|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2005.05.1458 on July 5 2006 © 2006 American Society of Clinical Oncology. Influence of Surgical Manipulation on Prostate Gene Expression: Implications for Molecular Correlates of Treatment Effects and Disease Prognosis
From the Divisions of Human Biology and Public Health Sciences, Fred Hutchinson Cancer Research Center; and the Departments of Urology and Epidemiology, University of Washington, Seattle, WA Address reprint requests to Daniel W. Lin, MD, Department of Urology, Box 356510, 1959 NE Pacific St, University of Washington, Seattle, WA 98195; e-mail: dlin{at}u.washington.edu
Purpose: Measurements of tissue gene expression are increasingly used for disease stratification, clinical trial eligibility, and assessment of neoadjuvant therapy response. However, the method of tissue acquisition alone could significantly influence the expression of specific transcripts or proteins. This study examines whether there are transcript alterations associated with surgical resection of the prostate gland by radical retropubic prostatectomy. Materials and Methods: Twelve patients with clinically localized prostate cancer underwent immediate in situ prostate biopsy after induction of anesthesia for radical prostatectomy. Ex vivo prostate biopsies were performed immediately after surgical removal. Prostate epithelium was acquired by laser-capture microdissection, and transcript abundance levels were quantitated by cDNA microarray hybridization and confirmed by quantitative polymerase chain reaction. Data were analyzed by paired, two-sample t test using Statistical Analysis of Microarray algorithms, and linear models were fit as a function of clinical characteristics.
Results: Of 5,753 cDNAs with measurable expression in prostate epithelium, 88 (1.5%) were altered as a result of surgery (false-discovery rate Conclusion: Surgical manipulation results in significant gene expression changes. Molecular analyses of surgical samples should recognize that transcript alterations occur rapidly, and these results are important when designing and analyzing molecular correlates of clinical studies.
Measurements of tissue gene expression by techniques such as cDNA microarray hybridization or quantitative polymerase chain reaction (PCR) -based assays are increasingly used to evaluate differential gene expression profiles.1-3 Gene expression signatures are used to classify disease severity, predict clinical outcomes, and define criteria for clinical trial eligibility.4-7 They can also be used to evaluate response to neoadjuvant therapies, on the basis of comparisons of pretreatment biopsy tissue with tissue from post-treatment surgical resection specimens.5,8-10 However, surgical manipulation may cause changes in gene expression and thus obscure measures of treatment effects or disease prognosis. This study examines whether there are transcript alterations associated with surgical resection of the prostate gland by radical retropubic prostatectomy. These results are important for the design of studies using gene expression or, more globally, gene expression signatures, as measures of disease prognosis, clinical trial eligibility and treatment response.
Study Design and Tissue Acquisition Participants were 12 patients with newly diagnosed, clinically localized prostate cancer and were selected as consecutive patients undergoing radical prostatectomy and consenting to study procedures. No participant received previous hormonal ablation, chemotherapy, or radiotherapy. The study design is shown in Figure 1. At time of radical prostatectomy and after induction of anesthesia, we obtained four in situ prostate biopsy cores using an 18-gauge prostate needle biopsy gun (Bard Inc, Murray Hill, NJ), and immediately embedded and froze them. Radical prostatectomy was performed as follows: The prostate was mobilized from surrounding tissue, and the dorsal venous complex was divided and oversewn. After dissecting the adjacent neurovascular bundle(s), where appropriate, the urethra was incised, the prostate was dissected free of the rectum, and the vascular pedicles were ligated. The seminal vesicles were then mobilized from the surrounding tissue, and the prostate was dissected free of the bladder neck and removed. We immediately obtained four ex vivo biopsy cores following the same protocol just described. The time interval from ligation of vascular pedicles to ex vivo biopsy was defined as the ischemia time. All protocols for tissue acquisition, processing, and analysis were approved by the University of Washington (Seattle, WA) internal review board.
Specimen Handling and Laser-Capture Microdissection Biopsy cores were embedded individually in polyethylene glycol freezing media (Tissue-Tek OCT Compound, Sakura Finetek, Torrance, CA) and placed in isopentane that was precooled in liquid nitrogen. Specimens were then placed at –80°C until laser-capture microdissection (LCM). At time of LCM, 8-µm frozen sections were obtained and immediately fixed in cold 95% ethanol. Sections were briefly (5 to 10 seconds) stained with hematoxylin using the Arcturus HistoGene Staining Solution (Arcturus Bioscience, Mountain View, CA) and dehydrated in 100% ethanol followed by xylenes (described in the Arcturus HistoGene LCM Frozen Section Staining Kit protocol). Five thousand epithelial cells from histologically benign glands were obtained by LCM using the Arcturus PixCell II instrument (modified from Emmert-Buck11). Digital photos were taken of tissue sections before, during, and after LCM and assessed by an experienced genitourinary pathologist (B.S.K.) to confirm the histology of the laser captured cells. Each slide capture session lasted no longer than 20 minutes to minimize RNA degradation. The laser capture settings were 55 mW beam, 1.5 ms pulse, and 15 µm spot size. Cells were lysed in Arcturus RNA Extraction Buffer, RNA was isolated using the Arcturus PicoPure RNA Isolation Kit, and the samples were treated with DNAse using the Qiagen RNase-Free DNAse Set (Qiagen Inc, Valencia, CA). RNA was amplified for two rounds using the Ambion MessageAmp aRNA Kit (Ambion Inc, Austin, TX), and sample quality and quantity were assessed by agarose gel electrophoresis and absorbance at A260. To provide a reference standard RNA for use on cDNA microarrays, we isolated total RNA from LNCaP, DU145, PC3, and CWR22 cell lines (American Type Culture Collection, Manassas, VA) growing at log phase in dye-free RPMI-1640-1640 medium supplemented with 10% fetal bovine serum (FBS; Life Technologies, Rockville, MD). RNA was purified using Trizol (Life Technologies) following the manufacturer's protocol.
Gene Expression Analysis by Microarray Hybridization
Statistical and Pathway Analyses We used Gene Microarray Pathway Profiler (GenMAPP) software14 for functional pathway analyses. Briefly, pathway maps were downloaded from www.GenMAPP.org, and microarray expression data incorporated. Gene expression data on pathways were color-coded by average fold-change for individual genes.
Quantitative Reverse-Transcription PCR Relative quantification of gene expression by quantitative PCR (40 cycles of 60°C annealing, 72°C extension, and 95°C melting) was performed on a 7700 Sequence Detector using SYBR Green Master mix and gene-specific primers following the manufacturer's recommendations. (Applied Biosystems)
Participant Demographic and Clinical Characteristics The demographic characteristics and operative details of the 12 participants are given in Table 1. Median participant age was 64 years, median preoperative serum PSA was 5.2 ng/mL, and most men were white. Median prostate volume was 43 mL (range, 31 to 149 mL), and nine procedures (75%) included cavernosal nerve-sparing. Median operative time was 196 minutes (range, 164 to 221), median ischemia time was 28 minutes (range 19 to 40 minutes), and median estimated blood loss was 875 mL.
Gene Expression Changes Associated With Surgical Manipulation Table 2 gives the distribution of differences in genes expression between pre- and postsurgical specimens for 5,753 cDNAs. Differences in this table are categorized symmetrically above and below zero in units of log2, which are also shown as relative expression on a linear scale. In more than 96% of transcripts, relative expression of pre- and postsurgical specimens was between 0.67 and 1.49. However, the distribution of differences in expression was not symmetric. Comparing post- with presurgical tissues, expression of 41 transcripts was higher by two-fold or more, whereas none was lower by 50% or less. Surgical manipulation was associated with significantly higher transcript levels in 2.3% of the cDNAs using a criterion of a 15% FDR, and with 1% of the cDNAs using a more conservative criteria of a 5% FDR. Even using a relatively high 15% FDR, no transcript levels were significantly lower in the postsurgical specimens.
Figure 2 shows the list of 62 unique genes (1.5%; from 88 cDNAs) that showed higher expression in postsurgical specimens with FDR 10% or lower. These include several genes involved in the acute phase response, IER2 and JUNB, and the regulation of cell proliferation, p21Cip1 and KLF6.
Confirmation With qRT-PCR We performed qRT-PCR for transcripts encoding DUSP1 and KLF6, which were significantly higher on the basis of microarray results, and for the transcript encoding alpha-methylacyl-CoA racemase (AMACR), which did not differ. Figure 3 shows median and range of threshold cycles for pre- and postsurgical specimens; the y-axis is log2 threshold cycle, and a lower threshold cycle of detection reflects higher message (ie, upregulation). On the basis of qRT-PCR, both DUSP1 and KLF6 were significantly upregulated (14.7 ± 5.5, P = .001 and 4.2 ± 1.1, P = .05, mean fold change ± quartile, respectively) compared with 3.7 and 3.6 mean fold changes found by microarray, respectively. There was no difference in qRT-PCR expression of AMACR (mean 0.8 ± 0.6 fold, P = .94), which was consistent with microarray results (1.1 mean fold change).
Effects of Clinical and Operative Characteristics on Differences in Gene Expression We examined effects of age, PSA, total operative time, gland volume, ischemia time and estimated blood loss on the differential expression of the 62 genes (with FDR < 10%) associated with surgical manipulation. For all characteristics except gland volume, most effect sizes were very small, and among each set of 62 genes evaluated for clinical characteristic, the number of P values that were less than.05 ranged from 0 for operative time to 3 for age. In contrast, gland volume was significantly and negatively associated with the magnitude of gene expression differences. Specifically, in 55 (89%) of 62 genes, differences between pre- and postsurgical expression were smaller as gland size increased. For eight (13%) of these genes, associations with prostate volume were statistically significant (P < .05) and did not differ whether the analyses included or excluded the outlier with a 149-g prostate volume. Among these eight genes, the median reduction in prepost surgery expression difference was 13% for each 10 mL of gland volume, with a range of 10% to 26%. For several genes, there were no differences in transcript expression in tissues from the largest prostates.
Molecular Pathway Analysis
In this study of the effects of surgical manipulation on gene expression in prostate tissue, we found that significant transcript alterations occur simply as a result of surgical excision. In particular, approximately 5% of the measured transcripts were changed greater than 1.5 fold, and 1.5% exhibited changes statistically associated with the process of surgical resection. Furthermore, no genes were downregulated more than 50%, and all genes that were significantly different (FDR  15%) were upregulated. Lastly, higher prostate gland volume was associated with smaller differences in pre-/post-transcript expression. Although we hypothesized that transcript levels would be affected by surgical manipulation, two findings were surprising. First, we found that no transcripts were significantly downregulated in postsurgical specimens. One explanation for this may be that downregulation would require active degradation of transcripts, and significant degradation is unlikely to occur during the relatively short operative time. Second, we found that differences in transcript expression decreased as prostate gland size increased. Larger prostate glands may have more robust blood supply, thus protecting them from ischemic damage; however, studies examining the association of prostate gland vascularity to volume have failed to reveal any significant correlation.15 We also found that neither age, PSA concentration, total operative time, ischemia time, nor estimated blood loss significantly affected differences in prepost surgical transcript expression; however, we cannot exclude these characteristics as possible modulators of pre-/postsurgical differences because statistical power to detect significant differences was modest. These results are important in at least two situations for which cDNA gene expression signatures are used. First, in the context of evaluating neoadjuvant therapy, particularly when comparing pretreatment biopsy tissue with post-treatment surgical tissue, procedure-associated transcript changes should be identified so that those genes affected by surgical manipulation will not be used to evaluate treatment outcomes. Second, in studies that use expression signatures to predict clinical outcomes or to determine clinical trial eligibility, expression of genes affected by surgery will no longer reflect the underlying tumor biology. For example, we found that surgical excision altered the expression of genes within the JNK pathway shown in Figure 4. JNK is required for in vitro and in vivo growth of prostate carcinoma cells and in resistance to Fas receptor-mediated apoptosis and apoptosis in response to chemotherapeutic drugs.16-18 Furthermore, the JNK pathway is a proposed novel target in the treatment of prostate carcinoma, and specific inhibitors of the this pathway are under investigation,16,18 underscoring the immediate clinical relevance of our study. Upregulated gene expression of the JUN transcripts in postsurgical tissue may be interpreted as elevated JNK activity and possible resistance to chemotherapeutic regimens; however, in our study, these genes are significantly upregulated by the process of tumor removal, independent of tumor biology. One important limitation to this study is that we did not evaluate whether gene expression differed between multiple biopsy cores obtained from the same individual at the same point in time. Thus, we not could evaluate the extent to which our findings could be due to random core-to-core variability. However, we did evaluate gene expression in the diagnostic biopsy obtained from study participants several weeks before surgery, and found that the expression of only three (0.05%) of the 5,753 genes differed significantly between the diagnostic and in situ presurgical biopsies (Fig A1, online only). We infer from these results that gene expression patterns within an individual are relatively invariant, and thus not likely to explain this study's results. Other potential limitations of our study deserve mention. First, we sampled approximately 6,000 genes of the prostate tissue transcriptome, and additional genes that were not evaluated could be affected to an equal or greater extent. Second, it remains to be determined whether these procedure-associated changes in transcript levels result in corresponding alterations in the cognate proteins, although several previously published studies examining gene expression signatures in clinical trials did not evaluate or require corresponding protein alterations.4-7,10 Lastly, it is possible that the ischemic effects of surgery could differentially affect healthy tissues compared with regions of cancer where robust neovascularization may attenuate compromised blood flow. However, we assume that the complete cessation of blood flow would result in significant changes, regardless of the vascular status. Several strengths of this study should also be noted. First, our study is the first to examine the effects of surgical manipulation alone on global gene expression. One previous study examined the effect of ischemia time after removal of the prostate, revealing that several stress response genes were upregulated with increasing time of tissue ischemia at room temperature.19 Second, by obtaining and freezing the tissue immediately on surgical excision, our study design represented the optimum conditions for tissue preservation. The time from tumor removal to tissue processing is rarely recorded or cited in the literature, and our results probably underestimate the effects of surgical manipulation on transcript changes. Furthermore, resections of many solid tumors are now routinely performed laparoscopically, a procedure that may increase tissue ischemia time, and importantly, the tissue in laparoscopic procedures often suffers ischemia time at body temperature rather than at room temperature after open surgery. Lastly, we examined all potential covariates that could be confounders and found only one, gland volume, that was associated with the gene expression findings. Of note, ischemia time and total operative time were not associated with magnitude of gene expression changes. In summary, the process of acquiring tissues for molecular analyses results in significant gene expression changes. Molecular analyses of surgical samples should recognize that transcript alterations occur rapidly, and these changes may be associated with the method of tissue acquisition and be mistaken for markers of disease severity or disease response to treatment. Because our study design used ideal tissue collection and processing conditions, it is unknown whether less optimal protocols would result in additional gene expression changes, and our results are likely overly conservative. However, most genes were not statistically affected, thus, depending on the study design and end points, surgical specimens may indeed prove adequate. These results are important when designing and analyzing molecular correlates of clinical studies, particularly when comparing samples obtained through disparate collection methods or when comparing tissue samples from different surgical centers where postsurgical processing methods are not standardized.
The authors indicated no potential conflicts of interest.
published online ahead of print at www.jco.org on July 5, 2006. Supported by Grant No. DK65083 (D.W.L.), the Pacific Northwest Prostate Cancer SPORE Grant No. CA97186 (D.W.L. and P.S.N.), and Grant No. CA85859 (P.S.N.). Terms in blue are defined in the glossary, found at the end of this article and online at www.jco.org. Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Ramaswamy S, Golub TR: DNA microarrays in clinical oncology. J Clin Oncol 20:1932-1941, 2002 2. Rhodes DR, Chinnaiyan AM: Integrative analysis of the cancer transcriptome. Nat Genet 37:S31-S37, 2005 (suppl)[CrossRef][Medline] 3. Golub TR, Slonim DK, Tamayo P, et al: Molecular classification of cancer: Class discovery and class prediction by gene expression monitoring. Science 286:531-537, 1999 4. Chen CN, Lin JJ, Chen JJ, et al: Gene expression profile predicts patient survival of gastric cancer after surgical resection. J Clin Oncol 23:7286-7295, 2005 5. Takata R, Katagiri T, Kanehira M, et al: Predicting response to methotrexate, vinblastine, doxorubicin, and cisplatin neoadjuvant chemotherapy for bladder cancers through genome-wide gene expression profiling. Clin Cancer Res 11:2625-2636, 2005 6. Jansen MP, Foekens JA, van Staveren IL, et al: Molecular classification of tamoxifen-resistant breast carcinomas by gene expression profiling. J Clin Oncol 23:732-740, 2005 7. Iwao-Koizumi K, Matoba R, Ueno N, et al: Prediction of docetaxel response in human breast cancer by gene expression profiling. J Clin Oncol 23:422-431, 2005 8. Chang JC, Wooten EC, Tsimelzon A, et al: Patterns of resistance and incomplete response to docetaxel by gene expression profiling in breast cancer patients. J Clin Oncol 23:1169-1177, 2005 9. Ochi K, Daigo Y, Katagiri T, et al: Prediction of response to neoadjuvant chemotherapy for osteosarcoma by gene-expression profiles. Int J Oncol 24:647-655, 2004[Medline] 10. Chang JC, Wooten EC, Tsimelzon A, et al: Gene expression profiling for the prediction of therapeutic response to docetaxel in patients with breast cancer. Lancet 362:362-369, 2003[CrossRef][Medline] 11. Moore S, Knudsen B, True LD, et al: Loss of stearoyl-CoA desaturase expression is a frequent event in prostate carcinoma. Int J Cancer 114:563-571, 2005[CrossRef][Medline] 12. Tusher VG, Tibshirani R, Chu G: Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 98:5116-5121, 2001 13. Dahlquist KD, Salomonis N, Vranizan K, et al: GenMAPP, a new tool for viewing and analyzing microarray data on biological pathways. Nat Genet 31:19-20, 2002[CrossRef][Medline] 14. Cetinkaya M, Gunce S, Ulusoy E, et al: Relationship between prostate specific antigen density, microvessel density and prostatic volume in benign prostatic hyperplasia and advanced prostatic carcinoma. Int Urol Nephrol 30:581-585, 1998[Medline] 15. Yang YM, Bost F, Charbono W, et al: C-Jun NH(2)-terminal kinase mediates proliferation and tumor growth of human prostate carcinoma. Clin Cancer Res 9:391-401, 2003 16. Potapova O, Anisimov SV, Gorospe M, et al: Targets of c-Jun NH(2)-terminal kinase 2-mediated tumor growth regulation revealed by serial analysis of gene expression. Cancer Res 62:3257-3263, 2002 17. Curtin JF, Cotter TG: JNK regulates HIPK3 expression and promotes resistance to Fas-mediated apoptosis in DU 145 prostate carcinoma cells. J Biol Chem 279:17090-17100, 2004 18. Dash A, Maine IP, Varambally S, et al: Changes in differential gene expression because of warm ischemia time of radical prostatectomy specimens. Am J Pathol 161:1743-1748, 2002 Submitted November 29, 2005; accepted March 10, 2006. This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
tibs/SAM/
