- Email: [email protected]
The Sensitivity of Initial Transurethral Resection or Biopsy of Bladder Tumor(s) for Detecting Bladder Cancer Variants on Radical Cystectomy
The Sensitivity of Initial Transurethral Resection or Biopsy of Bladder Tumor(s) for Detecting Bladder Cancer Variants on Radical Cystectomy Ahmed Abd el-Latif,* Katherine E. Watts,* Paul Elson, Amr Fergany and Donna E. Hansel† From the Glickman Urological and Kidney Institute (AAeL, AF), Pathology and Laboratory Medicine Institute (KEW), Taussig Cancer Institute (PE) and Genomic Medicine Institute, Cleveland Clinic (DEH), Cleveland, Ohio
Purpose: We determined the ability of bladder biopsy and transurethral resection of the bladder to accurately predict bladder cancer variants on radical cystectomy since certain variants may affect prognosis and treatment. Materials and Methods: We retrospectively evaluated the records of 302 patients who underwent biopsy and/or transurethral resection of the bladder followed by radical cystectomy from 2008 to 2010. The frequency of variant morphology and the sensitivity of the precystectomy material was determined using pathological findings at radical cystectomy as the final result. Results: Bladder cancer variants were identified in 159 patients (53%) on initial biopsy/transurethral resection and/or final pathological evaluation at radical cystectomy. The most common variant was urothelial carcinoma with squamous differentiation in 72 of 159 patients (45%), followed by micropapillary urothelial carcinoma in 41 (26%). In 9 patients (6%) variant morphology was identified only on biopsy/transurethral resection bladder and not on final radical cystectomy pathological assessment. The remaining 150 patients (94%) showed variant morphology on radical cystectomy with (79 or 53%) or without (71 or 47%) variant morphology on the preceding biopsy/transurethral resection. The sensitivity of variant detection showed a broad range by variant subtype. Overall, initial biopsy/transurethral resection sensitivity was 39% for predicting variant morphology on radical cystectomy. Conclusions: Overall sensitivity for predicting bladder cancer variants from biopsy/transurethral resection of the bladder sampling is relatively low. This is likely due to sampling and tumor heterogeneity rather than to an inaccurate pathological diagnosis. Additional predictive markers of variant morphology may be useful to determine which tumors contain aggressive variants that may alter outcomes or therapy.
Abbreviations and Acronyms RC ⫽ radical cystectomy TUR ⫽ transurethral resection TURB ⫽ bladder TUR UC ⫽ urothelial carcinoma Accepted for publication October 12, 2012. Study received institutional review board approval. * Equal study contribution. † Correspondence: 9500 Euclid Ave., Desk L25, Cleveland, Ohio 44195 (telephone: 216-444-5893; FAX: 216-445-3707; e-mail: [email protected]).
Key Words: urinary bladder, urothelium, carcinoma, cystectomy, neoplasm aggressiveness BLADDER cancer is a term often used synonymously with UC. However, in the last several decades it became clear that 1) there are distinct nonurothelial forms of bladder cancer, such as squamous cell carcinoma and adenocarci-
noma, 2) UC represents numerous different subtypes, termed variants, and 3) these variants often differ in behavior from common, conventional UC.1–3 UC variants may be present in isolation or be admixed with their conven-
0022-5347/13/1894-1263/0 THE JOURNAL OF UROLOGY® © 2013 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
http://dx.doi.org/10.1016/j.juro.2012.10.054 Vol. 189, 1263-1267, April 2013 RESEARCH, INC. Printed in U.S.A.
AND
www.jurology.com
1263
1264
SENSITIVITY OF TRANSURETHRAL RESECTION OR BIOPSY FOR DETECTING BLADDER CANCER
tional counterpart and they often show distinct immunohistochemical profiles, which likely reflect the different molecular pathways activated in these tumors.4 – 6 Some variants, such as micropapillary UC, may require different management to improve outcomes.7,8 The diagnosis of UC variants may be challenging, although aggressive course work and publications in the pathology community have led to increased awareness of these entities and improved diagnosis. From a pathology perspective the importance of recognizing UC variants is twofold. 1) The variant histology may be reflected in metastatic tumor deposits and, thus, it can reliably document metastasis from a urothelial primary site. 2) The distinction between an unusual UC variant and direct extension or metastasis from a secondary site should be considered.3,9,10 Most of these variants were recognized in the 2004 WHO classification but several rare variants that were not included are emerging in the literature.11,12 The supplementary table (http://jurology. com) lists the histological and clinical features, and prognostic significance of the variants identified in this study. For many less common or rare variants the relevance of documenting this pathological information for clinical management is still uncertain because evidence-based studies have been difficult to perform. However, based on extrapolation from well studied variants, such as micropapillary and small cell carcinoma, the importance of dissecting out less common variants may yet prove to hold significant clinical relevance.10 As an example of a recent management shift, a diagnosis of nonmuscle invasive micropapillary UC on biopsy or TUR may result in early RC, which improved survival in some series.7,8 Thus, as measured by sensitivity, we determined the ability of precystectomy biopsy or TURB specimens to detect distinct UC variants that may be present in the RC specimen.
MATERIALS AND METHODS This study was approved by the institutional review board. We retrospectively reviewed the pathology reports of patients diagnosed with a UC variant or other bladder cancer subtype at biopsy/TURB or at RC from 2008 to 2010. All cases were re-reviewed by a urological pathologist (DEH) to verify the initial diagnoses, which were originally rendered by fellowship trained urological pathologists in a subspecialty setting. Any amount of variant differentiation was reported and no significant discrepancies were identified. During this 3-year period, 302 patients underwent RC after a diagnosis of muscle invasive UC. Of these patients 159 (53%) were diagnosed with a nonconventional UC variant in pure or mixed form on RC, the matched preced-
ing biopsy or TURB. For study purposes TURB was defined as resection of all visible tumor vs partial sampling on biopsy. The supplementary table (http://jurology.com) lists the variants identified. We excluded from analysis patients who underwent only biopsy or TURB without subsequent RC as well as any with pure nonurothelial morphology of squamous cell carcinoma or adenocarcinoma. Sensitivity was calculated as the proportion of RC specimens with a particular variant that also had the variant in the biopsy or TURB specimen.
RESULTS Of the 159 cases identified for this analysis only 9 (6%) demonstrated variant(s) only on precystectomy biopsy or TURB. This suggests that in a surprisingly few cases differences in variant composition may be due to extensive sampling and complete resection of the variant on precystectomy specimens. All 9 cases showed a single variant pattern, including 4 with squamous differentiation, 2 with glandular differentiation, 2 with a small cell component and 1 with micropapillary features. The remaining 150 cases demonstrated some component of nonconventional UC on precystectomy pathological evaluation. Of these 150 RC cases 126 had a single morphological variant pattern, while the remaining 24 had multiple patterns on final pathological assessment. The most common variant at RC was UC with squamous differentiation (61 cases), followed by micropapillary differentiation (39), UC with glandular differentiation (13), small cell differentiation (24), nested carcinoma (19) and sarcomatoid differentiation (14). No change in the final diagnosis was rendered on pathological re-review. Depending on the frequency in this population, the sensitivity of biopsy vs TURB detection was 44% vs 90% for small cell, 54% vs 48% for micropapillary, 38% vs 53% for squamous, 33% vs 13% for sarcomatoid and 20% vs 25% for glandular differentiation, and 20% vs 22% for nested morphology. We did not identify any biopsy/TURB specimens that contained extremely rare variants, such as lipoid or trophoblastic differentiation, plasmacytoid features or microcystic features, which were found as small foci on the corresponding RC specimens. The table shows our results. The overall sensitivity of biopsy and TURB was 35% and 43%, respectively, to predict the final variant morphology on RC. This appears to have been only the result of tissue sampling and tumor heterogeneity, rather than differences in pathological diagnosis. The overall sensitivity of biopsy or TURB to predict variant morphology was greatest in UC cases with small cell (81%) and micropapillary (50%) differentiation, followed by tumors with glandular differentiation. Modest sensitivity was seen for sarco-
SENSITIVITY OF TRANSURETHRAL RESECTION OR BIOPSY FOR DETECTING BLADDER CANCER
Variants detected on biopsy and/or TURB and matched RC by sensitivity % Sensitivity Variant
No. Biopsy
No. TURB
No. RC
Biopsy
TURB
Overall
Small cell Micropapillary Squamous Glandular Nested Sarcomatoid Undifferentiated Plasmacytoid Trophoblastic Lipoid Microcystic
6 8 15 6 2 3 0 0 0 0 0
11 15 24 12 3 3 0 0 0 0 0
20 39 61 13 19 14 3 3 3 1 1
44 54 38 20 20 33 0 0 0 0 0
90 48 53 25 22 13 0 0 0 0 0
81 50 47 23 21 21 0 0 0 0 0
matoid differentiation and nested morphology (each 21%). However, early diagnosis of these aggressive variants was missed in at least half of biopsy and TUR specimens. Not surprisingly, nested UC was detected in only 21% of precystectomy specimens. This was a challenging diagnosis of a bland form of UC, which depends on identifying invasion into deeper levels of the bladder wall, and adequate sampling of the lamina propria and muscularis propria (detrusor muscle). Finally, rarer variants generally comprised only a minute focus of cancer on RC and included plasmacytoid, microcystic and trophoblastic differentiation. Immunohistochemical stains were used to diagnose a portion of the cases with small cell (83%), plasmacytoid (67%), trophoblastic (67%), nested (21%), sarcomatoid (20%), glandular (13%) and micropapillary (2%) differentiation.
DISCUSSION Identifying bladder cancer variant histology is important since for many variants potential prognostic and therapeutic implications accompany these diagnoses. Key examples include the recent discussion of early cystectomy for micropapillary UC even in the absence of muscularis propria (detrusor muscle) invasion as well as cisplatin targeted therapy in patients with small cell carcinoma.7,8,13–15 With increasing understanding of variant specific molecular changes and more detailed studies of the outcomes of less common variants, early diagnosis is likely to have a key role in patient directed therapy. However, early diagnosis is constrained by the material available to the pathologist at biopsy or TURB. Diagnosing bladder cancer variants depends on various factors, including the amount of tissue sampled, diagnosing pathologist skill and ancillary markers in a subset of variants. Generally, most variants occur in the presence of a background of morphologically recognizable, conventional UC and
1265
rarely in pure form. Due to the heterogeneity of these tumors, the morphological variant may only be recognized on RC when more tissue is available for assessment. In the few instances of pure variant morphology a differential diagnosis of secondary or metastatic tumors involving the bladder may be entertained, which can be often discriminated by selective immunohistochemical stains.16 Until recently, the diagnosis of UC variants was considered academic. However, our understanding of the different prognostic and diagnostic implications of some variants led to discussion of the relative importance of all variants.2,3,12 In our study the most common variant was UC with divergent differentiation, which includes squamous and glandular differentiation. Squamous differentiation was by far the most common variant identified with a frequency of approximately 24% in our entire population of patients with RC during the study period, consistent with other reports.17,18 In some studies this variant was associated with worsened outcomes, although stage-by-stage comparison suggested long-term outcomes similar to those of conventional UC.19,20 Glandular differentiation is less common and in our study it was present in approximately 4% of RC specimens. The significance of glandular differentiation is less clear and in many cases it occurs in conjunction with squamous differentiation.21 More aggressive UC variants diagnosed during our 3-year study period include small cell, micropapillary and nested UC. Small cell carcinoma may occur in pure form or be admixed with UC, squamous cell carcinoma or adenocarcinoma.22,23 It was proposed that small cell carcinoma represents clonal derivation from preexistent carcinoma and evidence to support this was described for conventional UC.24 Tumors with this form of neuroendocrine differentiation typically show poor outcomes due to extensive local invasion and distant spread.22,23 More than half of all small cell carcinomas were detected on precystectomy specimens in our study. Another aggressive variant that requires early diagnosis is micropapillary UC. Recent studies suggest that this variant behaves more aggressively than conventional UC with muscle invasion early in the disease course in many cases.7 To improve patient survival, some groups suggested early cystectomy after a diagnosis of micropapillary UC on biopsy or TUR even without documentation of muscle invasion.8 Results of this proactive intervention suggest improved long-term survival for patients with this disease.8 However, our study suggests that only a third of cases may be identified early in the disease course. Although this is beyond the scope of our study, differences in morphological features and consensus in the diagnosis of this variant may contribute to the frequency and accuracy of early diagnosis.
1266
SENSITIVITY OF TRANSURETHRAL RESECTION OR BIOPSY FOR DETECTING BLADDER CANCER
Nested UC consists of bland nests of virtually normal-appearing urothelial cells but the invasive nature of this lesion is evident on specimens that contain deeper elements of the bladder wall.25 Generally, nested UC is identified by invasion of these bland-appearing nests deep into the wall of the bladder with diagnosis often requiring clear-cut invasion.26 These tumors may carry a worsened prognosis, although this may be secondary to diagnosis at a late stage due to challenges in microscopic diagnosis.27 Approximately 21% of all carcinomas with a nested UC component were identified on precystectomy material, which required the diagnosis of deep lamina propria or muscularis propria invasion (40% and 60%, respectively) to render the diagnosis on biopsy or TURB. UC with sarcomatoid differentiation is a description of carcinoma with a spindled or mesenchymallike appearance.28 Rarely, heterologous elements may be present, such as chondrosarcomatous, osteosarcomatous and fibrosarcomatous differentiation.28 This variant appears to be more aggressive with only 37% 5-year survival.29 The sensitivity of identifying this variant on precystectomy specimens was only 21%, which may be due to the finding that the sarcomatoid component generally represents only a small proportion of the tumor volume on RC.28,29 Other variants identified only on RC included a few cases that contained plasmacytoid, trophoblastic and microcystic differentiation. Variants that were absent from our study include rhabdoid, undifferentiated giant cell and osteoclastic giant cell differentiation. Generally, only a few cases representing these rare variants have been reported in the literature. The absence of any of these variants during a 3-year study period was not unanticipated. Although we reviewed the records of all patients with a variant histology diagnosis seen in the last 3 years, our study is limited by its retrospective nature and the absence of some rare variants during the study period. These cases were also seen at original diagnosis by a group of expert genitourinary pathologists, which may be generalizable to other practices. However, extrapolating these results may
have a significant impact on discussions of the importance of early variant diagnosis and management. Briefly, our findings suggest that many UC variants may not be detected on precystectomy specimens due to tissue sampling and/or a small proportion of variants on RC. At least in a subspecialized setting, the sensitivity of detecting variant subtypes does not appear to be the effect of failure to diagnose these entities. To date the amount of variant component needed to drive outcomes remains unclear. This question was recently addressed for the micropapillary variant, which is especially pertinent in the context of early cystectomy for this disease. Samaratunga and Khoo reported that patients with moderate (10% to 50% of tumor volume) and extensive (greater than 50%) involvement by micropapillary UC are at a high risk for presenting with advanced stage disease, while patients with focal micropapillary UC (less than 10% of tumor volume) are more likely to be detected at an early stage.30 Ongoing investigation into the significance of other variants and their extent in tumor specimens is under way at many centers. Similar to micropapillary UC, it is anticipated that other variants may ultimately be treated by distinct methods. Future studies that can enhance variant detection and determine tumor behavior may ultimately be needed to better treat patients.
CONCLUSIONS The sensitivity of initial TURB to predict the final bladder cancer morphological variants was greatest for UC with a component of small cell differentiation (81%). There was lower sensitivity for other, less common forms of UC. Because many variants were recently described to portend more aggressive biological behavior, identifying these variants is important to guide patient treatment and determine final patient outcomes. These results suggest the need to develop additional surrogate markers to predict UC variants and/or aggressive tumor behavior irrespective of tumor heterogeneity.
REFERENCES 1. Young RH and Eble JN: Unusual forms of carcinoma of the urinary bladder. Hum Pathol 1991; 22: 948. 2. Amin MB, Murphy WM, Reuter VE et al: A symposium on controversies in the pathology of transitional cell carcinomas of the urinary bladder. Part I. Anat Pathol 1996; 1: 1. 3. Amin MB: Histological variants of urothelial carcinoma: diagnostic, therapeutic and prognostic implications. Mod Pathol, suppl., 2009; 22: S96.
4. Lin O, Cardillo M, Dalbagni G et al: Nested variant of urothelial carcinoma: a clinicopathologic and immunohistochemical study of 12 cases. Mod Pathol 2003; 16: 1289. 5. Sangoi AR, Higgins JP, Rouse RV et al: Immunohistochemical comparison of MUC1, CA125, and Her2Neu in invasive micropapillary carcinoma of the urinary tract and typical invasive urothelial carcinoma with retraction artifact. Mod Pathol 2009; 22: 660.
6. Ishii S, Ohbu M, Toomine Y et al: Immunohistochemical, molecular, and clinicopathological analyses of urothelial carcinoma, micropapillary variant. Pathol Int 2011; 12: 723. 7. Kamat AM, Dinney CP, Gee JR et al: Micropapillary bladder cancer: a review of the University of Texas M. D. Anderson Cancer Center experience with 100 consecutive patients. Cancer 2007; 110: 62. 8. Kamat AM, Gee JR, Dinney CP et al: The case for early cystectomy in the treatment of nonmuscle
SENSITIVITY OF TRANSURETHRAL RESECTION OR BIOPSY FOR DETECTING BLADDER CANCER
1267
invasive micropapillary bladder carcinoma. J Urol 2006; 175: 881.
rubicin and etoposide/cisplatin in small-cell urothelial cancer. J Clin Oncol 2009; 27: 2592.
clinicopathological study of 51 cases. Histopathology 2005; 46: 57.
9. Lopez-Beltran A and Cheng L: Histologic variants of urothelial carcinoma: differential diagnosis and clinical implications. Hum Pathol 2006; 37: 1371.
16. Hodges KB, Lopez-Beltran A, Emerson RE et al: Clinical utility of immunohistochemistry in the diagnoses of urinary bladder neoplasia. Appl Immunohistochem Mol Morphol 2010; 18: 401.
23. Choong NW, Quevedo JF and Kaur JS: Small cell carcinoma of the urinary bladder. The Mayo Clinic experience. Cancer 2005; 103: 1172.
10. Lopez-Beltran A, Requena MJ, Cheng L et al: Pathological variants of invasive bladder cancer according to their suggested clinical significance. BJU Int 2008; 101: 275.
17. Lopez-Beltran A, Requena MJ, Alvarez-Kindelan J et al: Squamous differentiation in primary urothelial carcinoma of the urinary tract as seen by MAC387 immunohistochemistry. J Clin Pathol 2007; 60: 332.
11. Eble JN, Epstein JI and Sesterhenn IA: Pathology and Genetics of Tumors of the Urinary System and Male Genital Organs. Lyon: IARC 2004. 12. Nigwekar P and Amin MB: The many faces of urothelial carcinoma: an update with an emphasis on recently described variants. Adv Anat Pathol 2008; 15: 218. 13. Ghoneim IA, Miocinovic R, Stephenson AJ et al: Neoadjuvant systemic therapy or early cystectomy? Single-center analysis of outcomes after therapy for patients with clinically localized micropapillary urothelial carcinoma of the bladder. Urology 2011; 77: 867. 14. Siefker-Ratdke AO, Dinney CP, Abrahams NA et al: Evidence supporting preoperative chemotherapy for small cell carcinoma of the bladder: a retrospective review of the M D Anderson cancer experience. J Urol 2004; 172: 481. 15. Siefker-Ratdke AO, Kamat AM, Grossman HB et al: Phase II clinical trial of neoadjuvant alternating doublet chemotherapy with ifosfamide/doxo-
18. Black PC, Brown GA and Dinney CP: The impact of variant histology on the outcome of bladder cancer treated with curative intent. Urol Oncol 2009; 27: 3. 19. Homma I, Masumori N, Sato E et al: Local recurrence after radical cystectomy for invasive bladder cancer: an analysis of predictive factors. Urology 2004; 64: 744. 20. El-Sebaie M, Zaghloul MS, Howard G et al: Squamous cell carcinoma of the bilharzial and nonbilharzial urinary bladder: A review of the etiological features, natural history, and management. Int J Clin Oncol 2005; 10: 20. 21. Wasco M, Daignault Y, Zhang L et al: Urothelial carcinoma with divergent histologic differentiation (mixed histologic features) predicts the presence of locally advanced bladder cancer when detected at transurethral resection. Urology 2007; 70: 69. 22. Abrahams NA, Moran C, Reyes AO et al: Small cell carcinoma of the bladder: a contemporary
24. Cheng L, Jones TD, McCarthy RP et al: Molecular genetic evidence for a common clonal origin of urinary bladder small cell carcinoma and coexisting urothelial carcinoma. Am J Pathol 2005; 166: 1533. 25. Talbert ML and Young RH: Carcinomas of the urinary bladder with deceptively benign-appearing foci. A report of three cases. Am J Surg Pathol 1989; 13: 374. 26. Dhall D, Al-Ahmadie H and Olgac S: Nested variant of urothelial carcinoma. Arch Pathol Lab Med 2007; 131: 1725. 27. Drew PA, Furman J, Civantos F et al: The nested variant of transitional cell carcinoma: an aggressive neoplasm with innocuous histology. Mod Pathol 1996; 9: 989. 28. Lopez-Beltran A, Pacelli A, Rothenberg HJ et al: Carcinosarcoma and sarcomatoid carcinoma of the bladder: clinicopathological study of 41 cases. J Urol 1998; 159: 1497. 29. Wright JL, Black PC, Brown GA et al: Differences in survival among patients with sarcomatoid carcinoma, carcinosarcoma and urothelial carcinoma of the bladder. J Urol 2007; 178: 2302. 30. Samaratunga H and Khoo K: Micropapillary variant of urothelial carcinoma of the urinary bladder; a clinicopathological and immunohistochemical study. Histopathology 2004; 45: 55.
Purpose: We determined the ability of bladder biopsy and transurethral resection of the bladder to accurately predict bladder cancer variants on radical cystectomy since certain variants may affect prognosis and treatment. Materials and Methods: We retrospectively evaluated the records of 302 patients who underwent biopsy and/or transurethral resection of the bladder followed by radical cystectomy from 2008 to 2010. The frequency of variant morphology and the sensitivity of the precystectomy material was determined using pathological findings at radical cystectomy as the final result. Results: Bladder cancer variants were identified in 159 patients (53%) on initial biopsy/transurethral resection and/or final pathological evaluation at radical cystectomy. The most common variant was urothelial carcinoma with squamous differentiation in 72 of 159 patients (45%), followed by micropapillary urothelial carcinoma in 41 (26%). In 9 patients (6%) variant morphology was identified only on biopsy/transurethral resection bladder and not on final radical cystectomy pathological assessment. The remaining 150 patients (94%) showed variant morphology on radical cystectomy with (79 or 53%) or without (71 or 47%) variant morphology on the preceding biopsy/transurethral resection. The sensitivity of variant detection showed a broad range by variant subtype. Overall, initial biopsy/transurethral resection sensitivity was 39% for predicting variant morphology on radical cystectomy. Conclusions: Overall sensitivity for predicting bladder cancer variants from biopsy/transurethral resection of the bladder sampling is relatively low. This is likely due to sampling and tumor heterogeneity rather than to an inaccurate pathological diagnosis. Additional predictive markers of variant morphology may be useful to determine which tumors contain aggressive variants that may alter outcomes or therapy.
Abbreviations and Acronyms RC ⫽ radical cystectomy TUR ⫽ transurethral resection TURB ⫽ bladder TUR UC ⫽ urothelial carcinoma Accepted for publication October 12, 2012. Study received institutional review board approval. * Equal study contribution. † Correspondence: 9500 Euclid Ave., Desk L25, Cleveland, Ohio 44195 (telephone: 216-444-5893; FAX: 216-445-3707; e-mail: [email protected]).
Key Words: urinary bladder, urothelium, carcinoma, cystectomy, neoplasm aggressiveness BLADDER cancer is a term often used synonymously with UC. However, in the last several decades it became clear that 1) there are distinct nonurothelial forms of bladder cancer, such as squamous cell carcinoma and adenocarci-
noma, 2) UC represents numerous different subtypes, termed variants, and 3) these variants often differ in behavior from common, conventional UC.1–3 UC variants may be present in isolation or be admixed with their conven-
0022-5347/13/1894-1263/0 THE JOURNAL OF UROLOGY® © 2013 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
http://dx.doi.org/10.1016/j.juro.2012.10.054 Vol. 189, 1263-1267, April 2013 RESEARCH, INC. Printed in U.S.A.
AND
www.jurology.com
1263
1264
SENSITIVITY OF TRANSURETHRAL RESECTION OR BIOPSY FOR DETECTING BLADDER CANCER
tional counterpart and they often show distinct immunohistochemical profiles, which likely reflect the different molecular pathways activated in these tumors.4 – 6 Some variants, such as micropapillary UC, may require different management to improve outcomes.7,8 The diagnosis of UC variants may be challenging, although aggressive course work and publications in the pathology community have led to increased awareness of these entities and improved diagnosis. From a pathology perspective the importance of recognizing UC variants is twofold. 1) The variant histology may be reflected in metastatic tumor deposits and, thus, it can reliably document metastasis from a urothelial primary site. 2) The distinction between an unusual UC variant and direct extension or metastasis from a secondary site should be considered.3,9,10 Most of these variants were recognized in the 2004 WHO classification but several rare variants that were not included are emerging in the literature.11,12 The supplementary table (http://jurology. com) lists the histological and clinical features, and prognostic significance of the variants identified in this study. For many less common or rare variants the relevance of documenting this pathological information for clinical management is still uncertain because evidence-based studies have been difficult to perform. However, based on extrapolation from well studied variants, such as micropapillary and small cell carcinoma, the importance of dissecting out less common variants may yet prove to hold significant clinical relevance.10 As an example of a recent management shift, a diagnosis of nonmuscle invasive micropapillary UC on biopsy or TUR may result in early RC, which improved survival in some series.7,8 Thus, as measured by sensitivity, we determined the ability of precystectomy biopsy or TURB specimens to detect distinct UC variants that may be present in the RC specimen.
MATERIALS AND METHODS This study was approved by the institutional review board. We retrospectively reviewed the pathology reports of patients diagnosed with a UC variant or other bladder cancer subtype at biopsy/TURB or at RC from 2008 to 2010. All cases were re-reviewed by a urological pathologist (DEH) to verify the initial diagnoses, which were originally rendered by fellowship trained urological pathologists in a subspecialty setting. Any amount of variant differentiation was reported and no significant discrepancies were identified. During this 3-year period, 302 patients underwent RC after a diagnosis of muscle invasive UC. Of these patients 159 (53%) were diagnosed with a nonconventional UC variant in pure or mixed form on RC, the matched preced-
ing biopsy or TURB. For study purposes TURB was defined as resection of all visible tumor vs partial sampling on biopsy. The supplementary table (http://jurology.com) lists the variants identified. We excluded from analysis patients who underwent only biopsy or TURB without subsequent RC as well as any with pure nonurothelial morphology of squamous cell carcinoma or adenocarcinoma. Sensitivity was calculated as the proportion of RC specimens with a particular variant that also had the variant in the biopsy or TURB specimen.
RESULTS Of the 159 cases identified for this analysis only 9 (6%) demonstrated variant(s) only on precystectomy biopsy or TURB. This suggests that in a surprisingly few cases differences in variant composition may be due to extensive sampling and complete resection of the variant on precystectomy specimens. All 9 cases showed a single variant pattern, including 4 with squamous differentiation, 2 with glandular differentiation, 2 with a small cell component and 1 with micropapillary features. The remaining 150 cases demonstrated some component of nonconventional UC on precystectomy pathological evaluation. Of these 150 RC cases 126 had a single morphological variant pattern, while the remaining 24 had multiple patterns on final pathological assessment. The most common variant at RC was UC with squamous differentiation (61 cases), followed by micropapillary differentiation (39), UC with glandular differentiation (13), small cell differentiation (24), nested carcinoma (19) and sarcomatoid differentiation (14). No change in the final diagnosis was rendered on pathological re-review. Depending on the frequency in this population, the sensitivity of biopsy vs TURB detection was 44% vs 90% for small cell, 54% vs 48% for micropapillary, 38% vs 53% for squamous, 33% vs 13% for sarcomatoid and 20% vs 25% for glandular differentiation, and 20% vs 22% for nested morphology. We did not identify any biopsy/TURB specimens that contained extremely rare variants, such as lipoid or trophoblastic differentiation, plasmacytoid features or microcystic features, which were found as small foci on the corresponding RC specimens. The table shows our results. The overall sensitivity of biopsy and TURB was 35% and 43%, respectively, to predict the final variant morphology on RC. This appears to have been only the result of tissue sampling and tumor heterogeneity, rather than differences in pathological diagnosis. The overall sensitivity of biopsy or TURB to predict variant morphology was greatest in UC cases with small cell (81%) and micropapillary (50%) differentiation, followed by tumors with glandular differentiation. Modest sensitivity was seen for sarco-
SENSITIVITY OF TRANSURETHRAL RESECTION OR BIOPSY FOR DETECTING BLADDER CANCER
Variants detected on biopsy and/or TURB and matched RC by sensitivity % Sensitivity Variant
No. Biopsy
No. TURB
No. RC
Biopsy
TURB
Overall
Small cell Micropapillary Squamous Glandular Nested Sarcomatoid Undifferentiated Plasmacytoid Trophoblastic Lipoid Microcystic
6 8 15 6 2 3 0 0 0 0 0
11 15 24 12 3 3 0 0 0 0 0
20 39 61 13 19 14 3 3 3 1 1
44 54 38 20 20 33 0 0 0 0 0
90 48 53 25 22 13 0 0 0 0 0
81 50 47 23 21 21 0 0 0 0 0
matoid differentiation and nested morphology (each 21%). However, early diagnosis of these aggressive variants was missed in at least half of biopsy and TUR specimens. Not surprisingly, nested UC was detected in only 21% of precystectomy specimens. This was a challenging diagnosis of a bland form of UC, which depends on identifying invasion into deeper levels of the bladder wall, and adequate sampling of the lamina propria and muscularis propria (detrusor muscle). Finally, rarer variants generally comprised only a minute focus of cancer on RC and included plasmacytoid, microcystic and trophoblastic differentiation. Immunohistochemical stains were used to diagnose a portion of the cases with small cell (83%), plasmacytoid (67%), trophoblastic (67%), nested (21%), sarcomatoid (20%), glandular (13%) and micropapillary (2%) differentiation.
DISCUSSION Identifying bladder cancer variant histology is important since for many variants potential prognostic and therapeutic implications accompany these diagnoses. Key examples include the recent discussion of early cystectomy for micropapillary UC even in the absence of muscularis propria (detrusor muscle) invasion as well as cisplatin targeted therapy in patients with small cell carcinoma.7,8,13–15 With increasing understanding of variant specific molecular changes and more detailed studies of the outcomes of less common variants, early diagnosis is likely to have a key role in patient directed therapy. However, early diagnosis is constrained by the material available to the pathologist at biopsy or TURB. Diagnosing bladder cancer variants depends on various factors, including the amount of tissue sampled, diagnosing pathologist skill and ancillary markers in a subset of variants. Generally, most variants occur in the presence of a background of morphologically recognizable, conventional UC and
1265
rarely in pure form. Due to the heterogeneity of these tumors, the morphological variant may only be recognized on RC when more tissue is available for assessment. In the few instances of pure variant morphology a differential diagnosis of secondary or metastatic tumors involving the bladder may be entertained, which can be often discriminated by selective immunohistochemical stains.16 Until recently, the diagnosis of UC variants was considered academic. However, our understanding of the different prognostic and diagnostic implications of some variants led to discussion of the relative importance of all variants.2,3,12 In our study the most common variant was UC with divergent differentiation, which includes squamous and glandular differentiation. Squamous differentiation was by far the most common variant identified with a frequency of approximately 24% in our entire population of patients with RC during the study period, consistent with other reports.17,18 In some studies this variant was associated with worsened outcomes, although stage-by-stage comparison suggested long-term outcomes similar to those of conventional UC.19,20 Glandular differentiation is less common and in our study it was present in approximately 4% of RC specimens. The significance of glandular differentiation is less clear and in many cases it occurs in conjunction with squamous differentiation.21 More aggressive UC variants diagnosed during our 3-year study period include small cell, micropapillary and nested UC. Small cell carcinoma may occur in pure form or be admixed with UC, squamous cell carcinoma or adenocarcinoma.22,23 It was proposed that small cell carcinoma represents clonal derivation from preexistent carcinoma and evidence to support this was described for conventional UC.24 Tumors with this form of neuroendocrine differentiation typically show poor outcomes due to extensive local invasion and distant spread.22,23 More than half of all small cell carcinomas were detected on precystectomy specimens in our study. Another aggressive variant that requires early diagnosis is micropapillary UC. Recent studies suggest that this variant behaves more aggressively than conventional UC with muscle invasion early in the disease course in many cases.7 To improve patient survival, some groups suggested early cystectomy after a diagnosis of micropapillary UC on biopsy or TUR even without documentation of muscle invasion.8 Results of this proactive intervention suggest improved long-term survival for patients with this disease.8 However, our study suggests that only a third of cases may be identified early in the disease course. Although this is beyond the scope of our study, differences in morphological features and consensus in the diagnosis of this variant may contribute to the frequency and accuracy of early diagnosis.
1266
SENSITIVITY OF TRANSURETHRAL RESECTION OR BIOPSY FOR DETECTING BLADDER CANCER
Nested UC consists of bland nests of virtually normal-appearing urothelial cells but the invasive nature of this lesion is evident on specimens that contain deeper elements of the bladder wall.25 Generally, nested UC is identified by invasion of these bland-appearing nests deep into the wall of the bladder with diagnosis often requiring clear-cut invasion.26 These tumors may carry a worsened prognosis, although this may be secondary to diagnosis at a late stage due to challenges in microscopic diagnosis.27 Approximately 21% of all carcinomas with a nested UC component were identified on precystectomy material, which required the diagnosis of deep lamina propria or muscularis propria invasion (40% and 60%, respectively) to render the diagnosis on biopsy or TURB. UC with sarcomatoid differentiation is a description of carcinoma with a spindled or mesenchymallike appearance.28 Rarely, heterologous elements may be present, such as chondrosarcomatous, osteosarcomatous and fibrosarcomatous differentiation.28 This variant appears to be more aggressive with only 37% 5-year survival.29 The sensitivity of identifying this variant on precystectomy specimens was only 21%, which may be due to the finding that the sarcomatoid component generally represents only a small proportion of the tumor volume on RC.28,29 Other variants identified only on RC included a few cases that contained plasmacytoid, trophoblastic and microcystic differentiation. Variants that were absent from our study include rhabdoid, undifferentiated giant cell and osteoclastic giant cell differentiation. Generally, only a few cases representing these rare variants have been reported in the literature. The absence of any of these variants during a 3-year study period was not unanticipated. Although we reviewed the records of all patients with a variant histology diagnosis seen in the last 3 years, our study is limited by its retrospective nature and the absence of some rare variants during the study period. These cases were also seen at original diagnosis by a group of expert genitourinary pathologists, which may be generalizable to other practices. However, extrapolating these results may
have a significant impact on discussions of the importance of early variant diagnosis and management. Briefly, our findings suggest that many UC variants may not be detected on precystectomy specimens due to tissue sampling and/or a small proportion of variants on RC. At least in a subspecialized setting, the sensitivity of detecting variant subtypes does not appear to be the effect of failure to diagnose these entities. To date the amount of variant component needed to drive outcomes remains unclear. This question was recently addressed for the micropapillary variant, which is especially pertinent in the context of early cystectomy for this disease. Samaratunga and Khoo reported that patients with moderate (10% to 50% of tumor volume) and extensive (greater than 50%) involvement by micropapillary UC are at a high risk for presenting with advanced stage disease, while patients with focal micropapillary UC (less than 10% of tumor volume) are more likely to be detected at an early stage.30 Ongoing investigation into the significance of other variants and their extent in tumor specimens is under way at many centers. Similar to micropapillary UC, it is anticipated that other variants may ultimately be treated by distinct methods. Future studies that can enhance variant detection and determine tumor behavior may ultimately be needed to better treat patients.
CONCLUSIONS The sensitivity of initial TURB to predict the final bladder cancer morphological variants was greatest for UC with a component of small cell differentiation (81%). There was lower sensitivity for other, less common forms of UC. Because many variants were recently described to portend more aggressive biological behavior, identifying these variants is important to guide patient treatment and determine final patient outcomes. These results suggest the need to develop additional surrogate markers to predict UC variants and/or aggressive tumor behavior irrespective of tumor heterogeneity.
REFERENCES 1. Young RH and Eble JN: Unusual forms of carcinoma of the urinary bladder. Hum Pathol 1991; 22: 948. 2. Amin MB, Murphy WM, Reuter VE et al: A symposium on controversies in the pathology of transitional cell carcinomas of the urinary bladder. Part I. Anat Pathol 1996; 1: 1. 3. Amin MB: Histological variants of urothelial carcinoma: diagnostic, therapeutic and prognostic implications. Mod Pathol, suppl., 2009; 22: S96.
4. Lin O, Cardillo M, Dalbagni G et al: Nested variant of urothelial carcinoma: a clinicopathologic and immunohistochemical study of 12 cases. Mod Pathol 2003; 16: 1289. 5. Sangoi AR, Higgins JP, Rouse RV et al: Immunohistochemical comparison of MUC1, CA125, and Her2Neu in invasive micropapillary carcinoma of the urinary tract and typical invasive urothelial carcinoma with retraction artifact. Mod Pathol 2009; 22: 660.
6. Ishii S, Ohbu M, Toomine Y et al: Immunohistochemical, molecular, and clinicopathological analyses of urothelial carcinoma, micropapillary variant. Pathol Int 2011; 12: 723. 7. Kamat AM, Dinney CP, Gee JR et al: Micropapillary bladder cancer: a review of the University of Texas M. D. Anderson Cancer Center experience with 100 consecutive patients. Cancer 2007; 110: 62. 8. Kamat AM, Gee JR, Dinney CP et al: The case for early cystectomy in the treatment of nonmuscle
SENSITIVITY OF TRANSURETHRAL RESECTION OR BIOPSY FOR DETECTING BLADDER CANCER
1267
invasive micropapillary bladder carcinoma. J Urol 2006; 175: 881.
rubicin and etoposide/cisplatin in small-cell urothelial cancer. J Clin Oncol 2009; 27: 2592.
clinicopathological study of 51 cases. Histopathology 2005; 46: 57.
9. Lopez-Beltran A and Cheng L: Histologic variants of urothelial carcinoma: differential diagnosis and clinical implications. Hum Pathol 2006; 37: 1371.
16. Hodges KB, Lopez-Beltran A, Emerson RE et al: Clinical utility of immunohistochemistry in the diagnoses of urinary bladder neoplasia. Appl Immunohistochem Mol Morphol 2010; 18: 401.
23. Choong NW, Quevedo JF and Kaur JS: Small cell carcinoma of the urinary bladder. The Mayo Clinic experience. Cancer 2005; 103: 1172.
10. Lopez-Beltran A, Requena MJ, Cheng L et al: Pathological variants of invasive bladder cancer according to their suggested clinical significance. BJU Int 2008; 101: 275.
17. Lopez-Beltran A, Requena MJ, Alvarez-Kindelan J et al: Squamous differentiation in primary urothelial carcinoma of the urinary tract as seen by MAC387 immunohistochemistry. J Clin Pathol 2007; 60: 332.
11. Eble JN, Epstein JI and Sesterhenn IA: Pathology and Genetics of Tumors of the Urinary System and Male Genital Organs. Lyon: IARC 2004. 12. Nigwekar P and Amin MB: The many faces of urothelial carcinoma: an update with an emphasis on recently described variants. Adv Anat Pathol 2008; 15: 218. 13. Ghoneim IA, Miocinovic R, Stephenson AJ et al: Neoadjuvant systemic therapy or early cystectomy? Single-center analysis of outcomes after therapy for patients with clinically localized micropapillary urothelial carcinoma of the bladder. Urology 2011; 77: 867. 14. Siefker-Ratdke AO, Dinney CP, Abrahams NA et al: Evidence supporting preoperative chemotherapy for small cell carcinoma of the bladder: a retrospective review of the M D Anderson cancer experience. J Urol 2004; 172: 481. 15. Siefker-Ratdke AO, Kamat AM, Grossman HB et al: Phase II clinical trial of neoadjuvant alternating doublet chemotherapy with ifosfamide/doxo-
18. Black PC, Brown GA and Dinney CP: The impact of variant histology on the outcome of bladder cancer treated with curative intent. Urol Oncol 2009; 27: 3. 19. Homma I, Masumori N, Sato E et al: Local recurrence after radical cystectomy for invasive bladder cancer: an analysis of predictive factors. Urology 2004; 64: 744. 20. El-Sebaie M, Zaghloul MS, Howard G et al: Squamous cell carcinoma of the bilharzial and nonbilharzial urinary bladder: A review of the etiological features, natural history, and management. Int J Clin Oncol 2005; 10: 20. 21. Wasco M, Daignault Y, Zhang L et al: Urothelial carcinoma with divergent histologic differentiation (mixed histologic features) predicts the presence of locally advanced bladder cancer when detected at transurethral resection. Urology 2007; 70: 69. 22. Abrahams NA, Moran C, Reyes AO et al: Small cell carcinoma of the bladder: a contemporary
24. Cheng L, Jones TD, McCarthy RP et al: Molecular genetic evidence for a common clonal origin of urinary bladder small cell carcinoma and coexisting urothelial carcinoma. Am J Pathol 2005; 166: 1533. 25. Talbert ML and Young RH: Carcinomas of the urinary bladder with deceptively benign-appearing foci. A report of three cases. Am J Surg Pathol 1989; 13: 374. 26. Dhall D, Al-Ahmadie H and Olgac S: Nested variant of urothelial carcinoma. Arch Pathol Lab Med 2007; 131: 1725. 27. Drew PA, Furman J, Civantos F et al: The nested variant of transitional cell carcinoma: an aggressive neoplasm with innocuous histology. Mod Pathol 1996; 9: 989. 28. Lopez-Beltran A, Pacelli A, Rothenberg HJ et al: Carcinosarcoma and sarcomatoid carcinoma of the bladder: clinicopathological study of 41 cases. J Urol 1998; 159: 1497. 29. Wright JL, Black PC, Brown GA et al: Differences in survival among patients with sarcomatoid carcinoma, carcinosarcoma and urothelial carcinoma of the bladder. J Urol 2007; 178: 2302. 30. Samaratunga H and Khoo K: Micropapillary variant of urothelial carcinoma of the urinary bladder; a clinicopathological and immunohistochemical study. Histopathology 2004; 45: 55.