| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cancer Diagnostics |
1 Gen-Probe Incorporated, San Diego, CA.
2 Urological Sciences Research Foundation, Culver City, CA.
aAddress correspondence to this author at: Gen-Probe Incorporated, 10210 Genetic Center Dr., San Diego, CA 92121. Fax 858-410-8113; e-mail jackg{at}gen-probe.com.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: Whole-urine specimens were collected after digital rectal examination from 3 groups: men scheduled for prostate biopsy (n = 70), healthy men (<45 years of age with no known prostate cancer risk factors; n = 52), and men who had undergone radical prostatectomy (n = 21). PCA3 and prostate-specific antigen (PSA) mRNAs were isolated, amplified, and quantified by use of Gen-Probe DTS400® Systems. Prostate biopsy results were correlated with the PCA3/PSA mRNA ratio, and PSA mRNA concentrations were used to normalize PCA3 signals and confirm the yield of prostate-specific RNA. Assay precision, specimen stability, and mRNA yield were also evaluated.
Results: The specimen informative rate (fraction of specimens yielding sufficient RNA for analysis) was 98.2%. In this clinical research study, ROC curve analysis of prebiopsy specimens yielded an area under the curve of 0.746; sensitivity was 69% and specificity 79%. Serum PSA assay specificity was 28% for this same group. PCA3 and PSA mRNAs were undetectable in postprostatectomy specimens except for one man with recurrent prostate cancer. Assay interrun CVs were 
12%. Both mRNAs were stable in processed urine up to 5 days at 4 °C and after 5 freeze–thaw cycles.
Conclusion: The APTIMA® PCA3 assay combines simple specimen processing with precise assays and existing instruments and could add specificity to the current algorithm for prostate cancer diagnosis.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Among the newer molecular biomarkers for prostate cancer diagnosis (2)(3)(4)(5) is the product of prostate cancer gene 3 (PCA3), 2 a prostate-specific noncoding mRNA (6). PCA3 is overexpressed in 95% of prostate cancers tested, with a median 66-fold up-regulation compared with adjacent nonneoplastic prostatic tissue (4). In contrast, KLK3, the gene that encodes for PSA, is not up-regulated in cancer (7)(8), and PSA mRNA concentrations may therefore be used to normalize for the amount of prostate-specific RNA in molecular test samples. PCA3 expression is mostly undetectable in other tissues, including bladder and testis (9). The feasibility of a quantitative PCA3-based molecular test for urine sediments has been demonstrated (4).
The paradigm of direct detection of cancer cells in biological fluids is attractive because of the expected improvement in specificity compared with the current immunodiagnostic tests for surrogate markers such as serum PSA. This approach presents several challenges, however: The procedure for collecting and stabilizing specimens under standardized conditions must be defined. In addition, the method must be able to detect low concentrations of RNA in clinical samples [in previous molecular marker studies, 
20% of the clinical specimens contained insufficient prostate cells for testing (4)(10)]. Finally, the test must use a robust and reproducible approach that can be implemented in a clinical laboratory.
We describe here the development of prototype transcription-mediated amplification assays to quantify PCA3 and PSA mRNA in whole urine.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The patient population consisted of 3 groups. The prebiopsy group consisted of 70 men scheduled for prostate biopsy because of increased serum PSA (
2.5 µg/L) and/or abnormal diagnostic digital rectal examination (DRE) and/or other risk factors: mean (SD) age, 67 (11) years (range, 45–93 years); serum PSA, 7.7 (14.1) µg/L (0.4–101.7 µg/L); prostate gland volume, 40.0 (17.2) cm3 (8.6–92.8 cm3). Biopsies were performed according to the procedure described by Marks et al. (11). The second group (normal) included 52 men <45 years of age with no known prostate cancer risk factors; prostate biopsies were not performed on these men. Finally, urine specimens were also collected from 21 men who had undergone radical prostatectomy (postprostatectomy) at least 3 months before specimen collection. As additional negative controls, whole-urine specimens were collected from 6 women.
Male urine samples (20- to 30-mL first catch) were collected after an attentive DRE. The DRE was performed by applying firm pressure (enough to depress the prostate surface 1 cm) from the base to apex and from the lateral to the median line for each lobe. Exactly 3 strokes per lobe were performed. Urine was kept on ice and processed within 4 h by mixing with an equal volume of detergent-based stabilization buffer (Urine Transport Medium; UTM), which lyses the cells and stabilizes the mRNA. The UTM was the same formulation that is found in the APTIMA Urine Specimen Collection Kit for Male and Female Urine Specimens (Gen-Probe; cat. no. 1040). The processed samples were shipped overnight on cold packs and stored at –70 °C for up to 8 months before testing. PCA3 and PSA mRNA stability in processed urine specimens was assessed by monitoring copy numbers after incubation at defined temperatures or after freeze–thaw cycles.
assay description
The quantitative APTIMA PCA3 assay uses Gen-Probe technologies (12)(13)(14)(15), including target capture (purification of target mRNA by hybridization to magnetic particles via target-specific oligonucleotides), transcription-mediated amplification (amplification of target RNA sequences), and a hybridization protection assay (specific detection of amplification products by use of target-specific acridinium ester–labeled probes). The components for the APTIMA PCA3 assay include analyte-specific (PCA3 and PSA) target capture, amplification, and probe reagents, as well as calibrators and controls. All other reagent formulations are identical to the corresponding Gen-Probe APTIMA Combo 2® reagents (package insert IN0037-05, Rev. A.1). The PCA3 and PSA assay protocols are identical and use the reagent addition volumes and incubation times and temperatures specified in IN0037-05 Rev. A.1 (16). Assays were run with Gen-Probe DTS400® Systems; this allowed both PCA3 and PSA mRNA assays to be run in parallel by a single operator.
primers and probes
The PCA3 primers target exons 3 and 4, giving an amplification product of 75 nucleotides that includes the splice junction (6). The PCA3 probe spans the exon 3/4 junction; as such, the assay does not detect the unspliced forms of the PCA3 transcript. The PSA primers target exons 2 and 3 in the PSA mRNA (17), and as in the PCA3 assay, unspliced mRNA is not detected.
calibrators, controls, and assay setup
Calibrators and controls consisted of PCA3 or PSA transcripts in detergent solution and were included in each assay run. Copy numbers were chosen to encompass the range of PCA3 and PSA mRNA concentrations observed in processed urine specimens (Table 1
). Calibrators and controls were value-assigned with the PCA3 or PSA assays by use of reference transcripts at known copy concentrations (based on A260 nm measurements). For assay precision studies, calibrators, controls, and test samples were run in triplicate. For determination of clinical sensitivity and specificity relative to prostate biopsy, processed urine specimens were tested in triplicate; these data were also analyzed to determine the impact of running a single replicate of each specimen. For calibrators and controls, each reaction contained 400 µL; for processed urine specimens, each reaction contained the equivalent of 200 µL of whole urine (200 µL whole urine + 200 µL of UTM).
|
target capture
In a preamplification area, PCA3 or PSA Target Capture Reagent (100 µL) was added to strips of ten 12 x 75 mm polypropylene tubes held in racks. Well-mixed processed urine specimens or calibrators/controls (400 µL) were added, and the tubes were covered with sealing cards, vortex-mixed thoroughly on a multitube vortex-mixer, and incubated in a 62 °C water bath for 30 min. The racks were cooled at room temperature for 30 min and then placed on a Gen-Probe Target Capture System for 5–10 min to allow the magnetic particles to migrate to the sides of the tubes. The sealing cards were removed, and the liquid in the tubes was aspirated while the magnetic particles remained on the side of the tubes. Wash Reagent (1 mL) was added to each tube, and the tubes were covered, vortex-mixed for 10 s, and placed back on the Target Capture System. After 5 min, the Wash Reagent was aspirated, and the racks were transferred to the amplification area.
transcription-mediated amplification
PCA3 or PSA Amplification Reagent (75 µL) was added to each tube with a repeat pipettor, followed by Oil Reagent (200 µL). The tubes were covered, vortex-mixed for 10 s, and incubated in a 62 °C water bath for 10 min. The racks were transferred to a 42 °C water bath for 5 min. Enzyme Reagent (25 µL) was pipetted into each tube within 90 s while the rack remained in the water bath. The tubes were covered, shaken by hand to completely mix the contents, immediately returned to the 42 °C water bath, and incubated for 1 h.
detection
The racks were transferred to a postamplification area, and PCA3 or PSA Probe Reagent (100 µL) was added to each tube. The racks were covered, vortex-mixed until thoroughly mixed (10–20 s), incubated in a 62 °C water bath for 20 min, and then allowed to cool at room temperature for 5 min. Selection Reagent (250 µL) was pipetted into each tube, after which the tubes were covered, vortex-mixed until thoroughly mixed (10–20 s), and incubated in the 62 °C water bath for 10 min. The tubes were cooled at room temperature for 15 min and then transferred to a Gen-Probe LEADER® HC+ Luminometer. The signal in relative light units (RLUs) was determined over a 3-s read time after addition of Auto Detect Reagents I and II.
data analysis
Both the PCA3 and PSA assays were run with corresponding 5-point calibration curves. The RLU data were transferred to an Excel spreadsheet for analysis. A weighted least-squares fit was used to generate a 4parameter logistic dose–response curve for the 5 calibrator points with TableCurve 2D software (Systat; Ver. 5.01). The weight values used for each calibrator point are shown in Table 1
. The curve [RLU vs (calibrator copies/reaction)] is defined by the equation y = a + b/[1 + (x/c)d], where x is the calibrator copies/reaction, y is RLU output, and a, b, c, and d are the 4 parameters defining the logistic dose–response curve as determined by the software. The values for a, b, c, and d were entered into the equation, which was then solved for copies/reaction (c/rxn) based on RLU values obtained for the controls and specimens. The normalized PCA3/PSA ratio result was calculated by use of the mean c/rxn of the 3 replicates of each target. Samples with copy numbers greater than the highest calibrator were diluted in APTIMA Transport Medium (detergent solution of the same formulation as the negative calibrator; also contained in the APTIMA Adaptor Kit; Gen-Probe cat. no. 1087) to bring the sample within the dynamic range of the assay and retested. The measured copy numbers for controls were compared with their known values to verify the results interpolated from the calibration curve.
Assay precision was assessed by use the method described in NCCLS EP5-A2 (18). Confidence intervals for the area under the ROC curve were determined by a nonparametric approach using SAS/STAT software.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, as are the measured values as a percentage of the expected values (c/rxn divided by assigned c/rxn) and intrarun, interrun, and total CVs. The measured values of all controls were within 11% of the target values, and interrun CVs ranged from 4% to 12%. The results obtained for the pooled processed urine specimens are shown in Table 3
. Interrun CVs for PCA3 and PSA mRNA were similar to those obtained for the assay controls, and interrun CVs for the PCA3/PSA ratio were 15%–24%.
|
|
specimen stability
The stability of PCA3 and PSA mRNA in processed urine was evaluated by use of 3 clinical specimens. Frozen processed urine specimens were thawed, then stored at 4 °C or 30 °C for 2 weeks and assayed at different time points (Fig. 1
). Data are shown as the mean of the c/rxn values obtained for the 3 specimens. At 4 °C, mRNA copy numbers remained constant for up to 5 days. Degradation of PCA3 and PSA mRNA was observed at 30 °C; the SD also increased, reflecting the variability in room temperature stability from specimen to specimen (Fig. 1, A and B
). The PCA3/PSA ratio remained within 20% of initial for all 3 samples up to 2 weeks at 4 °C (Fig. 1C
). PCA3 and PSA mRNAs were also stable in processed urine specimens (n = 4) after 5 freeze–thaw cycles (data not shown).
|
pca3 and psa MRNA yield from cultured cells and urine
On average, 200 µL of whole urine yielded 500 000 copies of PSA mRNA (median, 186 198). PCA3 mRNA copy numbers were 20- to 30-fold lower (mean and median of 22 000 and 5000 c/rxn, respectively). The informative rate from prebiopsy and healthy controls, defined as the fraction of specimens that yielded sufficient PSA mRNA for analysis, was 98.2% (110 of 112). As a point of reference, PCA3 and PSA mRNA concentrations in the LNCaP prostate carcinoma cell line were also measured. On average, LNCaP cells contained 1000 to 2000 copies of PSA mRNA/cell; these results are in good agreement with previously published data (19). PCA3 concentrations in LNCaP cells were much lower (1 to 10 copies/cell), which is consistent with the fact that PCA3 promoter activity has been shown to be relatively weak in this cell line (20).
mRNA yield from postprostatectomy samples was also determined. Twenty of the 21 specimens yielded PCA3 and PSA assay signals at or near background values, and mRNA copy numbers were below the amount required for analysis. The remaining postprostatectomy specimen yielded PCA3 and PSA mRNA c/rxn values of 2368 and 42 744, respectively, and a PCA3/PSA mRNA ratio of 55 x 10–3. Follow-up indicated that during the 4 months between the radical prostatectomy and collection of urine for PCA3 testing, this patient had a recurrence and became biopsy-positive. The analytical specificity of the assay was also confirmed by testing 6 female urine samples; all gave RLU values within background (not shown).
pca3/psa ratio and correlation with biopsy results
The prebiopsy population tested included 70 males. Of the 68 that yielded informative specimens, 16 biopsies were positive for prostate cancer, and the remaining 52 were biopsy-negative; 40 of the 52 biopsy-negative males had BPH and/or inflammation. The PCA3/PSA mRNA ratio values obtained for different subject groups are shown in Fig. 2
. The median ratios (PCA3/PSA x 10–3) for the healthy, biopsy-negative, and biopsy-positive groups were 4.5, 27.0, and 81.8, respectively. The mean PCA3/PSA mRNA ratios of these 3 groups were significantly different (P <0.01).
|
ROC curve analysis was performed on the prebiopsy specimens with PCA3/PSA mRNA ratio used as the diagnostic indicator and biopsy as the comparison method (Fig. 3
). The area under the curve (AUC) was 0.746 (95% confidence interval, 0.574–0.918). The use of PCA3 copy numbers alone gave an AUC of only 0.575 (not shown), demonstrating the value of normalizing for prostate-specific mRNA concentration by use of the concentration of PSA mRNA. The ROC curve was used to determine sensitivity and specificity relative to prostate biopsy at different PCA3/PSA mRNA ratio cutoffs. At a cutoff of 50 x 10–3, sensitivity was 69% and specificity 79% (Table 4
). Increasing the cutoff to 80 x 10–3 decreased sensitivity to 56% but increased specificity to 94%. For comparison, the serum PSA assay (cutoff, 2.5 µg/L) yielded a specificity of 28% and sensitivity of 81% for this same study group; at 69% sensitivity, serum PSA assay specificity was 60% vs 79% for the APTIMA PCA3. Importantly, the clinical performance of the APTIMA PCA3 assay for prebiopsy samples from men with serum PSA in the "gray zone" of 2.5 to 10 µg/L (n = 44) was similar to that obtained for the entire prebiopsy group (sensitivity, 69%; specificity, 83%). To evaluate the impact of including disease-free persons on clinical performance, we also performed ROC analysis by adding the healthy group to the biopsy-negative group. The AUC increased from 0.746 to 0.825, sensitivity was 69%, and specificity improved to 89%.
|
|
These clinical specimen results were obtained by use of the means of 3 replicates to determine the PCA3/PSA ratio. To determine whether accurate results could be obtained by use of a single replicate of each specimen, we randomly combined the 3 PCA3 and PSA replicates to generate all 9 possible single-replicate ratios. For 59 of the 68 informative specimens (87%), all 9 possible single-replicate PCA3/PSA ratios yielded the same overall interpretation (above or below the cutoff of 50 x 10–3) as the mean. The remaining 9 specimens (13%) had PCA3/PSA ratios near the cutoff or some PCA3 or PSA replicates that were outliers.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Preliminary evaluation of PCA3 and PSA mRNA stability in processed urine indicated that specimens may be held for 5 days at 4 °C. It should be noted that at 30 °C, mRNA stability varied significantly from specimen to specimen; therefore, all shipments from the collection site were overnight on cold packs and specimens were frozen at –70 °C on receipt to ensure integrity. The freeze–thaw stability data indicate that frozen specimens may be thawed and retested several times without a negative impact on assay results.
When we used prostate biopsy as the comparison method, the APTIMA PCA3 assay yielded an AUCROC of 0.746; sensitivity was 69% and specificity 79% at a PCA3/PSA mRNA ratio cutoff of 50 x 10–3. This specificity was higher than that for the serum PSA assay for the same study group (28% at cutoff of 2.5 µg/L). Mean PCA3/PSA ratios were significantly different for the healthy, biopsy-negative (including BPH), and biopsy-positive patients. The inclusion of healthy (presumed to be disease-free) persons in the ROC analysis increased the AUC to 0.825 and also improved specificity to 89%; this is consistent with the fact that the mean PCA3/PSA ratio was significantly lower for the healthy group. Although these results are encouraging, data from other ongoing studies are needed to fully establish the clinical utility of the assay.
APTIMA PCA3 clinical performance was similar to that observed in previous studies that used reverse transcription-PCR and urine sediments. Using a quantitative PCA3/PSA ratio, Hessels et al. (4) reported an AUCROC of 0.717 with a sensitivity of 67% and specificity 83% in a cohort of 108 prebiopsy patients. Fradet et al. (10) reported a sensitivity and specificity of 66% and 89%, respectively, although direct comparison with that study is less straightforward because their method was semiquantitative and used a more complex data reduction algorithm. The APTIMA PCA3 assay was able to detect recurrence of prostate cancer in 1 man after radical prostatectomy. PCA3 and PSA mRNA concentrations were nearly undetectable in the remaining postprostatectomy specimens (RNA concentrations less than the minimum for informative specimens); these data provide preliminary confirmation of the prostate tissue specificity of the urine assay, using a known negative study population without prostate glands. The results obtained with female urine specimens also verify the analytical specificity of the assay.
The APTIMA PCA3 assay provides several advantages over other PCA3 urine tests (4)(10). The ability to use whole urine (as opposed to urine sediments) simplifies the specimen-processing procedure. Processing is complete with the addition of UTM; specimens can then be frozen and stored at the collection site. Reverse transcription-PCR methods require stabilization of the cells in whole urine and immediate shipment to the testing site. The target-capture technology used in the APTIMA PCA3 assay is more quantitative and user-friendly than are the RNA extraction methods required for reverse transcription-PCR. Perhaps most importantly, the informative rate of 98.2% is significantly higher than those obtained with the other methods. This is a key point because patients must be called in for another sample if the first specimen is noninformative. The APTIMA PCA3 assay can be completed in <6 h with instrumentation currently used for several US Food and Drug Administration–cleared products. The robustness of PCA3 and PSA mRNA assays has been demonstrated across operators, reagent lots, and equipment/instrument sets and yielded interrun CVs of 20% or less for assay controls and processed urine specimens. The CV for PCA3/PSA ratio in specimen pools was 24%.
In summary, a relatively simple specimen-processing procedure and a robust assay that can be run on an existing platform have been developed. Clinical specimen testing to date has yielded specificity greater than that for the serum PSA assay. Studies are ongoing to evaluate larger and more defined populations and to compare the performance of the PCA3 molecular urine test with established prostate cancer markers, including total serum PSA and free PSA. The prototype APTIMA PCA3 assay yielded encouraging results in this study and warrants further evaluation as an aid in the diagnosis of prostate cancer.
|
Footnotes |
|---|
2 Human genes: PCA3, prostate cancer gene 3; KLK3, kallikrein 3 (prostate-specific antigen).
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
4.0 ng per milliliter. N Engl J Med 2004;350:2239-2246.The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  B. Laxman, D. S. Morris, J. Yu, J. Siddiqui, J. Cao, R. Mehra, R. J. Lonigro, A. Tsodikov, J. T. Wei, S. A. Tomlins, et al. A First-Generation Multiplex Biomarker Analysis of Urine for the Early Detection of Prostate Cancer Cancer Res., February 1, 2008; 68(3): 645 - 649. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Hanke, I. Kausch, G. Dahmen, D. Jocham, and J. M. Warnecke Detailed Technical Analysis of Urine RNA-Based Tumor Diagnostics Reveals ETS2/Urokinase Plasminogen Activator to Be a Novel Marker for Bladder Cancer Clin. Chem., December 1, 2007; 53(12): 2070 - 2077. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P.M.Q. van Gils, D. Hessels, O. van Hooij, S. A. Jannink, W. P. Peelen, S. L.J. Hanssen, J. A. Witjes, E. B. Cornel, H. F.M. Karthaus, G. A.H.J. Smits, et al. The Time-Resolved Fluorescence-Based PCA3 Test on Urinary Sediments after Digital Rectal Examination; a Dutch Multicenter Validation of the Diagnostic Performance Clin. Cancer Res., February 1, 2007; 13(3): 939 - 943. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
