| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Laboratory Medicine, Krankenhaus Nordwest, Steinbacher Hohl 2, D-60489 Frankfurt, Germany.
2
Urology Department, Technical University of Munich,
Ismaninger Strasse 22, 81675 Munich, Germany.
3
Urology Clinic, University of Vienna, Währinger
Gürtel 18-20, Vienna A-1090, Austria.
4
Nuclear Medicine Division, National Tumor Institute, Via
Venezian 1, I-20133 Milan, Italy.
5
Institute of Urology, University of Padova, Via
Giustiniani 2, 35128 Padova, Italy.
6
Fondazione Vincenzo Pansadoro, Via Aurelia 559, 00165
Rome, Italy.
7
Urology Clinic and
8
Nuclear Medicine,
Hôpital Bichat, 46 rue Henri Huchard, Paris Cedex 18, France.
9
Urology Clinic, Hôpital Pitie-Salpetriere, 47-83
bd de l'hopital, 75634 Paris Cedex 13, France.
10
Clinical Trials, Bion Diagnostic Sciences, 12277 134th
Avenue NE, Redmond, WA 98052.
a Address correspondence to this author at: Krankenhaus Nordwest, Laboratoriumsmedizin, Steinbacher Hohl 2, D-60489, Frankfurt, Germany. Fax 49 69 7601 3647; e-mail th-books{at}t-online.de
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Methods: The BTA TRAKTM assay, a quantitative enzyme immunoassay for the bladder tumor-associated antigen in urine, was compared with exfoliative cytology in 220 patients (155 men, 65 women; mean age, 64.2 years) presenting with signs, symptoms, or preliminary diagnostic results suggestive of this disease. Cystoscopy was the standard of detection.
Results: In the 100 patients found to have bladder cancer, the overall sensitivities of the BTA TRAK assay (at a previously determined decision threshold of 14 kilounits/L) and cytology were 66% (66 of 100) and 33% (33 of 100), respectively (P <0.001). The BTA TRAK assay proved to be statistically more sensitive than cytology for tumor grades I and II and for stage Ta and T1 tumors. In contrast, the overall specificity of the BTA TRAK assay in the 120 patients without cystoscopically confirmed bladder cancer was 69% (83 of 120) and that of cytology was 99% (119 of 120; P <0.001). The specificity of the BTA TRAK assay was higher in patients without benign or malignant genitourinary disease other than bladder cancer (76%; n = 89) than in patients with these conditions. When the BTA TRAK assay and cytology were used together such that a positive result in either test was scored as positive and the results compared with those of the BTA TRAK assay alone, increases in overall sensitivity and equivalent specificity were observed.
Conclusion: Because of its relatively high sensitivity, the BTA TRAK assay could complement cytology as an adjunct to cystoscopy in the diagnosis and follow-up of most patients with bladder cancer.© 1999 American Association for Clinical Chemistry
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Fortunately, 75–85% of newly diagnosed bladder tumors are confined to the bladder mucosa, are not life threatening, and can be treated relatively easily by transurethral resection. The risk of recurrence or new tumor growth in these patients, however, is ~75% (1). Recurrence at the original site may indicate incomplete removal of the initial tumor. New tumor growth at sites other than the original may be the result of carcinogens in the urine affecting different parts of the bladder mucosa at different times or seeding by viable tumor cells, either before or at the time of resection (2). Because of the risk of recurrence, patients must be examined for bladder cancer at frequent intervals, especially during the first few years after treatment (3)(4).
A definitive diagnosis of bladder cancer can be made only by inspection of the interior of the bladder, typically with the aid of a cystoscope, and pathologic examination of a resected tumor or biopsy specimen. However, when bladder tumors are small or flat (carcinoma in situ), they cannot always be visualized reliably by cystoscopy (5)(6). For this reason, many urologists use an adjunctive diagnostic procedure, such as exfoliative cytology. Cytology, however, has several disadvantages, including its reliance for accuracy on the skill of the pathologist, its relative insensitivity to grade I (well-differentiated) tumors, and a lack of standardization of methods (7)(8).
The BTA TRAKTM assay, a quantitative enzyme immunoassay for a bladder tumor-associated antigen, is performed on a urine specimen in a clinical laboratory. The antigen measured by the assay is human complement factor H-related protein (hCFHrp),1 which is similar in structure and function to human complement factor H (hCFH) (9)(10). In cell culture, hCFHrp has been shown to be produced by several human bladder cancer cell lines but by not healthy epithelial cells (9)(10). hCFH plays an important inhibitory role in the control of the alternative complement pathway, and the function of this pathway is the lysis of cells recognized by the host as foreign (9)(10). By interacting with complement factor C3 convertase and serving as a cofactor for complement factor I, hCFH inhibits the formation of a membrane attack complex, thereby preventing cell lysis. In vitro, hCFHrp interrupts the complement cascade and protects cells from lysis by complement (9)(10). Production of hCFHrp may confer a selective growth advantage to cancer cells in vivo by allowing the cells to evade the host immune system (11).
Here we present the results of a multicenter study comparing the sensitivity of the BTA TRAK assay for bladder cancer with those of voided urine cytology or bladder wash cytology, using cystoscopy or cystoscopy and biopsy as the standard for definitive diagnosis.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
patients
Eligible for participation in the study were patients with or
without a history of bladder cancer who were suspected of having
bladder cancer on the basis of signs, symptoms, or recent intravenous
urographic or cystoscopic results and who were scheduled to undergo
cystoscopy and cytologic examination. Written or verbal informed
consent was obtained from each patient before participation in any
study-related activity.
samples
All patients provided a single midstream or catheterized urine or
bladder wash specimen the same day as but before their diagnostic
workup. A portion of the specimen was frozen at -20 °C and sent to
a central laboratory for processing according to the instructions
supplied with the BTA TRAK assay kit (Bard Diagnostic Sciences, C.R.
Bard, Inc.). The remainder of the sample was sent to the
institution's cytology laboratory for processing according to standard
procedures. When a bladder wash cytology was performed instead of
voided urine cytology, it was done at the time of cystoscopy, and the
wash was also sent to the cytology laboratory for processing.
The BTA TRAK assay can be performed on fresh, refrigerated, or previously frozen urine samples. A US study (12) on the stability of the antigen showed that urine specimens can be stored refrigerated for up to 2 weeks without the addition of preservatives or inhibitors and they can be stored frozen at -20 or -80 °C for at least 24–26 weeks with as many as three cycles of thawing and refreezing without deterioration in antigen concentrations.
Urine specimens should be screened for the presence of blood. Because the presence of hematuria is sometimes associated with false-positive BTA TRAK results, it is recommended that upon detection of hematuria by dipstick, that voided urine cytology be performed as well as the BTA TRAK test.
the bta trak assay
The assay procedure is that of a standard enzyme immunoassay. The
assay uses a capture monoclonal antibody (MAb) and an enzyme-labeled
reporter MAb that bind specifically to hCFHrp. Patient urine and a
diluent are added to microwells coated with capture MAb and allowed to
incubate for 1 h at 37 °C. Bladder tumor-associated antigen
present in the urine is bound by the capture MAb, and unbound urine
components are subsequently removed by washing. The alkaline
phosphatase-labeled reporter MAb is then added to the microwells,
followed by an additional 1-h incubation at 37 °C. The plate is
again washed to remove unbound conjugate, and the bound reporter
conjugate is detected by reaction with substrate for 30 min at
37 °C. The amount of antigen is determined by comparing the
absorbances in the sample wells with those of calibrators
assayed simultaneously (range of quantification, 0–100
kilounits/L). Assay calibrators and controls were calibrated
with antigen purified from urine samples of patients with confirmed
bladder cancer. The calibration curve used antigen concentrations of 0,
5, 20, 50, and 100 kilounits/L. One unit in the BTA TRAK assay is
~4.7 ng of hCFHrp.
the bta trak assay decision threshold
Because only symptomatic patients with clinical signs of bladder
cancer or patients with known bladder cancer were included in this
study, the decision threshold of an earlier clinical evaluation of the
BTA TRAK assay was used to select patients with a bladder tumor. In the
earlier study (12), the decision threshold, established by
an ROC analysis corresponding to 97% specificity, was found to
be 14 kilounits/L.
performance characteristics of the bta trak assay
The precision of the BTA TRAK assay was determined according to
the procedures recommended by the NCCLS (13) by
testing two controls and five urine samples in duplicate in each of 20
independent analytical runs (1 analytical run/day) at three
different laboratories. The within-run, between-run, and between-site
CVs for the low control (9.2 kilounits/L) were 7.1%, 2.4%, and
0.3%, respectively. The CVs for the high control (35.6 kilounits/L)
were 6.8%, 2.9%, and 1.2%, respectively (14).
To determine the recovery, urine specimens from four patients with known hCFHrp concentrations were added to a nondiseased urine sample containing endogenous antigen at a concentration of 1.76 kilounits/L, and the resulting antigen concentrations were determined using the BTA TRAK assay. The recoveries for the four patients were 100.0%, 91.4%, 96.0%, and 98.3%.
The results of the BTA TRAK test were not normalized to urinary creatinine. Preliminary results have shown that the association between the creatinine concentration in urine and the BTA TRAK values is not strong. Creatinine does not explain the variability in the BTA TRAK values in the monitoring of a bladder cancer patient, and it does not appear to smooth the values. In the cases of bladder cancer patients, although there is variation in the BTA TRAK values, the values never drop below the threshold of 14 kilounits/L.
cytology
Cytology was performed by microscopic examination of a voided
urine specimen or bladder washing for exfoliated cells from cancerous
bladder epithelium. Bladder washings, when performed, were obtained by
irrigating the bladder with physiologic saline solution through a
catheter or rigid cystoscope. Cytologic preparations were made by
fixing exfoliated cells, placing them on glass slides, and staining
them. The cytologic examinations were performed by expert pathologists
who adopted common criteria for evaluation and were unaware of the
results of the BTA TRAK assay. A positive cytology result was defined
as a laboratory report of "positive" for or "highly suggestive"
of malignancy. All of the doubtful reports, which involved the need to
repeat cytology, such as "suspicious or atypical cells" were not
considered positive because they did not provide any definite diagnosis
of cancer.
study analysis
A positive final diagnosis was defined as bladder cancer confirmed
by cystoscopy and biopsy. Tumors were graded and staged according to
the classification of the International Union Against Cancer
(15), in which staging depends on the type of tumor and its
extent of penetration into the bladder wall and grading is based on the
degree of tumor differentiation. Grade I, II, and III tumors are well,
moderately well, and poorly differentiated, respectively. A negative
final diagnosis was defined as no bladder cancer ascertained by
cystoscopy or cystoscopy and biopsy. If a definitive final diagnosis
was not possible (suspicious results), the case was not included
in the determination of sensitivity or specificity.
The primary measures of diagnostic efficacy were sensitivity to bladder cancer, defined as the ratio of the number of true-positive results to the number of positive final diagnoses (based on cystoscopy and biopsy), and specificity for bladder cancer, defined as the ratio of the total number of true-negative results to the total number of negative final diagnoses. The number of study participants required to detect a difference of 0.15 or greater between the BTA TRAK assay and cytology sensitivities was computed using the method of Schork and Williams (16). Sensitivities and specificities were calculated using combined data from the six study institutions. The precision of the sensitivity and specificity estimates are shown as exact 95% confidence intervals (CIs). An exact test analogous to the McNemar test was used to compare the overall sensitivities and specificities of the BTA TRAK assay and cytology and the sensitivities for the different tumor grades and stages. Data are presented separately for the two patient types (history of bladder cancer vs no history of bladder cancer) because patients with a history of bladder cancer had received treatment and were being monitored for recurrence.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
comparison of sensitivities of the bta trak assay and cytology
As is shown in Table 1
, of the 100 patients in this study who were diagnosed with
bladder cancer by cystoscopy, the overall clinical sensitivity of the
BTA TRAK assay was 66% (66 of 100; Fig. 1
), whereas that of cytology was 33% (33 of 100; P
<0.001). There were no statistically significant differences between
the sensitivities in those patients who had a previous history of this
disease (n = 51) and those who were newly diagnosed (n = 49)
for either cytology or the BTA TRAK assay, nor did the positivities of
cytology differ between voided urine samples and lavage fluid samples
(not shown).
|
|
The sensitivities of the BTA TRAK assay and cytology according to tumor
grade and stage are given in Table 2
. In terms of sensitivity, the BTA TRAK assay was superior to
cytology in the 25 patients with grade I tumors (48% vs 4%;
P <0.001) and in the 41 patients with grade II tumors (59%
vs 22%; P = 0.001). Similarly, the BTA TRAK assay
exhibited a higher sensitivity than cytology in detecting the 55 stage
Ta tumors (51% vs 15%; P <0.001)
and in the 24 stage T1 tumors (88% vs 50%;
P = 0.011). There were no statistical differences
between the BTA TRAK assay and cytology in grade III tumors (n =
34) or in stage T2–T4 tumors (n = 16) and
Tis tumors (n = 5).
|
comparison of the specificities of the bta trak assay and cytology
Seventy-three of the 120 patients whose final diagnosis was
negative for bladder cancer had a history of this disease. In addition,
31 were found to have other genitourinary (GU) cancers or GU diseases
such as renal calculi and urinary tract infection. The other GU cancers
included one prostate cancer, three benign bladder tumors, and one
renal cancer. These conditions may cause interference because of the
presence of hCFH in the urine and leakage from bleeding because of the
disease. Thus, the positive BTA TRAK values in patients with benign
disorders and dipstick-positive hematuria would not be surprising.
In contrast to its rather low sensitivity, cytology exhibited excellent specificity in all patient categories. Only one false-positive result was observed in all 120 patients, and that was in a patient with a history of bladder cancer.
The overall specificity of the BTA TRAK assay in these 120 patients was
69% (83 of 120; 95% CI, 60–77%; Fig. 1
). As is shown in Table 3
, significantly better specificity (P = 0.006)
was observed in patients who had no apparent GU disease (76%, 68 of
89; 95% CI, 66–85%) than in the group with GU disease (48%, 15 of
31; 95% CI, 30–67%). In addition, a history of bladder cancer was
not related to the observed specificity in either group (Table 3
).
|
Because the group with no previous history of bladder cancer contained
a higher percentage of patients with GU disease (47%, 22 of 47) than
that with previous bladder cancer (12%, 9 of 73; Table 3
), the
specificity of the BTA TRAK assay appeared to be better in the latter
(80%, with 95% CI, 68–88% vs 53%, with 95% CI, 38–68%).
complementarity of the bta trak assay and cytology
The effects of using both the BTA TRAK and cytology data on
sensitivity and specificity are shown in Table 4
. Because cytology had an exquisite specificity, calling the
overall result positive if either the BTA TRAK assay or cytology were
positive yielded a specificity equivalent to that of the BTA TRAK
test alone. As anticipated, the use of both assays in this manner
produced an apparent increase in sensitivity of the BTA TRAK test alone
(71% vs 66%), although this increase was not statistically
significant.
|
As expected, the effect of using cytology as a confirmatory test for
the BTA TRAK assay, that is, requiring that both tests be positive to
report a positive result, produced a lower overall sensitivity than
using cytology alone with a specificity equivalent to that of cytology
(Table 4
).
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
In this study, the BTA TRAK assay was significantly more sensitive than cytology for tumor grades I and II (P <0.05) and for stage Ta and T1 tumors (P <0.05). The fact that the difference in sensitivity between the BTA TRAK assay and cytology in the patients with T2–T4 tumors was not statistically significant may be attributable to the small sample size (16 patients). The sensitivity of the BTA TRAK assay in a small group (n = 5) of Tis tumors (carcinoma in situ) was 60%, compared with an 80% sensitivity of cytology in the same group. Sensitivities of the BTA TRAK assay in this study were very similar to those reported in a 900-patient US multicenter study (13), both in terms of grade (grade I, 48% vs 53%; grade II, 59% vs 68%; and grade III, 88% vs 85%) as well as in stage (Ta, 51% vs 59%; T1, 88% vs 92%; T2–T4, 88% vs 88%; and Tis, 60% vs 67%).
On the other hand, the specificity of cytology (99%) was clearly
superior to that of the BTA TRAK assay (69%) in this patient
population. It should be noted, however, that the specificity of the
BTA TRAK assay reported in this study was artificially lower than
generally expected for the following two reasons. First, there were no
exclusions made for patients with infections or renal stones as
required in the package insert. These conditions may cause interference
because of leakage of endogenous hCFH into the urine from the disease
process. Thus, the positive BTA TRAK values in the patients with benign
urologic disorders would not be surprising. It was shown recently that
hCFH is produced in human proximal tubular epithelial cells and human
mesangial cells in response to interferon-
, an important mediator of
inflammation (19)(21). Second, follow-up
information was not available on non-bladder cancer patients who tested
positive by the BTA TRAK assay. Sixty-one percent of these patients had
a history of bladder cancer, and patients in this group should not
necessarily be considered falsely positive. Given the high tumor
recurrence rate in patients with a history of bladder cancer, it is
likely that many of these positive BTA TRAK tests not confirmed by
cystoscopy indicated early detection of subclinical disease.
Thus, in terms of utility, if the primary intent of the physician in using an adjunctive test to cystoscopy is to maximize the detection of cancer while minimizing cost, the BTA TRAK assay could replace cytology in the diagnosis and follow-up of patients with bladder cancer, if infectious GU disease or injury, which reduces specificity, can be ruled out. The BTA TRAK assay appeared to be especially useful in the detection of low-grade tumors because of its greater sensitivity than cytology. Our study population did not include enough patients to draw conclusions regarding the sensitivity of the BTA TRAK assay in patients with carcinoma in situ. Carcinoma in situ generally has a worse prognosis than papillary carcinoma (21). Until the BTA TRAK assay can be further evaluated, we recommend that cytology continue to be used to supplement cystoscopy in patients with a history or suspicion of carcinoma in situ. This approach has the additional advantage that it should detect a few more cancers than the BTA TRAK assay alone, without decreasing the specificity of the BTA TRAK assay.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
induces biosynthesis of complement components C2, C4, and factor H by human proximal tubular epithelial cells. Cytokine 1997;9:176-183.
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  Z.-Z. Cheng, M. J. Corey, M. Parepalo, S. Majno, J. Hellwage, P. F. Zipfel, R. J. Kinders, M. Raitanen, S. Meri, and T. S. Jokiranta Complement Factor H as a Marker for Detection of Bladder Cancer Clin. Chem., May 1, 2005; 51(5): 856 - 863. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Sanchez-Carbayo Use of High-Throughput DNA Microarrays to Identify Biomarkers for Bladder Cancer Clin. Chem., January 1, 2003; 49(1): 23 - 31. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. I. Martin-Subero, I. Chudoba, L. Harder, S. Gesk, W. Grote, F. J. Novo, M. J. Calasanz, and R. Siebert Multicolor-FICTION: Expanding the Possibilities of Combined Morphologic, Immunophenotypic, and Genetic Single Cell Analyses Am. J. Pathol., August 1, 2002; 161(2): 413 - 420. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. A. Sokolova, K. C. Halling, R. B. Jenkins, H. M. Burkhardt, R. G. Meyer, S. A. Seelig, and W. King The Development of a Multitarget, Multicolor Fluorescence in Situ Hybridization Assay for the Detection of Urothelial Carcinoma in Urine J. Mol. Diagn., August 1, 2000; 2(3): 116 - 123. [Abstract] [Full Text]
|
|
|
|
 |
|
 A. Schneider, S. Borgnat, H. Lang, O. Régine, V. Lindner, M. Kassem, C. Saussine, P. Oudet, D. Jacqmin, and M. P. Gaub Evaluation of Microsatellite Analysis in Urine Sediment for Diagnosis of Bladder Cancer Cancer Res., August 1, 2000; 60(16): 4617 - 4622. [Abstract] [Full Text]
|
|
|
|
|
|
M. Burchardt, T. Burchardt, A. Shabsigh, A. De La Taille, M. C. Benson, and I. Sawczuk Current Concepts in Biomarker Technology for Bladder Cancers Clin. Chem., May 1, 2000; 46(5): 595 - 605. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. S. Fedarko, B. Fohr, P. G. Robey, M. F. Young, and L. W. Fisher Factor H Binding to Bone Sialoprotein and Osteopontin Enables Tumor Cell Evasion of Complement-mediated Attack J. Biol. Chem., May 26, 2000; 275(22): 16666 - 16672. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
