Novel molecular markers that are associated with prostate cancer (PCa) progression will provide valuable information in the diagnosis and treatment of the disease. Extracellular matrix metalloproteinase inducer (CD147) has been demonstrated to be involved in tumor invasion, metastasis, growth and survival. In our study, we examined whether the expression of CD147 can be used as a prognostic marker for predicting PCa progression. Tissue samples from 240 patients who received radical prostatectomy for PCa were obtained. CD147 expression in these samples was evaluated using immunohistochemical staining with a monoclonal antibody specifically against CD147. Increased expression of CD147 was correlated with higher Gleason scores (GS), positive surgical margin, prostate-specific antigen (PSA) failure, metastasis and reduced overall survival. Both univariate Cox regression analysis and multivariate analysis including competing biological variables demonstrated that increased CD147 expression was associated with increased risk for reduced PSA failure-free, metastasis-free and overall survival. Kaplan–Meier survival curves showed that the CD147 overexpression was a significant predictor for the PSA failure-free, metastasis-free and the overall survival in both pT2 and pT3 PCa patients. More significantly, higher expression of CD147 can serve as an independent prognostic predictor for PSA failure-free survival in PCa patients when they are stratified by GS. Our study results demonstrate the involvement of CD147 in PCa progression and suggest its potential role as an independent predictor of biochemical recurrence, development of metastasis and reduced overall survival in PCa.
Prostate cancer (PCa) is one of the most common cancers and the second leading cause of cancer death in men.1 Although patients with localized PCa can often be successfully treated with radical prostatectomy or radiation therapy, PCa deaths are usually the results of hormone refractory and metastatic disease. It is often difficult to identify patients who will progress, recur and require additional treatments. PCa prognosis varies significantly in patients according to clinical stage and pathological grade. More sensitive PCa novel molecular markers that are associated with biological aggressiveness and providing valuable information in the diagnosis and treatment of the disease are of particular importance. Currently, effective treatment of metastatic disease is one of the major therapeutic challenges in PCa treatment.2 In recent years, many studies have focused on identifying nomograms to include various prognostic parameters to predict PCa outcome.3–5 PCa cells with high- and low-metastatic potential vary in their biological properties, such as proliferation, adhesiveness, invasiveness and motility. These variations are the results of both germ line variation between individuals and somatic alterations of genes and gene expressions in cancer cells. For these reasons, considerable efforts have been made to discover new molecular markers that can accurately predict the onset of disease recurrence and lead to better targeted and more effective treatment.
Extracellular matrix metalloproteinase inducer EMMPRIN (also known as CD147), a member of the immunoglobulin family, is a glycoprotein enriched on the surface of many types of tumor cells.6 CD147 has been demonstrated to be involved in tumor invasion and metastasis via stimulating MMP synthesis in neighboring fibroblasts, in malignant cell proliferation via the activation of ERK1/2 and p38 mitogen-activated protein kinases, in enhancing angiogenesis via vascular endothelial growth factor, in inducing chemoresistant tumor cells via the production of hyaluronan and in the resistance of cancer cells to anoikis through inhibition of Bim.7–11 Reduction of CD147 expression increased DNA fragmentation and decreases cellular viability, as demonstrated by activation of caspase-3.10 Moreover, CD147 expression is reported to correlate with the clinical prognosis of patients with breast carcinoma12 and other types of cancers.13, 14 Inhibition of CD147 gene expression with RNA interference could reduce tumor cell invasion and tumorigenicity, as well as increased chemosensitivity to paclitaxel.15 However, the distribution of CD147 in benign prostate tissues, metastatic lesions and different PCa stages, especially the correlation between the CD147 expression and the distant metastasis-free survival, and prostate-specific antigen (PSA) failure-free survival are yet to be evaluated.
In our study, we examined the CD147 expression in formalin-fixed and paraffin-embedded (FFPE) specimens from 240 PCa patients who were confirmed to have PCa with a range of Gleason scores (GS) and well characterized clinicopathological properties. The correlation of PSA failure-free survival, metastasis-free survivals and the overall survival of all the PCa patients were analyzed to evaluate the influence of CD147 expression in cancer progression. We found that positive expression rate of CD147 in patients with PCa was significantly higher than that in normal prostate. An increased expression of CD147 was correlated with patients with higher GS and positive surgical margin. Increased staining of CD147 was also associated significantly with PSA failure-free survival, distant metastasis-free survival and the overall survival. Moreover, we found that higher expression of CD147 may be able to serve as an independent prognostic predictor for PSA failure-free survival in PCa patients with different GS and for the overall survival for patients with lower GS. Our study results demonstrate the involvement of CD147 in PCa progression and suggest its potential role as an independent predictor of biochemical recurrence, development of metastasis and reduced overall survival in PCa.
Material and Methods
Two hundred and forty consecutive patients with PCa who underwent radical prostatectomy at the Massachusetts General Hospital (MGH, Boston, MA) from September 1993 to March 1995 were included in our study. Patients who received adjuvant or neoadjuvant hormonal or radiation treatment prior to cancer recurrence were excluded. All hematoxyolin–eosin (H&E)-stained sections from each case were reviewed, and the GS was reassigned based on the current grading recommendation provided by the International Society of Urological Pathology.16 The tissue blocks containing the index PCa (tumor focus with the highest GS) were selected for inclusion in the tissue microarray (TMA). Relevant clinicopathological data collected included age, preoperative PSA, GS, American Joint Committee on Cancer (AJCC) T stage, surgical margin status, PSA recurrence, postoperative metastasis and overall survival. The PSA failure-free survival was defined as the time interval between the initial surgery and the day of the first appearance of detectable PSA in patients with two consecutive rises of PSA. The distant metastasis-free survival was defined as the time interval between the initial surgery and the day of the detection of metastatic lesions. The overall survival was determined from the date of surgery to the time of the last follow-up or death. Our study received approval by the human study ethics committees at MGH, Boston, MA and of the Ministry of Public Health of PR China.
Tissue microarray construction
The TMAs were constructed by using a manual tissue-array instrument (Beecher Instruments, Silver Spring, MD). Briefly, the H&E-stained slides were reviewed for accuracy of GS. The index tumor defined as the largest and/or highest GS was marked on the slide. Tissue cylinders with a diameter of 0.6 mm were punched from selected area of each donor block and brought into a recipient paraffin block. Each block contained normal prostate tissue derived from normal peripheral zone significantly away from the tumor, and total of 20 normal prostate tissues were included as the control samples for the immunohistochemical analysis. After construction of the TMA blocks, H&E-stained sections were made for histologic evaluation.
HAb 18G/CD147 monoclonal antibody (#080109) was provided from the Fourth Military Medical University, Shanxi, China. Histostain-Plus Kit (HRP, #85-9043, Broad Spectrum) was purchased from Invitrogen Corporation, San Diego, CA.
The TMA PE tissues were cut at 5 μm and then deparaffinized with xylene and rehydrated for further H&E or peroxidase 3,3′-diaminobenzidine immunohistochemical staining using DAKO EnVision System (Dako Diagnostics, Zug, Switzerland). Briefly, after a brief proteolytic digestion and a peroxidase blocking of tissue slides, the slides were incubated overnight with the primary antibody against respective target protein at a dilution of 1:1,000 at 4°C. After washing, peroxidase labeled polymer and substrate–chromogen were then used to visualize the staining of the interested proteins. Normal prostate tissues were used as control for immunohistochemical staining. After a hematoxylin counterstaining, each stained slide was scored by two experienced pathologists independently in a blinded fashion without any information regarding the clinicopathological data and clinical outcomes of the patients. The scores of the two pathologists were compared and any discrepant scores were resolved through re-examining the slide by both pathologists simultaneously to achieve a consensus score. The numbers of positive-staining cells showing immunoreactivity on the cell membranes and cytoplasm in 10 representative microscopic fields (or the tumor cells in the entire core if tumor cells are limited) were counted, and the percentage of positive cells was calculated. Tumor specimens were scored as positive if greater than 5% of the tumor cells exhibited immunoreactivity.
The software of SAS 9.2 version (SAS Institute, Cary, NC) was used for statistical analysis. Continuous variables were expressed as . Statistical analysis was performed independently by two biostatisticians with Fisher's exact test for any 2 × 2 tables and Pearson χ2 test for non-2 × 2 tables. The Cox proportional hazards model was used to evaluate the association of CD147 expression with PSA failure-free survival, distant metastasis-free survival and overall survival in univariate and multivariate analyses. The underlying time-axis for the Cox model is years of survival. Survival curves were estimated based on the Kaplan–Meier method and assessed with log-rank test. The distribution of each clinical feature was compared between CD147-positive and CD147-negative group, with the use of two sample t-test for continuous variables and χ2 test for categorical variables. Differences were considered statistically significant when p value was less than 0.05 (two-sided).
In the univariate analyses with Cox proportional hazard model of CD147 expression, confidence interval, CD147 status (CD147 positive vs. CD147 negative), GS (6, 7, 8, 9 and 10), preoperative PSA levels ng/ml (continuous value PSA levels < 4 ng/ml is considered normal in clinical practice), AJCC pathological tumor stage (T2 and T3), surgical margin (positive vs. negative), PSA failure (biochemical recurrence) were included. The CD147 status, GS, preoperative PSA, AJCC pathological tumor stage, surgical margin and the status of PSA failure were included in the multivariate analysis. Values of p ≤ 0.05 are considered significant.
CD147 expression and localization in patients with PCa
To determine whether CD147 expression could be linked to PCa progression, we stained prostate tissue sections using the antibody that specifically recognizes CD147. In general, we found that this antibody stained the membrane and cytoplasm of PCa cells and gave evenly distributed staining patten with various intensities (Figs. 1a–1c). The positive expression rate of CD147 in tissue samples from PCa patients [113/240 (47.08%)] was significantly higher than that in normal prostate tissues [1/20 (5%)] (p = 0.006). We noticed that in some sections of normal prostate tissues very faint CD147 staining (Fig. 1d) was observed when compared to tumor tissues obtained from PCa patients, shown in Figures 1a–1c, and was considered in the negative range.
We analyzed 240 radical prostatectomy specimens represented in TMA from PCa patients. The clinical characteristics of these patients and the statistical analysis of CD147-positive expression and various clinicopathological features are shown in Table 1. We found that CD147 expression was significantly correlated with GS (p = 0.0002), positive surgical margin status of PCa patients (p < 0.0001), reduced PSA failure-free survival (p < 0.0001), reduced metastasis-free survival (p < 0.0001) and reduce overall survival (p < 0.0002). When Kaplan–Meier analysis was conducted, we found that overexpression of CD147 is significantly associated with reduced PSA failure-free survival, distant metastasis-free survival and overall survival (Figs. 2a–2c).
Table 1. CD147 expression status
CD147 expression and surgical outcome
Univariate analysis [(a) in Table 2) revealed that CD147 expression (p < 0.001, p = 0.0023 and p = 0.0006, respectively) and GS (p < 0.001, p = 0.0013 and 0.0006, respectively) were significant prognostic factors for PSA failure-free survival, distant metastasis-free survival and overall survival in patients with PCa. On the other hand, our analysis showed that preoperative PSA concentration (p < 0.001), clinical stage (p = 0.0023) and surgical margin (p = 0.0036) were significant prognostic factors for PSA failure-free survival but did not show any correlative relationship with the distant metastasis-free survival and overall survival in patients with PCa [(a) in Table 2]. We then used the Cox proportional hazards multivariate model to examine the association of clinicopathological factors and CD147 staining with PSA failure-free survival, distant metastasis-free survival and overall survival [(b) in Table 2]. Multivariate analysis indicated that CD147 expression and GS are both independent predictors of PSA failure-free survival, distant metastasis-free survival and overall survival in patients with PCa. We also found that preoperative PSA concentration was shown to be independent predictors of PSA failure-free survival. However, it could not be considered as independent predictor for distant metastasis-free survival or overall survival in patients with PCa, whereas pathological tumor stage and surgical margin preoperative were not independent predictors for PSA failure-free survival, the distant metastasis-free survival or the overall survival. These results suggested that CD147 expression directly contributes to the malignant potential of PCa. Moreover, the Kaplan–Meier survival curves also showed that the CD147 overexpression was not only a significant predictor of both PSA failure-free survival and overall survival in higher pathological stage PCa (p < 0.0001 and p = 0.0028, respectively, shown in Figs. 3c and 3d), but also in lower pathological stage PCa (p < 0.0001 and p = 0.0253, respectively, shown in Figs. 3a and 3b).
Table 2. Univariate analyses with Cox proportional hazard model of CD147 expression
CD147 expression and Gleason score
Thus far, GS is considered as one of the most reliable factors for prediction of PCa disease progression. The need for additional tests to complement and improve on the GS would be extremely helpful to identify patients who need to be treated aggressively and who can safely be monitored for disease progression. In our study, we explore the potential correlation of CD147 overexpression levels with the survival outcome in PCa patients with different GS. As shown in (a) and (b) of Table 2, both univariate and multivariate analyses revealed that CD147 expression and GS were both significant prognostic factors for PSA failure-free survival, metastatic-free survival and overall survival in patients with PCa. Kaplan–Meier survival curves indicated that CD147 expression is associated with reduced PSA failure-free survival in the GS 6 and 7 groups of patients (Figs. 4a and 4c) but not in the GS ≥ 8 group of patients. Interestingly, as shown in Figures 4d and 4f, even though CD147 was not significantly associated with the overall survival in patients with higher GS (GS = 7, p = 0.1757 and GS ≥ 8, p = 0.3151), the Kaplan–Meier survival curve indicated a correlative relationship between CD147 expression and the reduced overall survival in patients at lower GS (GS ≤ 6, p = 0.0160, shown in Fig. 4b).
It is known that the clinical course of PCa patients varies according to clinical stages and GS. Patients with the same disease stage and GS may also have different outcomes when treated with the same clinical therapy. This represents the limitation of the standard histological methods and current diagnostic modality. Therefore, it has been a major challenge to give an accurate prognosis to patients suffering from PCa. Recently, efforts are focused on finding and evaluating new molecular prognostic factors in the hope that they may contribute to better assessment of the survival probability. Consequently, the goal is to establish targeted and personalized treatment for each individual PCa patient.
In our study, we examined the expression of CD147 in PCa cancer patients. We found that the positive expression rate of CD147 in PCa tissue was significantly higher (p = 0.006) than that in normal prostate tissues (5.0%). We further evaluated any significant correlations between CD147 expression and age, preoperative prostate-specific antigen (PSA) concentrations, the GS, AJCC stage, surgical margin status, PSA failure recurrences, metastasis failure and the overall survival. Our data indicated a positive association between CD147 expression and PCa progression. When the Cox proportional hazards model was performed to evaluate the association of CD147 expression with surgery treatment outcome, the increased expression of this molecule was correlated with decreased PSA failure-free, metastasis failure-free and overall survival. We also found that the CD147 overexpression was not only a significant predictor of overall survival in higher-stage PCa (pT3) but also in lower AJCC stage PCa (pT2). Furthermore, we showed overexpression of CD147 was a disease progression predictor independently from GS by multivariate analysis. These findings extended our previous observations on the prognostic significance of CD14717–19 in cancer progression. Results from our study provided not only new insight for PCa disease progression but also useful information to help clinicians to design personalized treatment for PCa patients.
CD147 is a mono-transmembrane glycoprotein in Ig super family and has two IgSF domains. It has characteristics of an adhesive molecule and is able to mediate cell-to-cell or cell-to-matrix adhesion.20–22 CD147 expression in breast carcinomas is associated with risk factors such as poor histological grade, negative hormone status, the mitotic index and tumor size.23 Higher CD147 immunostaining scores in hepatocellular carcinomas correlate significantly with tumor grading and tumor-node metastasis stages.24 In gastric carcinoma, CD147 expression was positively correlated with tumor size, the depth of invasion and lymphatic invasion, but not with lymph node metastasis, staging or differentiation.25 However, CD147 protein expression patterns within esophageal squamous cell carcinoma and dysplastic lesions were not associated with any of these clinicopathologic factors.26 These discrepancies suggest that there are different regulatory mechanisms of CD147 expression in cells of different origins. Previously studies have shown that CD147 expression seems to correlate with poor prognosis in breast cancer, esophageal squamous cell carcinoma and ovarian serous cancers; however, it cannot be viewed as an independent prognostic factor in these cancers.18, 27, 28 We consider that CD147 may be a good target for diagnosis, prediction for disease progression and therapy of PCa for following reasons. First, CD147 is expressed in 47% of PCa samples tested. Second, the CD147 expression in PCa is significantly increased when compared to normal prostate tissues which have little to no expression, implying that anti-CD147 therapy would have minimal toxicity to normal prostate cells. Third, the CD147 expression itself showed a significant correlation with PSA failure-free survival, distant metastasis-free survival and overall survival in patients with PCa.
Among the various approaches using clinical parameters including PSA levels at the time of initial diagnosis has been explored to predict PCa progression. GS, which is assigned to a PCa based on its microscopic architectural appearance, has shown to be one of the most reliable parameters for diagnosis and prediction for PCa disease progression. The need for additional tests to complement and improve on these existing approaches would be extremely helpful to identify patients who need to be treated aggressively separately from patients who can safely be monitored for disease progression. Our study suggests that CD147 not only can serve as a significant predictor for PSA failure-free and overall survival in high-stage PCa but also can be helpful in assisting clinical PCa treatment decision making in lower-stage PCa. Mostly significantly, our data demonstrated that CD147 overexpression can serve as an indicator for PCa progression, independent from GS. The accurate analysis of biomarker expression on TMA is not a trivial undertaking, considering innate heterogeneity within tumors and abundant variation among different samples. In addition, the assay reproducibility when not analyzed in a microarray format but as individual tests on different patients in different batches is another concern when conducting work using TMA. Certainly, larger cohorts and multiple cohorts will positively impact the quality, significance and reliability of the resulting biomarker studies. It will be ideal to test the result obtained from our study in a separate patient cohort, especially in another institution. Continuing effort is in place to construct a second TMA from a new patient cohort for the confirmation and future study of the role of CD147 and its related pathway in PCa progression.
In conclusion, our data showed that CD147 may be involved in PCa progression. It may also play a role in the future assessment of newly diagnosed PCa, as it may add to the current prognostic model including preoperative PSA, GS, AJCC tumor stage and surgical margin status. Our findings provide further understanding of PCa which will be helpful for the development of the effective clinical diagnosis and therapeutic treatment. Additional studies of in vivo molecular signaling and mechanism of CD147 in PCa are likely to further highlight the advantage of the clinical diagnosis with CD147 in PCa patients.