Lymphovascular invasion predicts poor prognosis in high-grade pt1 bladder cancer patients who underwent transurethral resection in one piece

Transcription

1 JJCO Japanese Journal of Clinical Oncology Japanese Journal of Clinical Oncology, 2017, 47(5) doi: /jjco/hyx012 Advance Access Publication Date: 10 February 2017 Original Article Original Article Lymphovascular invasion predicts poor prognosis in high-grade pt1 bladder cancer patients who underwent transurethral resection in one piece Rinzo Ukai 1, *, Kunihiro Hashimoto 1, Hirofumi Nakayama 2, and Toshiyuki Iwamoto 3 1 Department of Urology, JR Hiroshima Hospital, Hiroshima, 2 Department of Pathology and Laboratory Medicine, JR Hiroshima Hospital, Hiroshima, and 3 Department of Pathology, Chuden Hospital, Hiroshima, Japan *For reprints and all correspondence: Rinzo Ukai, Department of Urology, JR Hiroshima Hospital, Futaba-no-sato, Higashi-ku, Hiroshima , Japan. ukai-r@sa2.so-net.ne.jp Received 23 October 2016; Editorial decision 23 January 2017; Accepted 27 January 2017 Abstract Background: Lymphovascular invasion (LVI) in high-grade clinical T1 bladder cancer is usually considered a poor prognostic factor, but it is often difficult to achieve correct staging of T1 bladder cancer and diagnose the presence of LVI because of the inadequacy of conventional transurethral resection specimens. The aims of this study were to evaluate the prognostic value of LVI in patients with correctly staged high-grade pathological T1 (pt1) bladder cancer who initially underwent transurethral resection in one piece (TURBO). Methods: Eighty-six high-grade pt1 bladder cancer patients who underwent TURBO were enrolled. Risk of tumor understaging was avoided by examining the vertical resection margin of the TURBO specimen. Immunohistochemical staining using D2-40 and CD31 was performed to confirm LVI. We examined the association of LVI with other clinicopathological factors and the impact of LVI on progression-free survival and cancer-specific survival. Results: The median follow-up period was 49 months (range, 6 142). In all patients, the tumors were accurately staged as pt1 at initial TURBO. LVI was detected in 15 patients (17%) and was significantly associated with tumor growth pattern (P = 0.001). Multivariate analysis identified LVI as the only independent predictor for reduced progression-free survival (HR, 4.48; 95% CI, ; P = 0.009) and cancer-specific survival (HR, 4.35; 95% CI, ; P = 0.029). Conclusions: The presence of LVI in TURBO specimens independently predicts poor clinical outcomes in patients with high-grade pt1 bladder cancer. This information may help urologists to counsel their patients when deciding whether to choose a bladder-preserving strategy or radical cystectomy. Key words: urinary bladder neoplasms, urothelial carcinoma, lymphovascular invasion, disease progression, survival Introduction In non-muscle-invasive bladder cancer (NMIBC), ~60% of patients present with pta, 30% with pt1, and 10% with carcinoma in situ (CIS) lesions (1), T1 tumors are mostly high grade, and high-grade clinical stage T1 (HGT1; formerly T1G3) urothelial carcinoma of the bladder (UCB) is biologically the most aggressive phenotype among NMIBCs (2). In a limited data set on the clinical significance of lymphovascular invasion (LVI) in NMIBC patients, the presence of LVI detected by conventional transurethral resection of bladder tumor (TURBT) in HGT1 UCB patients has been suggested to be predictive of poor The Author Published by Oxford University Press. All rights reserved. For Permissions, please journals.permissions@oup.com. 447

2 448 LVI in HG pt1 UCB prognosis (3 6). However, because of the poor quality of TURBT specimens, caused by fragmentation, disorientation and cautery burn, it is difficult to identify the presence of LVI by hematoxylin and eosin (H&E) staining alone (5). Furthermore, tumors diagnosed as HGT1 have a greater tendency to be understaged, especially when the TURBT specimens are devoid of muscle tissue (7), potentially leading to tumor progression and death. The presence of LVI in TURBT specimens has also been associated with understaging and a decrease in cancer-specific survival rates after radical cystectomy (RC) (8). To avoid such understaging, repeat or restaging TURBT has been recommended in guidelines (9), but a substantial risk of understaging still remains (10). Thus, the results of a prognostic study may be questionable if it is based on TURBT specimens in patients with HGT1 UCB harboring LVI. In this study, we used transurethral resection in one piece (TURBO) as a routine tool for complete tumor resection, accurate staging, and identification of LVI (11). On the basis of the pathological findings, we examined the influence of LVI on progressionfree survival (PFS) and cancer-specific survival (CSS) in patients with high-grade pathological T1 (HG pt1) UCB, and determined whether LVI is associated with other clinicopathological factors. Patients and methods Patient selection and data collection This single-institution study was approved by the ethics committee of JR Hiroshima Hospital (approval number: 20147). The ethics committee issued an informed consent waiver because this was a retrospective study. From April 2000 to March 2012, 559 consecutive patients were newly diagnosed with UCB at the Department of Urology in JR Hiroshima Hospital. All patients underwent transurethral resection (TURBT and/or TURBO), biopsy or both. Of these, 86 patients who underwent visually complete resection by TURBO and were histologically diagnosed with primary HG pt1 UCB were enrolled. Exclusion criteria were bladder tumors associated with upper tract carcinoma during the study period, tumors with non-urothelial histology, and a positive vertical resection margin on the TURBO specimen. Clinical data were collected by a retrospective review of the medical records. TURBO and pathological analysis Surgical indications for TURBO are basically identical to those for TURBT, as long as the aim of TURBO is to perform a complete resection of the target tumor. While the operation time is presumed to be roughly one and a half times longer for TURBO compared with TURBT, conventional spinal anesthesia is effective and adequate for the entire course of the operation. TURBO, by which all the tumor tissue including the lamina propria and part of the detrusor muscle can be obtained in one piece, was performed as previously reported (12). Briefly, first, a circumscribed incision was made around the index tumor using a needle electrode. Next, the lesion was dissected through the superficial muscularis propria under a predetermined resection area. The dissection was completed and the tumor specimen was grasped with forceps and pulled out via the resectoscope sheath or via the urethra directly. Here, TURBO was defined as an en bloc resection or a planned piecemeal resection of up to three pieces that could be reconstructed close to the original shape. Additionally, in some cases involving larger or solid tumors, only the exophytic portion was excised by TURBT prior to TURBO, to reduce the tumor volume for securing retrieval through the urethra. Resected specimens were fixed in 10% buffered formalin and then sectioned serially in 1- to 3-mm intervals. All slices were embedded in a paraffin block and then subjected to histological analysis including immunohistochemistry (Fig. 1). The depth of invasion, section margins and LVI were carefully observed. The tumor grade was assigned according to the 2004 World Health Organization classification (13,14), and tumor staging was performed according to the 2010 American Joint Committee on Cancer (15). Immunohistochemical staining Immunohistochemistry was performed using the Ventana BenchMark GX automated staining system (Roche Diagnostics, Tokyo, Japan). Serial sections (4-μm thick) from the same blocks that were found to show LVI on H&E staining were stained with podoplanin (clone D2-40, pre-diluted; Nichirei, Tokyo, Japan) and CD31 (clone JC/70A; 1:50 dilution; DAKO, Kyoto, Japan). Structures positive for only CD31 were considered to be blood vessels, while those showing immunoreactivity with both CD31 and D2-40 were considered to be lymphatic vessels. LVI was defined as the presence of both lymphatic and vascular invasion. Treatment All 86 patients initially underwent TURBO. The agents used for intravesical instillation after TURBO included bacillus Calmette- Guerin (BCG) in 40 patients and mitomycin-c in four patients; Figure 1. TURBO specimen with lymphovascular invasion confirmed by D2-40 immunostaining. A solid tumor, measuring 12 mm in the long-axis diameter, was resected en bloc (A). The en face section stained with hematoxylin and eosin (H&E, 20) (B). In the area with a black square in B, multiple foci of lymphatic invasion (arrowheads) are found in the lamina propria (D2-40 immunostaining, 100) (C). Inset: Highpower view of the lymphatic invasion indicated by a red arrowhead in C ( 200).

3 Jpn J Clin Oncol, 2017, Vol. 47, No however, a standardized instillation protocol was not strictly followed in all of the patients. Because the TURBO procedure principally removes an entire tumor with normal surrounding tissue in one piece, we assume that there is no residual tumor left in the bladder after TURBO, except in cases with macroscopically invisible CIS. Therefore, we use BCG instillation mainly in patients with pathologically confirmed concomitant CIS. RC was performed in seven patients. Of these, four with multiple tumors and/or deep lamina propria invasion underwent early cystectomy. Among these four patients, only one had developed LVI at the initial TURBO. Deferred cystectomy was performed in the other three patients who had invasive intravesical recurrence after bladder instillation or any confirmed disease progression. Among the three patients, one had developed LVI at the initial TURBO. The other treatments included systemic chemotherapy or chemoradiation after confirmation of cancer progression. Follow-up After the initial TURBO, patients were generally followed up at least every 3 4 months with cystoscopy and urine cytology for 3 years, and every 6 months thereafter. Diagnostic imaging, mainly chest and abdominal computed tomography, were performed annually or as clinically indicated. Early cystectomy was defined as one 6 months of the initial TURBO, and deferred cystectomy as one >6 months after a TURBO. Bladder recurrence was defined as tumor relapse in the bladder. Progression was defined as the development of muscle-invasive disease ( T2) or the presence of metastatic disease. PFS was defined as the time interval between TURBO and progression. CSS was defined as the time interval between TURBO and death from bladder cancer. When patients died, the cause of death was determined by the attending physicians, by chart review corroborated by death certificates, or by death certificates alone. Statistical analysis Clinicopathological factors were analyzed using the χ 2 test or Fisher s exact test to test for differences between the presence and absence of LVI. Kaplan Meier curves were used to estimate PFS and CSS, and the differences were assessed using the log-rank tests. Univariate and multivariate analyses were performed using Cox proportional hazards regression model. The multivariate Cox analyses were performed to control confoundings which are not necessarily significant in univariate analysis, and in order to avoid overfitting we incorporated as few variables as possible that are prognostically meaningful into the multivariate analyses. Hazard ratios (HRs) and their 95% confidence intervals (CIs) were calculated for significant predictors. A two-tailed P value <0.05 was considered statistically significant. All statistical analyses were performed with EZR version 1.24 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). Results TURBO and pathological evaluation TURBO was performed in all 86 patients without any serious complications, regardless of whether their tumors were resected en bloc or piecemeal (up to three pieces). Because the vertical margins of the lamina propria were distinctly cancer-free, the pathological staging as pt1 was correct in all cases, comprising 78 (91%) cases with the muscularis propria and 8 (9%) with no muscle beneath the tumor base. Association of LVI with clinicopathological factors Of the 86 patients, we performed immunohistochemical staining in 35 (41%) of possible LVI lesions on H&E staining. Among them, 17 (20%) were negative for LVI, as also expected on H&E staining. Of the remaining 18, 15 LVI-positive cases (17%) were presented in this series. Of the other three with suspected LVI on H&E stained specimens, whose results were negative on immunostaining, two were diagnosed with retraction artifacts and one was diagnosed as having no cancer invasion of the lymphatics or vasculature. Table 1 shows the association between LVI and clinicopathological factors. The median age of the patients was 74 years (range, years), and 87% (75/86) were men. The median size of the resected specimens was 18 mm (range, 5 46mm). ConcomitantCISwasfoundin44patients(51%), including those with CIS lesions found only histologically on most of the lateral margins around the tumor. LVI was significantly associated with tumor growth pattern (P = 0.001), but not with age, sex, tumor size, multiplicity, number of resection pieces, concomitant CIS, bladder recurrence or intravesical instillation. Table 1. Association between lymphovascular invasion and clinicopathological factors in 86 patients who underwent TURBO Characteristics No. pts (%) No. pts with LVI (%) No Yes P value Age group (years) <70 22 (26) (74) Sex Men 75 (87) Women 11 (13) 9 2 Tumor size (cm) <3 72 (84) (16) 9 5 Multiplicity Single 57 (66) (34) 21 8 Resection pieces En bloc 71 (83) or 3 piecemeal 15 (17) 13 2 Concomitant CIS Absent 42 (49) 38 4 Present 44 (51) Growth pattern Papillary 50 (58) 47 3 Solid 36 (42) Bladder recurrence No 43 (50) 35 8 Yes ( 1) 43 (50) 36 7 Intravesical instillation No 42 (49) 37 5 Yes 44 (51) Progression No 72 (84) 63 9 Yes 14 (16) 8 6 Death from bladder cancer No 76 (88) Yes 10 (12) 5 5 TURBO, transurethral resection of bladder tumor in one piece; LVI, lymphovascular invasion; No., number of; pts., patients; CIS, carcinoma in situ.

4 450 LVI in HG pt1 UCB Association of LVI with survival At a median follow-up period of 49 months (range, months), 14 (16%) patients showed disease progression, of whom six patients (7%) had already developed LVI at the initial TURBO. Thirteen patients (15%) died of any cause, and of these, 10 (12%) died of bladder cancer. Among the 10 patients who died of bladder cancer, five (6%) had developed LVI as seen in their TURBO specimens, one of whom had undergone deferred cystectomy. One patient with LVI at TURBO who underwent early cystectomy survived without disease progression for 10 years at the last follow-up. When survival was compared between patients with and without LVI, Kaplan Meier analysis showed a significant difference in PFS and CSS (P = and P = 0.006, respectively) (Fig. 2). The 5-year PFS rates are 85.0% in LVI-negative patients and 64.8% in LVI-positive patients, and the 5-year CSS rates are 91.9% in LVI-negative patients and 65.8% in LVI-positive patients. Univariate Cox proportional hazards analysis revealed that tumor growth pattern (HR, 3.39; 95% CI, ; P = 0.039) and LVI (HR, 4.95; 95% CI, ; P = 0.003) significantly affected PFS, and only LVI was identified as a predictor of reduced CSS (HR, 4.91; 95% CI, ; P = 0.013) (Table 2). Multivariate Cox proportional hazard analyses that included tumor size, tumor growth pattern and concomitant CIS and LVI revealed that LVI was the only independent predictor of both PFS (HR, 4.48; 95% CI, ; P = 0.009) and CSS (HR, 4.35; 95% CI, ; P = 0.029) (Table 3). Discussion The biological behavior of HGT1 UCB diagnosed at the time of the initial TURBT is difficult to predict, because HGT1, a subset of NMIBCs, is still a heterogeneous group with variable outcomes. In this study, we examined LVI in TURBO specimens from 86 patients with HGT1 UCB that was pathologically staged as pt1 tumor. Our results revealed that the presence of LVI was the only independent prognostic factor for both PFS and CSS in multivariate analyses. While we found that tumor growth pattern was also associated with LVI, it was not an independent predictor of PFS or CSS. Importantly, all cases of HG pt1 UCB in the present study were not only macroscopically completely resected but also correctly staged by examining the vertical resection margin on the entirely step-sectioned tumor specimen obtained by TURBO. Furthermore, the identification of LVI in the serial histology sections was confirmed by auxiliary immunohistochemical staining. Thus, with accurate pathological staging and a reliable immunostaining method for LVI, we demonstrated that the presence of LVI plays an important role in tumor progression and is predictive of poor prognosis in patients with HG pt1 UCB. In contrast to several studies on RC specimens, the clinical implication of LVI in TURBT specimens has been sparsely investigated in patients with NMIBC (3 6). One of the reasons for this is considered to be diagnostic difficulty owing to the poor quality of TURBT specimens. The reported LVI rates in patients with T1 any grade UCB, or T1G3 UCB or HGT1 UCB on TURBT were 10 36% (3 6,16). Previous studies also revealed that in the multivariate analyses, the presence of LVI in patients with T1 UCB was independently associated with PFS, CSS (3) and overall survival (4,6). These results are comparable to our findings of an LVI rate of 17% and the presence of LVI as an independent predictor of PFS and CSS. In the present study, we identified LVI in patients with HG pt1 UCB on H&E-stained slides at first and then confirmed the presence of LVI with immunohistochemical staining using the endothelial cell markers, i.e. D2-40 and CD31. In two cases with apparent LVI on H&E staining, the lesions were finally diagnosed as retraction artifacts mimicking vascular invasion, since the D2-40 and CD31 immunostaining could discern retraction artifacts that isolate nests of tumor cells from true tumor emboli in lymphovascular spaces. Moreover, a tumor embolus that completely obliterates the lumen of the lymphatic channel could not be visualized on H&E sections, even more in the fragmented TURBT specimens (17). Most previous studies on TURBT specimens identified the presence of LVI on H&E-stained slides alone (3 5), while some authors advocate the use of immunohistochemical staining, especially in equivocal cases Figure 2. Kaplan Meier estimates of progression-free survival (A) and cancer-specific survival (B) are shown for 86 patients who underwent TURBO, stratified by the absence vs presence of LVI.

5 Jpn J Clin Oncol, 2017, Vol. 47, No Table 2. Univariate Cox regression analysis of PFS and CSS in 86 patients who underwent TURBO Variable Progression-free survival Cancer-specific survival HR (95% CI) P value HR (95% CI) P value Age (<70 vs 70 years) 0.88 ( ) ( ) Sex (men vs women) 2.45 ( ) ( ) Tumor size, cm (<3 vs 3) 0.94 ( ) ( ) Multiplicity (single vs 2) 1.36 ( ) ( ) Resection pieces (1 vs 2 or 3) 0.28 ( ) ( ) Concomitant CIS (absent vs present) 1.86 ( ) ( ) Growth pattern (papillary vs solid) 3.39 ( ) ( ) Bladder recurrence (no vs yes; 1) 0.42 ( ) ( ) Intravesical instillation (no vs yes) 0.88 ( ) ( ) LVI (absent vs present) 4.95 ( ) ( ) PFS, progression-free survival; CSS, cancer-specific survival; HR, hazard ratio; CI, confidence interval. Table 3. Multivariate Cox regression analysis of PFS and CSS in 86 patients who underwent TURBO Variable Progression-free survival Cancer-specific survival HR (95% CI) P value HR (95% CI) P value Tumor size, cm (<3vs 3) 0.64 ( ) ( ) Concomitant CIS (absent vs present) 1.47 ( ) ( ) Growth pattern (papillary vs solid) 1.52 ( ) ( ) LVI (absent vs present) 4.48 ( ) ( ) (4,18). Furthermore, on TURBO specimens, we could detect the presence of LVI more efficiently because the foci of lymphatic invasion occurred predominantly at the invasive front into the lamina propria (17), which can be identified assuredly when the tumor is obtained by TURBO. Kunju et al. studied the concordance between LVI diagnoses of TURBT and RC specimens, and found that the concordance was good in muscle-invasive tumors but low in clinical T1 tumors (18). They reported that the overall sensitivity and specificity of LVI detection in TURBT specimens were 37% and 87%, respectively. These relatively low sensitivity values for the detection of LVI in TURBT specimens are unsurprising given that there were no standardized criteria to detect LVI and TURBT can provide only low quality and quantity of specimens. These findings make it a priority to apply the immunohistochemical staining for LVI detection in TURBO specimens (19). LVI in TURBT specimens was found to be associated with increased risk of advanced pathological stage and lymph node metastasis at the time of RC (18). The incidence of nodal involvement in patients treated with RC for HGT1 UCB is reported to be 12 15% (20 22). An important point is that some HGT1 UCB cases certainly harbor lymph node metastasis with upstaging on RC. Therefore, accurate and reliable detection of LVI at the initial TURBT in patients with HGT1 UCB is essential to predict the outcomes and to shift the clinical decision-making toward earlier aggressive treatment such as RC with or without neoadjuvant chemotherapy. Regarding the quality of TURBT, fragmentation and disorientation of the tumor specimens and the inability to pathologically confirm complete resection are the inherent drawbacks of TURBT. As a result, HGT1 UCB cases are frequently understaged, and this may be one of the reasons for the biological heterogeneity (23). Dutta et al. reported that 40% of high-risk NMIBCs were understaged and actually found to be stages pt2 or greater carcinoma after the final pathological reports (7). They attributed the main cause of understaging to the absence of muscle in the TURBT specimens. However, any muscle present is not necessarily derived from beneath the leading edge of cancer invasion. Therefore, even when muscle is present in the TURBT specimens, the tumor cannot be definitively staged. Although guidelines recommend a restaging TURBT or a repeat TURBT for better staging and tumor eradication (24), the resection still does not warrant complete resection, and only diminishes the probability of clinical understaging (25). In contrast to these disadvantages of TURBT, we can potentially avoid understaging by confirming histologically clear resection margins on TURBO specimen and, hence, can spare the repeat TURBT after TURBO. Even when there is no muscle in the TURBO specimen, if the vertical resection margins of the lamina propria are clear of cancer invasion, the tumor can definitively be staged as pt1 (11). Thus, we believe that TURBO can be superior to the initial TURBT and is an acceptable alternative to restaging TURBT. Our study has some limitations. First, because of the retrospective nature of our single-center study of highly selected patients, inherent selection bias cannot be excluded. Since visual judgment of the completeness of resection is often arbitrary and may differ from surgeon to surgeon, we placed marking dots around the proposed resection field, with the index tumor at the center, for more objective judgment. Moreover, compared with en bloc resection, the disadvantage of incorrect staging is greater after performing up to three piecemeal resections and then reconstructing the tumor tissues. Nevertheless, the planned division of a target tumor can preserve the spatial orientation of the tumor relative to the bladder wall and still have the advantage of more accurate pathologic assessment than TURBT. Another limitation is that we did not perform immunohistochemical staining in all enrolled patients, so some LVI-positive cases may have been missed. However, we believe that applying immunohistochemical staining to only equivocal cases on H&E staining is now a reasonable real-world practice. In addition, we could not provide a useful report of the pathology of the lateral margins of whole-mounted tumor sections. However, if standardized diagnostic criteria were to be established for the

6 452 LVI in HG pt1 UCB pathology of lateral and vertical margins, it would be possible to propose definitive terminology such as histologically complete resection of bladder tumor. Finally, because of the small sample size and limited number of events, the number of variables that could be included in the multivariate analysis was confined to established predictors of oncologic outcomes, diminishing the statistical power of the analysis. In conclusion, based on an accurate pathological staging and auxiliary immunohistochemical staining, the presence of LVI was found to be an independent prognostic factor associated with PFS and CSS in patients with HG pt1 UCB. Our results may help urologists to counsel their patients when deciding whether to choose a bladder-preserving strategy or RC. Acknowledgements We are grateful to our pathology laboratory technician, Mayumi Yasuhara, for her excellent technical assistance. Funding The authors have no funding sources to disclose in association with this manuscript. Conflict of interest statement The authors declare no conflict of interest for this study. References 1. Aldousari S, Kassouf W. Update on the management of non-muscle invasive bladder cancer. Can Urol Assoc J 2010;4: Boström PJ, Alkhateeb S, van Rhijn BW, Kuk C, Zlotta AR. Optimal timing of radical cystectomy in T1 high-grade bladder cancer. Expert Rev Anticancer Ther 2010;10: Andius P, Johansson SL, Holmäng S. Prognostic factors in stage T1 bladder cancer: tumor pattern (solid or papillary) and vascular invasion more important than depth of invasion. Urology 2007;70: Branchereau J, Larue S, Vayleux B, Karam G, Bouchot O, Rigaud J. Prognostic value of the lymphovascular invasion in high-grade stage pt1 bladder cancer. Clin Genitourin Cancer 2013;11: Cho KS, Seo HK, Joung JY, et al. Lymphovascular invasion in transurethral resection specimens as predictor of progression and metastasis in patients with newly diagnosed T1 bladder urothelial cancer. J Urol 2009; 182: Lopez JI, Angulo JC. The prognostic significance of vascular invasion in stage T1 bladder cancer. Histopathology 1995;27: Dutta SC, Smith JA Jr, Shappell SB, Coffey CS, Chang SS, Cookson MS. Clinical understaging of high risk nonmuscle invasive urothelial carcinoma treated with radical cystectomy. J Urol 2001;166: Streeper NM, Simons CM, Konety BR, et al. The significance of lymphovascular invasion in transurethral resection of bladder tumour and cystectomy specimens on the survival of patients with urothelial bladder cancer. BJU Int 2009;103: Brausi M, Witjes JA, Lamm D, et al. A review of current guidelines and best practice recommendations for the management of nonmuscle invasive bladder cancer by the international bladder cancer group. J Urol 2011; 186: Ark JT, Keegan KA, Barocas DA, et al. Incidence and predictors of understaging in patients with clinical T1 urothelial carcinoma undergoing radical cystectomy. BJU Int 2014;113: Ukai R, Hashimoto K, Iwasa T, Nakayama H. Transurethral resection in one piece (TURBO) is an accurate tool for pathological staging of bladder tumor. Int J Urol 2010;17: Ukai R, Kawashita E, Ikeda H. A new technique for transurethral resection of superficial bladder tumor in 1 piece. J Urol 2000;163: Epstein JI, Amin MB, Reuter VR, Mostofi FK. The World Health Organization/International Society of Urological Pathology consensus classification of urothelial (transitional cell) neoplasms of the urinary bladder. Bladder Consensus Conference Committee. Am J Surg Pathol 1998;22: Eble JN, Sauter G, Epstein JI, Sesterhenn IA. Tumours of the urinary system. In: Eble JN, Sauter G, Epstein JI, Sesterhenn IA, editors. World Health Organization: Pathology and Genetics of Tumours of the Urinary Systems and Male Genital Organs. Lyon: IARC Press, 2004; Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A, editors. American Joint Committee on Cancer (AJCC) Staging Manual. 7th edn. New York: Springer, Quek ML, Stein JP, Nichols PW, et al. Prognostic significance of lymphovascular invasion of bladder cancer treated with radical cystectomy. J Urol 2005;174: Arnaout-Alkarain A, Kahn HJ, Narod SA, Sun PA, Marks AN. Significance of lymph vessel invasion identified by the endothelial lymphatic marker D2-40 in node negative breast cancer. Mod Pathol 2007;20: Kunju LP, You L, Zhang Y, Daignault S, Montie JE, Lee CT. Lymphovascular invasion of urothelial cancer in matched transurethral bladder tumor resection and radical cystectomy specimens. JUrol2008; 180: Kim HS, Kim M, Jeong CW, Kwak C, Kim HH, Ku JH. Presence of lymphovascular invasion in urothelial bladder cancer specimens after transurethral resections correlates with risk of upstaging and survival: a systematic review and meta-analysis. Urol Oncol 2014;32: Daneshmand S. Determining the role of cystectomy for high-grade T1 urothelial carcinoma. Urol Clin North Am 2013;40: WiesnerC,PfitzenmaierJ,FaldumA,GillitzerR,MelchiorSW,ThüroffJW. Lymph node metastases in non-muscle invasive bladder cancer are correlated with the number of transurethral resections and tumour upstaging at radical cystectomy. BJU Int 2005;95: Bianco FJ Jr, Justa D, Grignon DJ, Sakr WA, Pontes JE, Wood DP Jr. Management of clinical T1 bladder transitional cell carcinoma by radical cystectomy. Urol Oncol 2004;22: Stein JP, Penson DF. Invasive T1 bladder cancer: indications and rationale for radical cystectomy. BJU Int 2008;102: Persad R, Lamm D, Brausi M, et al. Current approaches to the management of non-muscle invasive bladder cancer: comparison of current guidelines and recommendations. Eur Urol Suppl 2008;7: Herr HW. The value of a second transurethral resection in evaluating patients with bladder tumors. J Urol 1999;162:74 6.