|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2007.15.0441 on September 15 2008 © 2008 American Society of Clinical Oncology. Prognostic Significance of Disseminated Tumor Cells in the Bone Marrow of Prostate Cancer Patients Treated With Neoadjuvant Hormone Treatment
From the Institutes of Pathology, Tumor Biology, and Experimental and Clinical Pharmacology, University Medical Center Hamburg-Eppendorf; Department of Urology, Campus Benjamin Franklin, Charite' Universitätsmedizin Berlin; and the Department of Urology, Dr Horst-Schmidt-Kliniken, Wiesbaden, Germany Corresponding author: Klaus Pantel, MD, PhD, Institute of Tumor Biology, University Medical Center Hamburg-Eppendorf, Martinistr 52, D-20246 Hamburg, Germany; e-mail: pantel{at}uke.uni-hamburg.de
Purpose To explore whether the presence of occult disseminated tumor cells (DTCs) in the bone marrow before neoadjuvant hormone therapy influences the prognosis of patients with organ confined prostate cancer treated by radical prostatectomy. Patients and Methods Pretreatment bone marrow aspirates from 193 cT (1-4) pN0M0 prostate cancer patients submitted to neoadjuvant hormone therapy (mean, 8 months) followed by radical prostatectomy were immunohistochemically evaluated by anticytokeratin antibody A45-B/B3 previously validated for the detection of DTCs. Bone marrow status was compared with established clinical and histopathologic risk parameters. Patients’ outcome was evaluated using prostate-specific antigen (PSA) blood serum measurements as surrogate marker for recurrence over a median follow-up of 44 months. Results DTCs were detected in 44.6% of patients. Bone marrow status neither correlated with tumor grade and stage, nor with the pretreatment PSA risk category (all P values > .05). In the univariate Kaplan-Meier analysis, the presence of DTCs was a significant prognostic factor with respect to poor PSA progression-free survival (log-rank test P = .0035). Using a multivariable piecewise Cox regression model, the presence of DTCs was an independent predictor of PSA relapse (relative risk 1.82; P = .014). Conclusion The presence of DTCs in the bone marrow of patients with prostate cancer before neoadjuvant hormone therapy and subsequent surgery represents an independent prognostic parameter, suggesting that DTCs may contribute to the failure of current neoadjuvant hormone therapy regimens.
Prostate cancer, the most frequent malignancy in men, is a heterogeneous tumor with respect to its biologic aggressiveness and clinical course. This behavior taken into account, modern therapy spans a wide spectrum ranging from watchful waiting, different forms of radiation therapy, different forms of radical surgery, to hormone therapy and chemotherapy. Treatment decisions, however, are strongly dependent on a precise characterization of the tumor's biologic potential. Standard parameters used in clinical routine such as clinical T stage determination, transrectal ultrasound, prostate-specific antigen (PSA) measurements, and histologic tumor grading on prostate biopsies are each, however, characterized by insufficiencies in precisely reflecting the biologic potential of the tumor.1-4 By combining the results of these parameters in nomograms, diagnostic power has been improved5-9 but limitations remain. For example, nomograms have been developed on selected patient populations and transfer on other patient populations has been shown to be associated with loss of accuracy.10 PSA, which correlates with tumor volume, apparently loses its predictive power through the current detection of increasingly smaller prostate cancers.4,11 This underlines the need for additional parameters that can aid in a more precise patient management. The clinical relevance of the detection of disseminated tumor cells (DTCs) in bone marrow (BM) or in the peripheral blood of solid tumor patients free of overt metastasis is still under investigation.12-14 BM is the most prominent metastatic site in prostate and breast cancer and several research groups have focused on the detection of DTCs in this organ,14 which is easily accessible and therefore routinely used for tumor staging in hematopoietic malignancies. Thus far, only large breast cancer studies have confirmed the independent prognostic value of the BM status.15 In prostate cancer, we and others already reported on the presence of DTCs in BM.16-22 However, most of the studies included a relatively small number of patients and/or follow-up information was sparse or lacking. In addition, the influence of hormone treatment on the prognostic power is also lacking. In face of the aforementioned diagnostic shortcomings, larger studies with an adequate follow-up are needed. Herein we report on the prognostic relevance of DTCs in BM of 193 patients with clinically localized prostate cancer submitted to neoadjuvant hormone therapy followed by radical prostatectomy and a median follow-up of 44 months.
Patients The study included a total of 193 patients with previously untreated clinical stage cT1-3 prostate cancer, enrolled between November 1992 and August 2000. Clinical T stage was determined by digital rectal examination and transrectal ultrasound. Osseous metastases were excluded by bone scans in patients with PSA higher than 10 ng/mL. After written informed consent, all patients were submitted to pelvic BM aspiration under local anesthesia, followed by PSA (IMMULITE Third Generation PSA assay; Diagnostic Products Corp, Los Angeles, CA) monitored prolonged neoadjuvant endocrine treatment (PPNET). Median patient age was 67 years (mean, 65.6; range, 38 to 76 years). PPNET consisted of leuprorelin (3.57 mg/month) or goserelin (3.6 mg/month) in combination with flutamide (250 mg three times a day) or bicalutamide (50 mg once a day). Patients were monitored monthly by digital rectal examination, serum PSA, and transrectal ultrasound until the PSA nadir was reached. The mean duration of PPNET was 8 months. After having reached PSA nadir, all patients were submitted to radical prostatectomy. The median postoperative PSA follow-up was 44 months (range, 3 to 162). The patients were generally scheduled for digital rectal examination and serum PSA measurement every 3 months for the first postoperative year and semiannually to annually thereafter. Biochemical progression was defined as sustained elevation of PSA  0.1 ng/mL on two or more occasions.
BM Preparation and Immunocytochemical Staining
Statistical Analysis
Association of BM Status With Clinical and Histopathologic Characteristics Of the 193 patients with primary prostate cancer, initial BM biopsies were positive in 86 (44.6%) and negative in 107 patients (55.4%). For each patient, 2 x 106 MNCs were screened and between one to 50 DTCs (median, two DTCs/patient) were detected. According to our previous analysis,24 BM samples were recorded as positive if one cytokeratin-positive cell was detected. There was no statistically significant correlation with biopsy (pretreament) or radical prostatectomy tumor grade (Gleason), clinical (pretreatment) or pathologic T stage (pT), and PSA category as the most established risk factors in prostate cancer (Table 1).
Association of BM Status With Disease Progression Overall, 71 (36.8%) of 193 patients developed biochemical relapse after radical prostatectomy (median, 16 months; range, 3 to 60 months), including 41 patients with positive preoperative BM findings. The median follow-up for nonprogressing patients was 61 months, including 68 patients with at least 5 years of PSA progression-free follow-up. The PSA relapse rate in BM-positive patients was significantly higher compared with BM-negative patients (41 of 86 [47.7%] v 30 of 107 [28.0%]; P = .0067; Table 2). Sensitivity and specificity of a positive BM status to predict PSA relapse was 0.58 and 0.63, respectively. The positive and negative predictive value of the presence of DTCs was calculated as 0.48 and 0.72, respectively.
PSA relapse was also associated with a higher pT and a high radical prostatectomy Gleason score. No such association was seen for clinical T stage and biopsy Gleason score. Risk of PSA relapse in BM-positive patients was positively correlated with pT stage (P = .0002) but not with Gleason grade and PSA category (Table 3).
A positive BM status was significantly associated with a shorter PSA recurrence-free survival (log-rank test P = .0035; Fig 1A). Beside the BM status, univariate analysis also showed prognostic significance for the established prognostic factors pT stage (P < .0001), Gleason score (P < .0067), and preoperative PSA serum level (P = .0007; Figs 1B to 1D).
To investigate whether BM status adds clinical relevant prognostic information to the well-established standard prognostic parameters, we performed a multivariate analysis taken all risk factors that were significant in the univariate analysis into account. As presented in Table 4, the presence of DTCs remained a significant prognostic factor with respect to PSA progression-free survival when we controlled for pT stage, Gleason grade, and PSA serum level (relative risk, 1.82; P = .014). Besides BM status, pT stage was the only other independent prognostic factor (P = .020).
In patients with prostate cancer, current diagnostic/prognostic parameters only incompletely reflect tumor biology. This drawback is documented by signs of tumor progression in up to 40% of the patients submitted to putative curative therapy and by the detection of biologically insignificant cancers resulting in an increased risk of overtreatment. Therefore, additional diagnostic parameters to improve tumor characterization are urgently needed. In our present study, we found strong evidence of the independent adverse prognostic significance of the presence of DTCs at the time of initial diagnosis. The prognostic impact of tumor cells in the BM has already clearly been demonstrated in patients with breast cancer.15 In prostate cancer there are only few studies on the prognostic relevance of these cells. These studies include fewer patients with a shorter follow-up, and the results are in part contradictory.20-22,25,26 To our knowledge, our study is the first one on a larger series of patients with prostate cancer with sufficiently long follow-up which clearly demonstrates an independent adverse prognostic effect of the presence of DTCs at the time of initial diagnosis. Furthermore, our study showed, for the first time, that the prognostic significance of DTCs in BM is maintained inspite of neoadjuvant hormone treatment. Neoadjuvant hormone therapy in prostate cancer, at least in theory, represents an attractive therapeutic option for patients who are known to have an increased risk of tumor relapse despite radical prostatectomy. The local effect of hormone therapy is well documented in several studies, reporting on a decreased surgical margin rate, extensive regressive changes, and even nondetectable tumor (pT0) in subsets of patients.27-29 However, these morphological effects have been shown to result, if at all, in a prognostic benefit only in small subgroups of patients.30 For this reason this therapeutic concept has been widely abandoned. Our understanding of the lacking prognostic benefit, however, is still incomplete and more knowledge about it will be of interest for future therapeutic concepts. This study provides first evidence that the presence of DTCs may contribute to these disappointing results. To answer the question of whether maximum endocrine pretreatment is able to reverse the negative impact of DTCs in the BM, PSA relapse was elected as surrogate indicator for tumor progression. Because PSA expression is testosterone dependent and androgen production is known to be temporarily depressed after neoadjuvant endocrine therapy, PSA relapse may represent an inadequate monitoring tool. However, addressing this question in a previous study, we could show a testosterone recovery rate of 92% until 12 months postcessation of hormone therapy.31The median follow-up of our nonrelapsing patients was 61 months, far beyond this PSA blind period. In addition, we used a very sensitive PSA threshold of 0.1 ng/mL as criterion to diagnose biochemical progression. Thus, it appears to be very unlikely that the PSA relapse rate has been artificially altered by neoadjuvant hormone treatment. Gleason grading after hormone treatment in general is not recommended because of regressive changes that may lead to overgrading or make grading even impossible.32 We and others27,33-35 agree that grading of pretreated prostate cancers is demanding but it is possible in the vast majority of cases if the pathologist is familiar with the spectrum of regressive changes and if he does a very thorough examination to detect also unaltered foci of tumor, which are found in many specimens. Only in the few postneoadjuvant pT0 cases we referred to the initial pretreatment biopsy grading. Interestingly our data confirm the study by Weckermann et al21 on nonpretreated patients, who also were unable to detect a relationship between grading and BM status. Nevertheless, in order to include pretreatment tumor characteristics, grading data from the pretreatment biopsies as well as initial clinical staging data were added in Tables 1 and 2. However, this additional data did not add accuracy. As presented in Table 2, neither biopsy grading nor clinical T stage correlated with PSA relapse as opposed to pT stage and radical prostatectomy grading. Due to the better diagnostic accuracy only pT stage and radical prostatectomy grading were included in uni- and multivariate analysis. For the detection of DTCs, we used an immunocytochemical cytokeratin assay that allows the identification of one tumor cell in the background of 1 million normal BM cells.24 Cytokeratins are still the state-of-the-art markers for the detection of DTCs in various assay formats developed over the past years.13,14 The specificity of the CK assay used in this study was previously demonstrated by the analysis of almost 200 noncarcinoma control patients24 and the detection of genomic tumor-specific aberrations in DTCs.36-38 However, open questions related to the biologic meaning of DTCs in the preoperative setting remain, which may explain the suboptimal predictive value of our BM assay. In general, DTCs are anticipated to reflect early systemic tumor spread contributing to subsequent postoperative PSA relapse. But local tumor recurrence, most likely not related to DTCs in BM detected in the preoperative setting, also results in PSA relapse. Therefore, PSA recurrence, although accepted as a surrogate marker for disease progression in prostate cancer, is unable to differentiate between local and/or systemic recurrences. A more in-depth analysis of our data points to a heterogeneous biologic potential of DTC in BM. Only 40% to 50% of patients with DTCs relapsed within the observation time, which is reflected in the low specificity rates reported. Although more relapses can be anticipated over the next years, this finding also points to the conclusion that DTCs are inhomogeneous with respect to their biologic potential. Therefore, more sophisticated methods are urgently needed to improve the characterization of DTCs. With this respect, technical achievements, like the simultaneous determination of the viability,39 proliferative potential,25,38 and molecular genetic characteristics of DTCs36,40,41 are promising methods, which also harbor the potential to provide helpful information on new therapeutic targets. In conclusion, the detection of DTCs in the BM of patients with prostate cancer represents an independent prognostic parameter even in hormonally treated patients. DTCs appear to survive hormone therapy42,43 and this might be one explanation for the well-known failure of neoadjuvant hormone therapy. However, improved characterization of DTCs and larger multicenter studies followed by nomogram testing against the established risk parameters are required to introduce BM examination for detection of DTCs into the future clinical management of patients with prostate cancer.
The author(s) indicated no potential conflicts of interest.
Conception and design: Jens Köllermann, Steffen Weikert, Klaus Pantel Provision of study materials or patients: Klaus Kleinschmidt Collection and assembly of data: Jens Köllermann, Klaus Pantel Data analysis and interpretation: Jens Köllermann, Steffen Weikert, Martin Schostak, Carsten Kempkensteffen, Thomas Rau, Klaus Pantel Manuscript writing: Jens Köllermann, Klaus Pantel Final approval of manuscript: Jens Köllermann, Klaus Pantel
We thank the urologists of the Wiesbaden area, namely Helga Weidenfeld, MD, Michael Weidenfeld, MD, B. Heseding, Jörg Caprano, MD, Anke Budde, MD, and Walter Müller, MD, for the continuous supply with follow-up data.
published online ahead of print at www.jco.org on September 15, 2008. Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Allsbrook WC Jr, Mangold KA, Johnson MH, et al: Interobserver reproducibility of Gleason grading of prostatic carcinoma: General pathologist. Hum Pathol 32:81-88, 2001[CrossRef][Medline] 2. Djavan B: Gleason scores from prostate biopsies obtained with 18gauge biopsy needles poorly predict gleason scores of radical prostatectomy specimens. Curr Opin Urol 8:441, 1998 3. Gleason DF: Undergrading of prostate cancer biopsies: A paradox inherent in all biologic bivariate distributions. Urology 47:289-291, 1996[CrossRef][Medline] 4. Stamey TA, Johnstone IM, McNeal JE, et al: Preoperative serum prostate specific antigen levels between 2 and 22 ng/ml correlate poorly with postradical prostatectomy cancer morphology: Prostate specific antigen cure rates appear constant between 2 and 9 ng/ml. J Urol 167:103-111, 2002[CrossRef][Medline] 5. Graefen M, Augustin H, Karakiewicz PI, et al: Can nomograms derived in the U.S. applied to German patients? A study about the validation of preoperative nomograms predicting the risk of recurrence after radical prostatectomy. Urologe A 42:685-692, 2003[Medline] 6. 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 7. Partin AW, Kattan MW, Subong EN, et al: Combination of prostate specific antigen, clinical stage, and Gleason score to predict pathological stage of localized prostate cancer: A multi-institutional update. JAMA 277:1445-1451, 1997 8. Graefen M, Augustin H, Karakiewicz PI, et al: Can predictive models for prostate cancer patients derived in the United States of America be utilized in European patients? A validation study of the Partin tables. Eur Urol 43:6-10, 2003[CrossRef][Medline] 9. Kattan MW, Potters L, Blasko JC, et al: Pretreatment nomogram for predicting freedom from recurrence after permanent prostate brachytherapy in prostate cancer. Urology 58:393-399, 2001[CrossRef][Medline] 10. Penson DF, Albertsen PC: Lessons learnt about early prostate cancer from large scale databases: Population-based pearls of wisdom. Surg Oncol 11:3-11, 2002[CrossRef][Medline] 11. Stamey TA: The era of serum prostate specific antigen as a marker for biopsy of the prostate and detecting prostate cancer is now over in the USA. BJU Int 94:963-964, 2004[CrossRef][Medline] 12. Cristofanilli M, Mendelsohn J: Circulating tumor cells in breast cancer: Advanced tools for "tailored" therapy? Proc Natl Acad Sci U S A 103:17073-17074, 2006 13. Pantel K, Alix-Panabieres C: The clinical significance of circulating tumor cells. Nat Clin Pract Oncol 4:62-63, 2007[Medline] 14. Pantel K, Brakenhoff RH: Dissecting the metastatic cascade. Nat Rev Cancer 4:448-456, 2004[CrossRef][Medline] 15. Braun S, Vogl FD, Naume B, et al: A pooled analysis of bone marrow micrometastasis in breast cancer. N Engl J Med 353:793-802, 2005 16. Köllermann J, Heseding B, Helpap B, et al: Comparative immunocytochemical assessment of isolated carcinoma cells in lymph nodes and bone marrow of patients with clinically localized prostate cancer. Int J Cancer 84:145-149, 1999[CrossRef][Medline] 17. Mansi JL, Berger U, Wilson P, et al: Detection of tumor cells in bone marrow of patients with prostatic carcinoma by immunocytochemical techniques. J Urol 139:545-548, 1988[Medline] 18. Pantel K, Aignherr C, Köllermann J, et al: Immunocytochemical detection of isolated tumour cells in bone marrow of patients with untreated stage C prostatic cancer. Eur J Cancer 31A:1627-1632, 1995[CrossRef] 19. Pfitzenmaier J, Ellis WJ, Hawley S, et al: The detection and isolation of viable prostate-specific antigen positive epithelial cells by enrichment: A comparison to standard prostate-specific antigen reverse transcriptase polymerase chain reaction and its clinical relevance in prostate cancer. Urol Oncol 25:214-220, 2007[Medline] 20. Weckermann D, Muller P, Wawroschek F, et al: Disseminated cytokeratin positive tumor cells in the bone marrow of patients with prostate cancer: Detection and prognostic value. J Urol 166:699-703, 2001[CrossRef][Medline] 21. Weckermann D, Wawroschek F, Krawczak G, et al: Does the immunocytochemical detection of epithelial cells in bone marrow (micrometastasis) influence the time to biochemical relapse after radical prostatectomy? Urol Res 27:285-290, 1999[CrossRef][Medline] 22. Wood DP Jr, Banerjee M: Presence of circulating prostate cells in the bone marrow of patients undergoing radical prostatectomy is predictive of disease-free survival. J Clin Oncol 15:3451-3457, 1997 23. Fehm T, Braun S, Müller V, et al: A concept for the standardized detection of disseminated tumor cells in bone marrow from patients with primary breast cancer and its clinical implementation. Cancer 107:885-892, 2006[CrossRef][Medline] 24. Braun S, Pantel K, Müller P, et al: Cytokeratin-positive cells in the bone marrow and survival of patients with stage I, II, or III breast cancer. N Engl J Med 342:525-533, 2000 25. Bianco FJ Jr, Wood DP Jr, Gomes de Oliveira J, et al: Proliferation of prostate cancer cells in the bone marrow predicts recurrence in patients with localized prostate cancer. Prostate 49:235-242, 2001[CrossRef][Medline] 26. Gao CL, Dean RC, Pinto A, et al: Detection of circulating prostate specific antigen expressing prostatic cells in the bone marrow of radical prostatectomy patients by sensitive reverse transcriptase polymerase chain reaction. J Urol 161:1070-1076, 1999[CrossRef][Medline] 27. Köllermann J, Feek U, Muller H, et al: Nondetected tumor (pT0) after prolonged, neoadjuvant treatment of localized prostatic carcinoma. Eur Urol 38:714-720, 2000[CrossRef][Medline] 28. Köllermann MW, Pantel K, Enzmann T, et al: Supersensitive PSA monitored neoadjuvant hormone treatment of clinically localized prostate cancer: Effects on positive margins, tumor detection and epithelial cells in bone marrow. Eur Urol 34:318-324, 1998[CrossRef][Medline] 29. Köllermann MW, Sprenk P, Mühlfait V, et al: Downstaging of stage C prostatic carcinoma by inductive chemo-hormonal therapy? Eur Urol 24:337-341, 1993[Medline] 30. Gleave ME, Goldenberg SL, Chin JL, et al: Randomized comparative study of 3 versus 8-month neoadjuvant hormonal therapy before radical prostatectomy: Biochemical and pathological effects. J Urol 166:500-506, 2001[CrossRef][Medline] 31. Köllermann J, Caprano J, Budde A, et al: Prolonged neoadjuvant endocrine therapy of prostate cancer: PSA failure rate influenced by decreased testosterone production after cessation of endocrine therapy? J Urol 169:278, 2003 (suppl)[CrossRef][Medline] 32. Carroll PR, Kantoff PW, Balk SP, et al: Overview consensus statement: Newer approaches to androgen deprivation therapy in prostate cancer. Urology 60:1-6, 2002[Medline] 33. Civantos F, Marcial MA, Banks ER, et al: Pathology of androgen deprivation therapy in prostate carcinoma: A comparative study of 173 patients. Cancer 75:1634-1641, 1995[Medline] 34. Helpap B, Köllermann J: Therapy induced regressive changes of prostate cancer. Pathologe 25:461-468, 2004[CrossRef][Medline] 35. Smith DM, Murphy WM: Histologic changes in prostate carcinomas treated with leuprolide (luteinizing hormone-releasing hormone effect): Distinction from poor tumor differentiation. Cancer 73:1472-1477, 1994[CrossRef][Medline] 36. Klein CA, Blankenstein TJ, Schmidt-Kittler O, et al: Genetic heterogeneity of single disseminated tumour cells in minimal residual cancer. Lancet 360:683-689, 2002[CrossRef][Medline] 37. Schmidt-Kittler O, Ragg T, Daskalakis A, et al: From latent disseminated cells to overt metastasis: Genetic analysis of systemic breast cancer progression. Proc Natl Acad Sci U S A 100:7737-7742, 2003 38. Solakoglu O, Maierhofer C, Lahr G, et al: Heterogeneous proliferative potential of occult metastatic cells in bone marrow of patients with solid epithelial tumors. Proc Natl Acad Sci U S A 99:2246-2251, 2002 39. Alix-Panabieres C, Vendrell JP, Pelle O, et al: Detection and characterization of putative metastatic precursor cells in cancer patients. Clin Chem 53:537-539, 2007 40. Kraus J, Pantel K, Pinkel D, et al: High-resolution genomic profiling of occult micrometastatic tumor cells. Genes Chromosomes Cancer 36:159-166, 2003[CrossRef][Medline] 41. Meng S, Tripathy D, Shete S, et al: HER-2 gene amplification can be acquired as breast cancer progresses. Proc Natl Acad Sci U S A 101:9393-9398, 2004 42. Pantel K, Enzmann T, Köllermann J, et al: Immunocytochemical monitoring of micrometastatic disease: Reduction of prostate cancer cells in bone marrow by androgen deprivation. Int J Cancer 71:521-525, 1997[CrossRef][Medline] 43. Wood DP Jr, Beaman A, Banerjee M, et al: Effect of neoadjuvant androgen deprivation on circulating prostate cells in the bone marrow of men undergoing radical prostatectomy. Clin Cancer Res 4:2119-2123, 1998[Abstract] Submitted October 24, 2007; accepted June 6, 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
2 test, as appropriate, according to the number of patients. The influence of BM status on biochemical progression after radical prostatectomy was determined by Kaplan-Meier analysis using the log-rank test with significance at P < .05. Multivariable survival analysis was performed with the Cox proportional hazards regression model. The assumptions for performing the Cox regression analysis were assessed for each parameter using the Cox proportional hazards model. All calculations were done with the statistical program SPSS 9.0 (SPSS, Chicago, IL) and GraphPad Prism, 4.0 (San Diego, CA). 
