SEARCH

SEARCH BY CITATION

Keywords:

  • prostate-specific antigen;
  • quantitative RT-PCR;
  • prostate cancer;
  • circulating cells;
  • BPH

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENT, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

OBJECTIVE

To assess the presence of circulating prostate-specific antigen (PSA)-expressing cells in patients with prostate cancer or benign prostatic hyperplasia (BPH), and to determine their diagnostic usefulness using a highly sensitive quantitative real-time reverse-transcription polymerase chain reaction (qRT-PCR) method.

PATIENTS, SUBJECTS AND METHODS

Venous blood samples were obtained from 175 patients with prostate cancer (12 metastatic and 163 not metastatic), 49 with BPH, and 50 healthy volunteers. To improve the specificity and sensitivity of the qRT-PCR three innovative features were combined; a primer overlapping two adjacent exons to inhibit nonspecific amplification; a no-end-point first round amplification to increase the sensitivity; and a target-specific primer for the RT phase to increase the specificity.

RESULTS

The sensitivity of the method was 1 cell/mL of blood and the interassay coefficient of variation was 10.5%. None of the healthy subjects tested positively, while 9% of those with prostatic cancer and 14% with BPH had PSA-positive cells in the blood. There was a positive association between a positive test and the National Comprehensive Cancer Network classification in the patients with newly diagnosed prostate cancer (P = 0.022). There were no additional statistically significant associations.

CONCLUSION

Our results strongly indicate that although there were no false-positive results and the sensitivity of the method was increased to maximal levels, a low frequency of positive results in patients with prostatic cancer and a high frequency of positive results in those with BPH seems to discourage the use of PSA-positive circulating cells in the search for a clinical diagnosis of prostate cancer.


Abbreviations
(q)RT-PCR

(quantitative) real-time reverse transcription PCR

CK

cytokeratin

S-PCR

single-round qRT-PCR

20 N-PCR

nested qRT-PCR with 20 first-round cycles

PBMC

peripheral blood mononuclear cells

CV

coefficient of variation

NCCN

National Comprehensive Cancer Network

RFU

relative fluorescence unit

PSA+

PSA-positive.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENT, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Since the introduction of PSA for detecting prostate cancer, this marker has rapidly become the mainstay of diagnosis and follow-up. Recently, a need to redefine the role of PSA in detecting prostate cancer has been indicated; in particular, a great debate has developed around the adequacy of current thresholds due to the lack of sensitivity and the risk of missing the diagnosis of early organ-confined prostate cancer [1]. Although previous reports clearly showed that serum PSA level is directly proportional to clinical stages, and that PSA is an essential component of nomograms for predicting clinical and pathological stages, some studies show that PSA is not related to prostate cancer [2]. Many authors have promoted research into markers for prostate cancer to increase the sensitivity and specificity of PSA or to find a substitute. Since 1990, reverse transcription (RT)-PCR techniques have been widely used in research and clinical practice applied to the diagnosis and staging of prostate cancer, both on blood and bone marrow samples, but the large variety of analytical protocols and the divergence of results still does not allow a clear definition of the clinical value of this technology [3–5]. To increase the sensitivity and specificity, nested RT-PCR was also applied but this approach has methodological drawbacks that can easily cause false-positive results [6–9]. Moreover, the positivity range of the RT-PCR PSA test was very variable both in patients with metastatic disease (28–88%) and with localized disease (0–62%), and test positivity was not statistically correlated with pathological stage [5,10–12]. To improve the RT-PCR technique, quantitative RT-PCR (qRT-PCR) that uses specific fluorescent probes [13] was suggested as the most sensitive method to identify and detect minimal numbers of cancer cells in peripheral blood. In addition to enhanced sensitivity and exact quantification of the results, qRT-PCR also greatly reduces the risks inherent in the manipulation of large amounts of amplified DNA. This last advantage should decrease the occurrence of false-positive results that often discourage the use of super-sensitive PCR methods. Different authors used this technology to detect PSA-positive (PSA+) cells [14–19] but the interpretation of results remains uncertain because of the variability of detected cell number [17] and, above all, the presence of apparent false-positive results [14,15,17]. In our institution, the optimization of a new qRT-PCR method allowed the precise quantification of cytokeratin 20-positive (CK20+) cells in patients with colorectal and breast cancer, with no false-positive result [20]. This new method provides two novel features, a primer overlapping two adjacent exons to inhibit nonspecific amplifications and a no-end-point first-round amplification to increase the sensitivity. In the present report, we reassessed the role of qRT-PCR for detecting PSA+ cells in the blood, applying the new method to a large group of patients with prostate cancer or BPH, and healthy subjects.

PATIENT, SUBJECTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENT, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Between May 2003 and June 2005, venous blood samples were obtained from 175 patients with prostate cancer (12 metastatic and 163 not metastatic), 49 with BPH, and from 50 healthy volunteers. In the patients, blood specimens were collected ≥3 weeks after a DRE or prostate needle biopsy. Patients with BPH were diagnosed after a DRE, TRUS and PSA testing; none had any sign of cancer or prostatitis, and all subsequently had a TURP or transvesical prostatectomy, with histopathological confirmation of the diagnosis of BPH. Patients were informed about the study and had given informed consent. The 50 healthy volunteers (aged <40 years) served to check the specificity of the method, assuming that the underlying incidence of prostate cancer was very low. All patients with metastatic disease provided more samples, to verify the reliability of the results obtained on the first sample collected. In eight of these patients systemic blood was sampled three times at 5-min intervals.

For cDNA preparation, 10 mL of blood were collected in sodium-citrated tubes and used exclusively within 3 h. Peripheral blood mononuclear cells (PBMC) were separated using Ficoll-density separation (GE Healthcare, Milwaukee, WI, USA). The human prostate carcinoma cell line LNCaP (DSMZ German Collection, Braunschweig, Germany) were maintained in continuous culture in RPMI 1640 (Invitrogen, Paisley, UK) with 10% fetal bovine serum. RNA was extracted from 1 × 106 LNCaP and 3 × 106 PBMC as already described [20]. cDNA was prepared adding 30 µL of extracted RNA to 48 µL reaction mixture containing 600 nm target-specific primer E4 and 600 U M-MLV RT (Invitrogen).

For qRT-PCR, different primer pairs (Invitrogen) were used for PSA amplification (Table 1). Oligonucleotides were identified and designed as already described [20]. Serial dilutions (from 1:10 to 1:106) of cDNA from 1 × 106 LNCaP were used to generate a standard curve in each experiment. Each standard cDNA was obtained normalizing the new preparation with reference to LNCaP cDNA; LNCaP were always harvested at confluence.

Table 1.  Primer pair amplicons analysed by qRT-PCR for PSA gene (GenBank accession no. X05332)
Sequence of selected primer pairs*Length of amplicon, bpMeasured melting temperature of amplicon, °C
  • *

    Primers E3 and I7 are forward; E4 and I8 are reverse. All sequences are written 5′–3′.

Primer E3; AGTCTGCGGCGGTGTTCTGG, exon 2Primers E3-E4, 596 
Primer E4; TCGGGCAGGGCACATGGTTC, exon 5
Primer I7; TGCCCGCTGCATCAGGAACAAAA, exon 2/3Primers I7-I8, 12189.5 °C
Primer I8; CATATCGTAGAGCGGGTGTGG, exon 3

Single-round qRT-PCR (S-PCR; 40 cycles, 94 °C, 30 s; 67 °C, 30 s; 72 °C, 30 s) of PSA was carried out in 50 µL of reaction mixture on the iCycler instrument (Bio-Rad, Hercules, CA, USA) using serial dilutions of LNCaP cDNA. Then 10 µL of template were added to the amplification mixture containing 150 nm primers I7 and I8, 2.5 U platinum Taq DNA-polymerase (Invitrogen), 3 mm MgCl2, 200 µm dNTP mixture (Applied Biosystems, Foster City, CA, USA) and 3.25 µL SYBR-Green (Sigma Chem Co., St Louis, MO, USA) diluted 1:10 000.

Nested qRT-PCR with 20 first-round cycles (20 N-PCR) first-round amplification (20 cycles, 94 °C, 30 s; 65 °C, 30 s; 72 °C, 30 s) was carried out in 50 µL of reaction mixture; 10 µL of cDNA obtained from PBMC or LNCaP were added to the amplification mixture containing 150 nm primer E3 and 50 nm E4, 2.5 U platinum Taq DNA Polymerase, 3 mm MgCl2 and 200 µm dNTP mixture. Then 1 µL was re-amplified (40 cycles, 94 °C, 30 s; 67 °C, 30 s; 72 °C, 30 s) in 50 µL of the same reaction mixture using 200 nm primer pair I7 and I8, and 3.25 µL SYBR-Green diluted 1:10 000. A standard curve with four dilutions of LNCaP cDNA was included in each PCR to quantify the number of PSA cell-equivalents in clinical samples. Samples were analysed in triplicate. To assign the specificity of the PCR, the amplicon melting temperature was measured. Non-specific amplifications were never detected. PCR products from LNCaP and PSA+ cells of patients with prostate cancer were directly sequenced as already described [20] using primers I7 and I8. All samples originating a detectable amplification signal were considered positive. To avoid contamination, the precautions included separate rooms and laboratory accessories for blood sampling, RNA isolation, first- and second-round PCR.

For ‘spiking’ experiments, varying numbers of LNCaP (from 1 to 104 cells) were added to 3 mL of blood and separated with PBMC using Ficoll-density separation or immunocapture with BerEP4-coated magnetic beads (Dynal-Biotech, NY, USA). RNA was purified as described above or in combination with a DNAse treatment (Ambion, Austin, TX, USA).

For the statistical analysis, patients were classified as newly diagnosed with prostate cancer (group A), already treated (group B), and those with BPH. The sample size of the study was constrained by the possibility of recruiting patients accessing the urology department, and for whom blood samples and the necessary clinical information were available. However, based on the assumptions of: (i) a two-sided type I error rate of ≤0.05; and (ii) an 80% power to detect a difference in the mean number of PSA+ cells of ≈1.5 between group A and BPH, 1.0 between group B and BPH, and 1.0 between groups A + B and BPH, we estimated that sample sizes of 48 patients with BPH, 44 in group A and 96 in group B were sufficient.

We tested the association between the detection of PSA+ cells in blood samples and PSA serum level at diagnosis, PSA serum level at the moment of sample collection for patients in group B, and biopsy Gleason score. All patients were graded for clinical stage, after a physical examination and imaging findings, and stratified according to recurrence risk as defined by National Comprehensive Cancer Network (NCCN) ‘Practice Guidelines in Oncology’ for prostate cancer (2005; Table 2). Patients in group B, who were PCR-tested during the follow-up after treatments, were furthermore classified in three categories as shown in Table 2. The association between a qRT-PCR-positive test and the above variables was tested using Fisher’s exact test. Given that the PSA cell-equivalents in blood samples were not normally distributed, the Kruskal–Wallis one-way anova by ranks was used for testing the equality of population medians among the groups. To assess if the method was capable of distinguishing among patients with BPH or prostate cancer, the probability distributions of the number of PSA cell-equivalents were compared using the Wilcoxon-Mann–Whitney rank sum test (BPH vs A, BPH vs B, and BPH vs A + B). The mean number of PSA cell-equivalents was then log-transformed to approach normality, and its association to the above variables compared using univariate linear regression separately in the three groups. The effects of the considered clinical variables are shown as differences of means with 95% CIs.

Table 2.  Characteristics of patients with prostate cancer
CharacteristicAll patientsPSA+
Group AGroup BTotalGroup AGroup BTotal
  • *

    1, low risk (T1-T2a and Gleason score 2-6 and PSA <10 ng/mL); 2, intermediate risk (T2b-T2c or Gleason score 7 or PSA 10-20 ng/mL); 3, high risk (T3a or Gleason score 8-10 or PSA >20 ng/mL); 4, very high risk (T3b-T4, or M+ or N+).

  • †with no clinical or radiological sign of disease.

n or n (%)59 1161757 (12)8 (7)15 (9)
Mean (range) n PSA cell-equivalents, qRT-PCR   1.61 (0.04–8.4)4.25 (0.32–15.18) 3.11 (0.04–20.72)
NCCN class, n (%)*
 122 (37) 33 (28) 55 (31)2 (9)1 (3) 3 (5)
 219 (32) 47 (41) 66 (38)1 (5)3 (6) 4 (6)
 313 (22) 26 (22) 39 (22)1 (8)2 (8) 3 (8)
 4 5 (8) 10 (9) 15 (9)3 (60)2 (20) 5 (33)
 Total59 (100) 116 (100)175 (100)7 (12)8 (7)15 (9)
Current stage of the disease, n (%)
 No evidence of disease 81 (70) 81 (70)5 (6) 5 (6)
 Biochemical relapse 25 (22) 25 (22)1 (4) 1 (4)
 Metastatic  9 (8)  9 (8)2 (22) 2 (22)
 Not available  1 (1)  1 (1)
 Total 116 (100) 116 (100)8 (7) 8 (7)

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENT, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Because of its higher sensitivity and of the feasibility of confirming the obtained results by the melting curve, SYBR-Green was used to monitor qRT-PCR reactions instead of the Taqman probe [20]. S-PCR and 20 N-PCR were performed using serial dilution of cDNA from LNCaP. The number of first- round cycles for 20 N-PCR was chosen to reach the highest level of sensitivity while conserving quantitative characteristics [20]. The sensitivity of 20 N-PCR (Fig. 1B) was clearly higher than that of S-PCR (Fig. 1A). Indeed, the amplification curve started ≈10 cycles earlier for 20 N-PCR than for S-PCR (increase in sensitivity ≈1000 times), and the detection limit shifted from 10 to one cell, reaching the maximum theoretical sensitivity (see below). Also, the efficiency of the reaction was increased in 20 N-PCR, with an efficiency of 103.8(4.2)%, giving a slope of −3.24 (0.09), an intercept of 22.59 (2.38), and a correlation coefficient of 0.993 (0.005) in eight independent test runs, while the efficiency of S-PCR was 89.3 (3.94)%, with a slope of −3.61 (0.12), intercept 35.70 (0.56) and correlation coefficient of 0.993 (0.006) in three independent test runs. There was a further increase in sensitivity during the RNA transcription step; the detection limit of 20 N-PCR using target-specific primer sequence E4 (Fig. 2B) was at least 10 times lower than that obtained using oligo-dT (Fig. 2A). Using LNCaP and primer I7 designed to overlap two adjacent exons (Table 1) [20] a single peak, corresponding to the specific Tm, indicated absolute specificity of primers I7 and I8. The observed Tm was 89.5 (0.13) °C in eight independent test runs. Amplification specificity was confirmed by sequencing the PCR products. 20 N-PCR analysis on PBMC isolated from healthy donors and with no retro-transcription step constantly showed no amplification signals (not shown). The analytical reproducibility of 20 N-PCR was tested by varying LNCaP numbers, and PCR assays were performed on the same cDNA in eight triplicate experiments; the mean (sd) and coefficient of variation (CV) are shown in Table 3. Melting curve analysis was used on all PSA+ clinical samples (see below) and constantly showed one specific peak (Fig. 3).

image

Figure 1. Representative PSA amplification plots of serially diluted LNCaP cells. Serially diluted cDNA was obtained from 1 × 106 LNCaP cells (1, 1:103; 2, 1:104; 3, 1:105; 4, 1:106) and amplified by S-PCR (A) and 20 N-PCR (B) using SYBR Green intercalating dye. The optical system software generated a real-time amplification plot where the number of cycles was related to fluorescence expressed as relative fluorescence unit (RFU). The threshold cycle (Ct) reflect the cycle number at which the fluorescence generated within a reaction crosses the threshold line. The amplification curve started ≈10 cycles earlier for 20 N-PCR (Ct = 13) than for S-PCR (Ct = 23) (increase in sensitivity, ≈1000 times), and the detection limit shifted from 10 to one cell, reaching the maximal theoretical sensitivity.

Download figure to PowerPoint

image

Figure 2. Representative PSA amplification plots of serially diluted LNCaP cells using two different cDNA synthesis methods. Serially diluted cDNA was obtained from 1 × 106 LNCaP cells (1, 1:103; 2, 1:104; 3, 1:105; 4, 1:106) and amplified by 20 N-PCR using oligo-dT (A) or target-specific primer sequence E4 (B) and SYBR Green. Numbers of cycles were plotted against fluorescence expressed as the RFU. The detection limit of 20 N-PCR using target-specific primer (panel B number 4) was at least 10 times lower than the detection limit obtained using oligo-dT (panel A number 3).

Download figure to PowerPoint

Table 3.  Analytical variance for varying numbers of LNCaP cells
No. of LNCaP cellsMean (sd) no. of threshold cycles CV
  1. Nested qRT-PCR from 20 N-PCR assays on the same cDNA in eight triplicate experiments.

1000 11.5 (0.61)0.053
10014.7 (0.61)0.042
1018.1 (0.88)0.049
120.9 (0.92)0.044
image

Figure 3. Representative melting-peak analysis of 20 N-PCR products. Curves were obtained with pure LNCaP cells (A) and clinical samples (B). The melting curve was generated by slowly heating the amplicon heteroduplex and measuring the changes in fluorescence that result when the SYBR Green melts away from the amplicon. The analysis was done after amplification using a temperature ramp of 0.5 °C/10 s between 55 and 95 °C. The melting point at 89.5 °C and the absence of nonspecific peaks indicate the amplification specificity.

Download figure to PowerPoint

To mimic a diagnostic situation of metastatic cells in clinical samples, serial dilutions of LNCaP were spiked in 3 mL of blood obtained from healthy subjects. The detection limit of 20 N-PCR was one LNCaP cell, in pure LNCaP cells and spiked cells, indicating no interference of a large excess of blood. DNAse treatment or immunocapture of epithelial cells did not decrease the detection limit of 20 N-PCR, as already reported [21] (data not shown). S-PCR was at least 10 (in pure LNCaP cells) and 100 (in spiked cells) times less sensitive than 20 N-PCR. Blood containing a few spiked LNCaP was also used to assess the reproducibility of the used RNA extraction method. RNA was extracted from 3 mL of blood containing 10 LNCaP in three independent experiments, and LNCaP were quantified against a standard curve as described; the calculated interassay CV was 10.5%.

In all, 175 patients with cancer, 49 with BPH and 50 healthy subjects were assayed for PSA+ cells by qRT-PCR. All the healthy subjects and 42 of 49 of patients with BPH showed completely negative amplification plots. Conversely PSA+ samples were found in 15 patients with cancer (four metastatic) (Table 2) and seven with BPH. The results were confirmed in duplicate experiments. In metastatic patients the results were identical if three separate samples were drawn at 5-min intervals from three PSA+ and five PSA-negative patients, and if multiple samples were analysed. Analysed blood samples were collected in a period of 3 weeks to 6 months from transrectal procedures; there was no correlation between the time elapsed from the transrectal procedure and PSA positivity in patients with cancer or BPH. That there was no correlation might indirectly indicate that, during the chosen period, PSA positivity is not influenced by residual circulating cells released by prostatic interventions.

Healthy subjects were not included in further statistical analysis because their age (<40 years) was not comparable with the age range of the patients (49–87 years). Table 4 shows that there were no statistically significant associations between qRT-PCR positive tests and the clinical variables considered in any group. The only positive association was with NCCN in group A (P = 0.022). The only other apparent significant association (with PSA serum level at diagnosis) in group A (P = 0.047) was in the opposite direction to that expected, with a higher probability of testing positive for lower PSA serum levels.

Table 4.  Association between a qRT-PCR positive test and clinical variables (categorical) by patient group
Clinical variables vs PCR outcomeP, Fisher’s exact test
BPHGroup AGroup B
  1. Classes of the categorical variables: *PSA serum level, ng/mL, at diagnosis, ≤2, 2–4, 4–20 and ≥20, and †at blood sampling for group B, <0.5 and ≥0.5; ‡≤6 and >6; §no evidence of disease, biochemical relapse, with no clinical or radiological sign of disease; or metastatic.

PSA serum level:
 At diagnosis*0.9380.0470.843
 At blood sampling0.254
Gleason score0.4270.105
NCCN class0.0220.301
Current stage of the disease§0.239

The study (49 men with BPH, 59 in group A and 116 in group B) was under-powered to detect differences (δ) in the mean number of PSA+ cells (A vs BPH, δ = 0.02, power 0.05; B vs BPH, δ = 0.41, power 0.28; A + B vs BPH, δ = 0.28, power 0.15). Moreover, the distribution of the number of PSA+ cells was highly skewed, with more zero counts (>90% in the whole sample), thus making the initial power calculation unsuitable for detecting quantitative (PSA+ cell number) differences between groups. A statistically significant difference in the number of detected PSA+ cells among the three groups was also not supported by the nonparametric tests used (Kruskal–Wallis, P = 0.265). The PCR test was unable to discriminate between patients with cancer and BPH (Wilcoxon-Mann–Whitney rank sum test, BPH vs A + B, P = 0.247; BPH vs A, P = 0.788; BPH vs B, P = 0.126). The effect of PSA serum level at diagnosis, biopsy Gleason score and NCCN class on the mean number of PSA+ cells was not statistically significant in any group (Table 5). The statistical test for effect modification of the group was also not statistically significant.

Table 5.  Effect of clinical variables on the (log-transformed) mean number of detected LNCaP cell-equivalents by patient group
VariableDifference (95% CI) in (log-transformed) mean n of PSA cell-equivalentsP*
BPHGroup AGroup B
  • *

    Test for effect modification of the patient group.

  • †Effect per 1 unit of PSA;

  • ‡Effect per 1 unit of Gleason score.

PSA serum level at diagnosis, ng/mL0.662 (−0.119; 1.443)−0.005 (−0.017; 0.008)0.022 (−0.013; 0.056)0.096
Gleason score−0.005 (−0.782; 0.772)0.159 (−0.602; 0.919)0.747
NCCN class
 2 vs 1−0.755 (−5.123; 3.612)−0.250 (−4.882; 4.382)0.376
 3 vs 11.213 (−3.419; 5.845)−0.014 (−4.646; 4.619) 
 4 vs 1−1.730 (−6.362; 2.902)1.328 (−2.125; 4.781) 

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENT, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

RT-PCR has often been used to detect circulating prostate cancer cells in blood, but unfortunately, non-quantitative methods have yielded controversial results, and to date they have not been introduced into clinical practice. Technical drawbacks such as the lack of quantification, low reproducibility and uncertain sensitivity were indicated as the major reasons for the variable results. Renewed interest in PSA RT-PCR tests has arisen from the development of qRT-PCR techniques that allow precise quantification of mRNA levels, and a higher and well-defined sensitivity. Since 2001 a few reports have been published [14–19]; using qRT-PCR, low-grade amplification was detected in control subjects, as described in four reports, and threshold values were assigned to discriminate false-positive results [14–17]. No healthy subjects were analysed in two additional reports [18,19]. Unfortunately, the data presented in those reports were conflicting; the association between the results of the PCR test and clinical and pathological staging was variable [15–19]. Such variability could be reasonably due to the requirement to establish a threshold value to discriminate between negative and positive results. Considering those results, we were primarily interested in assessing a quantitative method specifically developed to provide a very high sensitivity (nested PCR, fluorescence detection) in the absence of any signal in healthy subjects.

A new recently described qRT-PCR method was shown to provide a higher specificity and sensitivity using a new primer design strategy and a pre-amplification step. With this technique we were able to completely abrogate nonspecific amplifications in healthy subjects and to detect the presence of a variable number of CK20+ cells in patients with breast or colon cancer [20]. Using the same approach, we optimized a new quantitative method that provided complete lack of amplification in healthy subjects and very high sensitivity (one cell/mL of blood) for detecting PSA-expressing cells in patients with prostate cancer. It appeared that increasing the specificity of the method caused no loss of sensitivity, because up to 60 (20 + 40) amplification cycles could be used, reaching the maximum theoretical sensitivity (one cell), without detecting nonspecific signals in samples from healthy subjects.

There was a positive association between qRT-PCR-positive tests and the NCCN classification in patients newly diagnosed with prostate cancer (P = 0.022). However, the frequency of PSA+ cells in patients with cancer was low (9%) and a significant percentage of those with BPH (14%) also had positive results. There was a low frequency of positive results also in severe disease, i.e. four of 12 metastatic patients were PSA+ and only one of 25 with biochemical relapse was PSA+.

Our results strongly indicate that cells expressing PSA are not present or detectable in the blood of healthy subjects even using a maximally sensitive assay. In addition to the analytical confirmation, the sensitivity of the method was sustained by the significant frequency of positive results in patients with BPH, where many circulating PSA+ cells could not be expected. Notably, after more than a year from the PSA test, no sign of malignant pathology was detected in patients with BPH, but although they will be constantly and carefully followed, we cannot exclude the presence of an initial malignant progression of the disease in those with BPH and PSA+ cells. The capability of the test to detect cancer cells is reflected by the observed association between qRT-PCR positivity and cancer progression. However, the low frequency of positive results even in advanced cancer, and positivity in BPH, indicate the unfeasibility of using PSA+ cells in blood for a general diagnostic application, unless pre-analytical enrichment techniques that allow the analysis of a larger volume of blood are applied [22]. Automation of the techniques of RNA extraction and further automation of PCR would increase the sensitivity and reproducibility. However, the choice and the adjustments (sensitivity and specificity) of the methods used could strongly influence the clinical results, and only the comparing the performance obtained by different methods can assist in this choice.

In conclusion, we cannot exclude that cancer cells were present in the circulation of more patients with prostate cancer, that the results could be affected by the low performance of PSA expression as a marker of circulating cancer cells, or due to the need to analyse larger blood volume. In view of its significant implications, research aimed at detecting circulating cancer cells is very active and is now exploiting new approaches such as surface proteomics and nanotechnologies to develop new diagnostic tools [22,23]. It is very likely that the present method described here will greatly improve its diagnostic value when coupled with new pre-analytical micromethods able to capture and enrich circulating metastatic cells [22].

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENT, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

We thank Dr Lisa Bonello for DNA sequencing and staff of AVIS, Blood Bank (Torino, Italy) for providing human blood from the healthy subjects. This study was supported by the ‘Oncology Special Project’, Compagnia di San Paolo/FIRMS and by the ‘Ricerca sanitaria finalizzata’, Regione Piemonte (Italy).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENT, SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES
  • 1
    Shah JB, Reese AC, McKiernan JM, Benson MC. PSA updated: still relevant in the new millennium? Eur Urol 2005; 47: 42732
  • 2
    Stamey TA, Caldwell M, McNeal JE, Nolley R, Hemenez M, Downs J. The prostate specific antigen era in the United States is over for prostate cancer: what happened in the last 20 years? J Urol 2004; 172: 1297301
  • 3
    Gomella LG, Raj GV, Moreno JG. Reverse transcriptase polymerase chain reaction for prostate specific antigen in the management of prostate cancer. J Urol 1997; 158: 32637
  • 4
    Schamhart DH, Maiazza R, Kurth KH. Identification of circulating prostate cancer cells: a challenge to the clinical implementation of molecular biology (review). Int J Oncol 2005; 26: 56577
  • 5
    Zach O, Lutz D. Tumor cell detection in peripheral blood and bone marrow. Curr Opin Oncol 2006; 18: 4856
  • 6
    Hara N, Kasahara T, Kawasaki T et al. Reverse transcription-polymerase chain reaction detection of prostate-specific antigen, prostate-specific membrane antigen, and prostate stem cell antigen in one milliliter of peripheral blood: value for the staging of prostate cancer. Clin Cancer Res 2002; 8: 17949
  • 7
    Llanes L, Ferruelo A, Paez A, Gomez JM, Moreno A, Berenguer A. The clinical utility of the prostate specific membrane antigen reverse-transcription/polymerase chain reaction to detect circulating prostate cells: an analysis in healthy men and women. BJU Int 2002; 89: 8825
  • 8
    Smith MR, Biggar S, Hussain M. Prostate-specific antigen messenger RNA is expressed in non-prostate cells: implications for detection of micrometastases. Cancer Res 1995; 55: 26404
  • 9
    Verhagen PC, Tilanus MG, De Weger RA, Van Moorselaar RJ, Van Den Tweel JG, Boon TA. Prognostic factors in localised prostate cancer with emphasis on the application of molecular techniques. Eur Urol 2002; 41: 36371
  • 10
    Ellis WJ, Vessella RL, Corey E et al. The value of a reverse transcriptase polymerase chain reaction assay in preoperative staging and followup of patients with prostate cancer. J Urol 1998; 159: 11348
  • 11
    Morgan TM, Lange PH, Vessella RL. Detection and characterization of circulating and disseminated prostate cancer cells. Front Biosci 2007; 112: 30009
  • 12
    Martinez-Pineiro L, Rios E, Martinez-Gomariz M et al. Molecular staging of prostatic cancer with RT-PCR assay for prostate-specific antigen in peripheral blood and lymph nodes: comparison with standard histological staging and immunohistochemical assessment of occult regional lymph node metastases. Eur Urol 2003; 43: 34250
  • 13
    Bustin SA, Mueller R. Real-time reverse transcription PCR (qRT-PCR) and its potential use in clinical diagnosis. Clin Sci (Lond) 2005; 109: 36579
  • 14
    Patel K, Whelan PJ, Prescott S et al. The use of real-time reverse transcription-PCR for prostate-specific antigen mRNA to discriminate between blood samples from healthy volunteers and from patients with metastatic prostate cancer. Clin Cancer Res 2004; 10: 75119
  • 15
    Zambon CF, Basso D, Prayer-Galetti T et al. Quantitative PSA mRNA determination in blood: a biochemical tool for scoring localized prostate cancer. Clin Biochem 2006; 39: 3338
  • 16
    Llanes L, Ferruelo A, Lujan M, Pascual C, Garcia-Mediero JM, Berenguer A. Quantitative real-time reverse transcription: polymerase chain reaction of prostate-specific antigen (PSA) for detection of circulating prostatic cells in patients with clinically localized prostate cancer. Prostate Cancer Prostatic Dis 2005; 8: 24852
  • 17
    Gelmini S, Tricarico C, Vona G et al. Real-Time quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) for the measurement of prostate-specific antigen mRNA in the peripheral blood of patients with prostate carcinoma using the taqman detection system. Clin Chem Laboratory Med 2001; 39: 38591
  • 18
    Straub B, Muller M, Krause H, Schrader M, Miller K. Quantitative real-time rt-PCR for detection of circulating prostate-specific antigen mRNA using sequence-specific oligonucleotide hybridization probes in prostate cancer patients. Oncology 2003; 65 (Suppl. 1): 127
  • 19
    Straub B, Muller M, Krause H et al. Detection of prostate-specific antigen RNA before and after radical retropubic prostatectomy and transurethral resection of the prostate using ‘Light-Cycler’-based quantitative real-time polymerase chain reaction. Urology 2001; 58: 81520
  • 20
    Giribaldi G, Procida S, Ulliers D et al. Specific detection of cytokeratin 20-positive cells in blood of colorectal and breast cancer patients by a high sensitivity real-time reverse transcriptase-polymerase chain reaction method. J Mol Diagn 2006; 8: 10512
  • 21
    Mejean A, Vona G, Nalpas B et al. Detection of circulating prostate derived cells in patients with prostate adenocarcinoma is an independent risk factor for tumor recurrence. J Urol 2000; 163: 20229
  • 22
    Nagrath S, Sequist LV, Maheswaran S et al. Isolation of rare circulating tumour cells in cancer patients by microchip technology. Nature 2007; 450: 12359
  • 23
    Cohen SJ, Alpaugh RK, Gross S et al. Isolation and characterization of circulating tumor cells in patients with metastatic colorectal cancer. Clin Colorectal Cancer 2006; 6: 12532