2005). An approach to the classification of spindle cell proliferations in the urinary bladder. Adv Anat Pathol 12:312–323.
    McKenney JK, Amin MB (2005). The role of immunohistochemistry in the diagnosis of urinary bladder neoplasms. Semin Diagn Pathol 22:69–87.
    Miyamoto H, Sharma RB, Illei PB, Epstein JI (2010). Pitfalls in the use of smoothelin to identify muscularis propria invasion by urothelial carcinoma. Am J Surg Pathol 34:418–422.
    Paner GP, Ro JY, Wojcik EM, Venkataraman G, Datta MW, Amin MB (2007). Further characterization of the muscle layers and lamina propria of the urinary bladder by systematic histologic mapping: implications for pathologic staging of invasive urothelial carcinoma. Am J Surg Pathol 31:1420–1429.
    Paner GP, Shen SS, Lapetino S, Venkataraman G, Barkan GA, Quek ML, et al.. (2009). Diagnostic utility of antibody to smoothelin in the distinction of muscularis propria from muscularis mucosae of the urinary bladder: a potential ancillary tool in the pathologic staging of invasive urothelial carcinoma. Am J Surg Pathol 33:91–98.
    Paner GP, Brown JG, Lapetino S, Nese N, Gupta R, Shen SS, et al.. (2010). Diagnostic use of antibody to smoothelin in the recognition of muscularis propria in transurethral resection of urinary bladder tumor (TURBT) specimens. Am J Surg Pathol 34:792–799.
    Saha K, Saha A, Datta C (2014). Does desmin immunohistochemistry have a role in assessing stage of urothelial carcinoma in transurethral resection of bladder tumor specimens? Clin Cancer Investig J 3:502–507.
    Vakar-Lopez F, Shen SS, Zhang S, Tamboli P, Ayala AG, Ro JY (2007). Muscularis mucosae of the urinary bladder revisited with emphasis on its hyperplastic patterns: a study of a large series of cystectomy specimens. Ann Diagn Pathol 11:395–401.
    ©2015Egyptian Journal of Pathology