Evidence Table Clinical Area:

Evidence Table
Clinical Area:
Reference:
UroVysion FISH test for the detection of bladder cancer.
1. Halling KC, King W, Sokolova I, et al. A comparison of cytology
and fluorescence in situ hybridization for the detection of
urothelial carcinoma. J Urol 2000;164:1768-1775.
2. Halling KC, King W, Sokolova I, et al. A comparison of BTA
stat, hemoglobin dipstick, telomerase and vysis assays for the
detection of urothelial carcinoma in urine. J Urol 2002;167:20012006.
Study Type: Comparison of diagnostic tests.
Study Aim: To compare FISH (UroVysion) test with cytology, and other noninvasive
tests for the detection of bladder cancer.
Outcomes
• Primary: Sensitivity, specificity.
Design
• Number of subjects: N=265.
• Description of study population: Patient ages ranged from 36 to 94 years with a mean
of 69.7 years. 75% were men, 150 (56.6%) had a history of urothelial carcinoma, and
115 (43.4%) did not.
• Inclusion criteria/Exclusion criteria: The study included patients referred to
cystoscopy for a history of urothelial carcinoma or the evaluation of genitourinary
symptoms/signs. Inclusion/exclusion criteria were not discussed.
• Procedure: A voided urine sample was collected from all patients immediately
before the scheduled cystoscopy. UroVysion analysis, BTA stat, and hemoglobin
dipstick assays were performed according to the manufacturer’s specifications Urine
specimens for cytology were collected no more than 2 weeks before the day of the
scheduled cystoscopy. 131 biopsies or surgical resections were performed on 121
(46%) patients and studied histologically. Biopsy was not performed among 144
(54%) patient (80 had no history of bladder cancer, and 64 had a history of bladder
cancer).
Validity:
• Independent blind comparison with a gold standard or follow-up of those not
receiving the gold standard test? The gold standard was the cystoscopic/histological
results of the bladder biopsy. Tests were evaluated blindly without knowledge of the
cystoscopy or cytology results, or clinical history.
• Was “normal” defined? Yes.
• Appropriate spectrum of disease? Yes.
• Consecutive patients? Not discussed.
• Methods described in enough detail to enable you to replicate the test? Yes.
• Reproducible results? Yes.
2
•
Conclusions regarding validity of methods:
The study had its advantages and limitations. It had a gold standard to compare FISH
assay results with, was conducted on a wide spectrum of patients, with or without a
history of bladder carcinoma, and the test results were evaluated by personnel blinded to
the histology, and cytology results. However, the authors did not discuss how patients
were selected to the study, if they were consecutive, and what the inclusion/exclusion
criteria were.
Results:
The histologic results of the bladder biopsy/resection performed among 120 (46%)
patients, were positive for urothelial cancer in 57%, negative in 36%, showed
atypia/dysplasia in 5%, and metastatic carcinoma in 2%.
The TNM tumor stage for those with cancer was Ta in 51% of the patients, Tis in 24%,
T1 in 12%, and T2 or greater in 13%. The tumor grade was G1 in 16% of cases, G2 in
33%, G3 in 51%.
Overall sensitivity and specificity of FISH test and cytology, and sensitivity of by stage
and grade of the transitional cell bladder carcinoma*
FISH
Cytology
(No/Total No)
Overall sensitivity 81% (59/73)
58% (40/69)
Overall specificity 96% (75/78)
98% (48/49)
p value
0.001
0.564
Sensitivity by stage
Ta
65% (24/37)
Tis
100% (17/17)
T1-4
95% (18/19)
47% (17/36)
78% (14/18)
60% ( 9/15)
0.058
0.046
0.025
Sensitivity by grade
1
36% ( 4/11)
2
76% (19/25)
3
97% (36/37)
27% ( 3/11)
54% (13/24)
71% (24/34)
0.564
0.059
0.003
*Sensitivity of FISH and cytology was determined for 75 patients with biopsy proven urothelial
carcinoma, and specificity was determined for 80 patients with no history of urothelial carcinoma and
negative cystoscopic findings.
Specificity was also calculated for 43 healthy donors older than 50 years, and was =100%
7 of the 11 patients with positive FISH but negative biopsy had a follow-up biopsy (after a range 3-12
months) that revealed recurrent urothelial carcinoma.
3
Overall Sensitivity and specificity of different noninvasive tests
Using cystoscopy/histopathology as gold standard
FISH test
BTA stat
Hemoglobin dipstick
Telomerase
Sensitivity
Specificity
81%*
78%
74%
46%
96%**
74%
51%
91%
* The difference in sensitivity of FISH compared to the other tests was only significant for telomerase
test (p=0.0001), which had the lowest sensitivity of all.
**FISH had significantly higher specificity than BTA stat (p=0.0001), and hemoglobin dipstick
(p<0.0001)
Authors’ Conclusions:
The authors concluded that the sensitivity of FISH test is superior to that of cytology,
BTA stat, hemoglobin dipstick and telomerase in detecting urothelial carcinoma. They
noted that it is a promising assay for the detection of urothelial carcinoma in urine
specimens and has the potential of improving the management of the disease. They
recommended further studies to determine the role of these tests in the treatment of
patients with superficial urothelial carcinoma.
Reviewer’s Conclusions:
The study had generally valid methodology, with some limitations. Its show that
compared to the gold standard, FISH test had an overall sensitivity of 81% and a
specificity of 96% in detecting urinary bladder carcinoma. It appears to have a low
sensitivity in detecting low grade, and early carcinomas, and a higher sensitivity for
detecting grade3, and T1-T4 carcinomas. Compared to the urine cytology, and other
noninvasive tests FISH assay had a significantly higher sensitivity than urine cytology,
and telomerase test, and similar sensitivity to BTA stat hemoglobin dipstick. The
specificity of FISH on the other hand was significantly higher than that of PTA stat and
hemoglobin dipstick but not to urine cytology or telomerase.