Int Urol Nephrol DOI 10.1007/s11255-014-0714-1
UROLOGY - ORIGINAL PAPER
FISH analysis of washing urine from the upper urinary tract for the detection of urothelial cancers Torsten Gruschwitz • Mieczyslaw Gajda • Astrid Enkelmann • Marc-Oliver Grimm • Heiko Wunderlich • Marcus Horstmann Kerstin Junker
•
Received: 12 January 2014 / Accepted: 5 April 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract Introduction To evaluate FISH analysis of washing urine from the upper urinary tract (UUT) in comparison with cytology (Cyt) for the detection of urothelial cancers. Patients and methods In 82 patients with symptoms or abnormalities of the UUT sampling of washing urine for FISH and Cyt and a stepwise diagnostic work-up (e.g. retrograde ureteropyelography, ureterorenoscopy and endoscopic biopsy) were performed. In case of endoscopically and/or histologically proven malignancy patients either underwent nephroureterectomy, partial ureterectomy or local treatment. Sensitivity and specificity for FISH and Cyt as well as its combination were determined. Results Urothelial cancer of the UUT was detected in 20 patients. Eleven patients underwent nephroureterectomy, six partial ureterectomy and three endoscopic tumour
Marcus Horstmann and Kerstin Junker contributed equally to this work. T. Gruschwitz A. Enkelmann M.-O. Grimm H. Wunderlich M. Horstmann (&) K. Junker Department of Urology, Jena University Hospital, Lessingstr. 1, 07743 Jena, Germany e-mail: [email protected] M. Gajda Institute of Pathology, Jena University Hospital, Ziegelmu¨hlenweg 1, 07743 Jena, Germany H. Wunderlich Clinic of Urology and Pediatric Urology, St. Georg Klinikum, Mu¨hlha¨user Straße 94, 99817 Eisenach, Germany K. Junker Clinic of Urology and Pediatric Urology, Saarland University Medical Center, Kirrberger Straße 1, 66424 Homburg/Saar, Germany
treatment. This revealed nine pTa, three pT1 and seven muscle-invasive tumours. Twelve tumours were classified as low and seven as high-grade tumours. In one patient with a macroscopic unequivocal finding of tumour, endoscopic laser ablation without histologic confirmation was performed. FISH was evaluable in 76 patients and detected 16 tumours with a sensitivity and specificity of 84.2 and 91.1 %, respectively. Cyt was performed in 79 and was evaluable in 78 patients. It detected ten tumours with a sensitivity and specificity of 52.6 and 91.4 %, respectively. Cyt and FISH together detected 19 tumours with (sensitivity 100 % and specificity 83.6 %). Conclusion FISH was more sensitive than and equally specific to Cyt in the detection of urothelial cancers of the UUT. Both markers in combination revealed the best sensitivity, making it a possible approach in future settings. Keywords Urothelial cancer FISH Cytology Upper urinary tract
Introduction The UroVisionTM FISH test is currently approved as a urinary-based tumour marker in the detection and followup of urothelial bladder cancer [1]. It detects aneuploidy of chromosomes 3, 7, 17 and loss of locus 9p21, which are known alterations in urothelial cancer. In clinical evaluation, FISH was highly reliable with a higher sensitivity than and an equal specificity to cytology (Cyt) [2–4]. Probably, due to its cytogenetic-based principle, FISH is in contrast to other urinary-based tumour markers less likely to become false positive [5, 6]. Because of its high reliability and the proven molecular similarity of urothelial cancers of the bladder and the upper urinary tract (UUT) [7], FISH is
123
Int Urol Nephrol
also considered as a potential marker for the detection of urothelial cancers in these anatomic regions. Clinically, this would be helpful, because diagnosis of UUT tumours is still hampered by limited endoscopic access and a high risk of non-diagnostic biopsies [8]. However, until to date, FISH is not approved as a marker for UUT tumours and studies evaluating FISH in the UUT remain scarce. They include either studies of voided [9–12] or of endoscopically collected urine from the UUT [13, 14] and evaluated only patients with known or suspected urothelial cancer. In the present study, we evaluated FISH as a diagnostic tool in the so far largest group of patients (n = 82). In contrast to other studies, we included patients who had either symptoms or abnormalities of the UUT that made urothelial cancer one differential diagnosis among others. In fact, 62 out of 82 patients were finally diagnosed to have no or other abnormalities of the UUT. During a stepwise diagnostic work-up, FISH analysis and Cyt were conducted endoscopically.
Patients and methods Between 2006 and 2012, washing urine form the UUT was analysed for Cyt and FISH in 82 patients in one single centre. Clinical data were evaluated retrospectively, and the study was approved by the local ethics committee (Approval No: 3693-02/13). Mean age of all patients was 66 years (range 36–89 years). Of the patients, 47 were men and 35 women. Seventeen had a history of urothelial cancer (bladder cancer: n = 15, urothelial cancer of the opposite UUT: n = 2 or both: n = 1). Samples of washing urine were collected from the ureter or the renal pelvis depending on the site of the suspected lesion. Saline was injected through the endoscope or a ureter catheter positioned below the suspected lesion. In case of no obvious lesion, urine was retrieved from the ureter. Samples of 5–10 ml were collected by syringe aspiration or dropwise urine backflow. Cyt was performed by one senior pathologist (M.G.) according to the cytopathologic classification of the Union International Contre le Cancer [15]. Samples with no or mild atypia were classified as negative and samples with moderate to severe as positive. The commercially available UroVysionTM test (Abbott Molecular/Vysis, Des Plaines, IL, USA) was performed according to the description of the manufacturer for FISH analysis [16]. A test was considered evaluable if at least 25 cells were present. Of them, 25—if present—morphologically atypical cells were selected for evaluation. If not present, morphologically normal cells were evaluated. For a positive test, the following criteria were applied: A test was considered positive if of 25 nuclei at least 5 exhibited 2
123
or more aneuploidies (gains) of chromosomes 3, 7 or 17, or if at least 10 nuclei exhibited 1 or more aneuploidies (gains) of chromosomes 3, 7 or 17, or if at least 10 nuclei exhibited a complete loss of the locus 9p21 [4, 6, 16]. If none of these criteria were met the test was negative. Neither the cytopathologist nor the FISH reviewer had knowledge of the clinical findings or the history of the patients. The reference standard for Cyt and FISH was the complete radiologic, endoscopic and histopathological work-up including final surgical treatment. Patients were first evaluated by standard radiologic imaging with intravenous pyelography or CT scan. Then, patients underwent retrograde pyelography and uretero-/or ureterorenoscopy with or without biopsy. Surgical treatment was based on histologically proven malignancy and in some cases on endoscopical findings only. Depending on their tumour characteristics and general condition patients either underwent nephroureterectomy, partial ureterectomy or local treatment. Sensitivity and specificity for Cyt, FISH and its combination were determined according to the final histopathological or, if not available, endoscopic and radiologic results. In the combination of Cyt and FISH, the marker panel was considered positive if either Cyt or FISH was positive. Patients’ characteristics were analysed by descriptive statistics. Sensitivity and specificity for Cyt and FISH were calculated and compared by a McNemar test. A p value of \0.05 was considered significant. Analyses were conducted with IBM SPSS Statistics for Windows, Version 21.0., Armonk, NY: IBM Corp., Released 2012.
Results Urothelial cancer of the UUT was found in 20/82 patients. Mean age of them was 72 years (range 55–89 years). Twelve were men and eight women. Eleven patients underwent nephroureterectomy, six partial ureterectomy and three endoscopic tumour treatment. Histopathological evaluation revealed nine pTa, three pT1 and seven muscleinvasive tumours. Twelve tumours were classified as low and seven as high-grade tumours. In one patient with a history of nephrectomy of his right kidney and a macroscopic unequivocal finding of tumour in the left renal pelvis endoscopic laser ablation was performed without histologic confirmation. None of the patients had concomitant urothelial cancer of the bladder at the time of urine collection. Sixty-two patients had no urothelial cancer of the investigated UUT or bladder. Their mean age was 64 years (range 55–89 years). Of these patients, twelve had a nephroureterolithiasis and nine a ureteral stricture.
Int Urol Nephrol Table 1 Sensitivity, specificity, positive and negative predictive values of FISH, Cyt and the combination of Cyt and FISH
Sensitivity
Cyt (%)
FISH (%)
p value
Cyt/FISH (%)
52.6
84.2
0.146
100.0
1.000
Table 2 Sensitivity of Cyt and FISH according to tumour grade in 18 tumour patients with available histopathology Patients (n)
Cyt sens (%)
FISH sens (%)
Low grade
11
54.5
81.8
7
57.1
85.7
Specificity
91.4
91.1
83.6
High grade
PPV
66.7
76.2
65.5
FISH fluorescence in situ hybridisation, Cyt cytology
NPV
85.5
94.4
83.9
FISH fluorescence in situ hybridisation, Cyt cytology, PPV positive predictive value, NPV negative predictive value
Table 3 Sensitivity of Cyt and FISH according to tumour stage (T) in 18 tumour patients with available histopathology
p values were evaluated for the difference between Cyt and FISH alone
Malignant diseases other than urothelial cancer were found in six patients. Of them, three had renal cell carcinoma, one a renal papilloma (not classified as urothelial cancer), one a recurrent extraureteral lymphoma and one a leiomyosarcoma. Cyt was performed in 79 patients and evaluable in 78 (99 %). In the group of patients with UUT tumours (n = 20), Cyt was performed and evaluable in 19 patients. It was not performed in one patient with a pT2G2 tumour. In these 19 patients, ten tumours were correctly diagnosed. However, 49 pTaG low-grade, 19 pT1 high-grade, 1 pT3 low-grade, 29 pT3-4 high-grade tumours and the only endoscopically confirmed tumour were not missed by Cyt (Table 4). Based on these 19 patients, Cyt had a sensitivity of 52.6 %. It was negative in 62 patients and correctly diagnosed 53 patients without cancer with a specificity of 91.4 % (Table 1). FISH was performed in 82 patients and evaluable in 76 (93 %). In six patients, it was not evaluable because of a cell count below 25. Of the 20 patients with UUT tumours, FISH was not evaluable in the same patient as in Cyt. FISH correctly diagnosed 16 tumours and was negative in two pTa-1 low-grade and one pT3 high-grade tumour (Table 4). Based on the calculation of 19 evaluable tests, its sensitivity was 84.2 % (Table 1). The specificity of FISH was 91.1 %. It correctly diagnosed 52 cancer-free patients and had five false-positive results. Of these patients, one had a history for bladder cancer, one a ureterolithiasis and three painless haematuria. Regarding tumour grade both tests revealed a slightly better sensitivity for high-grade tumours (Table 2). Whereas Cyt missed three high-grade tumours, FISH only missed one. Regarding T-category, Cyt had a sensitivity of 58.3 % and FISH of 83.3 % for non-muscle-invasive tumours (pTa–pT1), and for muscle-invasive tumours ([pT1) of 50.0 and 83.3 %, respectively (Table 3). Statistical comparison of Cyt and FISH revealed no significant differences regarding their sensitivity (McNemar: p = 0.146, n.s.) (Table 1). Furthermore, the subgroup of patients with inconclusive or negative standard radiologic
Patients (n)
Cyt sens (%)
FISH sens (%)
pTa-1
12
58.3
83.3
pT2-4
6
50.0
83.3
FISH fluorescence in situ hybridisation, Cyt cytology
imaging was evaluated. It included three patients with pTa low-grade tumours, one patient with a pT3 high-grade tumour and the patient with endoscopic laser ablation. In this subgroup, Cyt was positive in two patients (29 pTa low grade) and FISH in all. The best sensitivity was found by the combination of Cyt and FISH. Together, they detected 19 tumours with a sensitivity of 100.0 %. This was calculated without the patient in whom neither Cyt nor FISH results were available. The combination of both markers resulted in 10 falsepositive results and a specificity of 83.6 %. Only six of the patients with tumours had positive results in both tests, eight had a positive FISH and a negative Cyt and three a positive Cyt and a negative FISH, respectively (Table 4).
Discussion Both FISH and cytology (Cyt) can be used for the detection of urothelial cancer of the UUT in voided urine and washing urine form the UUT. Voided urine has the advantage of a non-invasive test but leaves the localisation of a tumour in case of a positive test open. If retrieved directly from the UUT, the cancer site can be encircled more precisely, but the test becomes invasive [8, 13]. For Cyt, which is the current gold standard [17], it was shown that its sensitivity increases if washing urine was taken from the UUT [13, 18]. Because of potential cell damage, however, these samples are also discussed to create more false-positive results [19, 20]. Even though Cyt has an excellent specificity, its main disadvantages remain its observer dependence and a limited sensitivity in low-grade tumours [21]. In the present study, we evaluated FISH in comparison with Cyt from washing urine of the UUT during an endoscopic work-up of patients symptomatic from or with
123
Int Urol Nephrol Table 4 Cyt and FISH results of patients with urothelial cancer of the upper urinary tract (n = 20) Pat
Sex
Age
Surgical intervention
T Stage
Grading
Cyt
Gain of 3 (cells)
Gain of 7 (cells)
Gain of 17 (cells)
Complete loss of 9p21 (cells)
FISH
1
F
62
Nephroureterectomy
pTa
Low
Positive
25
26
19
0
Positive
2
M
70
Nephroureterectomy
pT1
High
Positive
25
25
23
2
Positive
3
F
85
pTa
Low
Negative
25
25
25
0
Positive
4
M
86
Transurethral resection Endoscopic laser ablation
pTa
Low
Negative
23
21
17
5
Positive
5
F
78
Nephroureterectomy
pTa
Low
Positive
0
0
17
25
Positive
6
M
73
Endoscopic laser ablation
n.e.
n.e.
Negative
20
20
20
20
Positive
7
M
76
Nephroureterectomy
pT3
High
Positive
16
18
18
0
Positive
8
F
70
Nephroureterectomy
pT1
Low
Positive
0
1
3
7
Negative
9
F
70
Nephroureterectomy
pT1
High
Negative
16
18
18
0
Positive
10
M
76
Distal ureter resection
pTa
Low
Positive
3
4
2
0
Negative
11
M
67
Nephroureterectomy
pT2
Low
n.e.
n.e.
n.e.
n.e.
n.e.
Not enough cells
12
F
77
Nephroureterectomy
pT3
Low
Negative
0
0
0
23
Positive
13 14
F M
66 70
Nephroureterectomy Nephroureterectomy
pT3 pT3
High High
Positive Negative
0 1
0 0
0 0
0 18
Negative Positive
15
M
73
Palliative nephrectomy
pT4
High
Negative
0
0
0
10
Positive
16
M
56
Distal ureter resection
pTa
Low
Positive
8
8
6
0
Positive
17
M
76
Distal ureter resection
pTa
Low
Negative
3
6
3
25
Positive
18
M
73
Distal ureter resection
pTa
Low
Positive
11
11
11
n.e.
Positive
19
F
89
Distal ureter resection
pT3
High
Positive
25
25
25
5
Positive
20
M
55
Distal ureter resection
pTa
Low
Negative
15
15
15
0
Positive
n.e. not evaluated, Cyt cytology, FISH fluorescence in situ hybridisation
suspicious findings of the UUT. The excellent diagnostic value of FISH for UUT tumours was already shown in studies from voided urine in patients with known urothelial cancers in comparison with healthy persons [10, 12, 22] and in comparison with patients with benign urological diseases [11]. In these studies, reported sensitivities and specificities range from 73.5 to 85.7 % and 94 to 100 %, respectively [10, 12, 22]. FISH was, furthermore, evaluated from washing urine of the UUT in patients suspicious for UUT tumours. Data regarding this, however, remain scarce with to our knowledge only two published studies. They revealed— most probably because of a direct access to mobilised urothelial cells—similar to Cyt the highest sensitivities [13, 14]. Mian et al. reported a sensitivity of 100 % in a group of 55 patients including 24 tumour patients, [14] and
123
Akkad et al. [13] reported a sensitivity of 87 % in a group of 16 patients including nine tumour patients. Sensitivity in the present study was lower but at 84.2 % still high. One reason for that might be the use of different cut-off criteria for a positive FISH test in the cited studies [13, 14] and ours. Because the cut-off criteria were initially created for bladder cancer, the differences in diagnostic sensitivities imply that further research activity should be spent on the determination of the best, probably different cut-off for UUT tumours. Even though this was not possible in the present study due to the small number of tumour patients, such an evaluation is important, as for example in the present study two out of three patients with UUT tumours and a negative FISH test revealed chromosomal aberrations just below the chosen cut-off. In contrast to bladder cancer [2, 23], the FISH test also revealed in our
Int Urol Nephrol
series a good sensitivity for low-grade and non-muscleinvasive tumours and showed no association of the chromosomes 3, 7, and 17 with high-stage tumours as described by others [10]. As these aspects cannot be sufficiently evaluated by the present data, they need to be addressed in larger series. Besides the use of different cut-off criteria another explanation for the differences in the sensitivities of the FISH test is patient selection. Whereas Mian et al. and Akkad et al. [13, 14] performed their studies in patients with suspected urothelial cancer of the UUT and found tumours in 38 and 56 %, respectively, our study was conducted in patients with only the differential diagnosis of urothelial cancer. This not only explains the low number of detected tumours (n = 20, 24 %), but also the relatively high number of other pathological conditions. However, even though one might have expected that this would have substantially increased the number of false-positive results, this was not the case in our study. Specificity of FISH in the present study was at 91.1 % similar to results by Mian et al. (89.5 %) and even higher than those reported by Akkad et al. (80 %) [13, 14]. Our data therefore furthermore underline the robustness of the FISH test against possible confounders. Still five patients had a false-positive FISH test in the present study. One had a ureterolithiasis, and one was known for bladder cancer. The other three underwent evaluation for haematuria (29 microhaematuria, 19 gross haematuria) with negative ureterorenoscopies but no histopathological evaluation. In all of them, missed flat tumour lesions or anticipator positive tests [24] remain possible and cannot be excluded because no extended follow-up is available for them. However, because the cytogenetic principles of the FISH test per se do not permit ‘‘true’’ false-positive tests, the ongoing clinical evaluation with long-term follow-up of patients with positive FISH tests remains of high importance. In comparison with Cyt, FISH was more sensitive in the present study. The difference between both tests, however, did not reach statistical significance as in other studies [10, 12, 14, 22]. This can be explained by the relatively good performance of Cyt with a sensitivity of 52.6 % and the small number of overlapping positive results in the present study. In most other studies, the sensitivity of Cyt was less good and ranged from 20.8 to 40 % [10, 12, 14, 22]. Regarding specificity, our results of Cyt were excellent (91 %) and in accordance with the current literature [13, 14, 22]. Nevertheless, in synopsis, we consider Cyt alone as a marker for UUT tumours as not suitable because of its unacceptably low sensitivity. FISH in contrast was more reliable with only three missed tumours and an equally high specificity. Its value became especially obvious in the subgroup of five patients with inconclusive or negative standard radiologic imaging in which Cyt detected two tumours and FISH all. The most important limitation of
FISH in the present study was its unevaluability due to a cell count below 25 cells, which was also observed by Mian et al. and Akkad et al. [13, 14]. In our opinion, this problem needs to be addressed by a standardization of the washing procedure for optimal sampling in the future. In the present study, the combination of both markers revealed the best detection rate of urothelial cancers with a sensitivity of 100 % when compared to both markers alone. This was achieved only by a moderate reduction in the specificity to 83.6 % due to the fact that overlapping positive results were found in only six patients. So we consider this combination a possible approach in future settings. In conclusion, in the so far largest series, FISH was a robust and reliable method in the detection of urothelial cancer of the UUT with a higher sensitivity than Cyt. According to the present data, FISH analysis alone or in combination with Cyt from washing urine of the UUT seems to be of high value for the detection of UUT tumours even in patients with negative or inconclusive radiologic imaging. Due to the high specificity, further diagnostic measures are strongly recommended in case of a positive FISH or Cyt test. For the detection of UUT tumours, the combination of FISH and Cyt might ease clinical decisions, especially if standard diagnostic tools such as radiologic imaging or biopsies are inconclusive or not diagnostic. Acknowledgments We thank Bianka Ilse for performing and evaluating the FISH tests and Dr. Thomas Lehmann for his statistical advice. Conflict of interest
The authors have nothing to declare.
References 1. Tilki D, Burger M, Dalbagni G, Grossman HB, Hakenberg OW, Palou J et al (2011) Urine markers for detection and surveillance of non-muscle-invasive bladder cancer. Eur Urol 60(3):484–92 2. Bubendorf L, Grilli B, Sauter G, Mihatsch MJ, Gasser TC, Dalquen P (2001) Multiprobe FISH for enhanced detection of bladder cancer in voided urine specimens and bladder washings. Am J Clin Pathol 116(1):79–86 3. Sarosdy MF, Schellhammer P, Bokinsky G, Kahn P, Chao R, Yore L et al (2002) Clinical evaluation of a multi-target fluorescent in situ hybridization assay for detection of bladder cancer. J Urol 168(5):1950–1954 4. Halling KC, King W, Sokolova IA, Meyer RG, Burkhardt HM, Halling AC et al (2000) A comparison of cytology and fluorescence in situ hybridization for the detection of urothelial carcinoma. J Urol 164(5):1768–1775 5. Jones JS (2006) DNA-based molecular cytology for bladder cancer surveillance. Urology 67(3 Suppl 1):35–45 (discussion-7) 6. Halling KC, King W, Sokolova IA, Karnes RJ, Meyer RG, Powell EL et al (2002) A comparison of BTA stat, hemoglobin dipstick, telomerase and Vysis UroVysion assays for the detection of urothelial carcinoma in urine. J Urol 167(5):2001–2006
123
Int Urol Nephrol 7. Fadl-Elmula I, Gorunova L, Mandahl N, Elfving P, Lundgren R, Rademark C et al (1999) Cytogenetic analysis of upper urinary tract transitional cell carcinomas. Cancer Genet Cytogenet 115(2):123–127 8. Nison L, Roupret M, Bozzini G, Ouzzane A, Audenet F, Pignot G et al (2013) The oncologic impact of a delay between diagnosis and radical nephroureterectomy due to diagnostic ureteroscopy in upper urinary tract urothelial carcinomas: results from a large collaborative database. World J Urol 31(1):69–76 9. Huang WT, Li LY, Pang J, Ruan XX, Sun QP, Yang WJ et al (2012) Fluorescence in situ hybridization assay detects upper urinary tract transitional cell carcinoma in patients with asymptomatic hematuria and negative urine cytology. Neoplasma 59(4): 355–360 10. Marin-Aguilera M, Mengual L, Ribal MJ, Burset M, Arce Y, Ars E et al (2007) Utility of a multiprobe fluorescence in situ hybridization assay in the detection of superficial urothelial bladder cancer. Cancer Genet Cytogenet 173(2):131–135 11. Wang J, Wu J, Peng L, Tu P, Li W, Liu L et al (2012) Distinguishing urothelial carcinoma in the upper urinary tract from benign diseases with hematuria using FISH. Acta Cytol 56(5):533–538 12. Shan Z, Wu P, Zheng S, Tan W, Zhou H, Zuo Y et al (2010) Evaluation of upper urinary tract tumors by FISH in Chinese patients. Cancer Genet Cytogenet 203(2):238–246 13. Akkad T, Brunner A, Pallwein L, Gozzi C, Bartsch G, Mikuz G et al (2007) Fluorescence in situ hybridization for detecting upper urinary tract tumors—a preliminary report. Urology 70(4):753–757 14. Mian C, Mazzoleni G, Vikoler S, Martini T, Knuchel-Clark R, Zaak D et al (2010) Fluorescence in situ hybridisation in the diagnosis of upper urinary tract tumours. Eur Urol 58(2):288–292 15. Sobin LH, Wittekind C (eds) (2002) TNM classification of malignant tumours, 6th edn. Wiley, New York
123
16. Inc. V. UroVysion Protocols—(Bladder Cancer FISH Assay) 2000 17. Kobayashi Y, Saika T, Miyaji Y, Saegusa M, Arata R, Akebi N et al (2012) Preoperative positive urine cytology is a risk factor for subsequent development of bladder cancer after nephroureterectomy in patients with upper urinary tract urothelial carcinoma. World J Urol 30(2):271–275 18. Umiker W, Lapides J, Sourenne R (1962) Exfoliative cytology of papillomas and intra-epithelial carcinomas of the urinary bladder. Acta Cytol 6:255–266 19. Highman WJ (1986) Transitional carcinoma of the upper urinary tract: a histological and cytopathological study. J Clin Pathol 39(3):297–305 20. Guarnizo E, Pavlovich CP, Seiba M, Carlson DL, Vaughan ED Jr, Sosa RE (2000) Ureteroscopic biopsy of upper tract urothelial carcinoma: improved diagnostic accuracy and histopathological considerations using a multi-biopsy approach. J Urol 163(1):52–55 21. Lotan Y, Roehrborn CG (2003) Sensitivity and specificity of commonly available bladder tumor markers versus cytology: results of a comprehensive literature review and meta-analyses. Urology 61(1):109–118 (discussion 18) 22. Luo B, Li W, Deng CH, Zheng FF, Sun XZ, Wang DH et al (2009) Utility of fluorescence in situ hybridization in the diagnosis of upper urinary tract urothelial carcinoma. Cancer Genet Cytogenet 189(2):93–97 23. Hajdinjak T (2008) UroVysion FISH test for detecting urothelial cancers: meta-analysis of diagnostic accuracy and comparison with urinary cytology testing. Urol Oncol 26(6):646–651 24. Skacel M, Fahmy M, Brainard JA, Pettay JD, Biscotti CV, Liou LS et al (2003) Multitarget fluorescence in situ hybridization assay detects transitional cell carcinoma in the majority of patients with bladder cancer and atypical or negative urine cytology. J Urol 169(6):2101–2105
UROLOGY - ORIGINAL PAPER
FISH analysis of washing urine from the upper urinary tract for the detection of urothelial cancers Torsten Gruschwitz • Mieczyslaw Gajda • Astrid Enkelmann • Marc-Oliver Grimm • Heiko Wunderlich • Marcus Horstmann Kerstin Junker
•
Received: 12 January 2014 / Accepted: 5 April 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract Introduction To evaluate FISH analysis of washing urine from the upper urinary tract (UUT) in comparison with cytology (Cyt) for the detection of urothelial cancers. Patients and methods In 82 patients with symptoms or abnormalities of the UUT sampling of washing urine for FISH and Cyt and a stepwise diagnostic work-up (e.g. retrograde ureteropyelography, ureterorenoscopy and endoscopic biopsy) were performed. In case of endoscopically and/or histologically proven malignancy patients either underwent nephroureterectomy, partial ureterectomy or local treatment. Sensitivity and specificity for FISH and Cyt as well as its combination were determined. Results Urothelial cancer of the UUT was detected in 20 patients. Eleven patients underwent nephroureterectomy, six partial ureterectomy and three endoscopic tumour
Marcus Horstmann and Kerstin Junker contributed equally to this work. T. Gruschwitz A. Enkelmann M.-O. Grimm H. Wunderlich M. Horstmann (&) K. Junker Department of Urology, Jena University Hospital, Lessingstr. 1, 07743 Jena, Germany e-mail: [email protected] M. Gajda Institute of Pathology, Jena University Hospital, Ziegelmu¨hlenweg 1, 07743 Jena, Germany H. Wunderlich Clinic of Urology and Pediatric Urology, St. Georg Klinikum, Mu¨hlha¨user Straße 94, 99817 Eisenach, Germany K. Junker Clinic of Urology and Pediatric Urology, Saarland University Medical Center, Kirrberger Straße 1, 66424 Homburg/Saar, Germany
treatment. This revealed nine pTa, three pT1 and seven muscle-invasive tumours. Twelve tumours were classified as low and seven as high-grade tumours. In one patient with a macroscopic unequivocal finding of tumour, endoscopic laser ablation without histologic confirmation was performed. FISH was evaluable in 76 patients and detected 16 tumours with a sensitivity and specificity of 84.2 and 91.1 %, respectively. Cyt was performed in 79 and was evaluable in 78 patients. It detected ten tumours with a sensitivity and specificity of 52.6 and 91.4 %, respectively. Cyt and FISH together detected 19 tumours with (sensitivity 100 % and specificity 83.6 %). Conclusion FISH was more sensitive than and equally specific to Cyt in the detection of urothelial cancers of the UUT. Both markers in combination revealed the best sensitivity, making it a possible approach in future settings. Keywords Urothelial cancer FISH Cytology Upper urinary tract
Introduction The UroVisionTM FISH test is currently approved as a urinary-based tumour marker in the detection and followup of urothelial bladder cancer [1]. It detects aneuploidy of chromosomes 3, 7, 17 and loss of locus 9p21, which are known alterations in urothelial cancer. In clinical evaluation, FISH was highly reliable with a higher sensitivity than and an equal specificity to cytology (Cyt) [2–4]. Probably, due to its cytogenetic-based principle, FISH is in contrast to other urinary-based tumour markers less likely to become false positive [5, 6]. Because of its high reliability and the proven molecular similarity of urothelial cancers of the bladder and the upper urinary tract (UUT) [7], FISH is
123
Int Urol Nephrol
also considered as a potential marker for the detection of urothelial cancers in these anatomic regions. Clinically, this would be helpful, because diagnosis of UUT tumours is still hampered by limited endoscopic access and a high risk of non-diagnostic biopsies [8]. However, until to date, FISH is not approved as a marker for UUT tumours and studies evaluating FISH in the UUT remain scarce. They include either studies of voided [9–12] or of endoscopically collected urine from the UUT [13, 14] and evaluated only patients with known or suspected urothelial cancer. In the present study, we evaluated FISH as a diagnostic tool in the so far largest group of patients (n = 82). In contrast to other studies, we included patients who had either symptoms or abnormalities of the UUT that made urothelial cancer one differential diagnosis among others. In fact, 62 out of 82 patients were finally diagnosed to have no or other abnormalities of the UUT. During a stepwise diagnostic work-up, FISH analysis and Cyt were conducted endoscopically.
Patients and methods Between 2006 and 2012, washing urine form the UUT was analysed for Cyt and FISH in 82 patients in one single centre. Clinical data were evaluated retrospectively, and the study was approved by the local ethics committee (Approval No: 3693-02/13). Mean age of all patients was 66 years (range 36–89 years). Of the patients, 47 were men and 35 women. Seventeen had a history of urothelial cancer (bladder cancer: n = 15, urothelial cancer of the opposite UUT: n = 2 or both: n = 1). Samples of washing urine were collected from the ureter or the renal pelvis depending on the site of the suspected lesion. Saline was injected through the endoscope or a ureter catheter positioned below the suspected lesion. In case of no obvious lesion, urine was retrieved from the ureter. Samples of 5–10 ml were collected by syringe aspiration or dropwise urine backflow. Cyt was performed by one senior pathologist (M.G.) according to the cytopathologic classification of the Union International Contre le Cancer [15]. Samples with no or mild atypia were classified as negative and samples with moderate to severe as positive. The commercially available UroVysionTM test (Abbott Molecular/Vysis, Des Plaines, IL, USA) was performed according to the description of the manufacturer for FISH analysis [16]. A test was considered evaluable if at least 25 cells were present. Of them, 25—if present—morphologically atypical cells were selected for evaluation. If not present, morphologically normal cells were evaluated. For a positive test, the following criteria were applied: A test was considered positive if of 25 nuclei at least 5 exhibited 2
123
or more aneuploidies (gains) of chromosomes 3, 7 or 17, or if at least 10 nuclei exhibited 1 or more aneuploidies (gains) of chromosomes 3, 7 or 17, or if at least 10 nuclei exhibited a complete loss of the locus 9p21 [4, 6, 16]. If none of these criteria were met the test was negative. Neither the cytopathologist nor the FISH reviewer had knowledge of the clinical findings or the history of the patients. The reference standard for Cyt and FISH was the complete radiologic, endoscopic and histopathological work-up including final surgical treatment. Patients were first evaluated by standard radiologic imaging with intravenous pyelography or CT scan. Then, patients underwent retrograde pyelography and uretero-/or ureterorenoscopy with or without biopsy. Surgical treatment was based on histologically proven malignancy and in some cases on endoscopical findings only. Depending on their tumour characteristics and general condition patients either underwent nephroureterectomy, partial ureterectomy or local treatment. Sensitivity and specificity for Cyt, FISH and its combination were determined according to the final histopathological or, if not available, endoscopic and radiologic results. In the combination of Cyt and FISH, the marker panel was considered positive if either Cyt or FISH was positive. Patients’ characteristics were analysed by descriptive statistics. Sensitivity and specificity for Cyt and FISH were calculated and compared by a McNemar test. A p value of \0.05 was considered significant. Analyses were conducted with IBM SPSS Statistics for Windows, Version 21.0., Armonk, NY: IBM Corp., Released 2012.
Results Urothelial cancer of the UUT was found in 20/82 patients. Mean age of them was 72 years (range 55–89 years). Twelve were men and eight women. Eleven patients underwent nephroureterectomy, six partial ureterectomy and three endoscopic tumour treatment. Histopathological evaluation revealed nine pTa, three pT1 and seven muscleinvasive tumours. Twelve tumours were classified as low and seven as high-grade tumours. In one patient with a history of nephrectomy of his right kidney and a macroscopic unequivocal finding of tumour in the left renal pelvis endoscopic laser ablation was performed without histologic confirmation. None of the patients had concomitant urothelial cancer of the bladder at the time of urine collection. Sixty-two patients had no urothelial cancer of the investigated UUT or bladder. Their mean age was 64 years (range 55–89 years). Of these patients, twelve had a nephroureterolithiasis and nine a ureteral stricture.
Int Urol Nephrol Table 1 Sensitivity, specificity, positive and negative predictive values of FISH, Cyt and the combination of Cyt and FISH
Sensitivity
Cyt (%)
FISH (%)
p value
Cyt/FISH (%)
52.6
84.2
0.146
100.0
1.000
Table 2 Sensitivity of Cyt and FISH according to tumour grade in 18 tumour patients with available histopathology Patients (n)
Cyt sens (%)
FISH sens (%)
Low grade
11
54.5
81.8
7
57.1
85.7
Specificity
91.4
91.1
83.6
High grade
PPV
66.7
76.2
65.5
FISH fluorescence in situ hybridisation, Cyt cytology
NPV
85.5
94.4
83.9
FISH fluorescence in situ hybridisation, Cyt cytology, PPV positive predictive value, NPV negative predictive value
Table 3 Sensitivity of Cyt and FISH according to tumour stage (T) in 18 tumour patients with available histopathology
p values were evaluated for the difference between Cyt and FISH alone
Malignant diseases other than urothelial cancer were found in six patients. Of them, three had renal cell carcinoma, one a renal papilloma (not classified as urothelial cancer), one a recurrent extraureteral lymphoma and one a leiomyosarcoma. Cyt was performed in 79 patients and evaluable in 78 (99 %). In the group of patients with UUT tumours (n = 20), Cyt was performed and evaluable in 19 patients. It was not performed in one patient with a pT2G2 tumour. In these 19 patients, ten tumours were correctly diagnosed. However, 49 pTaG low-grade, 19 pT1 high-grade, 1 pT3 low-grade, 29 pT3-4 high-grade tumours and the only endoscopically confirmed tumour were not missed by Cyt (Table 4). Based on these 19 patients, Cyt had a sensitivity of 52.6 %. It was negative in 62 patients and correctly diagnosed 53 patients without cancer with a specificity of 91.4 % (Table 1). FISH was performed in 82 patients and evaluable in 76 (93 %). In six patients, it was not evaluable because of a cell count below 25. Of the 20 patients with UUT tumours, FISH was not evaluable in the same patient as in Cyt. FISH correctly diagnosed 16 tumours and was negative in two pTa-1 low-grade and one pT3 high-grade tumour (Table 4). Based on the calculation of 19 evaluable tests, its sensitivity was 84.2 % (Table 1). The specificity of FISH was 91.1 %. It correctly diagnosed 52 cancer-free patients and had five false-positive results. Of these patients, one had a history for bladder cancer, one a ureterolithiasis and three painless haematuria. Regarding tumour grade both tests revealed a slightly better sensitivity for high-grade tumours (Table 2). Whereas Cyt missed three high-grade tumours, FISH only missed one. Regarding T-category, Cyt had a sensitivity of 58.3 % and FISH of 83.3 % for non-muscle-invasive tumours (pTa–pT1), and for muscle-invasive tumours ([pT1) of 50.0 and 83.3 %, respectively (Table 3). Statistical comparison of Cyt and FISH revealed no significant differences regarding their sensitivity (McNemar: p = 0.146, n.s.) (Table 1). Furthermore, the subgroup of patients with inconclusive or negative standard radiologic
Patients (n)
Cyt sens (%)
FISH sens (%)
pTa-1
12
58.3
83.3
pT2-4
6
50.0
83.3
FISH fluorescence in situ hybridisation, Cyt cytology
imaging was evaluated. It included three patients with pTa low-grade tumours, one patient with a pT3 high-grade tumour and the patient with endoscopic laser ablation. In this subgroup, Cyt was positive in two patients (29 pTa low grade) and FISH in all. The best sensitivity was found by the combination of Cyt and FISH. Together, they detected 19 tumours with a sensitivity of 100.0 %. This was calculated without the patient in whom neither Cyt nor FISH results were available. The combination of both markers resulted in 10 falsepositive results and a specificity of 83.6 %. Only six of the patients with tumours had positive results in both tests, eight had a positive FISH and a negative Cyt and three a positive Cyt and a negative FISH, respectively (Table 4).
Discussion Both FISH and cytology (Cyt) can be used for the detection of urothelial cancer of the UUT in voided urine and washing urine form the UUT. Voided urine has the advantage of a non-invasive test but leaves the localisation of a tumour in case of a positive test open. If retrieved directly from the UUT, the cancer site can be encircled more precisely, but the test becomes invasive [8, 13]. For Cyt, which is the current gold standard [17], it was shown that its sensitivity increases if washing urine was taken from the UUT [13, 18]. Because of potential cell damage, however, these samples are also discussed to create more false-positive results [19, 20]. Even though Cyt has an excellent specificity, its main disadvantages remain its observer dependence and a limited sensitivity in low-grade tumours [21]. In the present study, we evaluated FISH in comparison with Cyt from washing urine of the UUT during an endoscopic work-up of patients symptomatic from or with
123
Int Urol Nephrol Table 4 Cyt and FISH results of patients with urothelial cancer of the upper urinary tract (n = 20) Pat
Sex
Age
Surgical intervention
T Stage
Grading
Cyt
Gain of 3 (cells)
Gain of 7 (cells)
Gain of 17 (cells)
Complete loss of 9p21 (cells)
FISH
1
F
62
Nephroureterectomy
pTa
Low
Positive
25
26
19
0
Positive
2
M
70
Nephroureterectomy
pT1
High
Positive
25
25
23
2
Positive
3
F
85
pTa
Low
Negative
25
25
25
0
Positive
4
M
86
Transurethral resection Endoscopic laser ablation
pTa
Low
Negative
23
21
17
5
Positive
5
F
78
Nephroureterectomy
pTa
Low
Positive
0
0
17
25
Positive
6
M
73
Endoscopic laser ablation
n.e.
n.e.
Negative
20
20
20
20
Positive
7
M
76
Nephroureterectomy
pT3
High
Positive
16
18
18
0
Positive
8
F
70
Nephroureterectomy
pT1
Low
Positive
0
1
3
7
Negative
9
F
70
Nephroureterectomy
pT1
High
Negative
16
18
18
0
Positive
10
M
76
Distal ureter resection
pTa
Low
Positive
3
4
2
0
Negative
11
M
67
Nephroureterectomy
pT2
Low
n.e.
n.e.
n.e.
n.e.
n.e.
Not enough cells
12
F
77
Nephroureterectomy
pT3
Low
Negative
0
0
0
23
Positive
13 14
F M
66 70
Nephroureterectomy Nephroureterectomy
pT3 pT3
High High
Positive Negative
0 1
0 0
0 0
0 18
Negative Positive
15
M
73
Palliative nephrectomy
pT4
High
Negative
0
0
0
10
Positive
16
M
56
Distal ureter resection
pTa
Low
Positive
8
8
6
0
Positive
17
M
76
Distal ureter resection
pTa
Low
Negative
3
6
3
25
Positive
18
M
73
Distal ureter resection
pTa
Low
Positive
11
11
11
n.e.
Positive
19
F
89
Distal ureter resection
pT3
High
Positive
25
25
25
5
Positive
20
M
55
Distal ureter resection
pTa
Low
Negative
15
15
15
0
Positive
n.e. not evaluated, Cyt cytology, FISH fluorescence in situ hybridisation
suspicious findings of the UUT. The excellent diagnostic value of FISH for UUT tumours was already shown in studies from voided urine in patients with known urothelial cancers in comparison with healthy persons [10, 12, 22] and in comparison with patients with benign urological diseases [11]. In these studies, reported sensitivities and specificities range from 73.5 to 85.7 % and 94 to 100 %, respectively [10, 12, 22]. FISH was, furthermore, evaluated from washing urine of the UUT in patients suspicious for UUT tumours. Data regarding this, however, remain scarce with to our knowledge only two published studies. They revealed— most probably because of a direct access to mobilised urothelial cells—similar to Cyt the highest sensitivities [13, 14]. Mian et al. reported a sensitivity of 100 % in a group of 55 patients including 24 tumour patients, [14] and
123
Akkad et al. [13] reported a sensitivity of 87 % in a group of 16 patients including nine tumour patients. Sensitivity in the present study was lower but at 84.2 % still high. One reason for that might be the use of different cut-off criteria for a positive FISH test in the cited studies [13, 14] and ours. Because the cut-off criteria were initially created for bladder cancer, the differences in diagnostic sensitivities imply that further research activity should be spent on the determination of the best, probably different cut-off for UUT tumours. Even though this was not possible in the present study due to the small number of tumour patients, such an evaluation is important, as for example in the present study two out of three patients with UUT tumours and a negative FISH test revealed chromosomal aberrations just below the chosen cut-off. In contrast to bladder cancer [2, 23], the FISH test also revealed in our
Int Urol Nephrol
series a good sensitivity for low-grade and non-muscleinvasive tumours and showed no association of the chromosomes 3, 7, and 17 with high-stage tumours as described by others [10]. As these aspects cannot be sufficiently evaluated by the present data, they need to be addressed in larger series. Besides the use of different cut-off criteria another explanation for the differences in the sensitivities of the FISH test is patient selection. Whereas Mian et al. and Akkad et al. [13, 14] performed their studies in patients with suspected urothelial cancer of the UUT and found tumours in 38 and 56 %, respectively, our study was conducted in patients with only the differential diagnosis of urothelial cancer. This not only explains the low number of detected tumours (n = 20, 24 %), but also the relatively high number of other pathological conditions. However, even though one might have expected that this would have substantially increased the number of false-positive results, this was not the case in our study. Specificity of FISH in the present study was at 91.1 % similar to results by Mian et al. (89.5 %) and even higher than those reported by Akkad et al. (80 %) [13, 14]. Our data therefore furthermore underline the robustness of the FISH test against possible confounders. Still five patients had a false-positive FISH test in the present study. One had a ureterolithiasis, and one was known for bladder cancer. The other three underwent evaluation for haematuria (29 microhaematuria, 19 gross haematuria) with negative ureterorenoscopies but no histopathological evaluation. In all of them, missed flat tumour lesions or anticipator positive tests [24] remain possible and cannot be excluded because no extended follow-up is available for them. However, because the cytogenetic principles of the FISH test per se do not permit ‘‘true’’ false-positive tests, the ongoing clinical evaluation with long-term follow-up of patients with positive FISH tests remains of high importance. In comparison with Cyt, FISH was more sensitive in the present study. The difference between both tests, however, did not reach statistical significance as in other studies [10, 12, 14, 22]. This can be explained by the relatively good performance of Cyt with a sensitivity of 52.6 % and the small number of overlapping positive results in the present study. In most other studies, the sensitivity of Cyt was less good and ranged from 20.8 to 40 % [10, 12, 14, 22]. Regarding specificity, our results of Cyt were excellent (91 %) and in accordance with the current literature [13, 14, 22]. Nevertheless, in synopsis, we consider Cyt alone as a marker for UUT tumours as not suitable because of its unacceptably low sensitivity. FISH in contrast was more reliable with only three missed tumours and an equally high specificity. Its value became especially obvious in the subgroup of five patients with inconclusive or negative standard radiologic imaging in which Cyt detected two tumours and FISH all. The most important limitation of
FISH in the present study was its unevaluability due to a cell count below 25 cells, which was also observed by Mian et al. and Akkad et al. [13, 14]. In our opinion, this problem needs to be addressed by a standardization of the washing procedure for optimal sampling in the future. In the present study, the combination of both markers revealed the best detection rate of urothelial cancers with a sensitivity of 100 % when compared to both markers alone. This was achieved only by a moderate reduction in the specificity to 83.6 % due to the fact that overlapping positive results were found in only six patients. So we consider this combination a possible approach in future settings. In conclusion, in the so far largest series, FISH was a robust and reliable method in the detection of urothelial cancer of the UUT with a higher sensitivity than Cyt. According to the present data, FISH analysis alone or in combination with Cyt from washing urine of the UUT seems to be of high value for the detection of UUT tumours even in patients with negative or inconclusive radiologic imaging. Due to the high specificity, further diagnostic measures are strongly recommended in case of a positive FISH or Cyt test. For the detection of UUT tumours, the combination of FISH and Cyt might ease clinical decisions, especially if standard diagnostic tools such as radiologic imaging or biopsies are inconclusive or not diagnostic. Acknowledgments We thank Bianka Ilse for performing and evaluating the FISH tests and Dr. Thomas Lehmann for his statistical advice. Conflict of interest
The authors have nothing to declare.
References 1. Tilki D, Burger M, Dalbagni G, Grossman HB, Hakenberg OW, Palou J et al (2011) Urine markers for detection and surveillance of non-muscle-invasive bladder cancer. Eur Urol 60(3):484–92 2. Bubendorf L, Grilli B, Sauter G, Mihatsch MJ, Gasser TC, Dalquen P (2001) Multiprobe FISH for enhanced detection of bladder cancer in voided urine specimens and bladder washings. Am J Clin Pathol 116(1):79–86 3. Sarosdy MF, Schellhammer P, Bokinsky G, Kahn P, Chao R, Yore L et al (2002) Clinical evaluation of a multi-target fluorescent in situ hybridization assay for detection of bladder cancer. J Urol 168(5):1950–1954 4. Halling KC, King W, Sokolova IA, Meyer RG, Burkhardt HM, Halling AC et al (2000) A comparison of cytology and fluorescence in situ hybridization for the detection of urothelial carcinoma. J Urol 164(5):1768–1775 5. Jones JS (2006) DNA-based molecular cytology for bladder cancer surveillance. Urology 67(3 Suppl 1):35–45 (discussion-7) 6. Halling KC, King W, Sokolova IA, Karnes RJ, Meyer RG, Powell EL et al (2002) A comparison of BTA stat, hemoglobin dipstick, telomerase and Vysis UroVysion assays for the detection of urothelial carcinoma in urine. J Urol 167(5):2001–2006
123
Int Urol Nephrol 7. Fadl-Elmula I, Gorunova L, Mandahl N, Elfving P, Lundgren R, Rademark C et al (1999) Cytogenetic analysis of upper urinary tract transitional cell carcinomas. Cancer Genet Cytogenet 115(2):123–127 8. Nison L, Roupret M, Bozzini G, Ouzzane A, Audenet F, Pignot G et al (2013) The oncologic impact of a delay between diagnosis and radical nephroureterectomy due to diagnostic ureteroscopy in upper urinary tract urothelial carcinomas: results from a large collaborative database. World J Urol 31(1):69–76 9. Huang WT, Li LY, Pang J, Ruan XX, Sun QP, Yang WJ et al (2012) Fluorescence in situ hybridization assay detects upper urinary tract transitional cell carcinoma in patients with asymptomatic hematuria and negative urine cytology. Neoplasma 59(4): 355–360 10. Marin-Aguilera M, Mengual L, Ribal MJ, Burset M, Arce Y, Ars E et al (2007) Utility of a multiprobe fluorescence in situ hybridization assay in the detection of superficial urothelial bladder cancer. Cancer Genet Cytogenet 173(2):131–135 11. Wang J, Wu J, Peng L, Tu P, Li W, Liu L et al (2012) Distinguishing urothelial carcinoma in the upper urinary tract from benign diseases with hematuria using FISH. Acta Cytol 56(5):533–538 12. Shan Z, Wu P, Zheng S, Tan W, Zhou H, Zuo Y et al (2010) Evaluation of upper urinary tract tumors by FISH in Chinese patients. Cancer Genet Cytogenet 203(2):238–246 13. Akkad T, Brunner A, Pallwein L, Gozzi C, Bartsch G, Mikuz G et al (2007) Fluorescence in situ hybridization for detecting upper urinary tract tumors—a preliminary report. Urology 70(4):753–757 14. Mian C, Mazzoleni G, Vikoler S, Martini T, Knuchel-Clark R, Zaak D et al (2010) Fluorescence in situ hybridisation in the diagnosis of upper urinary tract tumours. Eur Urol 58(2):288–292 15. Sobin LH, Wittekind C (eds) (2002) TNM classification of malignant tumours, 6th edn. Wiley, New York
123
16. Inc. V. UroVysion Protocols—(Bladder Cancer FISH Assay) 2000 17. Kobayashi Y, Saika T, Miyaji Y, Saegusa M, Arata R, Akebi N et al (2012) Preoperative positive urine cytology is a risk factor for subsequent development of bladder cancer after nephroureterectomy in patients with upper urinary tract urothelial carcinoma. World J Urol 30(2):271–275 18. Umiker W, Lapides J, Sourenne R (1962) Exfoliative cytology of papillomas and intra-epithelial carcinomas of the urinary bladder. Acta Cytol 6:255–266 19. Highman WJ (1986) Transitional carcinoma of the upper urinary tract: a histological and cytopathological study. J Clin Pathol 39(3):297–305 20. Guarnizo E, Pavlovich CP, Seiba M, Carlson DL, Vaughan ED Jr, Sosa RE (2000) Ureteroscopic biopsy of upper tract urothelial carcinoma: improved diagnostic accuracy and histopathological considerations using a multi-biopsy approach. J Urol 163(1):52–55 21. Lotan Y, Roehrborn CG (2003) Sensitivity and specificity of commonly available bladder tumor markers versus cytology: results of a comprehensive literature review and meta-analyses. Urology 61(1):109–118 (discussion 18) 22. Luo B, Li W, Deng CH, Zheng FF, Sun XZ, Wang DH et al (2009) Utility of fluorescence in situ hybridization in the diagnosis of upper urinary tract urothelial carcinoma. Cancer Genet Cytogenet 189(2):93–97 23. Hajdinjak T (2008) UroVysion FISH test for detecting urothelial cancers: meta-analysis of diagnostic accuracy and comparison with urinary cytology testing. Urol Oncol 26(6):646–651 24. Skacel M, Fahmy M, Brainard JA, Pettay JD, Biscotti CV, Liou LS et al (2003) Multitarget fluorescence in situ hybridization assay detects transitional cell carcinoma in the majority of patients with bladder cancer and atypical or negative urine cytology. J Urol 169(6):2101–2105
