- Email: [email protected]
Cytogenetic profiles as additional markers to pathological features in clinically localized prostate carcinoma
Cancer Letters 237 (2006) 76–82 www.elsevier.com/locate/canlet
Cytogenetic profiles as additional markers to pathological features in clinically localized prostate carcinoma Michele Galluccia, Roberta Merolab, Antonella Farsettic, Giulia Orlandib, Steno Sentinellid, Piero De Carlia, Costantino Leonardoa, Paolo Carlinie, Fiorella Guadagnib, Isabella Sperdutif, Anna Maria Cianciullib,* b
a Department of Urology, Regina Elena Cancer Institute, Rome, Italy Department of Clinical Pathology, Cytogenetic Unit, Regina Elena Cancer Institute, IFO, Via Elio Chianesi 53, 00144 Rome, Italy c Institute of Neurobiology and Molecular Medicine, National Research Council, Rome, Italy d Department of Pathology, Regina Elena Cancer Institute, Rome, Italy e Department of Oncology, Regina Elena Cancer Institute, Rome, Italy f Biostatistic Unit, Regina Elena Cancer Institute, Rome, Italy
Received 4 May 2005; received in revised form 20 May 2005; accepted 23 May 2005
Abstract Fluorescence in situ hybridization analysis for evaluation of 7, 8, X chromosomes and EGFR, LPL, MYC, AR genes in 79 neoplastic foci from 56 patients with clinically localized prostate cancer was performed. We found aneusomy for chromosome 7, 8 and X in 74/77 (96.1%), 56/76 (73.7%), 26/70 (37.1%) of examined foci respectively. No specimen was amplified for EGFR and AR genes, only 2/71 (2.8%) specimens showed MYC gene amplified. LPL deletion was present in 52/76 (68.4%) specimens. Statistically association between Gleason score and both chromosome 7 aneusomy and 8p21 deletion was present. The frequency of chromosome 7 aneusomy was statistically higher in T3-4 cases than T2c and T2a-T2b ones. We considered as unfavorable a genetic set if aneusomy for at least two chromosomes and one altered gene were present. The percentage of tumors, with unfavorable genetic pattern, increased from 36.4 to 75.0% in those with Gleason O7 and from 40.0 to 73.7% in those with stage T3 or more. These alterations could be considered potent genetic markers adjunctive to conventional prognostic parameters. Our objective was to establish specific genetic profiles which may discriminate favorable and unfavorable genetic prognosis tumors. q 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Genetic markers; Prostate cancer; FISH
1. Introduction
* Corresponding author. Tel./fax: C39 6 52665966. E-mail address: [email protected] (A.M. Cianciulli).
Current methods for assessing the prognosis of prostate carcinoma (CaP), such as clinical and pathological staging and histopathologic grading, provide not enough predictive information regarding
0304-3835/$ - see front matter q 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2005.05.033
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
the outcome and therapeutic strategy in individual cases. The broad application of cytogenetics and molecular genetic methods has led to the identification of tumor-associated chromosomal regions substantial for the tumorigenesis and progression of CaP [1,2]. Many of the chromosomal losses can be detected in the early stages of CaP, whereas gains and amplifications are mostly seen in hormone-refractory tumors, suggesting that oncogenes become activated at a late stage of the disease [3]. Specifically, loss of 8p21 is a common alteration in CaP. The loss of 8p21 has been reported to be up to 69 and 100% in clinically organconfined and metastatic CaP, respectively [4,5]. Furthermore, up to 50% of prostatic intraepithelial neoplasia (PIN) lesions have 8p21 loss [6]. The commonly deleted region of 8p21 includes the LPL (lipoprotein lipase) gene that is suggested to be responsible for the initiation or early event in prostate tumorigenesis. Another important genetic alteration in CaP is 8q24 overrepresentation, that is commonly found in advanced, metastatic and androgen-independent CaP [7,8]. The region contains an oncogene, MYC, which regulates cell proliferation and apoptosis and may be a target gene [9]. The molecular mechanisms of CaP resistance to the endocrine therapy is still poorly understood. It has been demonstrated that the AR (androgen receptor) gene is amplified in 30% of hormone-refractory CaP from patients treated with androgen withdrawal [10]. A functional cross-talk between growth factors and growth factor receptors of the EGFR family and AR-activated pathways has been shown in preclinical models in prostatic cancer cells [11]. Moreover, it has been demonstrated that EGFR expression increases during the natural history of the prostate, suggesting that EGFR-targeted drugs might be of therapeutic relevance in CaP [12]. On that basis, we evaluated EGFR (7p12), LPL (8p21), MYC (8q24), AR (Xq12) genes and the 7, 8, X chromosome status, where the genes are located, in 79 neoplastic foci from 56 patients with clinically localized prostate cancer by fluorescence in situ hybridization (FISH). In this work we intended to investigate the following two interrelated specific aims: (i) to evaluate the association of molecular cytogenetic characteristics with histopathological classification; (ii) to detect specific genetic signatures which may identify tumors with
77
unfavorable genetic prognosis in patients undergoing radical prostatectomy.
2. Materials and methods 2.1. Tumor sampling All clinical-pathological characteristics are illustrated in Table 1. Fresh tumor samples were prospectively collected from 56 men undergoing radical retropubic ascending prostatectomy at our institute (The National Cancer Institute, Rome) for clinically localized prostate cancer. An extended regional lymphadenectomy was performed in every case. None had received preoperative androgen deprivation or radiation therapy. Fresh tumor samples were macroscopically dissected from the prostatectomy specimen immediately after operation by a pathologist. A representative piece of the CaP sample was stained with hematoxylin–eosin. The samples were prepared by cytological imprints of the specimens from both lobes which were gently pressed onto the surface of sialinized slides. In this study, the Gleason grade (G) was determined for each examined focus of prostate carcinoma. Only one case presented two neoplastic foci with different grade. Pathologic stage was assigned for each case in accordance with the Union Internationale Contre le Cancer (UICC, 2002) TNM system [13]. All samples were reviewed Table 1 Patients characteristics Characteristic PSA at diagnosis 0–3.9 4–9.9 10–19.9 O20 Gleason grade !6 6 7 O7 Pathological stage T2a–2b T2c T3a–3b T4
No.
%
4 20 20 12
7.15 35.71 35.71 21.43
1 15 31 9
1.79 26.79 55.36 16.07
8 28 19 1
14.29 50.0 33.93 1.79
78
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
by a pathologist with experience in uropathology in order to evaluate cellular adequacy and only samples with O75% of tumor cells were examined. After histological evaluation, in tumors with bilateral involvement, the slides, obtained from both lobes, were selected for genetic evaluation which was consequently performed on 79 neoplastic foci. As a control population, normal prostatic tissue samples from patients undergoing cystectomy were used. In addition, as positive control we used prostatic cell lines (LNCaP). Informed consent was obtained from the patients, and tissue sampling was approved by the ethics committee of the National Cancer Institute, Rome. 2.2. FISH analysis
2.4. Statistical analysis The c2 test was used to evaluate the association of pathological characteristics (stage and Gleason score) with each genetic marker as with the anomalous genetic patterns which were identified. The accepted statistically significant difference was P!0.05.
3. Results 3.1. Genetic profiles (chromosomes 7, 8, X and EGFR, MYC, LPL status) in neoplastic examined foci On the basis of established criteria, the aneusomy cutoff levels were 3.3, 8.4 and 9.8% for chromosomes
The procedure of FISH analysis has been previously described [14]. The Vysis ProVysion Multi-color mixture (Vysis, Inc. Downers Grove, IL) was used for detection and quantification of chromosome 8 labeled with SpectrumAqua, the 8p21 labeled with SpectrumOrange and c-MYC gene located at 8q24, labeled with SpectrumGreen. We also employed specific probes for AR (Xq12), EGFR (7p12) genes and chromosome enumeration probes (CEP) specific for X and 7 chromosomes to adjust for the effects of aneuploidy and to establish the presence of true amplification. 2.3. Scoring FISH signals To validate our FISH assay, we evaluated the hybridization patterns in normal prostatic samples obtained from 10 patients undergoing cystectomy as control group. Tumoral foci were considered aneusomic, if the percentage of nondisomic nuclei exceeded the meanC3SD of any of the signal categories observed in the control group. To evaluate the genes status, at least 100 cells were scored for each case, tabulating the total number of genes and corresponding centromere signals for each patient on a standard worksheet. We used a ratio of more than 2 oncogene/centromere signals to define gene amplification, whereas gene deletion was determined by a ratio lower than 1. The FISH assay was not available for 10 genetic determinations for technical causes.
Fig. 1. (A) Distribution of ratios (centromere/gene signals) for EGFR, AR, MYC, LPL genes. (B) Fluorescence in situ hybridization image in tumor with MYC gene amplification [CEP 8: aqua signals; 8p21 (LPL gene): red signals; MYC gene: green signals].
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
79
Table 2 Summary of the combined chromosomes 7, 8, X and EGFR, AR, MYC, LPL gene anomaly patterns in 69 prostatic cancer foci
A, aneusomic; Amp, amplified; D, deleted.
7, 8 and X respectively. FISH was successfully performed on 69 out of 79 neoplastic foci (87.3%) for all considered genetic variables. By applying the cutoff values previously described, we defined 74/77 (96.1%), 56/76 (73.7%), 26/70 (37.1%) of examined foci as having an aneusomy for chromosome 7, 8 and X respectively. No sample was amplified for EGFR and AR gene (ratio!2), while 2/71 (2.8%) samples showed MYC gene amplification with ratio values of 4.3 and 2.2 (Fig. 1). As regards LPL deletion, 52/76 (68.4%) samples were deleted (ratio!1). As illustrated in Table 2, we subclassified FISH anomalies in 12 patterns, which describe each combination of all genetic alterations occurring in this cohort of patients. Patterns2 and patterns3, observed in 21 (30.4%)
and 15 (21.7%) samples respectively, showed the higher frequency. Moreover we considered unfavorable a genetic set, if at least two aneusomic chromosomes and one altered gene were present (see patterns 2, 3, 4, 6, 12 highlighted grey in Table 2). 3.2. Association of genetic alterations and unfavorable genetic set with clinicopathologic characteristics The FISH data obtained were compared with grade and stages. The Gleason grade of the 79 cancer foci was !7 in 18 foci,Z7 in 47 foci and O7 in 14 foci. As illustrated in Table 3, a statistically significant association between the Gleason score and both chromosome 7 aneusomy and 8p21 deletion
Table 3 Correlation of FISH abnormalities (% of examined cases) with Gleason score and stage Gleason score
No. Foci
CEP 7 Aneusomy
CEP 8 Aneusomy
CEP X Aneusomy
EGFR Amplified
AR Amplified
MYC Amplified
LPL(8p21) Deleted
!7 Z O7 P value Stage T2a–T2b T2ca T3–T4a P value
18 47 14
13 (81.3) 47 (100) 14 (100) 0.003
13 (81.3) 33 (72.7) 10 (71.4) 0.74
3 (27.3) 16 (34.0) 7 (58.3) 0.23
0 (0) 0 (0) 0 (0) –
0 (0) 0 (0) 0 (0) –
0 (0) 1 (2.2) 1 (7.1) 0.51
8 (50.0) 31 (67.4) 13 (92.9) 0.04
6 (75.0) 25 (96.2) 20 (100) 0.029
7 (87.5) 19 (76.0) 17 (85.0) 0.66
1 (14.3) 5 (23.8) 11 (55.0) 0.05
0 (0) 0 (0) 0 (0) –
0 (0) 0 (0) 0 (0) –
0 (0) 1 (4.8) 1 (5.0) 0.84
5 (71.4) 14 (66.7) 19 (95.0) 0.07
a
No. patients 8 28 20
The correlation was performed with the worst genetic value detected in bilaterally examined tumors.
80
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
Table 4 Association between combined genetic patterns with Gleason Score and with Stage Gleason Score
No foci evaluated
!7 Z7 (3C4) Z7 (4C3) O7 P value Stage T2a–T2b T2c T3–T4 P value
11 34 12 12
Poor genetic patternsa
4 (36.4) 17 (50.0) 8 (66.6) 9 (75.0) 0.21 No patients evaluated 7 4 (57.1) 20 8 (40.0) 19 14 (73.7) 0.08
Good genetic patterns 7 (63.6) 17 (50.0) 4 (33.3) 3 (25.0)
3 (42.9) 12 (60.0) 5 (26.3)
a Unfavorable genetic profile (at least two aneusomic chromosomes and one altered gene).
(P!0.003 and P!0.04, respectively) was present. Moreover, we noticed a positive trend when we compared chromosome X aneusomy and the Gleason score, even though this was not statistically significant. Comparing the FISH abnormalities with stage in evaluated bilateral tumors, we considered the worst detected genetic value. The frequency of chromosome 7 aneusomy was statistically higher in T3-4 than T2c and T2a-T2b (P!0.029). In addition, even if not statistically significant, a strong positive trend was also present for chromosome X aneusomy and 8p21 deletion, especially in the transition T2c versus T3-4 stage. Examining the different genetic patterns, we observed that the percentage of tumors, with unfavorable genetic set, increased from 36.4 to 75.0% with grade and from 40.0 to 73.7% with advanced stage (Table 4). In addition, taking into consideration intermediate grades (Gleason score 7), that is the most interesting group for urologists, we found poor genetic set in 17/34 (50%) tumors with G 3C4 and in 8/12 (66.6%) with G 4C3, even if the difference was not statistically significant (PZ0.32).
4. Discussion The natural course of human CaP is highly variable and we still lack reliable tools to predict the outcome in the individual cases. Radical
prostatectomy is generally regarded as an efficient way of curing the disease. Despite attempts to restrict radical prostatectomy to patients with clinically organ-confined disease, as many as 50% of the patients are found to have extraprostatic disease at the time of surgery [15]. To identify this patient subgroup with greater precision than achievable with current clinical and pathological means, additional prognostic information is needed. A large use of cytogenetics and molecular genetic has led to the identification of tumor-associated chromosomal regions critical for the tumorigenesis and progression of CaP [16,17]. Since prostate tumor progression is undoubtedly associated with particular somatic genetic alterations, one means of characterizing each patient would be to discern any significant genetic aberrations in tumors obtained during prostatectomy. Moreover it has been shown that, in a multifocal disease, some small low-grade tumor foci had a high frequency of genetic changes, whereas concurrent dominant high-grade tumor foci were normal, indicating that small cancers can have significant alterations [18]. Thus, the size of a cancer focus and its degree of histologic dedifferentiation may not reflect the extent of its genetic instability. Others studies [19,20] had examined overexpression and/or amplification of EGFR, MYC and AR genes. In our study we investigated these genes together with chromosomes where they are located, by fluorescence in situ hybridization, to adjust for the effects of aneuploidy and to establish the presence of true amplification. Analyzing the results by a ratio (centromeric/gene signals), we did not find amplified samples, save for two that had MYC gene amplification. These patients characterized by MYC amplification and poor genetic profile (Table 2) developed disease progression (biochemical failure) after six months from surgery (data not showed) even if differently classified by conventional histology, as illustrated in Table 3. These data are in accordance with literature: high-level amplifications are very rare in primary tumors (!2%), but are more common in advanced, metastatic and androgen-independent CaP [21]. On this basis, the presence of gene amplification in tumor, at the time of prostatectomy, could be considered an important prognostic variable.
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
Particular attention should be paid to the 7, X chromosome aneusomy and LPL gene deletion that not only revealed an accumulation of aberrations closely correlated with tumor and grade, but also, to chromosome 8 aneusomy, which were the most consistent aberrations found in the foci examined, according to previous studies [18,20]. When we considered the combined FISH results for the genetic variables, the dominant FISH anomaly patterns were the #2 and #3 profiles, as illustrated in Table 2, observed in 21 (30.4%) and 15 (21.7%) samples respectively. In this investigation the percentage of tumors, with unfavorable genetic profiles (at least two aneusomic chromosomes and one gene altered) increased in parallel with the grade and stage of tumor. Our results suggest that subgroups of CaP patients with poor- and goodprognosis genetic signatures, obtained with the assessment of evaluated chromosomes and genes status, may reflect the presence of genetically defined subtypes of CaP manifesting a biological different aggressiveness and consequently a distinct course of disease progression. Besides it is of clinical relevance to characterize each patient also in the same histological classification as in intermediate grade (Gleason score 7) that is the most interesting group for clinicians. In this class we found unfavorable genetic setting in 17/34 (50.0%) tumors with G 3C4 and in 8/12 (66.6%) tumors with G 4C3 confirming the presence of different degree of genetic heterogeneity in these patients (Table 4). Sato et al. evaluated CaP prognosis by genetic profiles considering only chromosome 8 and relatives genes [22]. In this work we assembled together various well-known CaP chromosomal and gene alterations to define a complete panel that could be employed by urologists for a better CaP stratification. Moreover, in our opinion, after opportune results validation, the stratification by these genetic patterns could be employed also in preoperative core biopsies as early adjunctive diagnostic measure. The goal is to recognize through emerging technologies, such as molecular cytogenetics, pathways driving the aggressiveness of CaP, as well as to identify specific and more accurate diagnostic markers. Only future larger studies, with long-term followup of these patients, should determine the validity
81
and clinical relevance of these genetic findings with resultant incorporation of molecular prognostic markers in clinical practice. Acknowledgements We thank Mrs. Paula Franke for the formal English revision of the manuscript and Mrs. Paola Canalini for technical support. Supported by grants from Italian Ministry of Health and Regional Italian Association for Cancer Research (AIRC). References [1] P. Cairns, K. Okami, N. Halachmi, M. Esteller, J.G. Herman, J. Jen, et al., Frequent inactivation of PTEN/MMAC1 in primary prostate cancer, Cancer Res. 57 (1997) 4997–5000. [2] C. Abate-Shen, M.M. Shen, Molecular genetics of prostate cancer, Genes Dev. 14 (2000) 2410–2434. [3] J.P. Elo, T. Visakorpi, Molecular genetics of prostate cancer, Ann. Med. 33 (2001) 130–141. [4] N. Tsuchiya, J.M. Slezak, M.M. Lieber, E.J. Bergstralh, R.B. Jenkins, Clinical significance of alterations of chromosome 8 detected by fluorescence in situ hybridization analysis in pathologic organ-confined prostate cancer, Genes Chromosomes Cancer 34 (2002) 363–371. [5] H. Matsuyama, Y. Pan, K. Oba, S. Yoshihiro, K. Matsuda, L. Hagarth, et al., Deletions on chromosome 8p21 may predict disease progression as well as pathological staging in prostate cancer, Clin. Cancer Res. 7 (2001) 3139–3143. [6] D.G. Bostwick, A. Shan, J. Qian, M. Darson, N.J. Mailhle, R. B. Jenkins, L. Cheng, Independent origin of multiple foci of prostatic intraepithelial neoplasia: comparison with matched foci of prostate carcinoma, Cancer 83 (1998) 1995–2002. [7] N.N. Nupponen, L. Kakkola, P. Koivisto, T. Visakorpi, Genetic alterations in hormone-refractory recurrent prostate carcinomas, Am. J. Pathol. 153 (1998) 141–148. [8] R. Jenkins, S. Takahashi, K. De Lacey, E. Bergstralh, M. Lieber, Prognostic significance of allelic imbalance of chromosome arms 7q, 8p,16q and 18q in stage T3N0M0 prostate cancer, Genes Chromosomes Cancer 21 (1998) 131–143. [9] B. Amati, K. Alevizopoulos, J. Vlach, Myc and the cell Cycle, Front. Biosci. 3 (1998) D250–D268. [10] C. Palmerg, P. Koivisto, L. Kakkola, T.L.J. Tammela, O.P. Kallioniemi, T. Visakorpi, Androgen receptor gene amplification at the time of primary progression predicts response to combined androgen blockade as a second-line therapy in advanced prostate cancer, J. Urol. 164 (2000) 1992–1995. [11] D. Ye, J. Mendelsohn, Z. Fan, Androgen and epidermal growth factor down-regulate cyclin-dependent kinase inhibitor p27 and costimulate proliferation of MDA PCA 2a and MDA PCA 2b prostate cancer cells, Clin. Cancer Res. 5 (1999) 2171–2177.
82
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
[12] G. Di Lorenzo, G. Tortora, F.P. D’Armiento, G. De Rosa, S. Staibano, R. Autorino, et al., Expression of epidermal growth factor receptor correlates with disease relapse and progression to androgen-independence in human prostate cancer, Clin. Cancer Res. 8 (2002) 3438–3444. [13] UICC, TNM Classification of Malignant Tumors, sixth ed., Wiley-Liss, New York, 2002. [14] A.M. Cianciulli, C. Leonardo, F. Guadagni, R. Marzano, F. Iori, C. De Nunzio, et al., Genetic instability in superficial bladder cancer and adjacent mucosa: an interphase cytogenetic study, Hum. Pathol. 34 (2003) 214–221. [15] J. Trapasso, J. deKernion, R. Smith, F. Dorey, The incidence and significance of detectable levels of serum prostate specific antigen after radical prostatectomy, J. Urol. 152 (1994) 1821–1825. [16] J.C. Alers, J. Rochat, P.J. Krijtenburg, W.C.P. Hop, R. Krance, C. Rosenberg, et al., Identification of genetic markers for prostate cancer progression, Lab. Invest. 80 (2000) 931–942. [17] J.C. Alers, P.J. Krijtenburg, A.N. Vis, R.F. Hoedemaeker, M. F. Wildhagen, W.C.P. Hop, et al., Molecular cytogenetic analysis of prostatic adenocarcinomas from screening studies: early cancers may contain aggressive genetic features, Am. J. Pathol. 158 (2001) 399–406.
[18] J. Qian, K. Hirasawa, D.G. Bostwick, E.J. Bergstralh, J.M. Slezak, K.L. Anderl, et al., Loss of p53 and c-myc overrepresentation in Stage T2–3 N1–3 M0 prostate cancer are potential markers for cancer progression, Mod. Pathol. 15 (2002) 35–44. [19] E. Hermes, S.D. Fossa, A.A. Berner, B. Otnes, J.M. Nesland, Expression of the epidermal growth factor receptor family in prostate carcinoma before and during androgen-independence, Br. J. Cancer 90 (2004) 449–454. [20] H. van Dekken, J.C. Alers, I.A. Damen, K.J. Vissers, P.J. Krijtenburg, R.F. Hoedemaeker, et al., Genetic evaluation of localized prostate cancer in a cohort of forty patients: gain of distal 8q discriminates between progressors and nonprogressors, Lab. Invest. 83 (2003) 789–796. [21] C. Kaltz-Wittmer, U. Klenk, A. Glaessgen, D.E. Aust, J. Diebold, U. Lohrs, G.B. Baretton, FISH analysis of gene aberrations (MYC, CCND1, ERBB2, RB and AR) in advanced prostatic carcinomas before and after androgen deprivation therapy, Lab. Invest. 80 (2000) 1455–1464. [22] K. Sato, J. Qian, J.M. Slezak, M.M. Lieber, D.G. Bostwick, E. J. Bergstralh, R.B. Jenkins, Clinical significance of alterations of chromosome 8 in high-grade, advanced, nonmetastatic prostate carcinoma, J. Natl Cancer Inst. 91 (1999) 1574–1580.
Cytogenetic profiles as additional markers to pathological features in clinically localized prostate carcinoma Michele Galluccia, Roberta Merolab, Antonella Farsettic, Giulia Orlandib, Steno Sentinellid, Piero De Carlia, Costantino Leonardoa, Paolo Carlinie, Fiorella Guadagnib, Isabella Sperdutif, Anna Maria Cianciullib,* b
a Department of Urology, Regina Elena Cancer Institute, Rome, Italy Department of Clinical Pathology, Cytogenetic Unit, Regina Elena Cancer Institute, IFO, Via Elio Chianesi 53, 00144 Rome, Italy c Institute of Neurobiology and Molecular Medicine, National Research Council, Rome, Italy d Department of Pathology, Regina Elena Cancer Institute, Rome, Italy e Department of Oncology, Regina Elena Cancer Institute, Rome, Italy f Biostatistic Unit, Regina Elena Cancer Institute, Rome, Italy
Received 4 May 2005; received in revised form 20 May 2005; accepted 23 May 2005
Abstract Fluorescence in situ hybridization analysis for evaluation of 7, 8, X chromosomes and EGFR, LPL, MYC, AR genes in 79 neoplastic foci from 56 patients with clinically localized prostate cancer was performed. We found aneusomy for chromosome 7, 8 and X in 74/77 (96.1%), 56/76 (73.7%), 26/70 (37.1%) of examined foci respectively. No specimen was amplified for EGFR and AR genes, only 2/71 (2.8%) specimens showed MYC gene amplified. LPL deletion was present in 52/76 (68.4%) specimens. Statistically association between Gleason score and both chromosome 7 aneusomy and 8p21 deletion was present. The frequency of chromosome 7 aneusomy was statistically higher in T3-4 cases than T2c and T2a-T2b ones. We considered as unfavorable a genetic set if aneusomy for at least two chromosomes and one altered gene were present. The percentage of tumors, with unfavorable genetic pattern, increased from 36.4 to 75.0% in those with Gleason O7 and from 40.0 to 73.7% in those with stage T3 or more. These alterations could be considered potent genetic markers adjunctive to conventional prognostic parameters. Our objective was to establish specific genetic profiles which may discriminate favorable and unfavorable genetic prognosis tumors. q 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Genetic markers; Prostate cancer; FISH
1. Introduction
* Corresponding author. Tel./fax: C39 6 52665966. E-mail address: [email protected] (A.M. Cianciulli).
Current methods for assessing the prognosis of prostate carcinoma (CaP), such as clinical and pathological staging and histopathologic grading, provide not enough predictive information regarding
0304-3835/$ - see front matter q 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2005.05.033
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
the outcome and therapeutic strategy in individual cases. The broad application of cytogenetics and molecular genetic methods has led to the identification of tumor-associated chromosomal regions substantial for the tumorigenesis and progression of CaP [1,2]. Many of the chromosomal losses can be detected in the early stages of CaP, whereas gains and amplifications are mostly seen in hormone-refractory tumors, suggesting that oncogenes become activated at a late stage of the disease [3]. Specifically, loss of 8p21 is a common alteration in CaP. The loss of 8p21 has been reported to be up to 69 and 100% in clinically organconfined and metastatic CaP, respectively [4,5]. Furthermore, up to 50% of prostatic intraepithelial neoplasia (PIN) lesions have 8p21 loss [6]. The commonly deleted region of 8p21 includes the LPL (lipoprotein lipase) gene that is suggested to be responsible for the initiation or early event in prostate tumorigenesis. Another important genetic alteration in CaP is 8q24 overrepresentation, that is commonly found in advanced, metastatic and androgen-independent CaP [7,8]. The region contains an oncogene, MYC, which regulates cell proliferation and apoptosis and may be a target gene [9]. The molecular mechanisms of CaP resistance to the endocrine therapy is still poorly understood. It has been demonstrated that the AR (androgen receptor) gene is amplified in 30% of hormone-refractory CaP from patients treated with androgen withdrawal [10]. A functional cross-talk between growth factors and growth factor receptors of the EGFR family and AR-activated pathways has been shown in preclinical models in prostatic cancer cells [11]. Moreover, it has been demonstrated that EGFR expression increases during the natural history of the prostate, suggesting that EGFR-targeted drugs might be of therapeutic relevance in CaP [12]. On that basis, we evaluated EGFR (7p12), LPL (8p21), MYC (8q24), AR (Xq12) genes and the 7, 8, X chromosome status, where the genes are located, in 79 neoplastic foci from 56 patients with clinically localized prostate cancer by fluorescence in situ hybridization (FISH). In this work we intended to investigate the following two interrelated specific aims: (i) to evaluate the association of molecular cytogenetic characteristics with histopathological classification; (ii) to detect specific genetic signatures which may identify tumors with
77
unfavorable genetic prognosis in patients undergoing radical prostatectomy.
2. Materials and methods 2.1. Tumor sampling All clinical-pathological characteristics are illustrated in Table 1. Fresh tumor samples were prospectively collected from 56 men undergoing radical retropubic ascending prostatectomy at our institute (The National Cancer Institute, Rome) for clinically localized prostate cancer. An extended regional lymphadenectomy was performed in every case. None had received preoperative androgen deprivation or radiation therapy. Fresh tumor samples were macroscopically dissected from the prostatectomy specimen immediately after operation by a pathologist. A representative piece of the CaP sample was stained with hematoxylin–eosin. The samples were prepared by cytological imprints of the specimens from both lobes which were gently pressed onto the surface of sialinized slides. In this study, the Gleason grade (G) was determined for each examined focus of prostate carcinoma. Only one case presented two neoplastic foci with different grade. Pathologic stage was assigned for each case in accordance with the Union Internationale Contre le Cancer (UICC, 2002) TNM system [13]. All samples were reviewed Table 1 Patients characteristics Characteristic PSA at diagnosis 0–3.9 4–9.9 10–19.9 O20 Gleason grade !6 6 7 O7 Pathological stage T2a–2b T2c T3a–3b T4
No.
%
4 20 20 12
7.15 35.71 35.71 21.43
1 15 31 9
1.79 26.79 55.36 16.07
8 28 19 1
14.29 50.0 33.93 1.79
78
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
by a pathologist with experience in uropathology in order to evaluate cellular adequacy and only samples with O75% of tumor cells were examined. After histological evaluation, in tumors with bilateral involvement, the slides, obtained from both lobes, were selected for genetic evaluation which was consequently performed on 79 neoplastic foci. As a control population, normal prostatic tissue samples from patients undergoing cystectomy were used. In addition, as positive control we used prostatic cell lines (LNCaP). Informed consent was obtained from the patients, and tissue sampling was approved by the ethics committee of the National Cancer Institute, Rome. 2.2. FISH analysis
2.4. Statistical analysis The c2 test was used to evaluate the association of pathological characteristics (stage and Gleason score) with each genetic marker as with the anomalous genetic patterns which were identified. The accepted statistically significant difference was P!0.05.
3. Results 3.1. Genetic profiles (chromosomes 7, 8, X and EGFR, MYC, LPL status) in neoplastic examined foci On the basis of established criteria, the aneusomy cutoff levels were 3.3, 8.4 and 9.8% for chromosomes
The procedure of FISH analysis has been previously described [14]. The Vysis ProVysion Multi-color mixture (Vysis, Inc. Downers Grove, IL) was used for detection and quantification of chromosome 8 labeled with SpectrumAqua, the 8p21 labeled with SpectrumOrange and c-MYC gene located at 8q24, labeled with SpectrumGreen. We also employed specific probes for AR (Xq12), EGFR (7p12) genes and chromosome enumeration probes (CEP) specific for X and 7 chromosomes to adjust for the effects of aneuploidy and to establish the presence of true amplification. 2.3. Scoring FISH signals To validate our FISH assay, we evaluated the hybridization patterns in normal prostatic samples obtained from 10 patients undergoing cystectomy as control group. Tumoral foci were considered aneusomic, if the percentage of nondisomic nuclei exceeded the meanC3SD of any of the signal categories observed in the control group. To evaluate the genes status, at least 100 cells were scored for each case, tabulating the total number of genes and corresponding centromere signals for each patient on a standard worksheet. We used a ratio of more than 2 oncogene/centromere signals to define gene amplification, whereas gene deletion was determined by a ratio lower than 1. The FISH assay was not available for 10 genetic determinations for technical causes.
Fig. 1. (A) Distribution of ratios (centromere/gene signals) for EGFR, AR, MYC, LPL genes. (B) Fluorescence in situ hybridization image in tumor with MYC gene amplification [CEP 8: aqua signals; 8p21 (LPL gene): red signals; MYC gene: green signals].
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
79
Table 2 Summary of the combined chromosomes 7, 8, X and EGFR, AR, MYC, LPL gene anomaly patterns in 69 prostatic cancer foci
A, aneusomic; Amp, amplified; D, deleted.
7, 8 and X respectively. FISH was successfully performed on 69 out of 79 neoplastic foci (87.3%) for all considered genetic variables. By applying the cutoff values previously described, we defined 74/77 (96.1%), 56/76 (73.7%), 26/70 (37.1%) of examined foci as having an aneusomy for chromosome 7, 8 and X respectively. No sample was amplified for EGFR and AR gene (ratio!2), while 2/71 (2.8%) samples showed MYC gene amplification with ratio values of 4.3 and 2.2 (Fig. 1). As regards LPL deletion, 52/76 (68.4%) samples were deleted (ratio!1). As illustrated in Table 2, we subclassified FISH anomalies in 12 patterns, which describe each combination of all genetic alterations occurring in this cohort of patients. Patterns2 and patterns3, observed in 21 (30.4%)
and 15 (21.7%) samples respectively, showed the higher frequency. Moreover we considered unfavorable a genetic set, if at least two aneusomic chromosomes and one altered gene were present (see patterns 2, 3, 4, 6, 12 highlighted grey in Table 2). 3.2. Association of genetic alterations and unfavorable genetic set with clinicopathologic characteristics The FISH data obtained were compared with grade and stages. The Gleason grade of the 79 cancer foci was !7 in 18 foci,Z7 in 47 foci and O7 in 14 foci. As illustrated in Table 3, a statistically significant association between the Gleason score and both chromosome 7 aneusomy and 8p21 deletion
Table 3 Correlation of FISH abnormalities (% of examined cases) with Gleason score and stage Gleason score
No. Foci
CEP 7 Aneusomy
CEP 8 Aneusomy
CEP X Aneusomy
EGFR Amplified
AR Amplified
MYC Amplified
LPL(8p21) Deleted
!7 Z O7 P value Stage T2a–T2b T2ca T3–T4a P value
18 47 14
13 (81.3) 47 (100) 14 (100) 0.003
13 (81.3) 33 (72.7) 10 (71.4) 0.74
3 (27.3) 16 (34.0) 7 (58.3) 0.23
0 (0) 0 (0) 0 (0) –
0 (0) 0 (0) 0 (0) –
0 (0) 1 (2.2) 1 (7.1) 0.51
8 (50.0) 31 (67.4) 13 (92.9) 0.04
6 (75.0) 25 (96.2) 20 (100) 0.029
7 (87.5) 19 (76.0) 17 (85.0) 0.66
1 (14.3) 5 (23.8) 11 (55.0) 0.05
0 (0) 0 (0) 0 (0) –
0 (0) 0 (0) 0 (0) –
0 (0) 1 (4.8) 1 (5.0) 0.84
5 (71.4) 14 (66.7) 19 (95.0) 0.07
a
No. patients 8 28 20
The correlation was performed with the worst genetic value detected in bilaterally examined tumors.
80
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
Table 4 Association between combined genetic patterns with Gleason Score and with Stage Gleason Score
No foci evaluated
!7 Z7 (3C4) Z7 (4C3) O7 P value Stage T2a–T2b T2c T3–T4 P value
11 34 12 12
Poor genetic patternsa
4 (36.4) 17 (50.0) 8 (66.6) 9 (75.0) 0.21 No patients evaluated 7 4 (57.1) 20 8 (40.0) 19 14 (73.7) 0.08
Good genetic patterns 7 (63.6) 17 (50.0) 4 (33.3) 3 (25.0)
3 (42.9) 12 (60.0) 5 (26.3)
a Unfavorable genetic profile (at least two aneusomic chromosomes and one altered gene).
(P!0.003 and P!0.04, respectively) was present. Moreover, we noticed a positive trend when we compared chromosome X aneusomy and the Gleason score, even though this was not statistically significant. Comparing the FISH abnormalities with stage in evaluated bilateral tumors, we considered the worst detected genetic value. The frequency of chromosome 7 aneusomy was statistically higher in T3-4 than T2c and T2a-T2b (P!0.029). In addition, even if not statistically significant, a strong positive trend was also present for chromosome X aneusomy and 8p21 deletion, especially in the transition T2c versus T3-4 stage. Examining the different genetic patterns, we observed that the percentage of tumors, with unfavorable genetic set, increased from 36.4 to 75.0% with grade and from 40.0 to 73.7% with advanced stage (Table 4). In addition, taking into consideration intermediate grades (Gleason score 7), that is the most interesting group for urologists, we found poor genetic set in 17/34 (50%) tumors with G 3C4 and in 8/12 (66.6%) with G 4C3, even if the difference was not statistically significant (PZ0.32).
4. Discussion The natural course of human CaP is highly variable and we still lack reliable tools to predict the outcome in the individual cases. Radical
prostatectomy is generally regarded as an efficient way of curing the disease. Despite attempts to restrict radical prostatectomy to patients with clinically organ-confined disease, as many as 50% of the patients are found to have extraprostatic disease at the time of surgery [15]. To identify this patient subgroup with greater precision than achievable with current clinical and pathological means, additional prognostic information is needed. A large use of cytogenetics and molecular genetic has led to the identification of tumor-associated chromosomal regions critical for the tumorigenesis and progression of CaP [16,17]. Since prostate tumor progression is undoubtedly associated with particular somatic genetic alterations, one means of characterizing each patient would be to discern any significant genetic aberrations in tumors obtained during prostatectomy. Moreover it has been shown that, in a multifocal disease, some small low-grade tumor foci had a high frequency of genetic changes, whereas concurrent dominant high-grade tumor foci were normal, indicating that small cancers can have significant alterations [18]. Thus, the size of a cancer focus and its degree of histologic dedifferentiation may not reflect the extent of its genetic instability. Others studies [19,20] had examined overexpression and/or amplification of EGFR, MYC and AR genes. In our study we investigated these genes together with chromosomes where they are located, by fluorescence in situ hybridization, to adjust for the effects of aneuploidy and to establish the presence of true amplification. Analyzing the results by a ratio (centromeric/gene signals), we did not find amplified samples, save for two that had MYC gene amplification. These patients characterized by MYC amplification and poor genetic profile (Table 2) developed disease progression (biochemical failure) after six months from surgery (data not showed) even if differently classified by conventional histology, as illustrated in Table 3. These data are in accordance with literature: high-level amplifications are very rare in primary tumors (!2%), but are more common in advanced, metastatic and androgen-independent CaP [21]. On this basis, the presence of gene amplification in tumor, at the time of prostatectomy, could be considered an important prognostic variable.
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
Particular attention should be paid to the 7, X chromosome aneusomy and LPL gene deletion that not only revealed an accumulation of aberrations closely correlated with tumor and grade, but also, to chromosome 8 aneusomy, which were the most consistent aberrations found in the foci examined, according to previous studies [18,20]. When we considered the combined FISH results for the genetic variables, the dominant FISH anomaly patterns were the #2 and #3 profiles, as illustrated in Table 2, observed in 21 (30.4%) and 15 (21.7%) samples respectively. In this investigation the percentage of tumors, with unfavorable genetic profiles (at least two aneusomic chromosomes and one gene altered) increased in parallel with the grade and stage of tumor. Our results suggest that subgroups of CaP patients with poor- and goodprognosis genetic signatures, obtained with the assessment of evaluated chromosomes and genes status, may reflect the presence of genetically defined subtypes of CaP manifesting a biological different aggressiveness and consequently a distinct course of disease progression. Besides it is of clinical relevance to characterize each patient also in the same histological classification as in intermediate grade (Gleason score 7) that is the most interesting group for clinicians. In this class we found unfavorable genetic setting in 17/34 (50.0%) tumors with G 3C4 and in 8/12 (66.6%) tumors with G 4C3 confirming the presence of different degree of genetic heterogeneity in these patients (Table 4). Sato et al. evaluated CaP prognosis by genetic profiles considering only chromosome 8 and relatives genes [22]. In this work we assembled together various well-known CaP chromosomal and gene alterations to define a complete panel that could be employed by urologists for a better CaP stratification. Moreover, in our opinion, after opportune results validation, the stratification by these genetic patterns could be employed also in preoperative core biopsies as early adjunctive diagnostic measure. The goal is to recognize through emerging technologies, such as molecular cytogenetics, pathways driving the aggressiveness of CaP, as well as to identify specific and more accurate diagnostic markers. Only future larger studies, with long-term followup of these patients, should determine the validity
81
and clinical relevance of these genetic findings with resultant incorporation of molecular prognostic markers in clinical practice. Acknowledgements We thank Mrs. Paula Franke for the formal English revision of the manuscript and Mrs. Paola Canalini for technical support. Supported by grants from Italian Ministry of Health and Regional Italian Association for Cancer Research (AIRC). References [1] P. Cairns, K. Okami, N. Halachmi, M. Esteller, J.G. Herman, J. Jen, et al., Frequent inactivation of PTEN/MMAC1 in primary prostate cancer, Cancer Res. 57 (1997) 4997–5000. [2] C. Abate-Shen, M.M. Shen, Molecular genetics of prostate cancer, Genes Dev. 14 (2000) 2410–2434. [3] J.P. Elo, T. Visakorpi, Molecular genetics of prostate cancer, Ann. Med. 33 (2001) 130–141. [4] N. Tsuchiya, J.M. Slezak, M.M. Lieber, E.J. Bergstralh, R.B. Jenkins, Clinical significance of alterations of chromosome 8 detected by fluorescence in situ hybridization analysis in pathologic organ-confined prostate cancer, Genes Chromosomes Cancer 34 (2002) 363–371. [5] H. Matsuyama, Y. Pan, K. Oba, S. Yoshihiro, K. Matsuda, L. Hagarth, et al., Deletions on chromosome 8p21 may predict disease progression as well as pathological staging in prostate cancer, Clin. Cancer Res. 7 (2001) 3139–3143. [6] D.G. Bostwick, A. Shan, J. Qian, M. Darson, N.J. Mailhle, R. B. Jenkins, L. Cheng, Independent origin of multiple foci of prostatic intraepithelial neoplasia: comparison with matched foci of prostate carcinoma, Cancer 83 (1998) 1995–2002. [7] N.N. Nupponen, L. Kakkola, P. Koivisto, T. Visakorpi, Genetic alterations in hormone-refractory recurrent prostate carcinomas, Am. J. Pathol. 153 (1998) 141–148. [8] R. Jenkins, S. Takahashi, K. De Lacey, E. Bergstralh, M. Lieber, Prognostic significance of allelic imbalance of chromosome arms 7q, 8p,16q and 18q in stage T3N0M0 prostate cancer, Genes Chromosomes Cancer 21 (1998) 131–143. [9] B. Amati, K. Alevizopoulos, J. Vlach, Myc and the cell Cycle, Front. Biosci. 3 (1998) D250–D268. [10] C. Palmerg, P. Koivisto, L. Kakkola, T.L.J. Tammela, O.P. Kallioniemi, T. Visakorpi, Androgen receptor gene amplification at the time of primary progression predicts response to combined androgen blockade as a second-line therapy in advanced prostate cancer, J. Urol. 164 (2000) 1992–1995. [11] D. Ye, J. Mendelsohn, Z. Fan, Androgen and epidermal growth factor down-regulate cyclin-dependent kinase inhibitor p27 and costimulate proliferation of MDA PCA 2a and MDA PCA 2b prostate cancer cells, Clin. Cancer Res. 5 (1999) 2171–2177.
82
M. Gallucci et al. / Cancer Letters 237 (2006) 76–82
[12] G. Di Lorenzo, G. Tortora, F.P. D’Armiento, G. De Rosa, S. Staibano, R. Autorino, et al., Expression of epidermal growth factor receptor correlates with disease relapse and progression to androgen-independence in human prostate cancer, Clin. Cancer Res. 8 (2002) 3438–3444. [13] UICC, TNM Classification of Malignant Tumors, sixth ed., Wiley-Liss, New York, 2002. [14] A.M. Cianciulli, C. Leonardo, F. Guadagni, R. Marzano, F. Iori, C. De Nunzio, et al., Genetic instability in superficial bladder cancer and adjacent mucosa: an interphase cytogenetic study, Hum. Pathol. 34 (2003) 214–221. [15] J. Trapasso, J. deKernion, R. Smith, F. Dorey, The incidence and significance of detectable levels of serum prostate specific antigen after radical prostatectomy, J. Urol. 152 (1994) 1821–1825. [16] J.C. Alers, J. Rochat, P.J. Krijtenburg, W.C.P. Hop, R. Krance, C. Rosenberg, et al., Identification of genetic markers for prostate cancer progression, Lab. Invest. 80 (2000) 931–942. [17] J.C. Alers, P.J. Krijtenburg, A.N. Vis, R.F. Hoedemaeker, M. F. Wildhagen, W.C.P. Hop, et al., Molecular cytogenetic analysis of prostatic adenocarcinomas from screening studies: early cancers may contain aggressive genetic features, Am. J. Pathol. 158 (2001) 399–406.
[18] J. Qian, K. Hirasawa, D.G. Bostwick, E.J. Bergstralh, J.M. Slezak, K.L. Anderl, et al., Loss of p53 and c-myc overrepresentation in Stage T2–3 N1–3 M0 prostate cancer are potential markers for cancer progression, Mod. Pathol. 15 (2002) 35–44. [19] E. Hermes, S.D. Fossa, A.A. Berner, B. Otnes, J.M. Nesland, Expression of the epidermal growth factor receptor family in prostate carcinoma before and during androgen-independence, Br. J. Cancer 90 (2004) 449–454. [20] H. van Dekken, J.C. Alers, I.A. Damen, K.J. Vissers, P.J. Krijtenburg, R.F. Hoedemaeker, et al., Genetic evaluation of localized prostate cancer in a cohort of forty patients: gain of distal 8q discriminates between progressors and nonprogressors, Lab. Invest. 83 (2003) 789–796. [21] C. Kaltz-Wittmer, U. Klenk, A. Glaessgen, D.E. Aust, J. Diebold, U. Lohrs, G.B. Baretton, FISH analysis of gene aberrations (MYC, CCND1, ERBB2, RB and AR) in advanced prostatic carcinomas before and after androgen deprivation therapy, Lab. Invest. 80 (2000) 1455–1464. [22] K. Sato, J. Qian, J.M. Slezak, M.M. Lieber, D.G. Bostwick, E. J. Bergstralh, R.B. Jenkins, Clinical significance of alterations of chromosome 8 in high-grade, advanced, nonmetastatic prostate carcinoma, J. Natl Cancer Inst. 91 (1999) 1574–1580.