|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 26, No 24 (August 20), 2008: pp. 3923-3929 © 2008 American Society of Clinical Oncology. DOI: 10.1200/JCO.2007.15.3155 Systems Pathology Approach for the Prediction of Prostate Cancer Progression After Radical Prostatectomy
From the Memorial Sloan-Kettering Cancer Center; Herbert Irving Comprehensive Cancer Center and Department of Pathology, Columbia University, New York; and Aureon Laboratories Inc, Yonkers, NY Corresponding author: Michael J. Donovan, MD, PhD, Aureon Laboratories Inc, 28 Wells Ave, Yonkers, NY 10701; e-mail: mdonovan{at}pathocon.com
Purpose For patients with prostate cancer treated by radical prostatectomy, no current personalized tools predict clinical failure (CF; metastasis and/or androgen-independent disease). We developed such a tool through integration of clinicopathologic data with image analysis and quantitative immunofluorescence of prostate cancer tissue. Patients and Methods A prospectively designed algorithm was applied retrospectively to a cohort of 758 patients with clinically localized or locally advanced prostate cancer. A model predicting distant metastasis and/or androgen-independent recurrence was derived from features selected through supervised multivariate learning. Performance of the model was estimated using the concordance index (CI). Results We developed a predictive model using a training set of 373 patients with 33 CF events. The model includes androgen receptor (AR) levels, dominant prostatectomy Gleason grade, lymph node involvement, and three quantitative characteristics from hematoxylin and eosin staining of prostate tissue. The model had a CI of 0.92, sensitivity of 90%, and specificity of 91% for predicting CF within 5 years after prostatectomy. Model validation on an independent cohort of 385 patients with 29 CF events yielded a CI of 0.84, sensitivity of 84%, and specificity of 85%. High levels of AR predicted shorter time to castrate prostate-specific antigen increase after androgen deprivation therapy (ADT). Conclusion The integration of clinicopathologic variables with imaging and biomarker data (systems pathology) resulted in a highly accurate tool for predicting CF within 5 years after prostatectomy. The data support a role for AR signaling in clinical progression and duration of response to ADT.
Among men diagnosed with prostate cancer, an increasing number elect local treatment, such as radical prostatectomy, with curative intent.1,2 Although the majority of patients seem to be cured by local treatment, approximately 15% to 40% will experience disease recurrence,3-5 generally first signaled by a post-therapy increase in prostate-specific antigen (PSA), which may be followed by metastasis or, for patients treated with androgen deprivation therapy (ADT), by a castrate increase in PSA. The ability to identify patients at high risk for disease progression after radical prostatectomy is critical for selecting the most appropriate therapeutic options and for rational design of clinical trials. Scientists have developed methods to predict the outcome of a particular patient's prostate cancer based on information accumulated during diagnosis, staging, and, in some models, surgery. These prognostic tools, including the Partin tables6 and the Kattan postoperative nomogram,7,8 although accurate, have limited ability to predict the prognosis for an individual patient. Furthermore, the models use clinical and pathologic variables as surrogates for biologic behavior and do not incorporate objective, quantitative features of the actual tumor sample or its molecular profile. We previously demonstrated that the systems pathology approach of integrating clinical data with cellular and biologic features associated with prostate cancer was able to accurately predict biochemical (PSA) recurrence with high accuracy.9 Here, we describe a similar approach but with enhanced histomorphologic attributes and quantitative immunofluorescence. The primary objective was to produce a time-dependent model to predict progression of prostate cancer with the clinically relevant end point of a castrate increase in PSA and/or metastasis within 5 years of radical prostatectomy. We also examined whether levels of androgen receptor (AR) predict time to disease progression after ADT.
Patients and Samples The study was approved by the institutional review board of Memorial Sloan-Kettering Cancer Center (MSKCC). Information was compiled on all patients (n = 971) treated with radical prostatectomy at MSKCC between 1985 and 2003 for localized and locally advanced prostate cancer and for whom tissue samples were available. We excluded patients who received treatment either before prostatectomy or immediately after but before biochemical recurrence, leaving 881 patients in the full cohort. The cohort was randomly assigned and evenly split between training and validation sets with similar numbers of clinical failure (CF) events. CF was prespecified as unequivocal radiographic or pathologic evidence of metastasis, castrate or noncastrate (including skeletal disease or soft tissue disease in lymph nodes or solid organs); an increasing PSA in a castrate state; or death attributed to prostate cancer. The time to CF was defined as the time from radical prostatectomy to the first of these events. If a patient did not experience CF as of his last visit or the patient outcome at the time of his most recent visit was unknown, then the patient's outcome was considered censored. Only patients with complete clinicopathologic, morphometric, and molecular data, as well as nonmissing outcome information, were further studied, leaving 373 assessable patients in the training set and 385 assessable patients in the validation set (Appendix Table A1, online only). The characteristics of these 758 patients were similar to those of both the 881 and 971 patients in the full cohorts (data not shown). Twelve tissue microarray blocks with triplicate 0.6-mm cores from 881 prostatectomy specimens were constructed at MSKCC. See Appendix (online only) for details on specimen handling and evaluation.
Image Analysis
Quantitative Multiplex Biomarker Immunofluorescence Because the majority of patients had multiple cores from which immunofluorescence features were extracted, a procedure was devised for aggregating values for each patient and feature. Four candidate functions (minimum, maximum, median, and mean) were considered (eg, minimum is the lowest value among the patient's cores). For a given feature, each of these functions was applied to aggregate the core values of each patient in the training set; then, the concordance index (CI) for each aggregation function as a predictor of CF was calculated. The CI for a feature is the probability of predicting in the correct order the feature values for two randomly chosen patients, where both patients experienced CF or one patient had CF before the last follow-up time of the censored patient. The best aggregating function for a feature (Table 1) was considered to be the one that had a CI that was farthest from random (0.5) and was then applied to all patients.
Statistical Analysis Support vector regression (SVR) for censored data (SVRc)9-12 was developed to take advantage of the ability of SVR to handle high dimensional data while adapting it for use with censored data. Our experience with SVRc has demonstrated that this approach can increase a model's predictive accuracy over that of the Cox model. See Appendix for calculation of sensitivity, specificity, hazard ratios, and CIs.
Patient Characteristics in the Training Set In the training set of 373 patients, 33 (9%) had CF after prostatectomy (24 with a positive bone scan and nine with a castrate increase in PSA). All but one of the CF patients had received hormonal therapy. The remaining patient underwent salvage radiotherapy. In addition, one patient received both salvage radiotherapy and hormonal therapy. These 373 patients were observed for a median of 76 months after prostatectomy; the overall median time to CF was not reached.
Features Derived From Quantitative Analysis of Prostate Tumor Samples
Quantitative immunofluorescence.
AR is pivotal for prostate cancer growth and progression, and the presence of AMACR has been reported to identify prostate cancer along with other pathologic processes, including high-grade prostatic intraepithelial neoplasia and atrophic glands.13 We quantified AR and AMACR in malignant epithelial cells with spectral imaging coupled with multiple antigen assessment and image analysis. This generated 18 immunofluorescent features related to AR and AMACR levels (Appendix Table A4, online only) including quantification of the (activated) nuclear form of AR within epithelial cells that express AMACR. Eleven of the immunofluorescence features displayed association with CF in univariate analysis (CI
Model Development The clinical features selected by the model were dominant prostatectomy Gleason grade and lymph node metastasis. Noteworthy is the role identified for tumor differentiation (ie, Gleason grade) and evidence of metastasis (ie, lymph node involvement) in predicting prostate cancer progression. The three imaging features selected by the model were the HE-stained properties of epithelial cytoplasm and area properties related to the size and shape of prostate cancer gland lumens. These features illustrate the importance of histochemical staining as a surrogate for biochemical or metabolic properties within the different cell types and overall tissue architecture associated with Gleason grade. Pathologists have traditionally acknowledged the importance of color when distinguishing benign from reactive or malignant cell types but have been unable to standardize and objectively quantify this observation. With the current model, the textural color properties of the epithelial cytoplasm were more strongly associated with CF than was the prostatectomy Gleason grade, suggesting the importance of cellular biochemical properties in progression of disease. Individual Kaplan-Meier curves for the three imaging features illustrate the ability to accurately stratify patients (Figs 1A to 1C).
From the 11 features quantified from immunofluorescence, the SVRc algorithm selected only the normalized average brightness/intensity of AR within AMACR-negative epithelial cells. Increasing amounts of AR were associated with a shorter time to CF. Figure 2A shows Kaplan-Meier curves for patients stratified according to this feature. Figures 2B and 2C illustrate the differing intensities observed for AR in AMACR-negative epithelial cells.
The training model had a CI of 0.92 (hazard ratio = 17.7), sensitivity of 90%, and specificity of 91% for predicting CF within 5 years of radical prostatectomy. Separate Kaplan-Meier curves were generated for patients whose SVRc score was greater than or less than 41 (corresponding to a 19.4% model-predicted probability of CF; Fig 3A); these curves illustrate the ability of the model to separate patients according to risk. See Appendix for cut point selection.
Validation The model was externally validated using data from 385 patients with a median follow-up time of 72 months. Twenty-nine patients (7.53%) had CF, 22 with positive bone scan and seven with castrate increase in PSA. All CF patients had received hormonal therapy, and seven had also received salvage radiotherapy. The six-feature model was applied with the outcome data blinded. The model's performance resulted in a CI of 0.84 (hazard ratio = 11.4), sensitivity of 84%, and specificity of 85% for predicting CF. Kaplan-Meier curves for patients with SVRc scores greater than or less than 40.94 are shown in Figure 3B. These high-risk and low-risk groups, as defined by the SVRc model cut point, showed a statistically significant difference in time to CF (log-rank test, P < .0001).
Association of AR With Response to ADT
One of the major challenges in the management of patients after radical prostatectomy is accurate assessment of the risk of recurrent disease. Given that approximately 219,000 new cases of prostate cancer were expected to be diagnosed in the United States in 200714 and that approximately 50% of these men will elect surgery,15 the scale of the problem is formidable. Although the existing predictive tools, including the updated Partin tables16 and Kattan nomogram,8 provide some benefit, the majority of patients are considered midrange, where predictive nomograms are neither accurate nor particularly helpful for decision making. Of note, because our model predicts outcome within 5 years, we evaluated the performance of the 5-year Kattan nomogram for PSA recurrence on the current validation cohort. The resulting CI was 0.75 compared with the CI of 0.84 with the current systems pathology model for CF. We also performed a Cox analysis for predicting CF using the 10 clinicopathologic parameters listed in Table 1. The Cox model had a CI of 0.81 (hazard ratio = 6.66; sensitivity and specificity = 0.79) on the validation cohort. Given that the demographic makeup of a randomly assigned group will affect predictive analyses, we further evaluated the robustness of our current model by comparing Cox analysis with the systems pathology approach on 100 balanced splits of the total patient cohort. Importantly, we found that the average CI on validation was 0.80 (average hazard ratio = 6.37) using the standard Cox approach and 0.83 (average hazard ratio = 9.11) using the systems approach, which was a statistically significant difference (P < .001; see Appendix). As established, the CI, hazard ratio, sensitivity, and specificity results confirm an incremental and significant increase in predictive accuracy when objective, patient tissue–specific features (ie, morphometry and antigen profiles) are included in a systems model predicting clinical outcome. To illustrate improved risk stratification, we calculated the hazard ratio for intermediate-risk patients (ie, Gleason score of 7 without evidence of lymph node involvement) in the validation cohort using all standard clinical pathologic parameters. The clinical model resulted in a hazard ratio of 2.7 compared with 5 for a systems approach, demonstrating a significant improvement in assigning risk for intermediate patients when additional tumor-specific features are incorporated. We have developed an integrative predictive test to determine the likelihood of a patient having clinically demonstrated recurrence within 5 years after prostatectomy. This model is quite different from our previous study, which focused on a PSA recurrence end point and used immunohistochemistry profiles for biomarker selection.9 The resultant patient risk profile derived from each model is distinct and represents a temporal and dynamic feature content that directly reflects the nature of the predicted clinical outcome. In contrast to standard clinical models, preoperative PSA and Gleason sum information were not selected. In the competition imposed by SVRc, these variables were supplanted by image analysis features that reflect differentiation and color characteristics linked to the biochemical properties of the prostate cancer. We believe that the highly accurate predictions of the model will allow the development of more informed and appropriate treatment plans. This would include the possibility of administering ADT, radiation therapy, and/or chemotherapy earlier if there is a high risk of recurrence. In addition, early identification of patients at high risk creates the opportunity to increase surveillance for progression. The model predictions may also be useful in guiding treatment decisions after biochemical recurrence. Some patients with biochemical recurrence may not require immediate intervention because the interval between biochemical recurrence and clinically significant events is typically years and also quite variable. Furthermore, the overall risk that a man will die of prostate cancer is approximately 3%.17 Thus, many men diagnosed with biochemical recurrence today are likely to have an indolent disease course. Knowledge of the probability of aggressive disease (ie, disease that will progress to CF within 5 years of surgery) may help the patient and his physician decide whether biochemical recurrence should trigger aggressive therapy. ADT, in use for more than 60 years,18 is currently the first line of systemic therapy against advanced prostate cancer. Although several studies have argued the benefits of early hormonal therapy in intermediate-risk (ie, locally advanced) and high-risk patients,19,20 the impact on overall survival continues to be debated.21-23 Moreover, the observed bone loss, anemia, and physical and psychosocial alterations from ADT have raised questions about its appropriate use, given that only a small percentage of patients will actually die from prostate cancer.23 Our data demonstrate an association between high levels of AR and shortened time to increase in PSA after ADT. Previously, we found that stronger immunohistochemical and quantitative immunofluorescent staining for AR in tumor epithelial cells was associated with a shorter time to PSA recurrence.9 AR was first evaluated as a predictor of response to hormonal therapy (with respect to progression of disease) 15 years ago by Sadi and Barrack.24 They later showed that variability of AR protein expression in prostate cancer metastases was correlated with poor response to hormonal therapy and, therefore, with progression of disease.25 AR by itself may be oncogenic, as evidenced by the observation that the majority of AR alterations are associated with gain as opposed to loss of function. In addition, Chen et al26 demonstrated that high levels of AR RNA and protein are both necessary and sufficient to convert cancer cells to castration resistance. Similarly, reduced AMACR expression has been associated with prostate cancer progression.27 Our studies support these initial analyses and suggest that AR content within this population of cells is the primary mechanism by which disease advances. Intriguing evidence from prostate cancer cell lines, including LNCaP, has demonstrated that a stable and intact AR facilitates apoptosis in response to radiation,28,29 suggesting an even more expanded biologic role for AR signaling. All of these findings illustrate the importance of AR and the androgen axis in prostate cancer growth and survival. With the evidence that AR plays a significant role in prostate cancer, we suggest that accurate quantitation of this receptor at the time of prostatectomy (and biopsy) might allow for more appropriate clinical management, analogous to the use of hormonal receptor status in breast cancer for predicting response to chemotherapy.30 By using the AR content of the tumor sample, the urologist may be able to predict utility of ADT and adjust therapeutic regimens accordingly. Patients with high scores in the current CF model and high levels of AR face a particular dilemma—an elevated probability of clinical progression accompanied by a decreased likelihood of durable response to ADT and possibly to salvage radiotherapy. These patients may be good candidates for therapies, such as histone deacetylase inhibitors, that target AR or its cofactors. Studies are underway to investigate a role for AR quantitation in therapeutic management. Such approaches could potentially improve patient selection for clinical trials and serve as a measure of treatment effect, specifically for therapies that target the AR signaling axis.
Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a ‘U’ are those for which no compensation was received; those relationships marked with a ‘C’ were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: Michael J. Donovan, Aureon Laboratories Inc; Stefan Hamann, Aureon Laboratories Inc (C); Mark Clayton, Aureon Laboratories Inc (C); Faisal M. Khan, Aureon Laboratories Inc (C); Marina Sapir, Aureon Laboratories Inc (C); Valentina Bayer-Zubek, Aureon Laboratories Inc (C); Gerardo Fernandez, Aureon Laboratories Inc (C); Ricardo Mesa-Tejada, Aureon Laboratories Inc (C); Mikhail Teverovskiy, Aureon Laboratories Inc (C); Carlos Cordon-Cardo, Aureon Laboratories Inc. Compensation: Board Member, Founder Consultant or Advisory Role: Michael J. Donovan, Aureon Laboratories Inc (C) Stock Ownership: Stefan Hamann, Aureon Laboratories Inc; Mark Clayton, Aureon Laboratories Inc; Faisal M. Khan, Aureon Laboratories Inc; Marina Sapir, Aureon Laboratories Inc; Valentina Bayer-Zubek, Aureon Laboratories Inc; Gerardo Fernandez, Aureon Laboratories Inc; Ricardo Mesa-Tejada, Aureon Laboratories Inc; Mikhail Teverovskiy, Aureon Laboratories Inc; Carlos Cordon-Cardo, Aureon Laboratories Inc Honoraria: None Research Funding: None Expert Testimony: None Other Remuneration: None
Conception and design: Michael J. Donovan, Peter T. Scardino, Carlos Cordon-Cardo Administrative support: Michael J. Donovan Provision of study materials or patients: Michael J. Donovan, Faisal M. Khan, Victor E. Reuter, Peter T. Scardino Collection and assembly of data: Michael J. Donovan, Stefan Hamann, Mark Clayton, Faisal M. Khan, Marina Sapir, Valentina Bayer-Zubek, Gerardo Fernandez, Ricardo Mesa-Tejada, Mikhail Teverovskiy Data analysis and interpretation: Michael J. Donovan, Stefan Hamann, Mark Clayton, Faisal M. Khan, Marina Sapir, Valentina Bayer-Zubek, Gerardo Fernandez, Ricardo Mesa-Tejada, Mikhail Teverovskiy, Carlos Cordon-Cardo Manuscript writing: Michael J. Donovan, Faisal M. Khan, Marina Sapir, Valentina Bayer-Zubek, Mikhail Teverovskiy, Carlos Cordon-Cardo Final approval of manuscript: Michael J. Donovan, Faisal M. Khan, Marina Sapir, Valentina Bayer-Zubek, Peter T. Scardino, Carlos Cordon-Cardo
We thank Angeliki Kotsianti, MD; David A. Verbel; Olivier Saidi, PhD; Vijay Aggarwal, PhD; Robert Shovlin; Charles DiComo, PhD; Ho-Yuen Pang, PhD; and Stephen Fogarsi for their contributions to the completion of this article.
Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Hull GW, Rabbani F, Abbas F, et al: Cancer control with radical prostatectomy alone in 1,000 consecutive patients. J Urol 167:528-534, 2002[CrossRef][Medline] 2. Johansson JE, Andren O, Andersson SO, et al: Natural history of early, localized prostate cancer. JAMA 291:2713-2719, 2004 3. Kattan MW, Eastham JA, Stapleton AM, et al: A preoperative nomogram for disease recurrence following radical prostatectomy for prostate cancer. J Natl Cancer Inst 90:766-771, 1998 4. Sun L, Gancarczyk K, Paquette E, et al: Introduction to Department of Defense Center for Prostate Disease Research Multicenter National Prostate Cancer Database, and analysis of changes in the PSA-era. Urol Oncol 6:203-209, 2001[CrossRef] 5. Ward JF, Blute ML, Slezak J, et al: The long-term clinical impact of biochemical recurrence of prostate cancer 5 or more years after radical prostatectomy. J Urol 170:1872-1876, 2003[CrossRef][Medline] 6. Pound CR, Partin AW, Eisenberger MA, et al: Natural history of progression after PSA elevation following radical prostatectomy. JAMA 281:1591-1597, 1999 7. Kattan MW, Wheeler TM, Scardino PT: Postoperative nomogram for disease recurrence after radical prostatectomy for prostate cancer. J Clin Oncol 17:1499-1507, 1999 8. Stephenson AJ, Scardino PT, Eastham JA, et al: Postoperative nomogram predicting the 10-year probability of prostate cancer recurrence after radical prostatectomy. J Clin Oncol 23:7005-7012, 2005 9. Cordon-Cardo C, Kotsianti A, Verbel DA, et al: Improved prediction of prostate cancer recurrence through systems pathology. J Clin Invest 117:1876-1883, 2007[CrossRef][Medline] 10. Cristianini N, Shawe-Taylor J: An Introduction to Support Vector Machines and Other Kernel-Based Learning Methods. Cambridge, United Kingdom, Cambridge University Press, 2000 11. Lee Y-J, Mangasarian OL, Wolberg WH: Breast cancer survival and chemotherapy: A support vector machine analysis. DIMACS Series Discret Math Theor C 55:1-10, 2000 12. Li H, Luan Y: Kernel Cox regression models for linking gene expression profiles to censored survival data. Pac Symp Biocomput 65-76, 2003 13. Hameed O, Humphrey PA: Immunohistochemistry in diagnostic surgical pathology of the prostate. Semin Diagn Pathol 22:88-104, 2005[CrossRef][Medline] 14. Jemal A, Siegel R, Ward E, et al: Cancer statistics, 2007. CA Cancer J Clin 57:43-66, 2007 15. Cooperberg MR, Broering JM, Litwin MS, et al: The contemporary management of prostate cancer in the United States: Lessons from the Cancer of the Prostate Strategic Urologic Research Endeavor (CapSURE), a national disease registry. J Urol 171:1393-1401, 2004[CrossRef][Medline] 16. Freedland SJ, Humphreys EB, Mangold LA, et al: Risk of prostate cancer-specific mortality following biochemical recurrence after radical prostatectomy. JAMA 294:433-439, 2005 17. Jemal A, Siegel R, Ward E, et al: Cancer statistics, 2006. CA Cancer J Clin 56:106-130, 2006 18. Huggins C, Hodges CV: Studies on prostate cancer: I. The effects of castration, of estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. Cancer Res 1:293-297, 1941 19. Messing EM, Manola J, Sarosdy M, et al: Immediate hormonal therapy compared with observation after radical prostatectomy and pelvic lymphadenectomy in men with node-positive prostate cancer. N Engl J Med 341:1781-1788, 1999 20. Wirth M, Tyrrell C, Delaere K, et al: Bicalutamide (Casodex) 150 mg in addition to standard care in patients with nonmetastatic prostate cancer: Updated results from a randomised double-blind phase III study (median follow-up 5.1 y) in the early prostate cancer programme. Prostate Cancer Prostatic Dis 8:194-200, 2005[CrossRef][Medline] 21. Messing EM, Thompson I Jr: Follow-up of conservatively managed prostate cancer: Watchful waiting and primary hormonal therapy. Urol Clin North Am 30:687-702, 2003[CrossRef][Medline] 22. Moul JW, Wu H, Sun L, et al: Early versus delayed hormonal therapy for prostate specific antigen only recurrence of prostate cancer after radical prostatectomy. J Urol 171:1141-1147, 2004[CrossRef][Medline] 23. Sharifi N, Gulley JL, Dahut WL: Androgen deprivation therapy for prostate cancer. JAMA 294:238-244, 2005 24. Sadi MV, Barrack ER: Androgen receptors and growth fraction in metastatic prostate cancer as predictors of time to tumour progression after hormonal therapy. Cancer Surv 11:195-215, 1991[Medline] 25. Sadi MV, Barrack ER: Image analysis of androgen receptor immunostaining in metastatic prostate cancer: Heterogeneity as a predictor of response to hormonal therapy. Cancer 71:2574-2580, 1993[CrossRef][Medline] 26. Chen CD, Welsbie DS, Tran C, et al: Molecular determinants of resistance to antiandrogen therapy. Nat Med 10:33-39, 2004[CrossRef][Medline] 27. Rubin MA, Bismar TA, Andren O, et al: Decreased alpha-methylacyl CoA racemase expression in localized prostate cancer is associated with an increased rate of biochemical recurrence and cancer-specific death. Cancer Epidemiol Biomarkers Prev 14:1424-1432, 2005 28. Harashima K, Akimoto T, Nonaka T, et al: Heat shock protein 90 (Hsp90) chaperone complex inhibitor, radicicol, potentiated radiation-induced cell killing in a hormone-sensitive prostate cancer cell line through degradation of the androgen receptor. Int J Radiat Biol 81:63-76, 2005[Medline] 29. Lin Y, Kokontis J, Tang F, et al: Androgen and its receptor promote Bax-mediated apoptosis. Mol Cell Biol 26:1908-1916, 2006 30. Berry DA, Cirrincione C, Henderson IC, et al: Estrogen-receptor status and outcomes of modern chemotherapy for patients with node-positive breast cancer. JAMA 295:1658-1667, 2006 Submitted November 13, 2007; accepted March 21, 2008.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2008 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
-methylacyl coenzyme A racemase (AMACR), cytokeratin 18 (CK18), TP73L (p63), and high molecular weight keratin were used, along with 4`-6-diamidino-2-phenylindole as nuclear stain, in a quint-plex assay (Appendix Table A2, online only). On the basis of the spectral profiles of the fluorochromes, antigen-specific gray-scale images were acquired. Software algorithms were used to localize the individual antigens. Using antigen distribution and intensity, the algorithms identified cell types and compartments (eg, luminal epithelial cells, epithelial/stromal nuclei) and quantified AR and AMACR in prostate tumor, benign glands, and stroma. Patients were excluded if their cores were missing from the array or if the sample contained only stroma. 
0.4 or CI 
0.6; 
