The Prostate 74:1043^1051 (2014)
Telomerase as aTumor Marker in Diagnosis of Prostatic Intraepithelial Neoplasia and Prostate Cancer P.V. Glybochko,1 E.G. Zezerov,2 A.I. Glukhov,2 Yu.G. Alyaev,1 S.E. Severin,2 K.A. Polyakovsky,3 V.A. Varshavsky,4 E.S. Severin,2 and A.Z. Vinarov1 1
Department of Urology, First Sechenov Moscow State Medical University, Moscow, Russia Department of Biochemistry, First Sechenov Moscow State Medical University, Moscow, Russia 3 Urology Clinic, First Sechenov Moscow State Medical University, Moscow, Russia 4 Department of Anatomic Pathology, First Sechenov Moscow State Medical University, Moscow, Russia 2
BACKGROUND. Early diagnosis of prostate cancer (CaP) can be addressed by studying prostatic intraepithelial neoplasia (PIN) as precancer (high-grade PIN or HGPIN). This article attempts to analyze the diagnostic role of telomerase as an early marker of carcinogenesis. METHODS. Complex urological patient evaluation and assessment of telomerase activity. RESULTS. Out of 92 patients 44% were diagnosed with CaP, 49% with low-grade PIN (LGPIN) in association with benign prostatic hyperplasia (BPH), and 7% with HGPIN in association with BPH. Active telomerase (AT) in prostate biopsy specimens was detected in 98% of patients with CaP, in 33% of patients with HGPIN, and in 20% of patients with LGPIN. In the event of simultaneous detection of AT and PIN in initial prostate biopsy specimens, further monitoring for 0.5–4.0 years revealed CaP development in 50–56% of cases. Further follow-up of patients with PIN and absent telomerase activity in initial biopsy specimens did not demonstrate the development of CaP. The PSA level was significantly higher in patients with active telomerase in the prostate tissue than in telomerase negative patients. CONCLUSIONS. Telomerase activity in the prostate tissue increases the risk of CaP development in patients with PIN. Detection of telomerase activity in prostate biopsy specimens from patients with PIN enables selection of a group of patients with high risk of CaP development and reduction of the number of prostate biopsies performed in other patients. Prostate 74:1043–1051, 2014. # 2014 Wiley Periodicals, Inc. KEY WORDS: prostate cancer; precancer; prostatic intraepithelial neoplasia; prostatespecific antigen; telomerase; dynamic process of prostate carcinogenesis
INTRODUCTION Over the course of our long-term and diverse research of prostate cancer pathology [1–4], about 10 years ago [4–8] our attention was drawn to telomerase, an enzyme, which is a marker of intensive and unlimited cell proliferation, including cancerous and precancerous cells in different kinds of oncopathology. Telomerase takes part in repair of telomeric DNA shortened during replication and increases the number of cell divisions thus allowing the cells to bypass the Hayflik limit of 35–50 divisions. As a result, the cells become conditionally immortal. Therefore, telomerase activity in the parenchyma of multiple organs is of diagnostic value as a marker of carcinogenesis and ß 2014 Wiley Periodicals, Inc.
even allows for detection of premalignant conditions [9]. Active telomerase (AT) is present in 47–92% of cases of prostate cancer (CaP) [10,11], in 16% of cases of high-grade prostatic intraepithelial neoplasia (HGPIN) [11,12], and is absent in patients with benign prostatic hyperplasia (BPH) [13]. As far back as 10 or more years ago, HGPIN was regarded as a morpho
Correspondence to: E.G. Zezerov, Department of Biochemistry, First Sechenov Moscow State Medical University, 8, Trubetskaya St, Building 2, Moscow, Russia. E-mail: [email protected] Received 28 December 2013; Accepted 17 April 2014 DOI 10.1002/pros.22823 Published online 22 May 2014 in Wiley Online Library (wileyonlinelibrary.com).
1044
Glybochko et al.
logical form of prostatic precancer [14]. These years of vast research involving both animal models and humans resulted in HGPIN being more confidently regarded as a precursor of prostate cancer [15]. The progression (usually a slow process) of HGPIN into CaP is studied using histological, biochemical and immunohistochemical techniques for the following purposes: to identify molecular mechanisms governing this transformation, predict the time of cancer onset, develop monitoring options and strategies aimed at delaying or stopping the process [16–21,22]. LGPIN is not classified as precancer [23]. However, works of other authors published over the last 5 years never consider the diagnostic value of comparison of telomerase activity in CaP, HGPIN, and LGPIN, especially during long-term follow-up of such categories of patients. For that reason we have been studying the role of telomerase in prostate gland diseases, specifically CaP [24,25], and also HGPIN and LGPIN since 2001. The present report discusses comparative results of AT assessment in different kinds of prostate gland oncopathology and the data of long-term (up to 4 years) monitoring for the purpose of estimation of diagnostic and/or prognostic value of AT. The article was made in the Department of Urology, Urology clinic and Department of Biochemistry of the Sechenov First Moscow State Medical University. MATERIALS AND METHODS The current work was based on the results of extensive examination of 92 urology clinic patients suspected of having CaP. The control group included eight patients with normal levels of prostate-specific antigen (PSA), with benign prostatic hyperplasia (BPH), and no AT in the tissue obtained during transurethral resection of the prostate (TURP). An additional control group included 20 younger males admitted to the Urology clinic for examination or treatment of other urological pathologies. These patients had no detected prostate pathology and normal PSA levels which will be given later. That is why prostate biopsy and—all the more so TURP with AT assessment—could not be performed in these 20 patents (unlike in eight aforementioned patients) due to ethical reasons as well as absence of medical indications. This approach is justified by a well-known fact that telomerase shows no enzyme activity in the cells of mature parenchymal organs that finished their ontogenesis and that show no signs of oncopathology [9]. Concrete evidence for this fact (using the prostate and kidneys as examples) was provided by works of several laboratories included in the review by Orlando et al. [11]. The Prostate
All patients underwent complete clinical examination and laboratory testing. The PSA of the blood serum was analyzed by means of immunochemiluminometric or enzyme-linked immunosorbent assays. The examination also included urinalysis, three-glass test, uroflowmetry, digital rectal examination (DRE), and transrectal ultrasound (TRUS) of the prostate followed by PSA density (PSAD) calculation. In some cases an additional examination—lesser pelvis magnetic resonance imaging (MRI)—was performed. Depending on the obtained results the patients underwent plurifocal transrectal ultrasound-guided needle prostate biopsy. Biopsies were taken using a “B&K 3535” (BK-Medical) diagnostic ultrasound system with a 6558/T/S bi-plane sector mechanical rectal probe operating at a frequency of 7.5 MHz. The material was obtained using transrectal needle prostate biopsies from standard sites and also sites where CaP foci were suspected on ultrasound (TRUS) or magnetic resonance imaging scans (MRI). Several (10–12) biopsy specimens were taken for histological examination. At least one biopsy specimen from each lobe of the gland and sites suspicious of CaP were used for AT assessment. The following segment describes the technique in more detail. In order to minimize the possibility of biopsy specimens being taken from different sites for histological analysis and AT assessment, we used sites with lower echogenicity. Having located a hypoechogenic area, we took a biopsy specimen for histological examination and then—from the same site—for AT evaluation. Due to the fact that in most cases the size of the hypoechogenic areas did not exceed 0.5 cm, the difference in sites used for tissue sampling was minimal. If the gland had homogeneous echotexture (in the minority of patients), two obligatory biopsy specimens for AT evaluation were taken from the central part of the peripheral zone of each prostate gland lobe. After that, a control biopsy specimen was taken for histological analysis without changing the position of the ultrasonic probe. The majority of patients underwent rebiopsy, however in some cases tissues obtained during TURP or radical prostatectomy (RPE) were used instead of re-biopsy specimens. The histological examination of prostate biopsy specimens and tissues obtained during TURP or RPE included standard tissue processing (paraffin wax embedding), 4 mm section preparation followed by hematoxylin, eosin, and picrofuscin staining. In complicated cases an immunohistochemical analysis using PSA and p63 antibodies (Dako) was performed. p63 protein is a p53 protein homologue, and it is frequently used in difficult cases to distinguish between CaP and PIN [14]. Universal criteria and the Gleason grading
TelomeraseçATumor Marker system were used for CaP evaluation. Criteria for differentiation between CaP and BPH, chronic prostatitis (ChP), and PIN were also used [26–28]. We consider it necessary to emphasize that precise differentiation between HGPIN and LGPIN was conducted in accordance with histological criteria developed by Bostwick and Brawer (1987) modified by Joniau et al. [14] or other researchers [28]. For AT assessment, the studied tissue was placed into a plastic test-tube, flash frozen in liquid nitrogen and stored at 70°C. Tissue extracts used for AT evaluation were obtained by the technique described by us earlier [24]: 6 intermittent homogenization cycles (each 1 min long) at 0°C with glass powder in a buffer solution containing Tris–HCl (pH 7.5), CHAPS, b-mercaptoethanol, and PMSF. After lysate centrifugation (12,000g, þ4°C, 30 min) the supernatant was analyzed for AT or frozen in liquid nitrogen and stored at 70°C for further use. To assess telomerase activity we used our own modification of TRAP (Telomeric Repeat Amplification Protocol). Other laboratories usually utilize TRAP to detect amplified telomerase reaction products radiolabelled with phosphorus P32. The detection process involves autoradiography of gel electropherograms [9]. . Our novel technique employed in the present work is a non-isotope modification of TRAP with the use of a highly sensitive SYBR Gold dye. This modified method was described in 1998 [8] and successfully utilized in our research [24,25]. Its sensitivity corresponds to active telomerase levels contained in approximately 15–16 telomerase-positive cells [6].
1045
Separation of the reaction products (TRAPproducts) was achieved with 10% non-denaturing polyacrylamide gel electrophoresis using 1 Trisborate-EDTA (TBE) buffer. PCR samples were then mixed with Orange G leading dye and 10 ml were added to the gel wells. The gel was exposed to a highly sensitive SYBR Gold dye for 30 min at room temperature in the dark. Visualization was performed with a UV transilluminator (l ¼ 300 nm). The stained gels were photographed with a Kodak digital camera. The digital photos were evaluated by means of IMAGE J 1.351 image processing program (“National Institutes of Health”, USA) which calculates the peak areas (in standard units), representing the luminescence intensity of discrete fragments of TRAP-products in the gel. The total peak area was regarded as the AT value in the studied sample. Relative AT was then calculated, by comparing the AT value in the studied sample to the positive control sample, that is, telomerase-positive LNCaP cell line extract containing 0.7 mg of protein. Telomerase activity in the control sample was considered to be 1.0 or 100%. The examined sample was regarded telomerasepositive if three or more horizontal strips of TRAPproducts were present. The distance between them amounted to the length of one telomeric repeat (six pairs of nucleotides). In case of negative control, pure CHAPS lysis buffer was added to the sample instead of cell extract. The statistical analysis was performed according to statistical guidelines for biological studies accepted in the USSR and Russia [29–33].
Telomerase Activity Analysis Design The first step was obtaining an extract from prostate tissues or LNCaP cells. The latter represent a transplantable cell line derived from human lymph nodes affected by metastatic prostate cancer (positive control for telomerase). The TS-primer elongation and subsequent amplification was carried out in 50 ml of reaction mixture containing 67 mM Tris–HCl (pH 8.8), 16.6 mM (NH4)2SO4, 0.01% Tween-20, 1.5 mM MgCl2, 1 mM EGTA, 50 mM of every dNTP, 0.1 mg TS-primer (50 ATTCCGTCGAGCAGAGTT-30 ), 1–10 ml of tissue extract containing 0.7, 2, 4, 8 mg protein or 1 ml (0.7 mg protein) of the LNCaP cell extract diluted with CHAPS lysis buffer and equivalent to 1,000 LNCaP cells (positive control). During the polymerase chain reaction (PCR) 0.1 mg of CX-primer (50 -CCCTTACCCTTACCCTTACCCTAA-30 ) and 2.5 units of SmarTaq DNA polymerase (“Dialat”, Russia) were added to the reaction mixture. The TS-primer elongation stage was followed by reaction mixture amplification during 35 PCR cycles.
RESULTS Based on histological examination of primary biopsy specimens of 92 patients (aged 47–81 years), CaP was diagnosed in 41 (44%) subjects/patients, HGPIN—in 6 (7%), and LGPIN—in 45 (49%) patients (Table I). Of 51 patients with PIN, HGPIN was detected in 12% of cases, and LGPIN in 88% of cases. The latter diagnosis was almost always associated with BPH. About 72% of PIN patients were also diagnosed with ChP. As a rule, total PSA level in the blood serum of patients with CaP, PIN, BPH, and ChP was higher than 4 ng/ml, and varied greatly, which will be discussed later. Taking a look at test subjects with PIN as a separate group, the mean total PSA level in 49 of 51 patients with PIN was 15.3 ng/ml (4.2–94.1 ng/ml). Two patients had total PSA less than 4.0 ng/ml. The free PSA index exceeded the prognostically favorable threshold of 15% in 15 patients (29%). It was lower than 15% in other patients. The PSA density (PSAD) The Prostate
1046
Glybochko et al.
TABLE I. ActiveTelomerase (AT) Detection in Prostate Biopsy Specimens of Patients WithVarious Urologic Diseases AT detected Disease CaP HGPIN and BPH LGPIN and BPH
Total number of patients
Number of patients
Percentage and mean percentage error (X m)
95 % confidence interval
41 6 45
40 2 9
98 2 33 21 20 6
87–100 4–78 10–35
The statistical analysis is carried out based on the technique and Table I developed by VS Genes.[23]
was below the prognostically favorable threshold of 0.1 ng/ml/cm3 in only 6 (12%) patients. We conducted preliminary testing of the sensitivity of our combined analysis (AT assessment þ histological examination) on 41 patients with CaP. As mentioned in Materials and Methods section, 10–12 biopsy specimens from both lobes of the gland underwent histological examination. At least one separate biopsy specimen from each lobe and, as a rule, several more tissue samples from hypoechogenic areas suspected of being cancer were sent for AT assessment (with strict adherence to the technique described in Materials and Methods section which lowers the risk of false conclusions). In 14 patients cancer was diagnosed based on histological examination in only one lobe of the prostate with 36% of these test subjects having AT in the same lobe. About 57% of patients from the aforementioned 14 test subjects had AT in both lobes of the prostate including the parts where no morphological changes were detected. AT appears to precede diagnosed epithelial cell cancer based on histological examination if one ignores the technical limitations preventing simultaneous analysis of the same biopsy specimen in terms of both histology and telomerase activity. It can be assumed that biopsy specimens undergoing AT evaluation without simultaneous histological analysis due to technical limitations may contain cancer cells which in theory always possess AT. In this case if AT is detected in two or more samples only assessed for AT while histological examination reveals no cancer in other 10–12 specimens, the statement concerning possible telomerase activation in cells which happens prior to typical cancer formation (e.g., in PIN) might prove wrong. However, if we consider the possibility of mistakes, we should specify the probability of error which does not exceed 14–17% (2/12 or 2/14). In other 27 patients of the aforementioned 41, prostate cancer and telomerase were detected in both lobes of the prostate. After the biopsy five patients with cancer in one lobe and two patients with CaP detected in both lobes underwent radical prostatectomy. AT assessment and histological analysis of The Prostate
prostatectomy specimens essentially matched the results of the same examinations conducted on preoperative biopsy specimens as well as differentiation parameters of various lobes of the gland (stages T1c, T2a, and T3a). We believe that this data lowers the probability of the error discussed above. The results of the analysis of primary biopsy specimens of patients with PIN are shown in Table I, Figure 1. AT was detected in biopsy samples in 11 (21%) of 51 patients with PIN. Both AT and LGPIN were detected in 9 of 45 patients (Fig. 2), that is, 20 6% (percentage and mean error) with a 10–35% confidence interval at the 95% probability level (Table I). Similarly, both AT and HGPIN were detected in 2 of six patients with HGPIN only (Fig. 3). The percentage and confidence interval were 33 21% and 4–78%, respectively (Table I). Although AT detection rates in patients with HGPIN were higher than in patients with LGPIN, the difference was not statistically significant due to the given number of observations. It is worth mentioning that AT was detected in 40 (98%) of 41 patients with CaP confirmed by histological examination (87–100% confidence interval). The difference between AT detection rates in CaP and PIN groups was statistically significant (Table I).
Fig. 1. Results ofmonitoring ofpatients with ATand PIN oninitial prostate biopsy (initial prostate biopsy specimens were examined histologically with simultaneous ATassessment).CaP, prostate cancer; AT, active telomerase; PSA, prostate-specific antigen; TURP, transurethralresection of theprostate.
TelomeraseçATumor Marker
Fig. 2. Results of monitoring of patients with LGPIN diagnosed via histological examination (initial prostate biopsy specimens were examined histologically with simultaneous AT assessment). CaP, prostate cancer; AT, active telomerase; LGPIN, low-grade prostatic intraepithelial neoplasia; HGPIN, high-grade prostatic intraepithelialneoplasia.
Further PIN patient monitoring included the blood level of PSA, PSAD, digital rectal examination (DRE), transrectal ultrasound (TRUS), and lesser pelvis magnetic resonance imaging (MRI). Some patients had indications for one or more prostate re-biopsies or, in rare cases, for TURP. The materials obtained underwent histological examination and AT assessment. The results given in Figure 1 show that 6 (60%) of 10 patients with high levels of AT and PIN on initial
Fig. 3. Results of monitoring of patients with HGPIN diagnosed via histological examination (initial prostate biopsy specimens were examined histologically with simultaneous AT assessment). CaP, prostate cancer; AT, active telomerase; HGPIN, high-grade prostatic intraepithelialneoplasia.
1047
biopsy were later diagnosed with CaP. Patients with no AT diagnosed with LGPIN (36 subjects) or HGPIN (4 subjects) on initial biopsy did not develop CaP during 0.5–3.0 years of follow-up (Figs. 2 and 3). One case of a very low AT level in association with LGPIN did not show CaP during four years of follow-up (Figs. 1 and 2). Three patients with AT and PIN did not undergo rebiopsy for the following reasons: improvement of the condition, based on PSA, DRE, TRUS, and MRI data (1 patient); refusal from follow-up (1 patient); refusal from re-biopsy upon detection of PSA level increase (1 patient) (Fig. 1). Figures 2 and 3 show the re-examination results of patients with LGPIN and HGPIN in more detail and separately. In the period of 0.5–3.0 years, four patients initially diagnosed with LGPIN in association with AT developed CaP, and two patients were diagnosed with HGPIN. In one case of HGPIN, CaP was detected in 2 years (Fig. 2). Overall, during 0.5–4.0 years LGPIN associated with AT transformed into CaP in five (56%) of nine cases as opposed to LGPIN without AT. However, 3 (16%) of 19 patients with LGPIN and no AT showed the transition of LGPIN to HGPIN on rebiopsy in 0.5–3.0 years (Fig. 2). In such and similar cases the terms “transformation” or “transition” from one type of histological changes into another do not necessarily imply “direct transformation” of LGPIN into HGPIN, for example. Such changes in histological structure of biopsy samples might be due to statistically random distribution of sites of morphological changes in prostatic tissues (mainly in the peripheral zone) and partly due to statistically random sampling during plurifocal biopsy and re-biopsy. One patient with an initial combination of HGPIN and AT showed no progression into CaP during 1-year follow-up (Fig. 3). The second patient with initial combination of HGPIN and AT underwent three rebiopsies showing no progression of the disease. However, in 1 year the 4th biopsy revealed CaP. Therefore, in patients with HGPIN associated with AT the CaP rate was 50%. All four patients with HGPIN and no AT were not diagnosed with CaP during 1–2 years of follow-up (Fig. 3). Thus, the data obtained allows for the first conclusion to be made: without AT neither LGPIN nor HGPIN transforms into CaP, whereas LGPIN could progress into HGPIN in 16% of cases. Secondly, during 0.5–4.0 years LGPIN and HGPIN associated with AT have a 56% and a 50% rate of progression into CaP respectively. Although given the insufficient quantity of studied patients (only six test subjects) with HGPIN, such comparison is imperfect. Figure 2 shows that of 45 patients with LGPIN only four (9%) were diagnosed with CaP on re-biopsy during 0.5–3.0 years. The Prostate
1048
Glybochko et al.
In all these four patients LGPIN was associated with active telomerase in the prostate tissue. This fact might prove to be of interest to other researchers. Another interesting observation is that the fifth patient (LGPIN plus AT) was diagnosed with HGPIN in 2 years which subsequently progressed into CaP in further two years. Other patients with LGPIN on initial biopsy (regardless of AT presence) retained the LGPIN phenotype during follow-up or acquired the HGPIN phenotype (one patient, Fig. 2). AT assessment is of obvious diagnostic and prognostic value in patients with PIN for it allows for selection and thorough monitoring of individuals with high risk of CaP development. Absence of AT in biopsy samples indicates low probability of malignant transformation which means that no re-biopsies will be required in cases of improvement (based on PSA, DRE, TRUS,PSAD, and MRI data). In terms of pathogenesis, the results suggest that telomerase activity in prostate tissue plays a role as a factor of carcinogenesis. The analysis of the experimental data also allowed for the investigation of the relationship between telomerase in prostate tissue and blood level of PSA. Prostate-specific antigen (PSA) is not only an antigen used for specific analysis, but also an enzyme, chymotrypsin-like serine protease, which is a member of the human kallikrein family [3]. The second enzyme, telomerase, is a factor of intense and/or malignant cellular proliferation [9]. Both of these protein enzymes are produced by epithelial secretory cells of the prostate gland and play a role in oncogenesis [2,4,7]. PSA acts as a catalyst for proteolysis and disintegration of insulin-like growth factor-binding protein 3 (IGFBP3). Similar to telomerase, insulin-like growth factor 1 (IGF-1) accelerates proliferation of epithelial cells of the prostate [2–4]. Apart from the endogenous function of PSA in the gland itself, this enzyme enters the urethra through the prostatic ducts during ejaculation, mixes with seminal fluid, and catalyzes proteolysis of gel matrix forming proteins of the seminal vesicles (semenogelins, fibronectin), decreases the viscosity of semen and improves sperm motility [3]. Only one millionth of PSA enters the bloodstream from a healthy gland as a defective inactive enzyme due to its binding with inhibitors in the blood (a1–antichymotrypsin, a2-macroglobulin and others) and due to its lys145 breakdown. It is possible due to microdamage of the basal cell layer and basement membranes of the gland. In prostate pathology these lesions intensified simultaneously with an increase in PSA-producing cells. In HGPIN, the basal cell layer is disrupted or fragmented, while in CaP damage is inflicted to the basal layer and basement membranes [3,14,34]. All of this leads to higher PSA concentration in the blood The Prostate
TABLE II. Correlation Between AT in Prostate Gland Tissue and PSA Level in Blood Serum Total PSA, ng/ml Patients with CaP, PIN and BPH
Number of patients
Average value— median
95 % confidence interval
51 41
18 10
13–27 7–12
Telomerase detected Telomerase absent
which correlates to the increased quantity of secretory prostatic epitheliocytes and the degree of damage to the cells of the basal layer and basement membranes. Our summarized clinical and laboratory data is given in Tables II and III. Of 92 patients with prostate pathology (CaP, PIN, BPH) in our study, subjects with active telomerase in the prostate had almost twice as high mean PSA concentration (median) than those without AT. Given the sufficient number of patients in our study (Table II), the difference was statistically significant (the confidence intervals did not overlap, P < 0.05). The data corresponds to the role of telomerase and PSA as factors of prostate carcinogenesis. A similar analysis (Table III) of various kinds of pathology showed that total PSA level is higher in CaP than in PIN and BPH. The difference also was statistically significant (P < 0.05). Table IV contains similar data concerning total PSA concentration in the blood serum in other patients of the four different groups. PSA level was normal in control group 4 which consisted of 20 relatively younger males without signs of prostate pathology. It was statistically
Telomerase as aTumor Marker in Diagnosis of Prostatic Intraepithelial Neoplasia and Prostate Cancer P.V. Glybochko,1 E.G. Zezerov,2 A.I. Glukhov,2 Yu.G. Alyaev,1 S.E. Severin,2 K.A. Polyakovsky,3 V.A. Varshavsky,4 E.S. Severin,2 and A.Z. Vinarov1 1
Department of Urology, First Sechenov Moscow State Medical University, Moscow, Russia Department of Biochemistry, First Sechenov Moscow State Medical University, Moscow, Russia 3 Urology Clinic, First Sechenov Moscow State Medical University, Moscow, Russia 4 Department of Anatomic Pathology, First Sechenov Moscow State Medical University, Moscow, Russia 2
BACKGROUND. Early diagnosis of prostate cancer (CaP) can be addressed by studying prostatic intraepithelial neoplasia (PIN) as precancer (high-grade PIN or HGPIN). This article attempts to analyze the diagnostic role of telomerase as an early marker of carcinogenesis. METHODS. Complex urological patient evaluation and assessment of telomerase activity. RESULTS. Out of 92 patients 44% were diagnosed with CaP, 49% with low-grade PIN (LGPIN) in association with benign prostatic hyperplasia (BPH), and 7% with HGPIN in association with BPH. Active telomerase (AT) in prostate biopsy specimens was detected in 98% of patients with CaP, in 33% of patients with HGPIN, and in 20% of patients with LGPIN. In the event of simultaneous detection of AT and PIN in initial prostate biopsy specimens, further monitoring for 0.5–4.0 years revealed CaP development in 50–56% of cases. Further follow-up of patients with PIN and absent telomerase activity in initial biopsy specimens did not demonstrate the development of CaP. The PSA level was significantly higher in patients with active telomerase in the prostate tissue than in telomerase negative patients. CONCLUSIONS. Telomerase activity in the prostate tissue increases the risk of CaP development in patients with PIN. Detection of telomerase activity in prostate biopsy specimens from patients with PIN enables selection of a group of patients with high risk of CaP development and reduction of the number of prostate biopsies performed in other patients. Prostate 74:1043–1051, 2014. # 2014 Wiley Periodicals, Inc. KEY WORDS: prostate cancer; precancer; prostatic intraepithelial neoplasia; prostatespecific antigen; telomerase; dynamic process of prostate carcinogenesis
INTRODUCTION Over the course of our long-term and diverse research of prostate cancer pathology [1–4], about 10 years ago [4–8] our attention was drawn to telomerase, an enzyme, which is a marker of intensive and unlimited cell proliferation, including cancerous and precancerous cells in different kinds of oncopathology. Telomerase takes part in repair of telomeric DNA shortened during replication and increases the number of cell divisions thus allowing the cells to bypass the Hayflik limit of 35–50 divisions. As a result, the cells become conditionally immortal. Therefore, telomerase activity in the parenchyma of multiple organs is of diagnostic value as a marker of carcinogenesis and ß 2014 Wiley Periodicals, Inc.
even allows for detection of premalignant conditions [9]. Active telomerase (AT) is present in 47–92% of cases of prostate cancer (CaP) [10,11], in 16% of cases of high-grade prostatic intraepithelial neoplasia (HGPIN) [11,12], and is absent in patients with benign prostatic hyperplasia (BPH) [13]. As far back as 10 or more years ago, HGPIN was regarded as a morpho
Correspondence to: E.G. Zezerov, Department of Biochemistry, First Sechenov Moscow State Medical University, 8, Trubetskaya St, Building 2, Moscow, Russia. E-mail: [email protected] Received 28 December 2013; Accepted 17 April 2014 DOI 10.1002/pros.22823 Published online 22 May 2014 in Wiley Online Library (wileyonlinelibrary.com).
1044
Glybochko et al.
logical form of prostatic precancer [14]. These years of vast research involving both animal models and humans resulted in HGPIN being more confidently regarded as a precursor of prostate cancer [15]. The progression (usually a slow process) of HGPIN into CaP is studied using histological, biochemical and immunohistochemical techniques for the following purposes: to identify molecular mechanisms governing this transformation, predict the time of cancer onset, develop monitoring options and strategies aimed at delaying or stopping the process [16–21,22]. LGPIN is not classified as precancer [23]. However, works of other authors published over the last 5 years never consider the diagnostic value of comparison of telomerase activity in CaP, HGPIN, and LGPIN, especially during long-term follow-up of such categories of patients. For that reason we have been studying the role of telomerase in prostate gland diseases, specifically CaP [24,25], and also HGPIN and LGPIN since 2001. The present report discusses comparative results of AT assessment in different kinds of prostate gland oncopathology and the data of long-term (up to 4 years) monitoring for the purpose of estimation of diagnostic and/or prognostic value of AT. The article was made in the Department of Urology, Urology clinic and Department of Biochemistry of the Sechenov First Moscow State Medical University. MATERIALS AND METHODS The current work was based on the results of extensive examination of 92 urology clinic patients suspected of having CaP. The control group included eight patients with normal levels of prostate-specific antigen (PSA), with benign prostatic hyperplasia (BPH), and no AT in the tissue obtained during transurethral resection of the prostate (TURP). An additional control group included 20 younger males admitted to the Urology clinic for examination or treatment of other urological pathologies. These patients had no detected prostate pathology and normal PSA levels which will be given later. That is why prostate biopsy and—all the more so TURP with AT assessment—could not be performed in these 20 patents (unlike in eight aforementioned patients) due to ethical reasons as well as absence of medical indications. This approach is justified by a well-known fact that telomerase shows no enzyme activity in the cells of mature parenchymal organs that finished their ontogenesis and that show no signs of oncopathology [9]. Concrete evidence for this fact (using the prostate and kidneys as examples) was provided by works of several laboratories included in the review by Orlando et al. [11]. The Prostate
All patients underwent complete clinical examination and laboratory testing. The PSA of the blood serum was analyzed by means of immunochemiluminometric or enzyme-linked immunosorbent assays. The examination also included urinalysis, three-glass test, uroflowmetry, digital rectal examination (DRE), and transrectal ultrasound (TRUS) of the prostate followed by PSA density (PSAD) calculation. In some cases an additional examination—lesser pelvis magnetic resonance imaging (MRI)—was performed. Depending on the obtained results the patients underwent plurifocal transrectal ultrasound-guided needle prostate biopsy. Biopsies were taken using a “B&K 3535” (BK-Medical) diagnostic ultrasound system with a 6558/T/S bi-plane sector mechanical rectal probe operating at a frequency of 7.5 MHz. The material was obtained using transrectal needle prostate biopsies from standard sites and also sites where CaP foci were suspected on ultrasound (TRUS) or magnetic resonance imaging scans (MRI). Several (10–12) biopsy specimens were taken for histological examination. At least one biopsy specimen from each lobe of the gland and sites suspicious of CaP were used for AT assessment. The following segment describes the technique in more detail. In order to minimize the possibility of biopsy specimens being taken from different sites for histological analysis and AT assessment, we used sites with lower echogenicity. Having located a hypoechogenic area, we took a biopsy specimen for histological examination and then—from the same site—for AT evaluation. Due to the fact that in most cases the size of the hypoechogenic areas did not exceed 0.5 cm, the difference in sites used for tissue sampling was minimal. If the gland had homogeneous echotexture (in the minority of patients), two obligatory biopsy specimens for AT evaluation were taken from the central part of the peripheral zone of each prostate gland lobe. After that, a control biopsy specimen was taken for histological analysis without changing the position of the ultrasonic probe. The majority of patients underwent rebiopsy, however in some cases tissues obtained during TURP or radical prostatectomy (RPE) were used instead of re-biopsy specimens. The histological examination of prostate biopsy specimens and tissues obtained during TURP or RPE included standard tissue processing (paraffin wax embedding), 4 mm section preparation followed by hematoxylin, eosin, and picrofuscin staining. In complicated cases an immunohistochemical analysis using PSA and p63 antibodies (Dako) was performed. p63 protein is a p53 protein homologue, and it is frequently used in difficult cases to distinguish between CaP and PIN [14]. Universal criteria and the Gleason grading
TelomeraseçATumor Marker system were used for CaP evaluation. Criteria for differentiation between CaP and BPH, chronic prostatitis (ChP), and PIN were also used [26–28]. We consider it necessary to emphasize that precise differentiation between HGPIN and LGPIN was conducted in accordance with histological criteria developed by Bostwick and Brawer (1987) modified by Joniau et al. [14] or other researchers [28]. For AT assessment, the studied tissue was placed into a plastic test-tube, flash frozen in liquid nitrogen and stored at 70°C. Tissue extracts used for AT evaluation were obtained by the technique described by us earlier [24]: 6 intermittent homogenization cycles (each 1 min long) at 0°C with glass powder in a buffer solution containing Tris–HCl (pH 7.5), CHAPS, b-mercaptoethanol, and PMSF. After lysate centrifugation (12,000g, þ4°C, 30 min) the supernatant was analyzed for AT or frozen in liquid nitrogen and stored at 70°C for further use. To assess telomerase activity we used our own modification of TRAP (Telomeric Repeat Amplification Protocol). Other laboratories usually utilize TRAP to detect amplified telomerase reaction products radiolabelled with phosphorus P32. The detection process involves autoradiography of gel electropherograms [9]. . Our novel technique employed in the present work is a non-isotope modification of TRAP with the use of a highly sensitive SYBR Gold dye. This modified method was described in 1998 [8] and successfully utilized in our research [24,25]. Its sensitivity corresponds to active telomerase levels contained in approximately 15–16 telomerase-positive cells [6].
1045
Separation of the reaction products (TRAPproducts) was achieved with 10% non-denaturing polyacrylamide gel electrophoresis using 1 Trisborate-EDTA (TBE) buffer. PCR samples were then mixed with Orange G leading dye and 10 ml were added to the gel wells. The gel was exposed to a highly sensitive SYBR Gold dye for 30 min at room temperature in the dark. Visualization was performed with a UV transilluminator (l ¼ 300 nm). The stained gels were photographed with a Kodak digital camera. The digital photos were evaluated by means of IMAGE J 1.351 image processing program (“National Institutes of Health”, USA) which calculates the peak areas (in standard units), representing the luminescence intensity of discrete fragments of TRAP-products in the gel. The total peak area was regarded as the AT value in the studied sample. Relative AT was then calculated, by comparing the AT value in the studied sample to the positive control sample, that is, telomerase-positive LNCaP cell line extract containing 0.7 mg of protein. Telomerase activity in the control sample was considered to be 1.0 or 100%. The examined sample was regarded telomerasepositive if three or more horizontal strips of TRAPproducts were present. The distance between them amounted to the length of one telomeric repeat (six pairs of nucleotides). In case of negative control, pure CHAPS lysis buffer was added to the sample instead of cell extract. The statistical analysis was performed according to statistical guidelines for biological studies accepted in the USSR and Russia [29–33].
Telomerase Activity Analysis Design The first step was obtaining an extract from prostate tissues or LNCaP cells. The latter represent a transplantable cell line derived from human lymph nodes affected by metastatic prostate cancer (positive control for telomerase). The TS-primer elongation and subsequent amplification was carried out in 50 ml of reaction mixture containing 67 mM Tris–HCl (pH 8.8), 16.6 mM (NH4)2SO4, 0.01% Tween-20, 1.5 mM MgCl2, 1 mM EGTA, 50 mM of every dNTP, 0.1 mg TS-primer (50 ATTCCGTCGAGCAGAGTT-30 ), 1–10 ml of tissue extract containing 0.7, 2, 4, 8 mg protein or 1 ml (0.7 mg protein) of the LNCaP cell extract diluted with CHAPS lysis buffer and equivalent to 1,000 LNCaP cells (positive control). During the polymerase chain reaction (PCR) 0.1 mg of CX-primer (50 -CCCTTACCCTTACCCTTACCCTAA-30 ) and 2.5 units of SmarTaq DNA polymerase (“Dialat”, Russia) were added to the reaction mixture. The TS-primer elongation stage was followed by reaction mixture amplification during 35 PCR cycles.
RESULTS Based on histological examination of primary biopsy specimens of 92 patients (aged 47–81 years), CaP was diagnosed in 41 (44%) subjects/patients, HGPIN—in 6 (7%), and LGPIN—in 45 (49%) patients (Table I). Of 51 patients with PIN, HGPIN was detected in 12% of cases, and LGPIN in 88% of cases. The latter diagnosis was almost always associated with BPH. About 72% of PIN patients were also diagnosed with ChP. As a rule, total PSA level in the blood serum of patients with CaP, PIN, BPH, and ChP was higher than 4 ng/ml, and varied greatly, which will be discussed later. Taking a look at test subjects with PIN as a separate group, the mean total PSA level in 49 of 51 patients with PIN was 15.3 ng/ml (4.2–94.1 ng/ml). Two patients had total PSA less than 4.0 ng/ml. The free PSA index exceeded the prognostically favorable threshold of 15% in 15 patients (29%). It was lower than 15% in other patients. The PSA density (PSAD) The Prostate
1046
Glybochko et al.
TABLE I. ActiveTelomerase (AT) Detection in Prostate Biopsy Specimens of Patients WithVarious Urologic Diseases AT detected Disease CaP HGPIN and BPH LGPIN and BPH
Total number of patients
Number of patients
Percentage and mean percentage error (X m)
95 % confidence interval
41 6 45
40 2 9
98 2 33 21 20 6
87–100 4–78 10–35
The statistical analysis is carried out based on the technique and Table I developed by VS Genes.[23]
was below the prognostically favorable threshold of 0.1 ng/ml/cm3 in only 6 (12%) patients. We conducted preliminary testing of the sensitivity of our combined analysis (AT assessment þ histological examination) on 41 patients with CaP. As mentioned in Materials and Methods section, 10–12 biopsy specimens from both lobes of the gland underwent histological examination. At least one separate biopsy specimen from each lobe and, as a rule, several more tissue samples from hypoechogenic areas suspected of being cancer were sent for AT assessment (with strict adherence to the technique described in Materials and Methods section which lowers the risk of false conclusions). In 14 patients cancer was diagnosed based on histological examination in only one lobe of the prostate with 36% of these test subjects having AT in the same lobe. About 57% of patients from the aforementioned 14 test subjects had AT in both lobes of the prostate including the parts where no morphological changes were detected. AT appears to precede diagnosed epithelial cell cancer based on histological examination if one ignores the technical limitations preventing simultaneous analysis of the same biopsy specimen in terms of both histology and telomerase activity. It can be assumed that biopsy specimens undergoing AT evaluation without simultaneous histological analysis due to technical limitations may contain cancer cells which in theory always possess AT. In this case if AT is detected in two or more samples only assessed for AT while histological examination reveals no cancer in other 10–12 specimens, the statement concerning possible telomerase activation in cells which happens prior to typical cancer formation (e.g., in PIN) might prove wrong. However, if we consider the possibility of mistakes, we should specify the probability of error which does not exceed 14–17% (2/12 or 2/14). In other 27 patients of the aforementioned 41, prostate cancer and telomerase were detected in both lobes of the prostate. After the biopsy five patients with cancer in one lobe and two patients with CaP detected in both lobes underwent radical prostatectomy. AT assessment and histological analysis of The Prostate
prostatectomy specimens essentially matched the results of the same examinations conducted on preoperative biopsy specimens as well as differentiation parameters of various lobes of the gland (stages T1c, T2a, and T3a). We believe that this data lowers the probability of the error discussed above. The results of the analysis of primary biopsy specimens of patients with PIN are shown in Table I, Figure 1. AT was detected in biopsy samples in 11 (21%) of 51 patients with PIN. Both AT and LGPIN were detected in 9 of 45 patients (Fig. 2), that is, 20 6% (percentage and mean error) with a 10–35% confidence interval at the 95% probability level (Table I). Similarly, both AT and HGPIN were detected in 2 of six patients with HGPIN only (Fig. 3). The percentage and confidence interval were 33 21% and 4–78%, respectively (Table I). Although AT detection rates in patients with HGPIN were higher than in patients with LGPIN, the difference was not statistically significant due to the given number of observations. It is worth mentioning that AT was detected in 40 (98%) of 41 patients with CaP confirmed by histological examination (87–100% confidence interval). The difference between AT detection rates in CaP and PIN groups was statistically significant (Table I).
Fig. 1. Results ofmonitoring ofpatients with ATand PIN oninitial prostate biopsy (initial prostate biopsy specimens were examined histologically with simultaneous ATassessment).CaP, prostate cancer; AT, active telomerase; PSA, prostate-specific antigen; TURP, transurethralresection of theprostate.
TelomeraseçATumor Marker
Fig. 2. Results of monitoring of patients with LGPIN diagnosed via histological examination (initial prostate biopsy specimens were examined histologically with simultaneous AT assessment). CaP, prostate cancer; AT, active telomerase; LGPIN, low-grade prostatic intraepithelial neoplasia; HGPIN, high-grade prostatic intraepithelialneoplasia.
Further PIN patient monitoring included the blood level of PSA, PSAD, digital rectal examination (DRE), transrectal ultrasound (TRUS), and lesser pelvis magnetic resonance imaging (MRI). Some patients had indications for one or more prostate re-biopsies or, in rare cases, for TURP. The materials obtained underwent histological examination and AT assessment. The results given in Figure 1 show that 6 (60%) of 10 patients with high levels of AT and PIN on initial
Fig. 3. Results of monitoring of patients with HGPIN diagnosed via histological examination (initial prostate biopsy specimens were examined histologically with simultaneous AT assessment). CaP, prostate cancer; AT, active telomerase; HGPIN, high-grade prostatic intraepithelialneoplasia.
1047
biopsy were later diagnosed with CaP. Patients with no AT diagnosed with LGPIN (36 subjects) or HGPIN (4 subjects) on initial biopsy did not develop CaP during 0.5–3.0 years of follow-up (Figs. 2 and 3). One case of a very low AT level in association with LGPIN did not show CaP during four years of follow-up (Figs. 1 and 2). Three patients with AT and PIN did not undergo rebiopsy for the following reasons: improvement of the condition, based on PSA, DRE, TRUS, and MRI data (1 patient); refusal from follow-up (1 patient); refusal from re-biopsy upon detection of PSA level increase (1 patient) (Fig. 1). Figures 2 and 3 show the re-examination results of patients with LGPIN and HGPIN in more detail and separately. In the period of 0.5–3.0 years, four patients initially diagnosed with LGPIN in association with AT developed CaP, and two patients were diagnosed with HGPIN. In one case of HGPIN, CaP was detected in 2 years (Fig. 2). Overall, during 0.5–4.0 years LGPIN associated with AT transformed into CaP in five (56%) of nine cases as opposed to LGPIN without AT. However, 3 (16%) of 19 patients with LGPIN and no AT showed the transition of LGPIN to HGPIN on rebiopsy in 0.5–3.0 years (Fig. 2). In such and similar cases the terms “transformation” or “transition” from one type of histological changes into another do not necessarily imply “direct transformation” of LGPIN into HGPIN, for example. Such changes in histological structure of biopsy samples might be due to statistically random distribution of sites of morphological changes in prostatic tissues (mainly in the peripheral zone) and partly due to statistically random sampling during plurifocal biopsy and re-biopsy. One patient with an initial combination of HGPIN and AT showed no progression into CaP during 1-year follow-up (Fig. 3). The second patient with initial combination of HGPIN and AT underwent three rebiopsies showing no progression of the disease. However, in 1 year the 4th biopsy revealed CaP. Therefore, in patients with HGPIN associated with AT the CaP rate was 50%. All four patients with HGPIN and no AT were not diagnosed with CaP during 1–2 years of follow-up (Fig. 3). Thus, the data obtained allows for the first conclusion to be made: without AT neither LGPIN nor HGPIN transforms into CaP, whereas LGPIN could progress into HGPIN in 16% of cases. Secondly, during 0.5–4.0 years LGPIN and HGPIN associated with AT have a 56% and a 50% rate of progression into CaP respectively. Although given the insufficient quantity of studied patients (only six test subjects) with HGPIN, such comparison is imperfect. Figure 2 shows that of 45 patients with LGPIN only four (9%) were diagnosed with CaP on re-biopsy during 0.5–3.0 years. The Prostate
1048
Glybochko et al.
In all these four patients LGPIN was associated with active telomerase in the prostate tissue. This fact might prove to be of interest to other researchers. Another interesting observation is that the fifth patient (LGPIN plus AT) was diagnosed with HGPIN in 2 years which subsequently progressed into CaP in further two years. Other patients with LGPIN on initial biopsy (regardless of AT presence) retained the LGPIN phenotype during follow-up or acquired the HGPIN phenotype (one patient, Fig. 2). AT assessment is of obvious diagnostic and prognostic value in patients with PIN for it allows for selection and thorough monitoring of individuals with high risk of CaP development. Absence of AT in biopsy samples indicates low probability of malignant transformation which means that no re-biopsies will be required in cases of improvement (based on PSA, DRE, TRUS,PSAD, and MRI data). In terms of pathogenesis, the results suggest that telomerase activity in prostate tissue plays a role as a factor of carcinogenesis. The analysis of the experimental data also allowed for the investigation of the relationship between telomerase in prostate tissue and blood level of PSA. Prostate-specific antigen (PSA) is not only an antigen used for specific analysis, but also an enzyme, chymotrypsin-like serine protease, which is a member of the human kallikrein family [3]. The second enzyme, telomerase, is a factor of intense and/or malignant cellular proliferation [9]. Both of these protein enzymes are produced by epithelial secretory cells of the prostate gland and play a role in oncogenesis [2,4,7]. PSA acts as a catalyst for proteolysis and disintegration of insulin-like growth factor-binding protein 3 (IGFBP3). Similar to telomerase, insulin-like growth factor 1 (IGF-1) accelerates proliferation of epithelial cells of the prostate [2–4]. Apart from the endogenous function of PSA in the gland itself, this enzyme enters the urethra through the prostatic ducts during ejaculation, mixes with seminal fluid, and catalyzes proteolysis of gel matrix forming proteins of the seminal vesicles (semenogelins, fibronectin), decreases the viscosity of semen and improves sperm motility [3]. Only one millionth of PSA enters the bloodstream from a healthy gland as a defective inactive enzyme due to its binding with inhibitors in the blood (a1–antichymotrypsin, a2-macroglobulin and others) and due to its lys145 breakdown. It is possible due to microdamage of the basal cell layer and basement membranes of the gland. In prostate pathology these lesions intensified simultaneously with an increase in PSA-producing cells. In HGPIN, the basal cell layer is disrupted or fragmented, while in CaP damage is inflicted to the basal layer and basement membranes [3,14,34]. All of this leads to higher PSA concentration in the blood The Prostate
TABLE II. Correlation Between AT in Prostate Gland Tissue and PSA Level in Blood Serum Total PSA, ng/ml Patients with CaP, PIN and BPH
Number of patients
Average value— median
95 % confidence interval
51 41
18 10
13–27 7–12
Telomerase detected Telomerase absent
which correlates to the increased quantity of secretory prostatic epitheliocytes and the degree of damage to the cells of the basal layer and basement membranes. Our summarized clinical and laboratory data is given in Tables II and III. Of 92 patients with prostate pathology (CaP, PIN, BPH) in our study, subjects with active telomerase in the prostate had almost twice as high mean PSA concentration (median) than those without AT. Given the sufficient number of patients in our study (Table II), the difference was statistically significant (the confidence intervals did not overlap, P < 0.05). The data corresponds to the role of telomerase and PSA as factors of prostate carcinogenesis. A similar analysis (Table III) of various kinds of pathology showed that total PSA level is higher in CaP than in PIN and BPH. The difference also was statistically significant (P < 0.05). Table IV contains similar data concerning total PSA concentration in the blood serum in other patients of the four different groups. PSA level was normal in control group 4 which consisted of 20 relatively younger males without signs of prostate pathology. It was statistically
