- Email: [email protected]
Influence of Renal Excretory Function on the Performance of Urine Based Markers to Detect Bladder Cancer
Influence of Renal Excretory Function on the Performance of Urine Based Markers to Detect Bladder Cancer Tilman Todenhöfer, Jörg Hennenlotter, Marc Witstruk, Georgios Gakis, Stefan Aufderklamm, Ursula Kuehs, Arnulf Stenzl and Christian Schwentner* From the Department of Urology, University Hospital Tübingen, Tübingen, Germany
Abbreviations and Acronyms BC ⫽ bladder cancer CKD ⫽ chronic kidney disease CYT ⫽ voided urine cytology FISH ⫽ fluorescence in situ hybridization GFR ⫽ glomerular filtration rate HU ⫽ hematuria NMP ⫽ nuclear matrix protein UTI ⫽ urinary tract infection Submitted for publication May 27, 2011. Study received institutional review board approval. * Correspondence: Department of Urology, Eberhard-Karls University, Hoppe-Seyler Str. 3, D-72076 Tübingen, Germany (telephone: ⫹497071-29-85092; FAX: ⫹49-7071-29-5092; e-mail: [email protected]).
Purpose: In hematuria cases urine based tests are used to detect bladder cancer, although the diagnostic yield remains insufficient due to influencing variables, including urinary tract infection. Many patients are elderly with renal insufficiency and have proteinuria as an additional influencing factor. To our knowledge no data are available on the accuracy of urine based bladder cancer tests in conjunction with renal function. Materials and Methods: Urine samples of 449 patients with hematuria and histology were included in analysis. Cytology, fluorescence in situ hybridization, immunocytology and nuclear matrix protein 22 assay were done. Renal function was classified as normal, impaired or severely impaired based on serum creatinine, the glomerular filtration rate and proteinuria. False-positive rates were statistically compared in regard to renal function. Results: A total of 382 patients did not have bladder cancer. There was an increased false-positive rate for creatinine and the glomerular filtration rate. The nuclear matrix protein 22 test showed a 22.0% and 46.7% false-positive rate in the normal and limited function cohorts, respectively (p ⫽ 0.05). Similar trends were noted for proteinuria. Indeterminate significance was detected, separating those with severely impaired function for immunocytology and those in the normal group for fluorescence in situ hybridization (p ⫽ 0.08 and 0.06, respectively). Proteinuria was a significant factor for urine cytology with increased false-positive results in the absence of urinary tract infection (p ⫽ 0.0017 and 0.05, respectively). Conclusions: To our knowledge this is the first study of renal function and the accuracy of urine based bladder cancer markers. Renal function influences the diagnostic yield. A decreased glomerular filtration rate was associated with increased false-positive nuclear matrix protein 22 results while proteinuria decreased urine cytology specificity. Renal function should be considered when urine based bladder cancer tests are interpreted. Key Words: urinary bladder; urinary bladder neoplasms; kidney function tests; tumor markers, biological; false positive reactions
68
www.jurology.com
UROTHELIAL carcinoma of the bladder is the fourth most common malignancy in men in the United States, accounting for more than 70,000 new cases and 14,000 deaths annually.1 Of the patients 70% initially present
with Ta/T1 tumors. Depending on risk stratification up to 15% of these tumors progress to a muscle invasive stage, which is associated with a limited 5-year survival rate.2 Early diagnosis is essential for timely treatment and
0022-5347/12/1871-0068/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
Vol. 187, 68-73, January 2012 Printed in U.S.A. DOI:10.1016/j.juro.2011.09.023
AND
RESEARCH, INC.
INFLUENCE OF RENAL EXCRETORY FUNCTION ON MARKERS TO DETECT BLADDER CANCER
subsequent decreased mortality. Cystoscopy with transurethral resection remains the gold standard for BC diagnosis.3 Several urine based tests are available for BC diagnosis and followup. Urine cytology is the most widely accepted, noninvasive marker for BC primary diagnosis and followup.3 Urine cytology has relatively low sensitivity for low grade tumors and, thus, it cannot replace cystoscopy.4 There is considerable interobserver variability and urine cytology depends highly on cytopathologist experience.5 Other noninvasive urine markers are gaining importance for diagnosis and followup. FISH uses a multitarget set of probes that specifically bind to centromeres of chromosomes 3, 7 and 17, and to the 9p21 locus of chromosome 9. Loss of the 9p21 locus, which is the site of the p16 tumor suppressor gene, was identified as the earliest, most frequent genetic aberration in patients with BC.6,7 The uCyt⫹™ immunocytological fluorescence assay uses 3 antibodies to detect carcinoembryonic antigen and mucinlike glycoprotein on the cell surface. NMP22, which regulates mitosis, is over expressed in malignant urothelial cells. NMP22 is released into urine by apoptotic cells and it can be detected and quantified.8 Generally these molecular tests show increased sensitivity but decreased specificity compared to CYT.4 However, several modifiable factors influence test performance. UTI can result in findings of falsepositive CYT9 and increased NMP22.8 Tests targeting NMP22 or other molecular markers yield increased false-positive findings in patients with previous instillation therapy, foreign bodies, benign prostate hyperplasia or stone disease.10,11 Changes in urine composition, such as HU and pyuria, can yield falsepositive results.12 Interference of urine constituents with urine based tests is widely acknowledged but to our knowledge there are no data on the influence of renal excretory function on test performance. This issue is of tremendous interest since patients with a primary diagnosis of BC are often elderly and increased age is associated with a higher CKD prevalence.13 Thus, we evaluated the impact of renal function on the performance of the 4 most widely used noninvasive urine tests.
PATIENTS AND METHODS Patients and Samples Enrolled in our study were 380 men and 69 women with a mean age of 65.5 years (range 18 to 93) with microscopic or macroscopic HU and without a BC history. The study received institutional review board approval. Voided urine was used for analysis and investigated by dipstick test and microscopy. Protein concentration was analyzed
69
quantitatively by a Miditron® M. CYT, UroVysion® for FISH, uCyt⫹ and the NMP22® Bladder Cancer Test were done. All patients underwent cystoscopy and histological assessment.
Urine Processing For CYT slides were cytospinned and stained according to Papanicolaou. Specimens were assessed by light microscopy according to defined criteria. FISH was done according to manufacturer instructions. Positive result criteria were 4 or greater nuclei of a total of 25 morphological suspicious cells showing 3 or greater signals of at least 2 of chromosomes 3, 7 or 17, or at least 12 nuclei showing no 9p21 signal.14 The uCyt⫹ test was done according to manufacturer instructions. The criterion for a positive result was at least 1 positive cell. NMP22 enzyme-linked immunosorbent assay was done according to the manufacturer protocol with concentrations greater than 10 U/ml classified as positive.8 Since UTI and previous mechanical manipulation such as catheterization are considered contraindications to the test, subgroup analysis was performed for patients without UTI or previous mechanical manipulation.
Urinary Tract Infection UTI was determined by dipstick test and urine microscopy. It was defined as at least 100 leukocytes per l and simultaneously as more than 1 erythrocyte or at least 100 leukocytes per l and the simultaneous presence of urine nitrite.
Renal Function Parameters Serum creatinine, GFR and urine protein15 served as renal function parameters. Creatinine was measured using standard laboratory equipment. GFR was calculated using the modification of diet in renal disease formula.16 Based on these parameters patients were divided into renal function groups, including group 1—normal, group 2—impaired and group 3—severely impaired. An algorithm respecting GFR and the urine protein concentration was used to subdivide patients into 3 renal function groups. Table 1 shows patient subclassification by these parameters.
Statistical Analysis Sensitivity and specificity were calculated after contingency analysis of the whole cohort and defined subgroups. False-positive rates were compared by contingency analysis depending on renal function according to the defined criteria. JMP® 7.2 and Pearson’s chi-square test were used with p ⬍0.05 considered significant. Values of p ⫽ 0.05 were considered borderline significant and are reported descriptively.
RESULTS Of the patients 382 patients (85.1%) did not have BC and 67 (14.9%) were diagnosed with BC by histopathology, including 42 (62.7%) with stage pTa, 13 (19.4%) with stage pT1 and 11 (16.4%) with muscle invasive stage pT2a or greater. In these 66 patients BC was G1 to G3 in 25 (37.9%), 24 (36.4%) and 17
70
INFLUENCE OF RENAL EXCRETORY FUNCTION ON MARKERS TO DETECT BLADDER CANCER
Table 1. Patient renal function by test results and applied algorithm Group 1 Renal function Serum creatinine: No. pts (%) Mg/dl GFR: No. pts Ml/min/1.73 m2 Proteinuria: No. pts Mg/dl Algorithm: No. pts Values
Group 2
Normal
Impaired 124 (28.6)
284 (65.4) 1.2–3.5
372 (85.7) Greater than 90
53 (12.2) 40–90
245 (55.3) Less than 12.5
167 (37.7) 12.5–149
Less than 1.2
229 (66.3) GFR greater than 90 ml/min/1.73 m2 or protein less than 12.5 mg/dl
109 (24.8) GFR 40 or greater-90 ml/min/1.73 m2 or less ⫹ protein 12.5 or greater-149 mg/dl or less
(26.0%), respectively. One patient was diagnosed with carcinoma in situ. Before retrieving urine samples mechanical manipulation of the lower urinary tract, including catheterization, digital rectal examination or cystoscopy, was done in 353 patients (78.6%). UTI was present in 84 patients (18.7%). Test Results Overall CYT, FISH, immunocytology and NMP22 sensitivity was 74.6%, 72.7%, 88.1% and 89.4%, respectively. Excluding those with prior mechanical manipulation or UTI NMP22 had 75.0% sensitivity. Overall CYT, FISH, immunocytology and NMP22 specificity was 90.2%, 86.3%, 80.1% and 34.1%, respectively. NMP22 specificity was improved to 67.2% by excluding patients with mechanical manipulation or UTI (table 2). Impact Serum creatinine. Serum creatinine, and corresponding CYT, FISH, immunocytology and NMP22 results were available for 434 patients. Table 1 shows patient renal function. Table 3 shows falsepositive rates by serum creatinine. CYT, FISH, immunocytology and NMP22 false-positive rates did not significantly differ among groups 1 to 3 (serum creatinine less than 1.2, 1.3 to 3.5 and greater than 3.5 mg/dl, respectively, table 3). Table 2. Overall test sensitivity, specificity, and negative and positive predictive values % Pos % Neg Predictive Predictive % Sensitivity % Specificity Value Value CYT FISH Immunocytology NMP: Overall No UTI or manipulation
74.6 72.7 88.1
90.2 86.3 80.1
58.1 49.0 44.4
95.1 94.6 97.4
89.4 75.0
34.1 67.2
19.3 12.0
94.8 97.8
Group 3 Severely Impaired 26 (6.0) Greater than 3.5 9 (2.1) Less than 40 31 (7.0) Greater than 149 39 (8.9) GFR less than 40 ml/min/1.73 m2 or protein greater than 149 mg/dl
Glomerular filtration rate. GFR was calculated in 434 cases. Table 3 shows GFR related false-positive rates. CYT, FISH and immunocytology did not significantly correlate with GFR in groups 1 to 3 (GFR 90 or greater, 40 to 89 and less than 40 ml/min/1.73 m2, respectively). The NMP22 false-positive rate in patients without manipulation or UTI was 20.0% vs 45.7% in those with GFR 90 or greater vs less than 90 ml/minute/1.73 m2 (p ⫽ 0.05). Proteinuria. Urine protein, and corresponding CYT, FISH, immunocytology and NMP22 results were available in 434 cases. Table 4 shows false-positive rates by proteinuria. The false-positive CYT rate was significantly increased in group 3 patients with proteinuria greater than 149 mg/dl and more than 4 times higher than in group 1 patients with protein less than 12.5 mg/dl. In group 2 proteinuria was 12.5 to 149 mg/dl. According to FISH or immunocytology false-positive rates results in patients with impaired and severely impaired renal function according to proteinuria grade were higher than in those with normal protein but no statistical significance was noted. In contrast, NMP22 false-positive findings depended significantly on urine protein (p ⬍0.05). After excluding patients with UTI or mechanical manipulation those with proteinuria greater than 149 mg/dl still had a 66.7% false-positive rate vs 30.7% in those with protein less than 149 mg/dl (p ⫽ 0.2). False-Positive Rates CYT in patients without UTI. Subgroup analysis was done in 35 patients without UTI but with significant proteinuria. Table 4 shows false-positive CYT rates in patients without UTI. According to the algorithm. Table 5 shows falsepositive results in groups with different renal function according to the algorithm, including group 1—GFR 90 ml/minute/1.73 m2 or greater, or protein less than 12.5 mg/dl, group 2—GFR 40 or greater to less than 90 ml/minute/1.73 m2 and protein 12.5 or
INFLUENCE OF RENAL EXCRETORY FUNCTION ON MARKERS TO DETECT BLADDER CANCER
71
Table 3. False-positive rate by serum creatinine and calculated GFR % NMP22 % CYT Overall (No. pts) Group: 1 vs 2 vs 3 p Value 1 ⫹ 2 vs 3 p Value 1 vs 2 ⫹ 3 p Value Overall (No. pts) Group: 1 vs 2 vs 3 p Value 1 ⫹ 2 vs 3 p Value 1 vs 2 ⫹ 3 p Value
% FISH
% Immunocytology Creatinine 20.5 (73)
10.0 (53)
13.7 (48)
10.8 vs 2.2 vs 22.2 0.1 9.7 vs 22.2 0.2 10.8 vs 5.6 0.2
14.6 vs 4.7 vs 28.6 0.1 13.4 vs 28.6 0.2 14.6 vs 8.0 0.2
10.0 (35)
13.7 (48)
11.4 vs 9.4 vs 8.7 0.8 10.1 vs 8.7 0.8 11.4 vs 9.4 0.6
14.0 vs 13.9 vs 9.5 0.8 13.9 vs 9.5 0.6 14.0 vs 13.5 0.9
Overall
No UTI or Manipulation
67.6 (242)
36.4 (20)
21.5 vs 11.1 vs 15.6 0.5 20.7 vs 11.1 0.5 21.4 vs 14.8 0.3 GFR 20.5 (73)
67.5 vs 67.4 vs 71.4 0.1 67.5 vs 71.4 0.8 67.5 vs 67.9 1.0
35.6 vs 37.5 vs 50.0 0.9 35.9 vs 50.0 0.7 35.6 vs 40 0.8
67.6 (242)
36.4 (20)
21.5 vs 21.2 vs 8.7 0.4 21.3 vs 8.7 0.1 21.5 vs 20.0 0.7
62.4 vs 69.4 vs 75.0 0.3 67.2 vs 75 0.5 62.4 vs 69.9 0.2
20.0 vs 45.5 vs 50.0 0.2 35.9 vs 50 0.7 20 vs 45.7 0.05
greater to less than 150 mg/dl, and group 3—GFR less than 40 ml/minute/1.73 m2, or protein 150 mg/dl or greater. False-positive CYT was significantly more common in group 3 patients with severely impaired renal function than in groups 1 and 2. Falsepositive NMP22 results significantly correlated with renal function, as determined by the algorithm. However, this correlation was not observed in patients without UTI or prior mechanical manipulation.
DISCUSSION Cystoscopy is a mainstay of BC diagnosis but CYT and urine markers are gaining importance for BC primary diagnosis and followup. CYT and urine markers such as FISH, immunocytology and NMP22 have inherent limitations that must be considered.17 We investigated the impact of renal function on CYT, FISH, immunocytology and NMP22 false-positive rates. False-positive NMP22 results were more common in patients with decreased GFR while NMP22 and CYT were less specific in patients with
proteinuria. FISH and immunocytology findings were widely independent of renal function. Overall test sensitivity and specificity were similar to those in previous studies.18,19 CYT sensitivity in our population must be considered superior to that in others but specificity was average. CYT shows interobserver variability and, thus, there is a broad spectrum of test accuracy.20 In our study FISH sensitivity and specificity were 72.7% and 86.3% but in a recent meta-analysis pooled sensitivity and specificity were 72.0% and 82.0%, respectively.21 In contrast to other studies,22 FISH sensitivity was not superior to that of CYT. This was a result of the relatively high sensitivity of CYT compared to that in other studies that investigated each test.22 Consistent with our and other studies immunocytology sensitivity outperformed CYT for sensitivity (88.0% vs 75.7%) but specificity was lower (80% vs 90%). The NMP22 sensitivity of 89.4% is in accordance with that in previous studies but our 34.1% specific-
Table 4. False-positive rates by proteinuria % CYT Overall Overall (No. pts) Group: 1 vs 2 vs 3 p Value 1 ⫹ 2 vs 3 p Value 1 vs 2 ⫹ 3 p Value
% NMP22 No UTI or Manipulation
% FISH
% Immunocytology
Overall
No UTI or Manipulation
9.7 (35)
9.3 (28)
13.6 (49)
19.6 (72)
65.8 (242)
32.3 (21)
7.3 vs 9.9 vs 30.4 0.0017 8.3 vs 30.4 0.0005 7.3 vs 13 0.06
7.9 vs 10.1 vs 25.0 0.13 8.7 vs 25 0.05 7.9 vs 11.7 0.27
10.7 vs 16.7 vs 22.7 0.1 13.0 vs 22.7 0.2 10.7 vs 17.5 0.06
16.8 vs 21.3 vs 33.3 0.1 18.6 vs 33.3 0.08 16.8 vs 23.1 0.1
59.5 vs 70.8 vs 95.2 0.0013 64.0 vs 95.2 0.0034 59.5 vs 74.1 0.0036
26.3 vs 37.5 vs 66.7 0.3 30.7 vs 66.7 0.2 26.3 vs 40.7 0.2
72
INFLUENCE OF RENAL EXCRETORY FUNCTION ON MARKERS TO DETECT BLADDER CANCER
Table 5. False-positive rates by algorithm % NMP22 % CYT Overall (No. pts) Group: 1 vs 2 vs 3 p Value 1 ⫹ 2 vs 3 p Value 1 vs 2 ⫹ 3 p Value
% FISH
% Immunocytology
Overall
No UTI or Manipulation
9.8 (35)
13.8 (49)
19.8 (72)
66.8 (243)
34.4 (21)
8.2 vs 9.8 vs 22.6 0.04 8.6 vs 22.6 0.0123 8.2 vs 13.3 0.1
12.4 vs 16.9 vs 16.7 0.5 13.5 vs 16.7 0.6 12.4 vs 16.8 0.3
17.8 vs 23.3 vs 25.8 0.4 19.2 vs 25.8 0.4 17.8 vs 23.9 0.2
60.6 vs 75.9 vs 92.9 0.0003 64.6 vs 92.9 0.0023 60.6 vs 80.0 0.0003
28.3 vs. 54.06 vs. 50.0 0.2 33.3 vs. 50.0 0.5 28.3 vs. 53.3 0.08
ity was relatively low. In patients without UTI or mechanical manipulations sensitivity was 69%. Results emphasize the need to strictly obey exclusion criteria for molecular urine tests.17 However, a specificity rate of greater than 90% after applying these exclusion criteria, which was previously reported, cannot be confirmed.10 The source of increased NMP22 in patients with impaired GFR remains unclear. Causes of increased NMP22 include any process that creates rapid turnover of epithelial cells. Increased leukocytes and red cells in urine cause false-positive NMP22 results.12 CKD is associated with HU and leukocyturia, which could serve as a source of NMP22 release. NMP22 is a 50 kDa protein and its filtration depends on glomerular filtration barrier function, which is often disturbed in CKD cases. Serum NMP22 has a noted cutoff of 110 U/ml in healthy controls.23 Increased NMP22 filtration due to glomerular defects may cause higher NMP22 without cancer. Increased epithelial markers are found in the urine of patients with CKD, indicating increased epithelial cell apoptosis.24 Prior studies showed increased apoptosis and necrosis in patients with kidney disease.25 These processes might also account for increased NMP22 in urine. To confirm that GFR has a significant impact on NMP22 performance a larger population of patients without prior manipulation or UTI is required since we observed borderline significance in our study. The increased NMP22 rate in patients with proteinuria is mainly caused by UTI and mechanical manipulation since excluding these patients eliminated proteinuria as a significant confounder. Each is a contraindication to NMP22 testing due to the high false-positive rate. To our knowledge the reason for decreased CYT specificity in patients with increased protein concentrations in urine remains unknown. There are possible explanations for changed CYT results due to proteinuria. Cell atypia could be caused by UTI. Proteinuria can be associated with UTI, although a systematic review did not confirm a causal role of asymptomatic UTI for proteinuria.26 Even when pa-
tients with UTI were excluded from analysis, the false-positive rate was higher and borderline significant in those with severe proteinuria. Thus, the association of proteinuria with false-positive CYT findings cannot be explained only by UTI. There is no clear-cut evidence that proteins influence urothelial cell morphology. However, microalbuminuria in patients without diabetes is associated with an increased risk of BC.27 To our knowledge the reason for this association has not yet been identified. Possible explanations are increased protein due to inflammation, which is a risk factor for cancer,28 and paraneoplastic protein excretion. In patients with CKD systemic inflammation is common due to uremia and increased cytokines, such as tumor necrosis factor-␣ or interleukin-6.29 Inflammation could also account for morphological changes in urothelial cells in patients with severe proteinuria due to CKD. However, GFR as a renal function parameter did not significantly affect CYT specificity. Regardless of mechanism our results clearly show that positive CYT findings in patients with a high protein concentration in urine must be interpreted differently than in patients with a normal urine composition. Since the immunocytology false-positive rate was not affected by urine protein, immunocytochemical detection of specific antigens may improve the accuracy of noninvasive diagnostic measurements in patients with high grade proteinuria. Also, since FISH is based on detecting chromosome alterations of cells in urine, the complete independence of this test from urine protein and GFR is not surprising. Some functional parameters had an inverse effect on test results. False-positive immunocytology findings were decreased in patients with high serum creatinine but high grade proteinuria led to more false-positive findings. This indicates that renal function parameters should be considered separately. A limitation of our study is that no differentiation was made between microscopic and macroscopic hematuria, which is known to influence BC test results.30 Also, 79% of patients underwent mechanical
INFLUENCE OF RENAL EXCRETORY FUNCTION ON MARKERS TO DETECT BLADDER CANCER
manipulation, which may have also influenced our results. External validation is required to further determine the influencing role of renal function on CYT and NMP22 performance.
73
The false-positive CYT rate was increased in patients with high urine protein concentrations, which led to decreased specificity regardless of UTI. Renal function should be considered when interpreting the results of noninvasive urine tests for detecting or monitoring BC.
CONCLUSIONS To our knowledge this is the first study of the effect of renal function on urine based tests for diagnosing BC. Our study revealed an increased NMP22 test false-positive rate in patients with decreased GFR.
ACKNOWLEDGMENTS Prof. Helmut Heinle, Institute of Physiology, University Tübingen, provided advice on renal function assessment.
REFERENCES 1. Kohler BA, Ward E, McCarthy BJ et al: Annual report to the nation on the status of cancer, 1975–2007, featuring tumors of the brain and other nervous system. J Natl Cancer Inst 2011; 103: 714. 2. Millan-Rodriguez F, Chechile-Toniolo G, SalvadorBayarri J et al: Primary superficial bladder cancer risk groups according to progression, mortality and recurrence. J Urol 2000; 164: 680. 3. Babjuk M, Oosterlinck W, Sylvester R et al: EAU guidelines on non-muscle-invasive urothelial carcinoma of the bladder, the 2011 update. Eur Urol 2011; 59: 997. 4. Budman LI, Kassouf W and Steinberg JR: Biomarkers for detection and surveillance of bladder cancer. Can Urol Assoc J 2008; 2: 212. 5. Raitanen MP, Aine R, Rintala E et al: Differences between local and review urinary cytology in diagnosis of bladder cancer. An interobserver multicenter analysis. Eur Urol 2002; 41: 284. 6. Daniely M, Rona R, Kaplan T et al: Combined analysis of morphology and fluorescence in situ hybridization significantly increases accuracy of bladder cancer detection in voided urine samples. Urology 2005; 66: 1354. 7. Bubendorf L: Multiprobe fluorescence in situ hybridization (UroVysion) for the detection of urothelial carcinoma—FISHing for the right catch. Acta Cytol 2011; 55: 113. 8. Shariat SF, Marberger MJ, Lotan Y et al: Variability in the performance of nuclear matrix protein 22 for the detection of bladder cancer. J Urol 2006; 176: 919.
11. Raitanen MP, Kaasinen E, Lukkarinen O et al: Analysis of false-positive BTA STAT test results in patients followed up for bladder cancer. Urology 2001; 57: 680. 12. Atsu N, Ekici S, Oge OO et al: False-positive results of the NMP22 test due to hematuria. J Urol 2002; 167: 555. 13. Coresh J, Selvin E, Stevens LA et al: Prevalence of chronic kidney disease in the United States. JAMA 2007; 298: 2038.
cell carcinoma of the bladder. BJU Int 2006; 97: 997. 21. Shariat SF, Karam JA, Lotan Y et al: Critical evaluation of urinary markers for bladder cancer detection and monitoring. Rev Urol 2008; 10: 120. 22. Kehinde EO, Al-Mulla F, Kapila K et al: Comparison of the sensitivity and specificity of urine cytology, urinary nuclear matrix protein-22 and multitarget fluorescence in situ hybridization assay in the detection of bladder cancer. Scand J Urol Nephrol 2011; 45: 113.
14. Kipp BR, Tyner HL, Campion MB et al: Chromosomal alterations detected by fluorescence in situ hybridization in urothelial carcinoma and rarer histologic variants of bladder cancer. Am J Clin Pathol 2008; 130: 552.
23. Miyanaga N, Akaza H, Ishikawa S et al: Clinical evaluation of nuclear matrix protein 22 (NMP22) in urine as a novel marker for urothelial cancer. Eur Urol 1997; 31: 163.
15. Chiarelli F, Verrotti A, Mohn A et al: The importance of microalbuminuria as an indicator of incipient diabetic nephropathy: therapeutic implications. Ann Med 1997; 29: 439.
24. Roth GA, Lebherz-Eichinger D, Ankersmit HJ et al: Increased total cytokeratin-18 serum and urine levels in chronic kidney disease. Clin Chim Acta 2011; 412: 713.
16. Levey AS, Bosch JP, Lewis JB et al: A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999; 130: 461.
25. Kumar D, Robertson S and Burns KD: Evidence of apoptosis in human diabetic kidney. Mol Cell Biochem 2004; 259: 67.
17. Sharma S, Zippe CD, Pandrangi L et al: Exclusion criteria enhance the specificity and positive predictive value of NMP22 and BTA stat. J Urol 1999; 162: 53. 18. Grossman HB, Soloway M, Messing E et al: Surveillance for recurrent bladder cancer using a point-of-care proteomic assay. JAMA 2006; 295: 299.
9. Bastacky S, Ibrahim S, Wilczynski SP et al: The accuracy of urinary cytology in daily practice. Cancer 1999; 87: 118.
19. Tritschler S, Scharf S, Karl A et al: Validation of the diagnostic value of NMP22 BladderChek test as a marker for bladder cancer by photodynamic diagnosis. Eur Urol 2007; 51: 403.
10. Ponsky LE, Sharma S, Pandrangi L et al: Screening and monitoring for bladder cancer: refining the use of NMP22. J Urol 2001; 166: 75.
20. Karakiewicz PI, Benayoun S, Zippe C et al: Institutional variability in the accuracy of urinary cytology for predicting recurrence of transitional
26. Carter JL, Tomson CR, Stevens PE et al: Does urinary tract infection cause proteinuria or microalbuminuria? A systematic review. Nephrol Dial Transplant 2006; 21: 3031. 27. Jorgensen L, Heuch I, Jenssen T et al: Association of albuminuria and cancer incidence. J Am Soc Nephrol 2008; 19: 992. 28. Coussens LM and Werb Z: Inflammation and cancer. Nature 2002; 420: 860. 29. Raj DS, Shah H, Shah VO et al: Markers of inflammation, proteolysis, and apoptosis in ESRD. Am J Kidney Dis 2003; 42: 1212. 30. Pesch B, Nasterlack M, Eberle F et al: The role of haematuria in bladder cancer screening among men with former occupational exposure to aromatic amines. BJU Int 2011; 108: 546.
Abbreviations and Acronyms BC ⫽ bladder cancer CKD ⫽ chronic kidney disease CYT ⫽ voided urine cytology FISH ⫽ fluorescence in situ hybridization GFR ⫽ glomerular filtration rate HU ⫽ hematuria NMP ⫽ nuclear matrix protein UTI ⫽ urinary tract infection Submitted for publication May 27, 2011. Study received institutional review board approval. * Correspondence: Department of Urology, Eberhard-Karls University, Hoppe-Seyler Str. 3, D-72076 Tübingen, Germany (telephone: ⫹497071-29-85092; FAX: ⫹49-7071-29-5092; e-mail: [email protected]).
Purpose: In hematuria cases urine based tests are used to detect bladder cancer, although the diagnostic yield remains insufficient due to influencing variables, including urinary tract infection. Many patients are elderly with renal insufficiency and have proteinuria as an additional influencing factor. To our knowledge no data are available on the accuracy of urine based bladder cancer tests in conjunction with renal function. Materials and Methods: Urine samples of 449 patients with hematuria and histology were included in analysis. Cytology, fluorescence in situ hybridization, immunocytology and nuclear matrix protein 22 assay were done. Renal function was classified as normal, impaired or severely impaired based on serum creatinine, the glomerular filtration rate and proteinuria. False-positive rates were statistically compared in regard to renal function. Results: A total of 382 patients did not have bladder cancer. There was an increased false-positive rate for creatinine and the glomerular filtration rate. The nuclear matrix protein 22 test showed a 22.0% and 46.7% false-positive rate in the normal and limited function cohorts, respectively (p ⫽ 0.05). Similar trends were noted for proteinuria. Indeterminate significance was detected, separating those with severely impaired function for immunocytology and those in the normal group for fluorescence in situ hybridization (p ⫽ 0.08 and 0.06, respectively). Proteinuria was a significant factor for urine cytology with increased false-positive results in the absence of urinary tract infection (p ⫽ 0.0017 and 0.05, respectively). Conclusions: To our knowledge this is the first study of renal function and the accuracy of urine based bladder cancer markers. Renal function influences the diagnostic yield. A decreased glomerular filtration rate was associated with increased false-positive nuclear matrix protein 22 results while proteinuria decreased urine cytology specificity. Renal function should be considered when urine based bladder cancer tests are interpreted. Key Words: urinary bladder; urinary bladder neoplasms; kidney function tests; tumor markers, biological; false positive reactions
68
www.jurology.com
UROTHELIAL carcinoma of the bladder is the fourth most common malignancy in men in the United States, accounting for more than 70,000 new cases and 14,000 deaths annually.1 Of the patients 70% initially present
with Ta/T1 tumors. Depending on risk stratification up to 15% of these tumors progress to a muscle invasive stage, which is associated with a limited 5-year survival rate.2 Early diagnosis is essential for timely treatment and
0022-5347/12/1871-0068/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
Vol. 187, 68-73, January 2012 Printed in U.S.A. DOI:10.1016/j.juro.2011.09.023
AND
RESEARCH, INC.
INFLUENCE OF RENAL EXCRETORY FUNCTION ON MARKERS TO DETECT BLADDER CANCER
subsequent decreased mortality. Cystoscopy with transurethral resection remains the gold standard for BC diagnosis.3 Several urine based tests are available for BC diagnosis and followup. Urine cytology is the most widely accepted, noninvasive marker for BC primary diagnosis and followup.3 Urine cytology has relatively low sensitivity for low grade tumors and, thus, it cannot replace cystoscopy.4 There is considerable interobserver variability and urine cytology depends highly on cytopathologist experience.5 Other noninvasive urine markers are gaining importance for diagnosis and followup. FISH uses a multitarget set of probes that specifically bind to centromeres of chromosomes 3, 7 and 17, and to the 9p21 locus of chromosome 9. Loss of the 9p21 locus, which is the site of the p16 tumor suppressor gene, was identified as the earliest, most frequent genetic aberration in patients with BC.6,7 The uCyt⫹™ immunocytological fluorescence assay uses 3 antibodies to detect carcinoembryonic antigen and mucinlike glycoprotein on the cell surface. NMP22, which regulates mitosis, is over expressed in malignant urothelial cells. NMP22 is released into urine by apoptotic cells and it can be detected and quantified.8 Generally these molecular tests show increased sensitivity but decreased specificity compared to CYT.4 However, several modifiable factors influence test performance. UTI can result in findings of falsepositive CYT9 and increased NMP22.8 Tests targeting NMP22 or other molecular markers yield increased false-positive findings in patients with previous instillation therapy, foreign bodies, benign prostate hyperplasia or stone disease.10,11 Changes in urine composition, such as HU and pyuria, can yield falsepositive results.12 Interference of urine constituents with urine based tests is widely acknowledged but to our knowledge there are no data on the influence of renal excretory function on test performance. This issue is of tremendous interest since patients with a primary diagnosis of BC are often elderly and increased age is associated with a higher CKD prevalence.13 Thus, we evaluated the impact of renal function on the performance of the 4 most widely used noninvasive urine tests.
PATIENTS AND METHODS Patients and Samples Enrolled in our study were 380 men and 69 women with a mean age of 65.5 years (range 18 to 93) with microscopic or macroscopic HU and without a BC history. The study received institutional review board approval. Voided urine was used for analysis and investigated by dipstick test and microscopy. Protein concentration was analyzed
69
quantitatively by a Miditron® M. CYT, UroVysion® for FISH, uCyt⫹ and the NMP22® Bladder Cancer Test were done. All patients underwent cystoscopy and histological assessment.
Urine Processing For CYT slides were cytospinned and stained according to Papanicolaou. Specimens were assessed by light microscopy according to defined criteria. FISH was done according to manufacturer instructions. Positive result criteria were 4 or greater nuclei of a total of 25 morphological suspicious cells showing 3 or greater signals of at least 2 of chromosomes 3, 7 or 17, or at least 12 nuclei showing no 9p21 signal.14 The uCyt⫹ test was done according to manufacturer instructions. The criterion for a positive result was at least 1 positive cell. NMP22 enzyme-linked immunosorbent assay was done according to the manufacturer protocol with concentrations greater than 10 U/ml classified as positive.8 Since UTI and previous mechanical manipulation such as catheterization are considered contraindications to the test, subgroup analysis was performed for patients without UTI or previous mechanical manipulation.
Urinary Tract Infection UTI was determined by dipstick test and urine microscopy. It was defined as at least 100 leukocytes per l and simultaneously as more than 1 erythrocyte or at least 100 leukocytes per l and the simultaneous presence of urine nitrite.
Renal Function Parameters Serum creatinine, GFR and urine protein15 served as renal function parameters. Creatinine was measured using standard laboratory equipment. GFR was calculated using the modification of diet in renal disease formula.16 Based on these parameters patients were divided into renal function groups, including group 1—normal, group 2—impaired and group 3—severely impaired. An algorithm respecting GFR and the urine protein concentration was used to subdivide patients into 3 renal function groups. Table 1 shows patient subclassification by these parameters.
Statistical Analysis Sensitivity and specificity were calculated after contingency analysis of the whole cohort and defined subgroups. False-positive rates were compared by contingency analysis depending on renal function according to the defined criteria. JMP® 7.2 and Pearson’s chi-square test were used with p ⬍0.05 considered significant. Values of p ⫽ 0.05 were considered borderline significant and are reported descriptively.
RESULTS Of the patients 382 patients (85.1%) did not have BC and 67 (14.9%) were diagnosed with BC by histopathology, including 42 (62.7%) with stage pTa, 13 (19.4%) with stage pT1 and 11 (16.4%) with muscle invasive stage pT2a or greater. In these 66 patients BC was G1 to G3 in 25 (37.9%), 24 (36.4%) and 17
70
INFLUENCE OF RENAL EXCRETORY FUNCTION ON MARKERS TO DETECT BLADDER CANCER
Table 1. Patient renal function by test results and applied algorithm Group 1 Renal function Serum creatinine: No. pts (%) Mg/dl GFR: No. pts Ml/min/1.73 m2 Proteinuria: No. pts Mg/dl Algorithm: No. pts Values
Group 2
Normal
Impaired 124 (28.6)
284 (65.4) 1.2–3.5
372 (85.7) Greater than 90
53 (12.2) 40–90
245 (55.3) Less than 12.5
167 (37.7) 12.5–149
Less than 1.2
229 (66.3) GFR greater than 90 ml/min/1.73 m2 or protein less than 12.5 mg/dl
109 (24.8) GFR 40 or greater-90 ml/min/1.73 m2 or less ⫹ protein 12.5 or greater-149 mg/dl or less
(26.0%), respectively. One patient was diagnosed with carcinoma in situ. Before retrieving urine samples mechanical manipulation of the lower urinary tract, including catheterization, digital rectal examination or cystoscopy, was done in 353 patients (78.6%). UTI was present in 84 patients (18.7%). Test Results Overall CYT, FISH, immunocytology and NMP22 sensitivity was 74.6%, 72.7%, 88.1% and 89.4%, respectively. Excluding those with prior mechanical manipulation or UTI NMP22 had 75.0% sensitivity. Overall CYT, FISH, immunocytology and NMP22 specificity was 90.2%, 86.3%, 80.1% and 34.1%, respectively. NMP22 specificity was improved to 67.2% by excluding patients with mechanical manipulation or UTI (table 2). Impact Serum creatinine. Serum creatinine, and corresponding CYT, FISH, immunocytology and NMP22 results were available for 434 patients. Table 1 shows patient renal function. Table 3 shows falsepositive rates by serum creatinine. CYT, FISH, immunocytology and NMP22 false-positive rates did not significantly differ among groups 1 to 3 (serum creatinine less than 1.2, 1.3 to 3.5 and greater than 3.5 mg/dl, respectively, table 3). Table 2. Overall test sensitivity, specificity, and negative and positive predictive values % Pos % Neg Predictive Predictive % Sensitivity % Specificity Value Value CYT FISH Immunocytology NMP: Overall No UTI or manipulation
74.6 72.7 88.1
90.2 86.3 80.1
58.1 49.0 44.4
95.1 94.6 97.4
89.4 75.0
34.1 67.2
19.3 12.0
94.8 97.8
Group 3 Severely Impaired 26 (6.0) Greater than 3.5 9 (2.1) Less than 40 31 (7.0) Greater than 149 39 (8.9) GFR less than 40 ml/min/1.73 m2 or protein greater than 149 mg/dl
Glomerular filtration rate. GFR was calculated in 434 cases. Table 3 shows GFR related false-positive rates. CYT, FISH and immunocytology did not significantly correlate with GFR in groups 1 to 3 (GFR 90 or greater, 40 to 89 and less than 40 ml/min/1.73 m2, respectively). The NMP22 false-positive rate in patients without manipulation or UTI was 20.0% vs 45.7% in those with GFR 90 or greater vs less than 90 ml/minute/1.73 m2 (p ⫽ 0.05). Proteinuria. Urine protein, and corresponding CYT, FISH, immunocytology and NMP22 results were available in 434 cases. Table 4 shows false-positive rates by proteinuria. The false-positive CYT rate was significantly increased in group 3 patients with proteinuria greater than 149 mg/dl and more than 4 times higher than in group 1 patients with protein less than 12.5 mg/dl. In group 2 proteinuria was 12.5 to 149 mg/dl. According to FISH or immunocytology false-positive rates results in patients with impaired and severely impaired renal function according to proteinuria grade were higher than in those with normal protein but no statistical significance was noted. In contrast, NMP22 false-positive findings depended significantly on urine protein (p ⬍0.05). After excluding patients with UTI or mechanical manipulation those with proteinuria greater than 149 mg/dl still had a 66.7% false-positive rate vs 30.7% in those with protein less than 149 mg/dl (p ⫽ 0.2). False-Positive Rates CYT in patients without UTI. Subgroup analysis was done in 35 patients without UTI but with significant proteinuria. Table 4 shows false-positive CYT rates in patients without UTI. According to the algorithm. Table 5 shows falsepositive results in groups with different renal function according to the algorithm, including group 1—GFR 90 ml/minute/1.73 m2 or greater, or protein less than 12.5 mg/dl, group 2—GFR 40 or greater to less than 90 ml/minute/1.73 m2 and protein 12.5 or
INFLUENCE OF RENAL EXCRETORY FUNCTION ON MARKERS TO DETECT BLADDER CANCER
71
Table 3. False-positive rate by serum creatinine and calculated GFR % NMP22 % CYT Overall (No. pts) Group: 1 vs 2 vs 3 p Value 1 ⫹ 2 vs 3 p Value 1 vs 2 ⫹ 3 p Value Overall (No. pts) Group: 1 vs 2 vs 3 p Value 1 ⫹ 2 vs 3 p Value 1 vs 2 ⫹ 3 p Value
% FISH
% Immunocytology Creatinine 20.5 (73)
10.0 (53)
13.7 (48)
10.8 vs 2.2 vs 22.2 0.1 9.7 vs 22.2 0.2 10.8 vs 5.6 0.2
14.6 vs 4.7 vs 28.6 0.1 13.4 vs 28.6 0.2 14.6 vs 8.0 0.2
10.0 (35)
13.7 (48)
11.4 vs 9.4 vs 8.7 0.8 10.1 vs 8.7 0.8 11.4 vs 9.4 0.6
14.0 vs 13.9 vs 9.5 0.8 13.9 vs 9.5 0.6 14.0 vs 13.5 0.9
Overall
No UTI or Manipulation
67.6 (242)
36.4 (20)
21.5 vs 11.1 vs 15.6 0.5 20.7 vs 11.1 0.5 21.4 vs 14.8 0.3 GFR 20.5 (73)
67.5 vs 67.4 vs 71.4 0.1 67.5 vs 71.4 0.8 67.5 vs 67.9 1.0
35.6 vs 37.5 vs 50.0 0.9 35.9 vs 50.0 0.7 35.6 vs 40 0.8
67.6 (242)
36.4 (20)
21.5 vs 21.2 vs 8.7 0.4 21.3 vs 8.7 0.1 21.5 vs 20.0 0.7
62.4 vs 69.4 vs 75.0 0.3 67.2 vs 75 0.5 62.4 vs 69.9 0.2
20.0 vs 45.5 vs 50.0 0.2 35.9 vs 50 0.7 20 vs 45.7 0.05
greater to less than 150 mg/dl, and group 3—GFR less than 40 ml/minute/1.73 m2, or protein 150 mg/dl or greater. False-positive CYT was significantly more common in group 3 patients with severely impaired renal function than in groups 1 and 2. Falsepositive NMP22 results significantly correlated with renal function, as determined by the algorithm. However, this correlation was not observed in patients without UTI or prior mechanical manipulation.
DISCUSSION Cystoscopy is a mainstay of BC diagnosis but CYT and urine markers are gaining importance for BC primary diagnosis and followup. CYT and urine markers such as FISH, immunocytology and NMP22 have inherent limitations that must be considered.17 We investigated the impact of renal function on CYT, FISH, immunocytology and NMP22 false-positive rates. False-positive NMP22 results were more common in patients with decreased GFR while NMP22 and CYT were less specific in patients with
proteinuria. FISH and immunocytology findings were widely independent of renal function. Overall test sensitivity and specificity were similar to those in previous studies.18,19 CYT sensitivity in our population must be considered superior to that in others but specificity was average. CYT shows interobserver variability and, thus, there is a broad spectrum of test accuracy.20 In our study FISH sensitivity and specificity were 72.7% and 86.3% but in a recent meta-analysis pooled sensitivity and specificity were 72.0% and 82.0%, respectively.21 In contrast to other studies,22 FISH sensitivity was not superior to that of CYT. This was a result of the relatively high sensitivity of CYT compared to that in other studies that investigated each test.22 Consistent with our and other studies immunocytology sensitivity outperformed CYT for sensitivity (88.0% vs 75.7%) but specificity was lower (80% vs 90%). The NMP22 sensitivity of 89.4% is in accordance with that in previous studies but our 34.1% specific-
Table 4. False-positive rates by proteinuria % CYT Overall Overall (No. pts) Group: 1 vs 2 vs 3 p Value 1 ⫹ 2 vs 3 p Value 1 vs 2 ⫹ 3 p Value
% NMP22 No UTI or Manipulation
% FISH
% Immunocytology
Overall
No UTI or Manipulation
9.7 (35)
9.3 (28)
13.6 (49)
19.6 (72)
65.8 (242)
32.3 (21)
7.3 vs 9.9 vs 30.4 0.0017 8.3 vs 30.4 0.0005 7.3 vs 13 0.06
7.9 vs 10.1 vs 25.0 0.13 8.7 vs 25 0.05 7.9 vs 11.7 0.27
10.7 vs 16.7 vs 22.7 0.1 13.0 vs 22.7 0.2 10.7 vs 17.5 0.06
16.8 vs 21.3 vs 33.3 0.1 18.6 vs 33.3 0.08 16.8 vs 23.1 0.1
59.5 vs 70.8 vs 95.2 0.0013 64.0 vs 95.2 0.0034 59.5 vs 74.1 0.0036
26.3 vs 37.5 vs 66.7 0.3 30.7 vs 66.7 0.2 26.3 vs 40.7 0.2
72
INFLUENCE OF RENAL EXCRETORY FUNCTION ON MARKERS TO DETECT BLADDER CANCER
Table 5. False-positive rates by algorithm % NMP22 % CYT Overall (No. pts) Group: 1 vs 2 vs 3 p Value 1 ⫹ 2 vs 3 p Value 1 vs 2 ⫹ 3 p Value
% FISH
% Immunocytology
Overall
No UTI or Manipulation
9.8 (35)
13.8 (49)
19.8 (72)
66.8 (243)
34.4 (21)
8.2 vs 9.8 vs 22.6 0.04 8.6 vs 22.6 0.0123 8.2 vs 13.3 0.1
12.4 vs 16.9 vs 16.7 0.5 13.5 vs 16.7 0.6 12.4 vs 16.8 0.3
17.8 vs 23.3 vs 25.8 0.4 19.2 vs 25.8 0.4 17.8 vs 23.9 0.2
60.6 vs 75.9 vs 92.9 0.0003 64.6 vs 92.9 0.0023 60.6 vs 80.0 0.0003
28.3 vs. 54.06 vs. 50.0 0.2 33.3 vs. 50.0 0.5 28.3 vs. 53.3 0.08
ity was relatively low. In patients without UTI or mechanical manipulations sensitivity was 69%. Results emphasize the need to strictly obey exclusion criteria for molecular urine tests.17 However, a specificity rate of greater than 90% after applying these exclusion criteria, which was previously reported, cannot be confirmed.10 The source of increased NMP22 in patients with impaired GFR remains unclear. Causes of increased NMP22 include any process that creates rapid turnover of epithelial cells. Increased leukocytes and red cells in urine cause false-positive NMP22 results.12 CKD is associated with HU and leukocyturia, which could serve as a source of NMP22 release. NMP22 is a 50 kDa protein and its filtration depends on glomerular filtration barrier function, which is often disturbed in CKD cases. Serum NMP22 has a noted cutoff of 110 U/ml in healthy controls.23 Increased NMP22 filtration due to glomerular defects may cause higher NMP22 without cancer. Increased epithelial markers are found in the urine of patients with CKD, indicating increased epithelial cell apoptosis.24 Prior studies showed increased apoptosis and necrosis in patients with kidney disease.25 These processes might also account for increased NMP22 in urine. To confirm that GFR has a significant impact on NMP22 performance a larger population of patients without prior manipulation or UTI is required since we observed borderline significance in our study. The increased NMP22 rate in patients with proteinuria is mainly caused by UTI and mechanical manipulation since excluding these patients eliminated proteinuria as a significant confounder. Each is a contraindication to NMP22 testing due to the high false-positive rate. To our knowledge the reason for decreased CYT specificity in patients with increased protein concentrations in urine remains unknown. There are possible explanations for changed CYT results due to proteinuria. Cell atypia could be caused by UTI. Proteinuria can be associated with UTI, although a systematic review did not confirm a causal role of asymptomatic UTI for proteinuria.26 Even when pa-
tients with UTI were excluded from analysis, the false-positive rate was higher and borderline significant in those with severe proteinuria. Thus, the association of proteinuria with false-positive CYT findings cannot be explained only by UTI. There is no clear-cut evidence that proteins influence urothelial cell morphology. However, microalbuminuria in patients without diabetes is associated with an increased risk of BC.27 To our knowledge the reason for this association has not yet been identified. Possible explanations are increased protein due to inflammation, which is a risk factor for cancer,28 and paraneoplastic protein excretion. In patients with CKD systemic inflammation is common due to uremia and increased cytokines, such as tumor necrosis factor-␣ or interleukin-6.29 Inflammation could also account for morphological changes in urothelial cells in patients with severe proteinuria due to CKD. However, GFR as a renal function parameter did not significantly affect CYT specificity. Regardless of mechanism our results clearly show that positive CYT findings in patients with a high protein concentration in urine must be interpreted differently than in patients with a normal urine composition. Since the immunocytology false-positive rate was not affected by urine protein, immunocytochemical detection of specific antigens may improve the accuracy of noninvasive diagnostic measurements in patients with high grade proteinuria. Also, since FISH is based on detecting chromosome alterations of cells in urine, the complete independence of this test from urine protein and GFR is not surprising. Some functional parameters had an inverse effect on test results. False-positive immunocytology findings were decreased in patients with high serum creatinine but high grade proteinuria led to more false-positive findings. This indicates that renal function parameters should be considered separately. A limitation of our study is that no differentiation was made between microscopic and macroscopic hematuria, which is known to influence BC test results.30 Also, 79% of patients underwent mechanical
INFLUENCE OF RENAL EXCRETORY FUNCTION ON MARKERS TO DETECT BLADDER CANCER
manipulation, which may have also influenced our results. External validation is required to further determine the influencing role of renal function on CYT and NMP22 performance.
73
The false-positive CYT rate was increased in patients with high urine protein concentrations, which led to decreased specificity regardless of UTI. Renal function should be considered when interpreting the results of noninvasive urine tests for detecting or monitoring BC.
CONCLUSIONS To our knowledge this is the first study of the effect of renal function on urine based tests for diagnosing BC. Our study revealed an increased NMP22 test false-positive rate in patients with decreased GFR.
ACKNOWLEDGMENTS Prof. Helmut Heinle, Institute of Physiology, University Tübingen, provided advice on renal function assessment.
REFERENCES 1. Kohler BA, Ward E, McCarthy BJ et al: Annual report to the nation on the status of cancer, 1975–2007, featuring tumors of the brain and other nervous system. J Natl Cancer Inst 2011; 103: 714. 2. Millan-Rodriguez F, Chechile-Toniolo G, SalvadorBayarri J et al: Primary superficial bladder cancer risk groups according to progression, mortality and recurrence. J Urol 2000; 164: 680. 3. Babjuk M, Oosterlinck W, Sylvester R et al: EAU guidelines on non-muscle-invasive urothelial carcinoma of the bladder, the 2011 update. Eur Urol 2011; 59: 997. 4. Budman LI, Kassouf W and Steinberg JR: Biomarkers for detection and surveillance of bladder cancer. Can Urol Assoc J 2008; 2: 212. 5. Raitanen MP, Aine R, Rintala E et al: Differences between local and review urinary cytology in diagnosis of bladder cancer. An interobserver multicenter analysis. Eur Urol 2002; 41: 284. 6. Daniely M, Rona R, Kaplan T et al: Combined analysis of morphology and fluorescence in situ hybridization significantly increases accuracy of bladder cancer detection in voided urine samples. Urology 2005; 66: 1354. 7. Bubendorf L: Multiprobe fluorescence in situ hybridization (UroVysion) for the detection of urothelial carcinoma—FISHing for the right catch. Acta Cytol 2011; 55: 113. 8. Shariat SF, Marberger MJ, Lotan Y et al: Variability in the performance of nuclear matrix protein 22 for the detection of bladder cancer. J Urol 2006; 176: 919.
11. Raitanen MP, Kaasinen E, Lukkarinen O et al: Analysis of false-positive BTA STAT test results in patients followed up for bladder cancer. Urology 2001; 57: 680. 12. Atsu N, Ekici S, Oge OO et al: False-positive results of the NMP22 test due to hematuria. J Urol 2002; 167: 555. 13. Coresh J, Selvin E, Stevens LA et al: Prevalence of chronic kidney disease in the United States. JAMA 2007; 298: 2038.
cell carcinoma of the bladder. BJU Int 2006; 97: 997. 21. Shariat SF, Karam JA, Lotan Y et al: Critical evaluation of urinary markers for bladder cancer detection and monitoring. Rev Urol 2008; 10: 120. 22. Kehinde EO, Al-Mulla F, Kapila K et al: Comparison of the sensitivity and specificity of urine cytology, urinary nuclear matrix protein-22 and multitarget fluorescence in situ hybridization assay in the detection of bladder cancer. Scand J Urol Nephrol 2011; 45: 113.
14. Kipp BR, Tyner HL, Campion MB et al: Chromosomal alterations detected by fluorescence in situ hybridization in urothelial carcinoma and rarer histologic variants of bladder cancer. Am J Clin Pathol 2008; 130: 552.
23. Miyanaga N, Akaza H, Ishikawa S et al: Clinical evaluation of nuclear matrix protein 22 (NMP22) in urine as a novel marker for urothelial cancer. Eur Urol 1997; 31: 163.
15. Chiarelli F, Verrotti A, Mohn A et al: The importance of microalbuminuria as an indicator of incipient diabetic nephropathy: therapeutic implications. Ann Med 1997; 29: 439.
24. Roth GA, Lebherz-Eichinger D, Ankersmit HJ et al: Increased total cytokeratin-18 serum and urine levels in chronic kidney disease. Clin Chim Acta 2011; 412: 713.
16. Levey AS, Bosch JP, Lewis JB et al: A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999; 130: 461.
25. Kumar D, Robertson S and Burns KD: Evidence of apoptosis in human diabetic kidney. Mol Cell Biochem 2004; 259: 67.
17. Sharma S, Zippe CD, Pandrangi L et al: Exclusion criteria enhance the specificity and positive predictive value of NMP22 and BTA stat. J Urol 1999; 162: 53. 18. Grossman HB, Soloway M, Messing E et al: Surveillance for recurrent bladder cancer using a point-of-care proteomic assay. JAMA 2006; 295: 299.
9. Bastacky S, Ibrahim S, Wilczynski SP et al: The accuracy of urinary cytology in daily practice. Cancer 1999; 87: 118.
19. Tritschler S, Scharf S, Karl A et al: Validation of the diagnostic value of NMP22 BladderChek test as a marker for bladder cancer by photodynamic diagnosis. Eur Urol 2007; 51: 403.
10. Ponsky LE, Sharma S, Pandrangi L et al: Screening and monitoring for bladder cancer: refining the use of NMP22. J Urol 2001; 166: 75.
20. Karakiewicz PI, Benayoun S, Zippe C et al: Institutional variability in the accuracy of urinary cytology for predicting recurrence of transitional
26. Carter JL, Tomson CR, Stevens PE et al: Does urinary tract infection cause proteinuria or microalbuminuria? A systematic review. Nephrol Dial Transplant 2006; 21: 3031. 27. Jorgensen L, Heuch I, Jenssen T et al: Association of albuminuria and cancer incidence. J Am Soc Nephrol 2008; 19: 992. 28. Coussens LM and Werb Z: Inflammation and cancer. Nature 2002; 420: 860. 29. Raj DS, Shah H, Shah VO et al: Markers of inflammation, proteolysis, and apoptosis in ESRD. Am J Kidney Dis 2003; 42: 1212. 30. Pesch B, Nasterlack M, Eberle F et al: The role of haematuria in bladder cancer screening among men with former occupational exposure to aromatic amines. BJU Int 2011; 108: 546.