Imaging, Diagnosis, Prognosis

Chemokine Markers Predict Biochemical Recurrence of Prostate Cancer following Prostatectomy David L. Blum,1 Ta t s u k i Ko y a m a , 2 Amosy E. M’Koma,1Juan M. Iturregui,3 Magaly Martinez-Ferrer,1 Consolate Uwamariya,1Joseph A. Smith, Jr.,1Peter E. Clark,1andNeilA.Bhowmick1

Abstract Purpose: Stratifying patients who have a high risk of prostate cancer recurrence following prostatectomy can potentiate the use of adjuvant therapy at an early stage. Inflammation has emerged as a mediator of prostate cancer metastatic progression. We hypothesized that chemokines can be biomarkers for distinguishing patients with high risk for biochemical recurrence of prostate cancer. Experimental Design: In a nested case-control study, 82 subjects developed biochemical recurrence within 5 years of prostatectomy. Prostate tissues from 98 age-matched subjects who were recurrence-free following prostatectomy in the same period were the controls. A high-throughput lectin-based enrichment of prostate tissue enabled multiplex ELISA to identify the expression of three chemokines to discriminate the two patient populations. Results: The expression of CX3CL1and IL-15 in prostate tissue was associated with 5-year biochemical recurrence-free survival following prostatectomy. However, the expression of chemokine ligand 4 (CCL4) was associated with biochemical recurrence. Multivariable logistic regression model combining preoperative prostate-specific antigen, Gleason score, surgical margin, and seminal vesicle status with the three chemokines doubled the specificity of prediction at 90% sensitivity compared with use of the clinicopathologic variables alone (P < 0.0001). Survival analysis yielded a nomogram that supported the use of CX3CL1, IL-15, and CCL4 in predicting 1-, 3-, and 5-year recurrence-free survival after prostatectomy. Conclusions: Each of the three chemokines can serve as independent predictors of biochemical recurrence. However, the combination of chemokine biomarkers plus clinicopathologic variables discriminated prostatectomy subjects for the probability of biochemical recurrence significantly better than clinicopathologic variables alone.

There is a large disparity between the number of newly surveillance. Traditionally, a combinatorial assessment of diagnosed cases of prostate cancer every year and the number various clinical parameters is used to risk-stratify patients. of men who die of metastatic progression of the disease (1). As Although these assessments have their utility and have a consequence, although prostate cancer is the second leading undergone many different revisions and modifications over cause of cancer-related mortality in men in the United States, time, they are still unable to distinguish the 80% of patients overtreatment of the disease is a concern. The challenge has that may not have any clinical consequences from their been to determine which patients harbor high-risk disease prostate cancer (2). Nonetheless, monotherapy such as surgical requiring aggressive/curative therapy and which patients or radiation ablation of the prostate may not be curative for harbor indolent disease that could be managed with active manypatients.Insuchpatientswithhigh-riskdisease, adjuvant therapy (e.g., antiandrogens, chemotherapy) is common following monotherapy. However, because adjuvant therapies often provide only temporary suspension of disease 1 2 Authors’Affiliations: Departments of Urologic Surgery, Biostatistics, and progression, the poor efficacy may be due to their administra- 3Pathology, Vanderbilt-Ingram Cancer Center, Vanderbilt University, Nashville, Tennessee tion after the development of frank biochemical prostate Received 7/3/08; accepted 8/13/08. cancer recurrence (3–5). Because a majority of patients remain Grant support: NIH Institutional Research and Academic Career Development disease-free after prostatectomy (75-80%), the identification of Award (5K12GM068543; A.E. M’Koma) and theT.J. Martel Foundation. candidates potentially benefiting from adjuvant therapy is The costs of publication of this article were defrayed in part by the payment of page important given the potential for decreased quality of life and charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. added morbidity from such intervention (6). Nomograms, Note: D.L. Blum,T. Koyama, and A.E. M’Koma contributed equally to this work. such as the one described by Kattan et al., are commonly used Requests for reprints: Neil A. Bhowmick, Department of Urologic Surgery, to predict disease recurrence after prostatectomy and involve Vanderbilt-Ingram Cancer Center, Vanderbilt University, A-1302 Medical Center multiple criteria that include the pretreatment prostate-specific North, 1161 21st Avenue South, Nashville, TN 37232. Phone: 615-343-7140; Fax: 615-322-5869; E-mail: [email protected]. antigen (PSA), prostate capsule invasion, pathologic Gleason F 2008 American Association for Cancer Research. score, surgical margin status, seminal vesicle involvement, and doi:10.1158/1078-0432.CCR-08-1716 lymph node involvement (7). However, the sensitivity of

ClinCancerRes2008;14(23)December1,2008 7790 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2008 American Association for Cancer Research. Chemokine Markers Predict Recurrence of Prostate Cancer

availability of archived fresh frozen peripheral zone tissue, and records Translational Relevance of presurgical hormone ablation therapy. Patients who had undergone hormone ablation therapy at any point before surgery or the Current therapeutic options for prostate cancer patients demonstration of biochemical recurrence were excluded. Biochemical involve surgical or radiation-based ablation of the prostate. recurrence following prostatectomy was defined as PSA z0.2 ng/mL Although this can be curative for many patients, up to 35% confirmed at least once with another PSA at least 2 wk apart, and of prostatectomy patients develop biochemical recurrence associated with two consecutive subsequent increases in PSA level. of prostate cancer. The high-risk prostate cancer patient Ultimately, for this nested study, we focused on 82 subjects who developed biochemical recurrence within 5 y of prostatectomy and an population often dies from the disease due to metastatic age-matched control group of 98 subjects who were free of recurrence progression. Early detection of prostate cancer likely to within the same time frame. The mean age for the subjects was 60 y develop biochemical recurrence can lead to proactive use (43-72 y). All subjects were annotated based on age, race, presurgical of adjuvant therapeutic options before frank biochemical serum PSA, pathologic Gleason score, pathologic stage, extracapsular recurrence. As tumor-associated inflammatory changes involvement, seminal vesicle involvement, surgical margin status, and can regulate prostate cancer metastatic progression, we detection of biochemical recurrence (Table 1). looked to chemokines in prostatectomy tissue as potential Sample preparation and analysis indicators of future biochemical recurrence. We identified three chemokines that independently predict biochemical The tissue samples were derived from a tumor bank of frozen cores of recurrence status within 5 years of prostatectomy in a the prostate from patients after radical prostatectomy for adenocarci- noma of the prostate. Eight 4-mm-diameter cores were taken from fresh nested case-control study. The differentially expressed prostates removed during prostatectomy and snap frozen in liquid chemokines further supplemented known clinicopatholog- nitrogen. Four of these cores were from the peripheral zone, where the ic variables to provide high sensitivity and specificity. The majority of the prostate cancer originates. The frozen cores, determined multivariable prediction models generated resulted in a by gross assessment of tumor involvement, were dissected longitudi- nomogram that provides superior prediction capacity. nally into three sections with the outside sections used for duplicate protein isolation (Fig. 1A). The central piece was paraformaldehyde fixed and paraffin embedded for histologic evaluation by a pathologist. Samples were homogenized in lysis buffer [100 mmol/L Tris predicting biochemical recurrence by such criteria can be (pH 7.2), 500 mmol/L NaCl], sonicated, and centrifuged for 5 min at improved (7). 1,000 Â g. Glycosylated proteins were purified using the glycoprotein Inflammatory chemokines in the tumor microenvironment isolation kit, WGA (Pierce Biotechnology, Inc.), according to the can regulate the fate of tumor progression (8, 9). We manufacturer’s instructions. All procedures except for binding to the hypothesized that such tissue chemokines can be strong wheat germ agglutinin resin (Pierce Biotechnology, Inc.) were done on biomarker candidates for distinguishing patients with a high ice in siliconized microcentrifuge tubes. The expression levels of A risk for biochemical recurrence or metastatic progression of 30 chemokines of the resulting samples (100 L) were measured by LINCO Research, Inc., using the human chemokine multiplex prostate cancer. Inflammatory cell recruitment in prostate antibody-array for TGF-h1, IL-1a, IL-1h, IL1a receptor, IL-2, IL-4, IL- cancer has emerged as a modulator of metastatic progression 5, IL-6, IL-7, IL-8, IL-10, IL-12p40, IL-12p70, IL-13, IL-15, IL-17, EGF, (10, 11). We sought to identify factors that mediate TGF-a, CX3CL1, CCL2, sCD40L, IP-10, VEGF, RANTES, GM-CSF, inflammatory cell recruitment because the immune response G-CSF, IFN-g, MIP1a, MIP1h, and Eotaxin. The expression levels were to cancer can result in tumor cell ablation and provide growth normalized to total protein. factors to further stimulate tumor progression and motility. The nested case-control study described identify the differen- Statistical analysis tial expression of two chemokine biomarkers in prostatic Baseline patient characteristics. Baseline demographic and clinical tissue that support greater detection sensitivity for the variables for all patients were assessed using Wilcoxon rank sum biochemical recurrence of prostate cancer alone following tests for continuous variables and Fisher’s exact tests for categorical prostatectomy. However, incorporating clinicopathologic variables. Biomarker expression levels underwent logarithmic transfor- mation to stabilize variance. Values below the detection limit were parameters plus three chemokine biomarkers provided supe- imputed with half of the minimum detected value for the particular rior prediction of biochemically recurrent disease than biomarker. clinicopathologic parameters alone. The nomogram developed Candidate chemokine selection. As there were 31 candidate chemo- based on the modeling results illustrate the power of the kines, it was necessary to reduce the number of candidate variables combined survival prediction. The model has the potential to before regression models could be considered. Exploratory data improve patient risk stratification after primary local treatment analyses were conducted on 36 (21 recurrent and 15 recurrence-free) for prostate cancer and enable earlier decision-making for prostatectomy subjects. The strength of marginal relationship to the possible secondary therapeutic options. response was used to eliminate variables that showed only a very weak relation. Spearman’s rank correlations and Wilcoxon rank sum tests were used for initial screening. A receiver operating characteristics Patients, Materials, and Methods (ROC) curve was constructed for each chemokine and the area under the ROC curve (AUC) was used to compare how strongly chemokines Patient selection were related to the recurrence. Finally, a backward elimination model This study was conducted in accordance with the Vanderbilt building strategy on 1,000 bootstrapped data was used to choose University Institutional Review Board. The digital medical record of variables for further consideration. Using this bootstrap sample, a full 660 subjects was retrospectively examined using the Vanderbilt model with all the variables in consideration was constructed. Then, the University Urologic Surgery registry of radical prostatectomies done variable with the largest P value (Wald test) was dropped and a new between 1998 and 2002. Several of these patients were excluded for model was fitted with one fewer variable. Variable elimination reasons that included availability of at least 5-year follow-up data, continued until all the remaining variables showed a P value of <0.5.

www.aacrjournals.org 7791 Clin Cancer Res 2008;14(23) December 1, 2008 Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2008 American Association for Cancer Research. Imaging, Diagnosis, Prognosis

Table 1. Clinical and pathologic stratification of the biochemical recurrent and recurrence-free subject groups (n = 180)

Overall (n = 180) Recurrence-free (n = 98) Recurrent (n = 82) P

Age 60.0 (56.0,66.0) 60.0 (55.3,66.0) 60.5 (56.0,66.0) 0.834* Race White 92% (165) 88% (86) 96% (79) 0.093c Black 7% (13) 10% (10) 4% (3) Others 1% (2) 2% (2) 0% (0) Extracapsular involvement 42% (76) 29% (28) 59% (48) <0.001c Pos. margin 28% (50) 11% (11) 48% (39) <0.001c Seminal vesicle involvement 16% (28) 2% (2) 32% (26) <0.001c Lymph node involvement 5% (9) 0% (0) 11% (9) <0.001c Preoperative PSA 6.3 (4.8,9.1) 5.7 (4.6,7.4) 7.3 (5.1,13.4) <0.001c Gleason 5-6 24% (44) 34% (33) 13% (11) <0.001c 7 58% (105) 59% (58) 57% (47) 8-9 18% (31) 7% (7) 30% (24) Clinical stage

T1c 63% (114) 68% (67) 57% (47) 0.092 T2a 25% (45) 24% (24) 26% (21) T2b+ 12% (21) 8% (7) 17% (14)

NOTE: For continuous variables, a (b, c ) represent the median a,lower quartile b,and the upper quartile c. Numbers after percentages are frequencies in parentheses. *Wilcoxon test. cFisher’s test.

The number of retentions of each variable in 1,000 iterations of the had focal adenocarcinoma and HGPIN as shown in Fig. 1. final model were used to select the variables for further consideration. However, the histologic patterns were not predictive of future Pair-wise Spearman’s rank correlations among the candidates were also progression to biochemical recurrence. There was no statistical considered in selecting these variables. All the information from these difference in the age, race, and clinical stage of the subjects in multiple analyses was used to select the candidate chemokines for the the recurrent and recurrence-free groups (Table 1). Because next phase of the data collection and model building. Logistic regression. Demographic and clinical variables for the logistic regression model were selected based on their relation to the outcome variable (biochemical recurrence) and interrelations among the covariates. Two models were constructed: (a) The three selected chemokine variables with the clinicopathologic variables made up one model. (b) We also fitted a clinical variable–only simpler model without any chemokine variables. These two models were compared using a likelihood ratio test. From each model, the predicted probabilities of recurrence were computed and the ROC curves were constructed. Comparison of the ROC curves was conducted with the integrated discrimination improvement (12). An estimate of the odds ratio with a confidence interval was reported for each variable based on the developed model. Survival analysis. To visualize the association between biochemical recurrence-free survival and each chemokine marker, the product limit estimator was computed for the two groups defined as above and below the median, and the log-rank tests were used to assess the difference of the recurrence-free survival between the two groups illustrated by Kaplan-Meier plots. A Cox proportional hazard regression was used to model the recurrence-free survival. The proportionality of the hazard ratio was assessed graphically and numerically using Schoenfeld’s partial residuals (13). The effects of predictors in the model were presented with individual hazard ratios. A nomogram equating each of the predictors to the probabilities of 1-, 3-, and 5-year recurrence-free survival is also presented. All analyses were carried out with R version 2.7.0 (14).

Results Fig. 1. The histologic evaluation of the frozen prostate cores. A, each tissue core was cut longitudinally into thirds for duplicate chemokine enrichment and histology Chemokines support prediction of biochemical recurrence. The analysis, respectively. Subjects B and C developed biochemical recurrence, but histology of prostatic tissue cores from prostatectomy subjects subject D was free of recurrence for the 5 y following prostatectomy.

Clin Cancer Res 2008;14(23) December 1, 2008 7792 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2008 American Association for Cancer Research. Chemokine Markers Predict Recurrence of Prostate Cancer

Table 2. Odds ratios and confidence intervals for the logistic regression model

Reference Odds ratio (95% CI) P

Gleason score 6 8 2.60 (1.01-6.71) 0.048 Preoperative PSA 5 9 1.54 (1.02-2.34) 0.041 Surgical margin Negative Positive 3.64 (1.41-9.42) 0.001 Seminal vesicle involvement Negative Positive 7.15 (1.44-35.5) 0.016 CCL4 0.003 0.04 2.16 (1.19-3.91) 0.011 CX3CL1 0.01 0.07 0.35 (0.21-0.58) <0.0001 IL-15 0.004 0.06 0.88 (0.51-1.55) 0.66

NOTE: Odds ratios for continuous variables represent change from the lower quartile to upper quartile except for Gleason score whose lower and upper quartiles were both 7. Abbreviation: 95% CI,95% confidence interval.

chemokines and growth factors that influence metastatic only 46 cases (25%) for which these two variables do not agree. progression are commonly in low abundance, we developed The limited number of subjects having positive lymph node a methodology to enrich such factors from tissue lysates. tumor involvement (n = 9) provided little information and was Following lectin enrichment of 0.03 to 0.05 g (wet weight) also not considered in further analysis. A fitted logistic regression tissue, 31 chemokines were screened by multiplex ELISA for a model of our population suggested two chemokines, CCL4 panel of inflammatory chemokines. Importantly, little to no (P = 0.011) and CX3CL1 (P < 0.0001), along with surgical signal was detected for these markers if wheat germ margin (P = 0.008), seminal vesicle involvement (P = 0.016), agglutinin–mediated enrichment process was not done on preoperative PSA (P = 0.041), and Gleason score (P = 0.048) to the tissue extracts. The use of another lectin, concanavalin A– be significant factors, whereas IL-15 (P = 0.66) was not. Table 2 mediated enrichment was ineffective in chemokine detection summarizes the fitted logistic regression model in terms of the (data not shown). The results of the initial screening of 36 odds ratios. prostatectomy subjects (21 recurrent and 15 recurrence-free) The logistic regression model that used only the clinical were analyzed by Spearman’s rank correlation and Wilcoxon variables (i.e., Gleason score, preoperative PSA, seminal vesicle rank sum tests. The common inflammatory factors found to involvement, and surgical margin status) was compared with be differentially expressed by both tests in biochemically the model combining clinical and chemokine variables. The recurrent and recurrence-free subjects included CX3CL1, IL-12, IL-15, IL-4, and CCL4. Of this group, CCL4 (macrophage inflammatory protein-1h) was up-regulated in patients that developed biochemical recurrence within 5 years following prostatectomy. However, CX3CL1 (fractalkine), IL-12, IL-15, and IL-4 were predominantly down-regulated in such high- risk patients. Interestingly, each of the factors has been implicated biologically to the progression of prostate cancer in the past (15–19). The AUC for individual chemokines and backward elimination bootstrap methods were correlated to biochemical recurrence in finally selecting CX3CL1, CCL4, and IL-15 for further analyses (data not shown). CX3CL1 was the best predictor of recurrence with all the selection methods, and the other two were among the favorable ones with multiple methods. Validation of chemokine biomarkers in model development. We sought to develop a multivariable logistic regression model for the prediction of the probability of biochemical recurrence following prostatectomy. We considered all predictor variables available to us that were either numerical or categories for logistic regression analysis. In addition to the three chemokine markers, the candidate clinical and surgical variables we considered were pathologic Gleason score, preoperative PSA, surgical margin status, seminal vesicle involvement, clinical stage, extracapsular involvement, and lymph node metastasis. However, for our Fig. 2. ROC curves for the prediction models affected by the chemokines (CCL4, population, the clinical stage categorized to T1c,T2a, and T2b or CX3CL1, and IL-15). Predicted probability of recurrence for each subject was greater did not provide any discrimination regarding the computed from logistic regression models, including preoperative PSA, surgical recurrence status of the subjects. Extracapsular involvement margin status, seminal vesicle invasion status, and pathologic Gleason score, with and without chemokines. Specificity and sensitivity were computed at each was also dropped as a covariate as it was highly correlated with possible cutoff on the predicted probability for the two models.The AUC values surgical margin (Spearman’s rank correlation = 0.47); there were were compared for the two models.

www.aacrjournals.org 7793 Clin Cancer Res 2008;14(23) December 1, 2008 Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2008 American Association for Cancer Research. Imaging, Diagnosis, Prognosis

improvement by the three biomarkers was highly significant (likelihood ratio test P <0.0001). Figure 2 shows the ROC curves obtained by the predicted values using the two models. Improvement in area under the curve (AUC) was 7.1 percentage points (from 80.6% to 87.7%). Integrated discrimination improvement was estimated to be 0.116 (P < 0.0001), supporting the statistical significance of the improvement (12). The addition of the chemokine bio- markers to the clinical variables provided little improvement in predictive ability up to f80% sensitivity. However, given a sensitivity of 90%, the clinical variables alone provided a specificity of only 36% (95% confidence interval, 20-58%) compared with the addition of the chemokines that provided a specificity of 72% (95% confidence interval, 55-84%). The addition of the chemokine markers to the clinical variables doubled the specificity at 90% sensitivity (P =0.02) according to the ROC analysis. The chemokines, particularly CCL4 and CX3CL1, supported the dichotomous prediction of biochemical recurrence and recurrence-free survival following prostatectomy. Analysis of recurrence-free survival. To define the efficacy of the markers in predicting recurrence-free survival, the same clinical and chemokine variables as in the logistic regression were considered for a Cox proportional hazard regression analysis. CCL4, CX3CL1, and IL-15 proved to individually serve as highly significant markers for biochemical recurrence status by the Kaplan-Meier method (Fig. 3). Using a dichotomous median split for the upper and lower concentration range for tissue chemokine expression, CX3CL1 exhibited the best prediction ability (P < 0.0001) followed by CCL4 (P < 0.001) and IL-15 (P = 0.003). The proportional hazard assumption was tested with scaled Schoenfeld residuals (20). There was no evidence of violation, as the m2 tests for trend were not significant for any of the seven variables (surgical margin status, seminal vesicle involvement, Gleason score, preoperative PSA, CCL4, and CX3CL1, and IL-15; P values ranging from 0.46 to 0.90). The effect of the covariates in the multivariable Cox regression model on recurrence-free survival was summarized with the hazard ratios with 95% confidence intervals in Table 3. CCL4, CX3CL1, preoperative PSA, and surgical margin were significant factors (Fig. 4). However, Gleason score, seminal vesicle involvement, and IL-15 were not significant in the overall survival multivariable model. We computed a nomogram from the Cox proportional hazard regression model that connects each predictor and the probabilities of 1, 3, and 5-year recurrence-free survival (Fig. 5). The contributions of CX3CL1 and preoperative PSA were further shown to be important in predicting recurrence- free survival. We also noted that IL-15, although not statistically significant on multivariable analysis, contributed

Fig. 3. The Kaplan-Meier estimates of the recurrence-free survival based on chemokine expression. The patients were separated into two groups, divided at median tissue level for CCL4 (A), CX3CL1 (B), and IL-15 (C).The two groups were discriminated by the median respective chemokine expression concentration indicated: Those above the median were termed upper half, whereas those below the median were termed lower half.The recurrence-free survival probabilities were estimated by the Kaplan-Meier method and the differences were tested using the log-rank test. Each of the dichotomous chemokine expression levels supported statistically significant differences in biochemical recurrence-free survival.

ClinCancerRes2008;14(23)December1,2008 7794 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2008 American Association for Cancer Research. Chemokine Markers Predict Recurrence of Prostate Cancer

Table 3. Cox proportional hazard regression

Reference Hazard ratio (95% CI) P

Gleason score 6 8 1.63 (0.91-2.92) 0.10 Preoperative PSA 5 9 1.35 (1.11-1.63) 0.0025 Surgical margins Negative Positive 1.81 (1.09-1.63) 0.023 Seminal vesicle involvement Negative Positive 1.70 (0.91-3.16) 0.095 CCL4 0.003 0.04 1.34 (1.01-1.78) 0.040 CX3CL1 0.01 0.07 0.60 (0.47-0.76) <0.0001 IL-15 0.004 0.06 0.77 (0.57-1.05) 0.095

NOTE: Multivariable Cox proportional hazard ratios were computed to determine predictors of biochemical recurrence-free survival. The hazard ratios were computed for a change from the lower quartile to upper quartile in continuous variables. For both surgical margins and seminal vesicle involvement,negative is the reference group.

to the prediction of recurrence-free survival comparable with parameters alone. The logistic regression model that included CCL4, seminal vesicle involvement, and surgical margins the three chemokines improved the specificity from 36% to within the nomogram. Together, the survival model suggested 72% at 90% sensitivity, when compared with clinicopatho- the use of surgical margin status, seminal vesicle involvement, logic parameters alone (Fig. 2). This corresponded with a Gleason score, preoperative PSA, CCL4, and CX3CL1, and significant improvement in the AUC for the model that IL-15 for predicting biochemical recurrence following included all three chemokines to 87.7% versus 80.6% for only prostatectomy. clinicopathologic variables. The chemokines identified through the lectin-based enrich- Discussion ment method have specific positive and negative biological roles in prostate cancer progression. All eukaryotic organisms In most cases, patients with clinically localized prostate glycosylate proteins exposed to the extracellular space. Because cancer treated locally with modalities such as surgery or lectins specifically bind such glycosylation groups, we pro- radiation therapy will be cured of their disease. However, a cessed the prostatectomy specimens with wheat germ agglu- proportion of men will harbor microscopic localized or tinin resin in batch and were able to observe the expression of metastatic residual disease. These patients will ultimately a number chemokines otherwise not detectable. Inflammatory develop biochemical recurrence of disease and, eventually, cells are emerging as potential mediators of cancer metastasis are at risk of developing clinical metastatic progression and (11). Both CX3CL1 and CCL4 can recruit natural killer cells, death from their prostate cancer. The risk of progression after T cells, and monocytes. Based on our data, however, the two local therapy is generally estimated based on available chemokines seem to reflect an opposite status of prostate clinicopathologic variables. For example, after radical prosta- cancer recurrence. Accordingly, apart from similar inflamma- tectomy, it is estimated based on both clinical data, such as tory recruitment characteristics, CCL4 has direct proliferative presurgical PSA, as well as pathologic data such as extrac- and migration effects on prostate cancer cells in vitro (28). apsular extension, seminal vesicle involvement, surgical Conversely, CX3CL1 is reported to reduce migration of margin status, and the Gleason score. These approaches prostate cancer cells in culture (29). Finally, IL-15 can prevent described by Amico et al., Kattan et al., and others have improved our ability to risk-stratify patients through a continuous variable, taking into account these types of variables with high specificity yet with relatively low sensitivity (21–23). Due to the lower sensitivity of such available risk- stratifying measures, it has been difficult to justify the routine use of adjuvant therapy before frank biochemical recurrence. Several randomized phase III trials have tested the use of routine, adjuvant radiation therapy after prostatectomy in men with high-risk disease estimated using only clinicopath- ologic parameters (24–26). These studies found significant improvement in biochemical recurrence-free survival in men with adjuvant radiation therapy but found no improvement in overall survival (27). An inability to identify men at high risk Fig. 4. Cox proportional hazard regression. Multivariable Cox proportional hazard may have contributed to the equivocal outcome. This was in regression for biochemical recurrence-free survival showed that preoperative PSA, no small part due to deficiencies in our ability to accurately surgical margin, CCL4, and CX3CL1were significant predictors of recurrence-free survival. For each predictor variable, the vertical bars illustrate the hazard ratio risk-stratify patients using only clinicopathologic variables. estimate and the gray horizontal bars represent the respective 95% confidence With the addition of the chemokines identified in this study intervals described inTable 3. The hazard ratios were computed for a change from to more standard clinicopathologic variables, we were able to the lower quartile to upper quartile in continuous variables, namely Gleason score 6 to 8, preoperative PSA 5 to 9, CCL4 0.003to 0.04, CX3CL10.01to 0.07, and predict the risk of recurrence among men who after IL-15 0.004 to 0.06. For both surgical margins and seminal vesicle involvement, prostatectomy with greater accuracy than by clinicopathologic negative is the reference group.

www.aacrjournals.org 7795 Clin Cancer Res 2008;14(23) December 1, 2008 Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2008 American Association for Cancer Research. Imaging, Diagnosis, Prognosis

Fig. 5. Nomogram from the Cox proportional hazard regression model. The Cox proportional hazard regression model was used to create a prediction model for 1-, 3-, and 5-y recurrence-free survival. A value in each predictor variable corresponds to a point scale (top). The sum of the individual predictor variable points corresponds to the probability of 1-, 3-, and 5-y recurrence-free survival (bottom).

prostate cancer progression by supporting natural killer cell surgery. These apparent early recurrence subjects likely had function in vivo (15, 30). CX3CL1 and CCL4 overwhelmingly already developed local or distant metastasis before surgery. In supported prediction of prostate cancer biochemical recur- an era where there is growing concern that many prostate rence under all univariant and multivariant criteria tested. cancers are overtreated, the use of this model could support the Although the univariable analysis suggested IL-15 to be clinical determination of those prostate cancers that would significant, as illustrated by Kaplan-Meier plot (Fig. 3), the progress to symptomatic disease and ultimately death, as well overall multivariable models (logistic regression and Cox as potentially provide novel biological targets to inhibit the proportional hazard) did not indicate statistical significance in metastatic progression. At the same time, it has the potential to predicting recurrence-free survival (Tables 2 and 3). Neverthe- identify those patients who harbor more indolent disease that less, specific multivariable 1-, 3-, and 5-year survival analysis could be managed by active surveillance and may be spared the supported the importance of IL-15 as a predictive factor, as potential morbidity of aggressive local therapy. In principle, the shown in the nomogram (Fig. 5). The significant contribution same model that was generated in this study could be done on of each of the chemokine biomarkers needs to be taken in diagnostic biopsy specimens and incorporated into a model context of clinicopathologic parameters. The well-recognized using clinical variables such as the PSA, clinical Gleason score, predictors for biochemical recurrence, like Gleason scores and and clinical stage to more accurately predict a patient’s disease seminal vesicle involvement, were greater in the population risk before any therapy. It should be noted, however, that this studied (Table 1); however, their predictive significance was idea was not formally tested in this study as the specimens were diminished when compared with the chemokine biomarkers obtained on freshly removed prostates after prostatectomy. in a multivariable analysis (Table 3; Fig. 4). Together, the data Nevertheless, the cores used here were of similar nature as argue for considering multiple biomarkers in the prediction of would be obtained at the time of diagnostic biopsy of the those prostate cancers likely to have the greatest clinical prostate. Notably, many of the tissues processed had no significance, especially when coupled to clinicopathologic evidence of adenocarcinoma in the sample (Fig. 1). As the parameters. biomarkers are secreted factors, it seems that actual tumor We have shown in this report that a cumulative evaluation of sampling is not particularly necessary. However, more studies the expression by a continuous variable was able to accurately are needed to know the extent of adjacency required for positive predict biochemical recurrence after radical prostatectomy. prediction ability. Therefore, the same techniques used in this Using a model incorporating chemokines and clinicopatholog- study can be readily transferred to the clinical setting at the time ic variables developed in this study, men could be identified of prostate biopsy. Previously reported blood or tissue who were at substantially higher risk of recurrence after biomarker analyses have not approached the sensitivity and prostatectomy. Whereas local recurrence can potentially be specificity of the prediction achieved by the model described treated with prostatic bed salvage radiation, systemic disease is here (31–37). treated with hormonal therapy, of which the exact timing is In summary, we have shown that prostate tissue levels of controversial. However, of the 82 subjects with biochemical three chemokines—CCL4, CX3CL1, and IL-15—are predictive recurrence, 44 (54%) showed elevated PSA within 1 year of of biochemical recurrence in men who have undergone

ClinCancerRes2008;14(23)December1,2008 7796 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2008 American Association for Cancer Research. Chemokine Markers Predict Recurrence of Prostate Cancer

radical prostatectomy for adenocarcinoma of the prostate Disclosure of Potential Conflicts of Interest with high specificity and sensitivity. Further, we have shown that a nomogram that incorporates these three chemokines No potential conflicts of interest were disclosed. with other established clinicopathologic variables is a better predictor of outcome than using clinicopathologic variables Acknowledgments

alone. This suggests that CCL4, CX3CL1, and IL-15 are We thank Brenda Hughes, Indrani Bhowmick, and Dr. Jheelam Banerjee for their biomarkers of prostate cancer recurrence after radical efforts in gathering patient information; Dr. Bruce Roth for critical discussions; and prostatectomy. Susan and Luke Simons for supporting the study (N.A. Bhowmick).

References 1. Jemal A,Tiwari RC, MurrayT, et al. Cancer statistics, teracted by shedding in prostate cancer. J Clin Invest positive prostate cancer: a systematic review and 2004. CA CancerJ Clin 2004;54:8 ^ 29. 2004;114:560^8. meta-analysis. Radiother Oncol 2008;88:1 ^ 9. 2. Thompson KE, Hernandez J, Canby-Hagino ED, 16. Jamieson WL, Shimizu S, D’Ambrosio JA, Meucci 28. AkashiT,KoizumiK,NagakawaO,FuseH,SaikiI. Troyer D, Thompson IM. Prognostic features in O, Fatatis A. CX3CR1 is expressed by prostate epi- Androgen receptor negatively influences the expres- men who died of prostate cancer. J Urol 2005;174: thelial cells and androgens regulate the levels of sion of chemokine receptors (CXCR4, CCR1) and li- 553^ 6; discussion 6. CX3CL1/fractalkine in the bone marrow: potential gand-mediated migration in prostate cancer DU-145. 3. Glode LM.The case for adjuvant therapy for prostate role in prostate cancer bone tropism. Cancer Res Oncol Rep 2006;16:831 ^ 6. cancer. J Urol 2006;176:S30 ^3. 2008;68:1715^22. 29. Shulby SA, Dolloff NG, Stearns ME, Meucci O, 4. Gomella LG, Zeltser I, Valicenti RK. Use of neoadju- 17. Wang H,ThompsonTC. Gene-modified bone marrow Fatatis A. CX3CR1-fractalkine expression regulates vant and adjuvant therapy to prevent or delay recur- cell therapy for prostate cancer. Gene Ther 2008;15: cellular mechanisms involved in adhesion, migration, rence of prostate cancer in patients undergoing 787 ^ 96. and survival of human prostate cancer cells. Cancer surgical treatment for prostate cancer. Urology 2003; 18. TakeshiU,SadarMD,SuzukiH,etal.Interleukin-4in Res 2004;64:4693^8. 62 Suppl 1:46 ^ 54. patients with prostate cancer. Anticancer Res 2005; 30. Suzuki K, Nakazato H, Matsui H, et al. NK cell- 5. Akduman B, Crawford ED. The management of high 25:4595^8. mediated anti-tumor immune response to human risk prostate cancer. J Urol 2003;169:1993 ^ 8. 19. Lee SO, Pinder E, Chun JY, Lou W, Sun M, Gao AC. prostate cancer cell, PC-3: immunogene therapy us- 6. Freedland SJ, Humphreys EB, Mangold LA, et al. Interleukin-4 stimulates androgen-independent ing a highly secretable form of interleukin-15 gene Risk of prostate cancer-specific mortality following growth in LNCaP human prostate cancer cells. Pros- transfer. J Leukoc Biol 2001;69:531 ^ 7. biochemical recurrence after radical prostatectomy. tate 2008;68:85 ^91. 31. Gonzalez CM, Roehl KA, Antenor JV, Blunt LW, Han JAMA 2005;294:433^ 9. 20. Grambsch P,TherneauT. Proportional hazards tests M, Catalona WJ. Preoperative PSA level significantly 7. Stephenson AJ, Scardino PT, Eastham JA, et al. and diagnostics based on weighted residuals. Biome- associated with interval to biochemical progression Postoperative nomogram predicting the 10-year trika 1994;81:515^ 26. after radical retropubic prostatectomy. Urology 2004; probability of prostate cancer recurrence after radical 21. D’Amico AV, Whittington R, Malkowicz SB, et al. 64:723^8. prostatectomy. J Clin Oncol 2005;23:7005 ^ 12. Combination of the preoperative PSA level, biopsy 32. Henshall SM, Horvath LG, Quinn DI, et al. Zinc-a2- 8. KakinumaT, Hwang ST. Chemokines, chemokine Gleason score, percentage of positive biopsies, and glycoprotein expression as a predictor of metastatic receptors, and cancer metastasis. JLeukoc Biol 2006; MRI T-stage to predict early PSA failure in men with prostate cancer following radical prostatectomy. 79:639 ^ 51. clinically localized prostate cancer. Urology 2000;55: J Natl Cancer Inst 2006;98:1420^ 4. 572 ^ 7. 33. Jayachandran J, Banez LL, Levy DE, et al. Risk 9. Tenta R, Sotiriou E, Pitulis N, Thyphronitis G, Koutsilieris M. Prostate cancer cell survival pathways 22. Kattan MW. Nomograms are superior to staging stratification for biochemical recurrence in men with positive surgical margins or extracapsular disease activated by bone metastasis microenvironment. and risk grouping systems for identifying high-risk J Musculoskelet Neuronal Interact 2005;5:135^ 44. patients: preoperative application in prostate cancer. after radical prostatectomy: results from the SEARCH database. J Urol 2008;179:1791 ^ 6; n Curr Opin Urol 2003;13:111^ 6. 10. Karin M. Nuclear factor- B in cancer development discussion 6. and progression. Nature 2006;441:431 ^ 6. 23. Chun FK, SteuberT, Erbersdobler A, et al. Develop- ment and internal validation of a nomogram predicting 34. Paris PL,WeinbergV, SimkoJ, et al. Preliminary eval- 11. Luo JL, Tan W, Ricono JM, et al. Nuclear cytokine- the probability of prostate cancer Gleason sum uation of prostate cancer metastatic risk biomarkers. activated IKK controls prostate cancer metastasis by a upgrading between biopsy and radical prostatectomy Int J Biol Markers 2005;20:141 ^ 5. repressing maspin. Nature 2007;446:690 ^ 4. pathology. Eur Urol 2006;49:820^ 6. 35. Schmidt H, DeAngelis G, Eltze E, Gockel I, Semjonow 12. Pencina MJ, D’Agostino RB, Sr., D’Agostino RB, Jr., 24. Bolla M, van Poppel H, Collette L, et al. Postopera- A, Brandt B. Asynchronous growth of prostate Vasan RS. Evaluating the added predictive ability of a tive radiotherapy after radical prostatectomy: a rando- cancer is reflected by circulating tumor cells deliv- new marker: from area under the ROC curve to reclas- mised controlled trial (EORTC trial 22911). Lancet ered from distinct, even small foci, harboring loss sification and beyond. Stat Med 2008;27:157 ^ 72; 2005;366:572^8. of heterozygosity of the PTEN gene. Cancer Res discussion 207 ^12. 25. Thompson IM, Jr., Tangen CM, Paradelo J, et al. 2006;66:8959^65. 13. Schoenfeld DA. Sample-size formula for the propor- Adjuvant radiotherapy for pathologically advanced 36. Simmons MN, Stephenson AJ, Klein EA. Natural tional-hazards regression model. Biometrics 1983;39: prostate cancer: a randomized clinical trial. JAMA history of biochemical recurrence after radical prosta- 499^503. 2006;296:2329^ 35. tectomy: risk assessment for secondary therapy. Eur 14. Foundation for Statistical Computing. A language 26. Bottke D, Wiegel T. Adjuvant radiotherapy after Urol 2007;51:1175^ 84. and environment for statistical computing. Vienna radical prostatectomy: indications, results and side 37. Ward JF, Zincke H, Bergstralh EJ, Slezak JM, Blute (Austria); 2008. effects. Urol Int 2007;78:193^7. ML. Prostate specific antigen doubling time subse- 15. WuJD,HigginsLM,SteinleA,CosmanD,HaugkK, 27. Morgan SC, Waldron TS, Eapen L, Mayhew LA, quent to radical prostatectomy as a prognosticator of Plymate SR. Prevalent expression of the immunosti- Winquist E, Lukka H. Adjuvant radiotherapy following outcome following salvage radiotherapy. J Urol 2004; mulatory MHC class I chain-related molecule is coun- radical prostatectomy for pathologic T3or margin- 172:2244 ^8.

www.aacrjournals.org 7797 Clin Cancer Res 2008;14(23) December 1, 2008 Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2008 American Association for Cancer Research. Chemokine Markers Predict Biochemical Recurrence of Prostate Cancer following Prostatectomy

David L. Blum, Tatsuki Koyama, Amosy E. M'Koma, et al.

Clin Cancer Res 2008;14:7790-7797.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/14/23/7790

Cited articles This article cites 36 articles, 4 of which you can access for free at: http://clincancerres.aacrjournals.org/content/14/23/7790.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/14/23/7790.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/14/23/7790. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2008 American Association for Cancer Research.