- Email: [email protected]
Is transurethral resection alone enough for the diagnosis of histological variants? A single-center study
Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
Original article
Is transurethral resection alone enough for the diagnosis of histological variants? A single-center study Marco Moschini, M.D.a,b,*, Shahrokh F. Shariat, M.D.c, Massimo Freschi, M.D.d, Francesco Soria, M.D.c, David D’Andrea, M.D.c, Mohammad Abufaraj, M.D.c, Beat Foerster, M.D.c, Paolo Dell’Oglio, M.D.a, Emanuele Zaffuto, M.D.a, Agostino Mattei, M.D.b, Andrea Salonia, M.D., Ph.Da, Francesco Montorsi, M.D.a, Alberto Briganti, M.D., Ph.Da, Andrea Gallina, M.D.a, Renzo Colombo, M.D.a a
Unit of Urology/Division of Oncology, IRCCS Ospedale San Raffaele, Urological Research Institute, Milan, Italy b Department of Urology, Klinik für Urologie, Luzerner Kantonsspital, Lucerne, Switzerland c Department of Urology, Medical University of Vienna, Vienna, Austria d Department of Pathology, Ospedale San Raffaele, Milan, Italy Received 23 December 2016; received in revised form 23 February 2017; accepted 21 March 2017
Abstract Introduction: To evaluate incidence of histological variants and grade agreement between transurethral resection (TUR) and radical cystectomy (RC) in patients with bladder cancer. Methods: A total of 779 patients treated with TUR and subsequently with RC between 1990 and 2013 at a single center were analyzed retrospectively. Variant histology classifications used in our analyses were sarcomatoid, small cell, squamous, or micropapillary. Grade agreement was calculated using the Cohen kappa coefficient. Logistic regression analyses were built to predict adverse pathologic features from histological variants at TUR. Results: Considering TUR, 213 (27.3%) patients were diagnosed with histological variants. Of these, 2.1% (n ¼ 16) were found with sarcomatoid variant, 1.7% (n ¼ 13) with small cell, 7.1% (n ¼ 55) with squamous, 12.5% (n ¼ 97) with micropapillary. Considering RC, 212 (27.2%) patients were diagnosed with histological variants. Poor agreement was found considering micropapillary variant and the presence of a histological variant in general (0.11 and 0.27, respectively). Intermediate agreement was found analyzing the presence of sarcomatoid, small cell, and squamous variants (0.43, 0.61, and 0.61, respectively). Small cell carcinoma at TUR was found associated with an increased risk of harboring positive soft tissue surgical margin (odds ratio ¼ 2.08; CI: 1.27–3.41; P ¼ 0.03). Conclusions: One out of our patients with bladder cancer was diagnosed with a histological variant either at TUR and RC. We found poor agreement between TUR and RC. Our findings highlight that TUR alone is not sufficient to accurately evaluate the presence of histological variants that may have an effect on treatment and survival outcomes. r 2017 Elsevier Inc. All rights reserved.
Keywords: Bladder cancer; Radical cystectomy; Histological variants; Transurethral resection
1. Introduction Urothelial carcinoma of the bladder presents often morphological features that differ from the urothelial Marco Moschini is supported by the EUSP Scholarship—European Association of Urology. * Corresponding author. Tel.: þ39-02-0643-5664; fax: þ39-02-2643-5659. E-mail addresses: [email protected], marco. [email protected] (M.-A. Moschini). http://dx.doi.org/10.1016/j.urolonc.2017.03.024 1078-1439/r 2017 Elsevier Inc. All rights reserved.
common aspect. The recent WHO 2016 [1,2] consensus highlighted the importance of a careful morphological description of bladder pathologic specimens as the presence of histological variants may drive management of disease and change survival expectations [3–7]. In this perspective, the correct diagnosis of histological variants at the time of transurethral resection (TUR) and radical cystectomy (RC) is pivotal to optimize therapeutic strategies (Fig.)
2
M. Moschini et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
Fig. Histological variant diagnosed at transurethral resection confirmed in the radical cystectomy specimen. (Color version of figure is available online.)
However, the correct diagnosis of histological variants has several difficulties. In perspective, a higher incidence of histological variants has been reported when high-trained uropathologists working in referral centers were asked to evaluate this aspect [7]. Considering TUR in fact, the diagnosis can be challenging considering the small amount of the specimen that can be eventually damaged by the resection itself [1,2]. However, at the time data about concordance between TUR and RC in respect of histological variants diagnosis are lacking. Assessing this can be of paramount importance to fully understand a phenomenon that can influence severely patients’ treatment. Moreover, finding poor concordance between TUR and RC may represent a strong argument in favor of an increased diagnostic awareness at the time of TUR suggesting the introduction of new biomarkers and therefore testing new techniques that can potentially avoid misdiagnosis. The aim of our study is to evaluate the incidence of histological variants at TUR submitted subsequently to RC in a single tertiary referral center. Moreover, we evaluated the concordance of these findings using RC as reference in a selection of patients treated without neoadjuvant chemotherapy (NAC) to minimize confounding effects. 2. Materials and methods A total of 779 patients treated with TUR and subsequently with RC and pelvic lymph node dissection between 1990 and 2013 at a single tertiary referral center were retrospectively evaluated. The procedures were approved by the institutional review board (Vescica/2010). Patients treated with NAC and pT0 at RC were excluded from the cohort due to the aim of this study. Pathological stages were classified according to the 2009 TNM classification [8]. Adverse pathologic stage was defined as pT3–T4 disease. Positive soft tissue surgical margin (STSM) was defined as the presence of tumor at inked areas of soft tissue on the RC specimen. Tumor grade was assessed according to 1998
WHO/International Society of Urologic Pathology (ISUP) consensus classification [9]. Dedicated uropathologists evaluated the presence of histological variants. Specifically, we stratified variant histology as follows: sarcomatoid, small cell, squamous, or micropapillary. Glandular, nested, lymphoepithelial, adenocarcinoma, or plasmacytoid variants were included in a group defined “others” as a consequence of the rarity of these findings in our series. We did not use a percentage of thresholds for variant histology, as we assumed in conformity with previous findings that any component of variant histology would drive outcomes [6]. 2.1. Outcomes definitions The primary end was to evaluate concordance between last TUR before RC and RC findings regarding histological variants. The secondary end point is to evaluate the incidence of histological variants at TUR and RC. 2.2. Statistical analyses Descriptive statistics of categorical variables focused on frequencies and proportions. Means, medians, and interquartile ranges were reported for continuously coded variables. The Mann-Whitney test and chi-square test were used to compare the statistical significance of differences in medians and proportions, respectively. Grade agreement was calculated using the Cohen kappa coefficient. Absolute value ranges between 0 and 1, where 0 represents pure chance agreement, and 0.1 to 0.4, 0.4 to 0.75, and 0.75 to 1.0, respectively, represent poor, intermediate, and good agreement. Multivariable logistic regression tested the effect of histological variants at the time of TUR and the subsequent risk of harboring adverse pathologic stage, node metastasis, or positive STSM at RC. Statistical significance was considered at P o 0.05. Statistical analyses were performed using SPSS v.22.0 (IBM Corp., Armonk, NY, USA) and STATA 13 (Stata Corp., College Station, TX, USA).
M. Moschini et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
3
26.6% (n ¼ 207), and 38.8% (n ¼ 303) patients, respectively.
3. Results 3.1. Baseline characteristics
3.2. Histological variants at TUR and RC In total, 779 patients were included in the study. Demographics and pathologic characteristics of patients are shown in Table 1. The median age was 68 years and most patients were men (n ¼ 647, 83.1%). Considering TUR, 69 (5.0%), 46 (5.9%), and 193 (24.8%) patients were found with carcinoma in situ, lymphovascular invasion, and preoperative hydronephrosis, respectively. Overall, 38% (n ¼ 298), 40% (n ¼ 412), and 22% (n ¼ 169) of patients were found with pT1–T2 vs. pT3 vs. pT4 bladder cancer (BCa) disease. Lymphovascular invasion, carcinoma in situ, and node metastases at RC were found in 25.9% (n ¼ 202),
Table 1 Descriptive characteristics of 779 patients treated with transurethral resection and subsequently with radical cystectomy without neoadjuvant chemotherapy for bladder cancer at a single institution between 1990 and 2013 Variables Age, y Mean Median (IQR) Sex Male Female CIS at TUR LVI at TUR Preop IDN BMI, kg/m2 Mean Median (IQR) CCI 0 1 Z2 Nodes removed Mean Median (IQR) Pathologic T stage pT1–pT2 pT3 pT4 Pathologic N stage pN0 pN1 pN2 pN3 Grade G1–G2 G3 Concomitant CIS Positive STSM LVI Adjuvant CT
Overall (n ¼ 779, 100%) 68 69 (62–75) 647 (83.1%) 132 (16.9%) 69 (5.0%) 46 (5.9%) 193 (24.8%) 25.7 25.2 (23.2–28.1) 295 (37.9%) 222 (28.5%) 175 (22.5%) 20 19 (13–27) 298 (38.3%) 312 (40.1%) 169 (21.7%) 476 (61.1%) 92 (11.8%) 178 (22.8%) 33 (4.2%) 88 (11.3%) 637 (81.8%) 207 (26.6%) 82 (10.5%) 202 (25.9%) 121 (15.5%)
BMI, body mass index; CCI ¼ Charlson comorbidity index; CIS ¼ carcinoma in situ; CT ¼ chemotherapy; IDN ¼ hydronephrosis; IQR ¼ interquartile range; LVI ¼ lymphovascular invasion.
Incidence of histological variance and concordance between TUR and RC are depicted in Table 2. Considering TUR, 213 (27.3%) patients were diagnosed with histological variants. Of these, 2.1% (n ¼ 16) were found with sarcomatoid variant, 1.7% (n ¼ 13) with small cell, 7.1% (n ¼ 55) with squamous, 12.5% (n ¼ 97) with micropapillary and 4.1% (n ¼ 32) defined as others. Considering RC, 212 (27.2%) patients were diagnosed with histological variants. Of these, 2.1% (n ¼ 16) were found with sarcomatoid variant, 1.7% (n ¼ 13) with small cell, 3.9% (n ¼ 30) with squamous, 10.2% (n ¼ 78) with micropapillary. Cohen kappa concordance was used to analyze agreement between TUR and RC considering histological variants. In general, poor agreement was found considering micropapillary variant and the presence of a histological variant in general (0.11 and 0.27, respectively). On the contrary, intermediate agreement was found analyzing the presence of sarcomatoid, small cell, and squamous variants (0.43, 0.61, and 0.61, respectively). 3.3. Predictors of adverse pathologic features at RC Logistic regression analyses assessing the effect of histological variants at TUR evaluating the risk of harboring adverse pathologic stage at RC, node-positive disease, and positive STSM are presented in Table 3. At multivariable analyses, none of the histological variants evaluated at TUR were found to be associated to adverse pathologic stage or node-positive disease at RC (all P 4 0.06). Conversely, diagnosis of small cell carcinoma at TUR was found associated with an increased risk of harboring positive STSM (odds ratio ¼ 2.08; CI: 1.27–3.41; P ¼ 0.03). Other type of histological variants was found not associated with the prediction of positive STSM (all P 4 0.2). 4. Discussion Assessing histological variants at the time of TUR is fundamental to define the correct management of patients affected by BCa [3,4]. However, at the time it is not clear if TUR alone can accurately diagnose the presence of histological variants. To evaluate this aspect, we retrospectively analyzed a series of patients treated with TUR and subsequently with RC excluding patients treated with NAC. We made several findings. First, we found that 1 of 4 patients has been diagnosed with a histological variant either at TUR and RC. Of these, most patients were diagnosed with squamous and micropapillary differentiations. Our findings confirm previous literature on this topic [6,10,11]. Monn et al. [6] analyzed a cohort of patients
4
M. Moschini et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
Table 2 Cohen kappa analyses assessing concordance between histological variants at transurethral resection and radical cystectomy
RC urothelial (n ¼ 567, 72.8%) RC sarcomatoid (n ¼ 16, 2.1%) RC small cell (n ¼ 13, 1.7%) RC squamous (n ¼ 30, 3.9%) RC micropapillary (n ¼ 78, 10.0%) Kappa-concordance index
TUR urothelial (n ¼ 566, 72.7%)
TUR sarcomatoid (n ¼ 16, 2.1%)
TUR small cell (n ¼ 13, 1.7%)
TUR squamous (n ¼ 55, 7.1%)
TUR micropapillary (n ¼ 97, 12.5%)
440/567 7/16 2/13 0/30 59/78 0.27
2/567 7/16 1/13 0/30 3/78 0.43
3/567 0/16 8/13 0/30 2/78 0.61
23/567 1/16 0/13 30/30 1/78 0.61
73/567 1/16 1/13 0/30 10/78 0.11
stage, node-positive disease, or positive STSM. Considering previous literature, small cell variants are described as invasive at presentation, displaying small cells with overlapping nuclei missing prominent nuclear details or prominent nucleoli. The presence of this variant is characterized by an aggressive clinical course, with advanced stage at presentation and a high propensity for metastasis [13,14]. The clinical implications of our study are several-fold. Given the high incidence of histological variants in the context of BCa and considering the poor concordance between TUR and RC, our findings suggest the need of new biomarkers to increase the accuracy in the diagnosis of histological variants at the time of TUR. In this regard, liquid biopsy performed on urine samples can help in this field to analyze microRNA and to assess specific mutation pattern and therefore histological variant [15]. Considering technical aspects, en bloc TUR should be evaluated in the context of diagnosis of variant histology. In fact one of the reason for the poor concordance between histological variant at the time of TUR is probably due to the small amount of the specimen available in some TUR that can be eventually damaged by the resection [1,2]. In this perspective, misdiagnosis is probably superior in patients with small tumor treated with TUR only, although at the time no data exist in literature supporting this theory. Our study is not devoid of limitations. First, our study was not prospective or randomized, as it was a retrospective medical record review, and our findings should be interpreted in this context. Second, all patients included in our
treated at a single institution with RC between 2008 and 2013. Authors found an incidence of 26%, whereas squamous and micropapillary variants were found as the most common variants. In this regard, Xylinas et al. [11] assessed in a multicentric series data of 1,984 patients treated with RC. They found a rate of 24.6% of patients with histological variants. In TUR series, scarce data has been presented assessing the presence of histological variants. In this context, we observed similar rates as RC. Second, we found poor agreement between the presence of histological variants at TUR and RC. We used Cohen kappa coefficient to evaluate agreement between TUR and RC in evaluating the presence of histological variants. Considering the presence of histological variants in general and micropapillary variant, we found poor agreement. On the contrary, considering sarcomatoid, small cell, and squamous variants we found intermediate agreement. Our findings confirm previous literature assessing poor concordance between TUR and RC considering the diagnosis histological variants. Cai et al. [12] analyzed data of 4,110 RC specimens collected between 2000 and 2009 at 17 centers. They observed poor concordance between TUR and RC. Third, we found that the presence of histological variants at TUR can only partially predict adverse pathological features at RC. This was probably a consequence of the poor agreement observed between TUR and RC. Specifically, only patients found with small cell carcinoma at TUR were found with increased risk of harboring positive STSM, whereas other histological variants at TUR were not associated with an increased risk of adverse pathologic
Table 3 Multivariable logistic regression analyses assessing adverse pathologic stage, node-positive disease, and positive surgical margin
Urothelial Sarcomatoid Small cell Squamous Micropapillary Others Variants overall
Adverse pathologic stage at RC
Node-positive disease
OR (95% CI)
P value
OR (95% CI)
P value
OR (95% CI)
P value
0.96 0.49 3.80 1.83 0.70 1.17 0.96
0.8 0.2 0.09 0.07 0.1 0.7 0.8
1.26 1.55 1.29 0.79 0.90 0.48 0.80
0.2 0.4 0.6 0.4 0.6 0.1 0.2
1.22 0.53 2.08 0.66 0.68 0.96 0.82
0.4 0.5 0.03 0.4 0.3 0.9 0.5
(0.68–1.36) (0.17–1.38) (0.80–18.08) (0.96–3.52) (0.44–1.09) (0.55–2.48) (0.68–1.36)
(0.90–1.76) (0.56–4.27) (0.42–4.00) (0.44–1.43) (0.57–1.41) (0.34–1.23) (0.56–1.14)
Positive surgical margin
(0.71–2.10) (0.11–4.31) (1.27–3.41) (0.23–1.89) (0.31–1.48) (0.28–3.31) (0.48–1.41)
OR ¼ odds ratio. Adjusted for: age, sex, year of surgery, carcinoma in situ at transurethral resection, lymphovascular invasion at transurethral resection, and preoperative hydronephrosis.
M. Moschini et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
cohort underwent RC at a high-volume tertiary referral center. Therefore, incidence of histological variants obtained in this clinical scenario might not be applicable to other settings, in particular histological variants diagnosed in noninstitutional center have been found to be underestimated [7]. However, we found poor concordance between TUR and RC even analyzing data derived from a high-volume tertiary referral center. Third, in the present cohort we were unable to evaluate the percentage of variant histology in BCa specimen after RC that is potentially related to the survival outcomes. Fourth, glandular, nested, lymphoepithelial, adenocarcinoma or plasmacytoid variants were grouped in the same group as were found rarely in our cohort. Larger, multicentric studies are required to evaluate survival outcomes of this cohort. Fifth, we were not able to find any association between preoperative sarcomatoid or small cell carcinoma variants and the subsequent risk of developing adverse pathologic features at RC. In this regard, it is possible that our results were derived by the small number of patients included with these characteristics.
5. Conclusion The presence of histological variants is a common finding in patients with BCa. However, considering this aspect we found poor concordance between TUR and RC. Our findings highlight the necessity of developing new biomarkers to increase the diagnostic value of TUR, which may change the therapeutic indication for RC or the necessity of neoadjuvant treatment. References [1] Humphrey PA, Moch H, Cubilla AL, Ulbright TM, Reuter VE. The 2016 WHO classification of tumours of the urinary system and male genital organs-Part B: prostate and bladder tumours. Eur Urol 2016. http://dx.doi.org/10.1016/j.eururo.2016.02.028. [2] Moch H, Humphrey PA, Ulbright TM. WHO classification of tumours of the urinary system and male genital organs. 2016. [3] Willis D, Kamat AM. Nonurothelial bladder cancer and rare variant histologies. Hematol Oncol Clin North Am 2015;29:237–52, http: //dx.doi.org/10.1016/j.hoc.2014.10.011.
5
[4] Seisen T, Compérat E, Léon P, Roupret M. Impact of histological variants on the outcomes of nonmuscle invasive bladder cancer after transurethral resection. Curr Opin Urol 2014;24:524–31, http://dx.doi. org/10.1097/MOU.0000000000000086. [5] Moschini M, Shariat SF, Luciano R, D’Andrea D, Foerster B, Abufaraj M, et al. Pure but not mixed histological variants are associated with poor survival at radical cystectomy in bladder cancer patients. Clin Genitourin Cancer 2016. http://dx.doi.org/10.1016/j.clgc.2016.12.006. [6] Monn MF, Kaimakliotis HZ, Pedrosa JA, Cary KC, Bihrle R, Cheng L, et al. Contemporary bladder cancer: variant histology may be a significant driver of disease. Urol Oncol 2015;33:18, http://dx.doi.org/ 10.1016/j.urolonc.2014.10.001. [7] Shah RB, Montgomery JS, Montie JE, Kunju LP. Variant (divergent) histologic differentiation in urothelial carcinoma is under-recognized in community practice: impact of mandatory central pathology review at a large referral hospital. Urol Oncol 2013;31:1650–5, http://dx.doi. org/10.1016/j.urolonc.2012.04.009. [8] Edge SB, Compton CC. The American Joint Committee on Cancer: the 7th edition of the AJCC cancer staging manual and the future of TNM. Ann Surg Oncol 2010;17:1471–4, http://dx.doi.org/10.1245/ s10434-010-0985-4. [9] 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;22:1435–48. [10] Willis DL, Porten SP, Kamat AM. Should histologic variants alter definitive treatment of bladder cancer? Curr Opin Urol 2013;23:435– 43, http://dx.doi.org/10.1097/MOU.0b013e328363e415. [11] Xylinas E, Rink M, Robinson BD, Lotan Y, Babjuk M, Brisuda A, et al. Impact of histological variants on oncological outcomes of patients with urothelial carcinoma of the bladder treated with radical cystectomy. Eur J Cancer 2013;49:1889–97, http://dx.doi.org/ 10.1016/j.ejca.2013.02.001. [12] Cai T, Tiscione D, Verze P, Pomara G, Racioppi M, Nesi G, et al. Concordance and clinical significance of uncommon variants of bladder urothelial carcinoma in transurethral resection and radical cystectomy specimens. Urology 2014;84:1141–6, http://dx.doi.org/ 10.1016/j.urology.2014.06.032. [13] Mukesh M, Cook N, Hollingdale AE, Ainsworth NL, Russell SG. Small cell carcinoma of the urinary bladder: a 15-year retrospective review of treatment and survival in the Anglian Cancer Network. BJU Int 2009;103:747–52, http://dx.doi.org/10.1111/j.1464-410X.2008.08241.x. [14] Trias I, Algaba F, Condom E, Español I, Seguí J, Orsola I, et al. Small cell carcinoma of the urinary bladder. Presentation of 23 cases and review of 134 published cases. Eur Urol 2001;39:85–90. [15] Birkenkamp-Demtröder K, Nordentoft I, Christensen E, Høyer S, Reinert T, Vang S, et al. Genomic alterations in liquid biopsies from patients with bladder cancer. Eur Urol 2016;70:75–82, http://dx.doi. org/10.1016/j.eururo.2016.01.007.
Original article
Is transurethral resection alone enough for the diagnosis of histological variants? A single-center study Marco Moschini, M.D.a,b,*, Shahrokh F. Shariat, M.D.c, Massimo Freschi, M.D.d, Francesco Soria, M.D.c, David D’Andrea, M.D.c, Mohammad Abufaraj, M.D.c, Beat Foerster, M.D.c, Paolo Dell’Oglio, M.D.a, Emanuele Zaffuto, M.D.a, Agostino Mattei, M.D.b, Andrea Salonia, M.D., Ph.Da, Francesco Montorsi, M.D.a, Alberto Briganti, M.D., Ph.Da, Andrea Gallina, M.D.a, Renzo Colombo, M.D.a a
Unit of Urology/Division of Oncology, IRCCS Ospedale San Raffaele, Urological Research Institute, Milan, Italy b Department of Urology, Klinik für Urologie, Luzerner Kantonsspital, Lucerne, Switzerland c Department of Urology, Medical University of Vienna, Vienna, Austria d Department of Pathology, Ospedale San Raffaele, Milan, Italy Received 23 December 2016; received in revised form 23 February 2017; accepted 21 March 2017
Abstract Introduction: To evaluate incidence of histological variants and grade agreement between transurethral resection (TUR) and radical cystectomy (RC) in patients with bladder cancer. Methods: A total of 779 patients treated with TUR and subsequently with RC between 1990 and 2013 at a single center were analyzed retrospectively. Variant histology classifications used in our analyses were sarcomatoid, small cell, squamous, or micropapillary. Grade agreement was calculated using the Cohen kappa coefficient. Logistic regression analyses were built to predict adverse pathologic features from histological variants at TUR. Results: Considering TUR, 213 (27.3%) patients were diagnosed with histological variants. Of these, 2.1% (n ¼ 16) were found with sarcomatoid variant, 1.7% (n ¼ 13) with small cell, 7.1% (n ¼ 55) with squamous, 12.5% (n ¼ 97) with micropapillary. Considering RC, 212 (27.2%) patients were diagnosed with histological variants. Poor agreement was found considering micropapillary variant and the presence of a histological variant in general (0.11 and 0.27, respectively). Intermediate agreement was found analyzing the presence of sarcomatoid, small cell, and squamous variants (0.43, 0.61, and 0.61, respectively). Small cell carcinoma at TUR was found associated with an increased risk of harboring positive soft tissue surgical margin (odds ratio ¼ 2.08; CI: 1.27–3.41; P ¼ 0.03). Conclusions: One out of our patients with bladder cancer was diagnosed with a histological variant either at TUR and RC. We found poor agreement between TUR and RC. Our findings highlight that TUR alone is not sufficient to accurately evaluate the presence of histological variants that may have an effect on treatment and survival outcomes. r 2017 Elsevier Inc. All rights reserved.
Keywords: Bladder cancer; Radical cystectomy; Histological variants; Transurethral resection
1. Introduction Urothelial carcinoma of the bladder presents often morphological features that differ from the urothelial Marco Moschini is supported by the EUSP Scholarship—European Association of Urology. * Corresponding author. Tel.: þ39-02-0643-5664; fax: þ39-02-2643-5659. E-mail addresses: [email protected], marco. [email protected] (M.-A. Moschini). http://dx.doi.org/10.1016/j.urolonc.2017.03.024 1078-1439/r 2017 Elsevier Inc. All rights reserved.
common aspect. The recent WHO 2016 [1,2] consensus highlighted the importance of a careful morphological description of bladder pathologic specimens as the presence of histological variants may drive management of disease and change survival expectations [3–7]. In this perspective, the correct diagnosis of histological variants at the time of transurethral resection (TUR) and radical cystectomy (RC) is pivotal to optimize therapeutic strategies (Fig.)
2
M. Moschini et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
Fig. Histological variant diagnosed at transurethral resection confirmed in the radical cystectomy specimen. (Color version of figure is available online.)
However, the correct diagnosis of histological variants has several difficulties. In perspective, a higher incidence of histological variants has been reported when high-trained uropathologists working in referral centers were asked to evaluate this aspect [7]. Considering TUR in fact, the diagnosis can be challenging considering the small amount of the specimen that can be eventually damaged by the resection itself [1,2]. However, at the time data about concordance between TUR and RC in respect of histological variants diagnosis are lacking. Assessing this can be of paramount importance to fully understand a phenomenon that can influence severely patients’ treatment. Moreover, finding poor concordance between TUR and RC may represent a strong argument in favor of an increased diagnostic awareness at the time of TUR suggesting the introduction of new biomarkers and therefore testing new techniques that can potentially avoid misdiagnosis. The aim of our study is to evaluate the incidence of histological variants at TUR submitted subsequently to RC in a single tertiary referral center. Moreover, we evaluated the concordance of these findings using RC as reference in a selection of patients treated without neoadjuvant chemotherapy (NAC) to minimize confounding effects. 2. Materials and methods A total of 779 patients treated with TUR and subsequently with RC and pelvic lymph node dissection between 1990 and 2013 at a single tertiary referral center were retrospectively evaluated. The procedures were approved by the institutional review board (Vescica/2010). Patients treated with NAC and pT0 at RC were excluded from the cohort due to the aim of this study. Pathological stages were classified according to the 2009 TNM classification [8]. Adverse pathologic stage was defined as pT3–T4 disease. Positive soft tissue surgical margin (STSM) was defined as the presence of tumor at inked areas of soft tissue on the RC specimen. Tumor grade was assessed according to 1998
WHO/International Society of Urologic Pathology (ISUP) consensus classification [9]. Dedicated uropathologists evaluated the presence of histological variants. Specifically, we stratified variant histology as follows: sarcomatoid, small cell, squamous, or micropapillary. Glandular, nested, lymphoepithelial, adenocarcinoma, or plasmacytoid variants were included in a group defined “others” as a consequence of the rarity of these findings in our series. We did not use a percentage of thresholds for variant histology, as we assumed in conformity with previous findings that any component of variant histology would drive outcomes [6]. 2.1. Outcomes definitions The primary end was to evaluate concordance between last TUR before RC and RC findings regarding histological variants. The secondary end point is to evaluate the incidence of histological variants at TUR and RC. 2.2. Statistical analyses Descriptive statistics of categorical variables focused on frequencies and proportions. Means, medians, and interquartile ranges were reported for continuously coded variables. The Mann-Whitney test and chi-square test were used to compare the statistical significance of differences in medians and proportions, respectively. Grade agreement was calculated using the Cohen kappa coefficient. Absolute value ranges between 0 and 1, where 0 represents pure chance agreement, and 0.1 to 0.4, 0.4 to 0.75, and 0.75 to 1.0, respectively, represent poor, intermediate, and good agreement. Multivariable logistic regression tested the effect of histological variants at the time of TUR and the subsequent risk of harboring adverse pathologic stage, node metastasis, or positive STSM at RC. Statistical significance was considered at P o 0.05. Statistical analyses were performed using SPSS v.22.0 (IBM Corp., Armonk, NY, USA) and STATA 13 (Stata Corp., College Station, TX, USA).
M. Moschini et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
3
26.6% (n ¼ 207), and 38.8% (n ¼ 303) patients, respectively.
3. Results 3.1. Baseline characteristics
3.2. Histological variants at TUR and RC In total, 779 patients were included in the study. Demographics and pathologic characteristics of patients are shown in Table 1. The median age was 68 years and most patients were men (n ¼ 647, 83.1%). Considering TUR, 69 (5.0%), 46 (5.9%), and 193 (24.8%) patients were found with carcinoma in situ, lymphovascular invasion, and preoperative hydronephrosis, respectively. Overall, 38% (n ¼ 298), 40% (n ¼ 412), and 22% (n ¼ 169) of patients were found with pT1–T2 vs. pT3 vs. pT4 bladder cancer (BCa) disease. Lymphovascular invasion, carcinoma in situ, and node metastases at RC were found in 25.9% (n ¼ 202),
Table 1 Descriptive characteristics of 779 patients treated with transurethral resection and subsequently with radical cystectomy without neoadjuvant chemotherapy for bladder cancer at a single institution between 1990 and 2013 Variables Age, y Mean Median (IQR) Sex Male Female CIS at TUR LVI at TUR Preop IDN BMI, kg/m2 Mean Median (IQR) CCI 0 1 Z2 Nodes removed Mean Median (IQR) Pathologic T stage pT1–pT2 pT3 pT4 Pathologic N stage pN0 pN1 pN2 pN3 Grade G1–G2 G3 Concomitant CIS Positive STSM LVI Adjuvant CT
Overall (n ¼ 779, 100%) 68 69 (62–75) 647 (83.1%) 132 (16.9%) 69 (5.0%) 46 (5.9%) 193 (24.8%) 25.7 25.2 (23.2–28.1) 295 (37.9%) 222 (28.5%) 175 (22.5%) 20 19 (13–27) 298 (38.3%) 312 (40.1%) 169 (21.7%) 476 (61.1%) 92 (11.8%) 178 (22.8%) 33 (4.2%) 88 (11.3%) 637 (81.8%) 207 (26.6%) 82 (10.5%) 202 (25.9%) 121 (15.5%)
BMI, body mass index; CCI ¼ Charlson comorbidity index; CIS ¼ carcinoma in situ; CT ¼ chemotherapy; IDN ¼ hydronephrosis; IQR ¼ interquartile range; LVI ¼ lymphovascular invasion.
Incidence of histological variance and concordance between TUR and RC are depicted in Table 2. Considering TUR, 213 (27.3%) patients were diagnosed with histological variants. Of these, 2.1% (n ¼ 16) were found with sarcomatoid variant, 1.7% (n ¼ 13) with small cell, 7.1% (n ¼ 55) with squamous, 12.5% (n ¼ 97) with micropapillary and 4.1% (n ¼ 32) defined as others. Considering RC, 212 (27.2%) patients were diagnosed with histological variants. Of these, 2.1% (n ¼ 16) were found with sarcomatoid variant, 1.7% (n ¼ 13) with small cell, 3.9% (n ¼ 30) with squamous, 10.2% (n ¼ 78) with micropapillary. Cohen kappa concordance was used to analyze agreement between TUR and RC considering histological variants. In general, poor agreement was found considering micropapillary variant and the presence of a histological variant in general (0.11 and 0.27, respectively). On the contrary, intermediate agreement was found analyzing the presence of sarcomatoid, small cell, and squamous variants (0.43, 0.61, and 0.61, respectively). 3.3. Predictors of adverse pathologic features at RC Logistic regression analyses assessing the effect of histological variants at TUR evaluating the risk of harboring adverse pathologic stage at RC, node-positive disease, and positive STSM are presented in Table 3. At multivariable analyses, none of the histological variants evaluated at TUR were found to be associated to adverse pathologic stage or node-positive disease at RC (all P 4 0.06). Conversely, diagnosis of small cell carcinoma at TUR was found associated with an increased risk of harboring positive STSM (odds ratio ¼ 2.08; CI: 1.27–3.41; P ¼ 0.03). Other type of histological variants was found not associated with the prediction of positive STSM (all P 4 0.2). 4. Discussion Assessing histological variants at the time of TUR is fundamental to define the correct management of patients affected by BCa [3,4]. However, at the time it is not clear if TUR alone can accurately diagnose the presence of histological variants. To evaluate this aspect, we retrospectively analyzed a series of patients treated with TUR and subsequently with RC excluding patients treated with NAC. We made several findings. First, we found that 1 of 4 patients has been diagnosed with a histological variant either at TUR and RC. Of these, most patients were diagnosed with squamous and micropapillary differentiations. Our findings confirm previous literature on this topic [6,10,11]. Monn et al. [6] analyzed a cohort of patients
4
M. Moschini et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
Table 2 Cohen kappa analyses assessing concordance between histological variants at transurethral resection and radical cystectomy
RC urothelial (n ¼ 567, 72.8%) RC sarcomatoid (n ¼ 16, 2.1%) RC small cell (n ¼ 13, 1.7%) RC squamous (n ¼ 30, 3.9%) RC micropapillary (n ¼ 78, 10.0%) Kappa-concordance index
TUR urothelial (n ¼ 566, 72.7%)
TUR sarcomatoid (n ¼ 16, 2.1%)
TUR small cell (n ¼ 13, 1.7%)
TUR squamous (n ¼ 55, 7.1%)
TUR micropapillary (n ¼ 97, 12.5%)
440/567 7/16 2/13 0/30 59/78 0.27
2/567 7/16 1/13 0/30 3/78 0.43
3/567 0/16 8/13 0/30 2/78 0.61
23/567 1/16 0/13 30/30 1/78 0.61
73/567 1/16 1/13 0/30 10/78 0.11
stage, node-positive disease, or positive STSM. Considering previous literature, small cell variants are described as invasive at presentation, displaying small cells with overlapping nuclei missing prominent nuclear details or prominent nucleoli. The presence of this variant is characterized by an aggressive clinical course, with advanced stage at presentation and a high propensity for metastasis [13,14]. The clinical implications of our study are several-fold. Given the high incidence of histological variants in the context of BCa and considering the poor concordance between TUR and RC, our findings suggest the need of new biomarkers to increase the accuracy in the diagnosis of histological variants at the time of TUR. In this regard, liquid biopsy performed on urine samples can help in this field to analyze microRNA and to assess specific mutation pattern and therefore histological variant [15]. Considering technical aspects, en bloc TUR should be evaluated in the context of diagnosis of variant histology. In fact one of the reason for the poor concordance between histological variant at the time of TUR is probably due to the small amount of the specimen available in some TUR that can be eventually damaged by the resection [1,2]. In this perspective, misdiagnosis is probably superior in patients with small tumor treated with TUR only, although at the time no data exist in literature supporting this theory. Our study is not devoid of limitations. First, our study was not prospective or randomized, as it was a retrospective medical record review, and our findings should be interpreted in this context. Second, all patients included in our
treated at a single institution with RC between 2008 and 2013. Authors found an incidence of 26%, whereas squamous and micropapillary variants were found as the most common variants. In this regard, Xylinas et al. [11] assessed in a multicentric series data of 1,984 patients treated with RC. They found a rate of 24.6% of patients with histological variants. In TUR series, scarce data has been presented assessing the presence of histological variants. In this context, we observed similar rates as RC. Second, we found poor agreement between the presence of histological variants at TUR and RC. We used Cohen kappa coefficient to evaluate agreement between TUR and RC in evaluating the presence of histological variants. Considering the presence of histological variants in general and micropapillary variant, we found poor agreement. On the contrary, considering sarcomatoid, small cell, and squamous variants we found intermediate agreement. Our findings confirm previous literature assessing poor concordance between TUR and RC considering the diagnosis histological variants. Cai et al. [12] analyzed data of 4,110 RC specimens collected between 2000 and 2009 at 17 centers. They observed poor concordance between TUR and RC. Third, we found that the presence of histological variants at TUR can only partially predict adverse pathological features at RC. This was probably a consequence of the poor agreement observed between TUR and RC. Specifically, only patients found with small cell carcinoma at TUR were found with increased risk of harboring positive STSM, whereas other histological variants at TUR were not associated with an increased risk of adverse pathologic
Table 3 Multivariable logistic regression analyses assessing adverse pathologic stage, node-positive disease, and positive surgical margin
Urothelial Sarcomatoid Small cell Squamous Micropapillary Others Variants overall
Adverse pathologic stage at RC
Node-positive disease
OR (95% CI)
P value
OR (95% CI)
P value
OR (95% CI)
P value
0.96 0.49 3.80 1.83 0.70 1.17 0.96
0.8 0.2 0.09 0.07 0.1 0.7 0.8
1.26 1.55 1.29 0.79 0.90 0.48 0.80
0.2 0.4 0.6 0.4 0.6 0.1 0.2
1.22 0.53 2.08 0.66 0.68 0.96 0.82
0.4 0.5 0.03 0.4 0.3 0.9 0.5
(0.68–1.36) (0.17–1.38) (0.80–18.08) (0.96–3.52) (0.44–1.09) (0.55–2.48) (0.68–1.36)
(0.90–1.76) (0.56–4.27) (0.42–4.00) (0.44–1.43) (0.57–1.41) (0.34–1.23) (0.56–1.14)
Positive surgical margin
(0.71–2.10) (0.11–4.31) (1.27–3.41) (0.23–1.89) (0.31–1.48) (0.28–3.31) (0.48–1.41)
OR ¼ odds ratio. Adjusted for: age, sex, year of surgery, carcinoma in situ at transurethral resection, lymphovascular invasion at transurethral resection, and preoperative hydronephrosis.
M. Moschini et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
cohort underwent RC at a high-volume tertiary referral center. Therefore, incidence of histological variants obtained in this clinical scenario might not be applicable to other settings, in particular histological variants diagnosed in noninstitutional center have been found to be underestimated [7]. However, we found poor concordance between TUR and RC even analyzing data derived from a high-volume tertiary referral center. Third, in the present cohort we were unable to evaluate the percentage of variant histology in BCa specimen after RC that is potentially related to the survival outcomes. Fourth, glandular, nested, lymphoepithelial, adenocarcinoma or plasmacytoid variants were grouped in the same group as were found rarely in our cohort. Larger, multicentric studies are required to evaluate survival outcomes of this cohort. Fifth, we were not able to find any association between preoperative sarcomatoid or small cell carcinoma variants and the subsequent risk of developing adverse pathologic features at RC. In this regard, it is possible that our results were derived by the small number of patients included with these characteristics.
5. Conclusion The presence of histological variants is a common finding in patients with BCa. However, considering this aspect we found poor concordance between TUR and RC. Our findings highlight the necessity of developing new biomarkers to increase the diagnostic value of TUR, which may change the therapeutic indication for RC or the necessity of neoadjuvant treatment. References [1] Humphrey PA, Moch H, Cubilla AL, Ulbright TM, Reuter VE. The 2016 WHO classification of tumours of the urinary system and male genital organs-Part B: prostate and bladder tumours. Eur Urol 2016. http://dx.doi.org/10.1016/j.eururo.2016.02.028. [2] Moch H, Humphrey PA, Ulbright TM. WHO classification of tumours of the urinary system and male genital organs. 2016. [3] Willis D, Kamat AM. Nonurothelial bladder cancer and rare variant histologies. Hematol Oncol Clin North Am 2015;29:237–52, http: //dx.doi.org/10.1016/j.hoc.2014.10.011.
5
[4] Seisen T, Compérat E, Léon P, Roupret M. Impact of histological variants on the outcomes of nonmuscle invasive bladder cancer after transurethral resection. Curr Opin Urol 2014;24:524–31, http://dx.doi. org/10.1097/MOU.0000000000000086. [5] Moschini M, Shariat SF, Luciano R, D’Andrea D, Foerster B, Abufaraj M, et al. Pure but not mixed histological variants are associated with poor survival at radical cystectomy in bladder cancer patients. Clin Genitourin Cancer 2016. http://dx.doi.org/10.1016/j.clgc.2016.12.006. [6] Monn MF, Kaimakliotis HZ, Pedrosa JA, Cary KC, Bihrle R, Cheng L, et al. Contemporary bladder cancer: variant histology may be a significant driver of disease. Urol Oncol 2015;33:18, http://dx.doi.org/ 10.1016/j.urolonc.2014.10.001. [7] Shah RB, Montgomery JS, Montie JE, Kunju LP. Variant (divergent) histologic differentiation in urothelial carcinoma is under-recognized in community practice: impact of mandatory central pathology review at a large referral hospital. Urol Oncol 2013;31:1650–5, http://dx.doi. org/10.1016/j.urolonc.2012.04.009. [8] Edge SB, Compton CC. The American Joint Committee on Cancer: the 7th edition of the AJCC cancer staging manual and the future of TNM. Ann Surg Oncol 2010;17:1471–4, http://dx.doi.org/10.1245/ s10434-010-0985-4. [9] 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;22:1435–48. [10] Willis DL, Porten SP, Kamat AM. Should histologic variants alter definitive treatment of bladder cancer? Curr Opin Urol 2013;23:435– 43, http://dx.doi.org/10.1097/MOU.0b013e328363e415. [11] Xylinas E, Rink M, Robinson BD, Lotan Y, Babjuk M, Brisuda A, et al. Impact of histological variants on oncological outcomes of patients with urothelial carcinoma of the bladder treated with radical cystectomy. Eur J Cancer 2013;49:1889–97, http://dx.doi.org/ 10.1016/j.ejca.2013.02.001. [12] Cai T, Tiscione D, Verze P, Pomara G, Racioppi M, Nesi G, et al. Concordance and clinical significance of uncommon variants of bladder urothelial carcinoma in transurethral resection and radical cystectomy specimens. Urology 2014;84:1141–6, http://dx.doi.org/ 10.1016/j.urology.2014.06.032. [13] Mukesh M, Cook N, Hollingdale AE, Ainsworth NL, Russell SG. Small cell carcinoma of the urinary bladder: a 15-year retrospective review of treatment and survival in the Anglian Cancer Network. BJU Int 2009;103:747–52, http://dx.doi.org/10.1111/j.1464-410X.2008.08241.x. [14] Trias I, Algaba F, Condom E, Español I, Seguí J, Orsola I, et al. Small cell carcinoma of the urinary bladder. Presentation of 23 cases and review of 134 published cases. Eur Urol 2001;39:85–90. [15] Birkenkamp-Demtröder K, Nordentoft I, Christensen E, Høyer S, Reinert T, Vang S, et al. Genomic alterations in liquid biopsies from patients with bladder cancer. Eur Urol 2016;70:75–82, http://dx.doi. org/10.1016/j.eururo.2016.01.007.