Clinical Cancer Research The Future of Cancer Research: Science and Patient Impact
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
Clinical Cancer Research 13, 161-169, January 1, 2007. doi: 10.1158/1078-0432.CCR-06-0774
© 2007 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yao, X.
Right arrow Articles by Teh, B. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yao, X.
Right arrow Articles by Teh, B. T.

Imaging, Diagnosis, Prognosis

Two Distinct Types of Blood Vessels in Clear Cell Renal Cell Carcinoma Have Contrasting Prognostic Implications

Xin Yao1,4, Chao-Nan Qian1,5, Zhong-Fa Zhang1, Min-Han Tan6,7, Eric J. Kort2, Ximing J. Yang8, James H. Resau3 and Bin Tean Teh1

Authors' Affiliations: 1 Laboratory of Cancer Genetics, 2 Laboratory of Molecular Epidemiology, and 3 Laboratory of Analytical, Cellular and Molecular Microscopy, Laboratory of Microarray Technology, Van Andel Research Institute, Grand Rapids, Michigan; 4 Department of Urological Oncology, Tianjin Medical University Cancer Institute and Hospital, Tianjin, People's Republic of China; 5 Department of Nasopharyngeal Carcinoma, Sun Yat-sen University Cancer Center, Guangzhou, People's Republic of China; 6 Department of Medical Oncology, National Cancer Centre, Singapore; 7 Department of Molecular Epidemiology and Biostatistics, Karolinska Institute, Stockholm, Sweden; and 8 Surgical Pathology, Northwestern University Feinberg School of Medicine, Feinberg, Chicago, Illinois

Requests for reprints: Bin Tean Teh, Laboratory of Cancer Genetics, Van Andel Research Institute, 333 Bostwick Avenue Northeast, Grand Rapids, MI 49503. Phone: 616-234-5296; Fax: 616-234-5297; E-mail: Bin.Teh{at}vai.org.


   Abstract
Top
 Abstract
Materials and Methods
 Results
 Discussion
 References
 
Purpose: Intratumoral microvascular density (MVD) has been controversial as an indicator of prognosis in clear cell renal cell carcinoma (CCRCC). Classification of the intratumoral blood vessels based on differential expressions of blood vessel markers has not been correlated with patient prognosis in CCRCC. In this study, we aimed to evaluate the association of different categories of blood vessels with the patients' outcomes.

Experimental Design: Seventy-eight CCRCC patients who underwent nephrectomy alone were enrolled. Paraffin-embedded CCRCC tissues, together with 16 nonmalignant kidney cortex tissues, were used in tissue microarray analyses and conventional section analyses. The characteristics of intratumoral blood vessels were identified by multiple blood vessel markers and pericyte markers. A computerized image analysis program was used to quantitatively calculate the vascular density.

Results: Two distinct types of microvessels were identified in CCRCC: undifferentiated (CD31+/CD34–) and differentiated (CD34+) vessels. A higher undifferentiated MVD significantly correlated with higher tumor grades and shorter patient survival. In contrast, a higher differentiated MVD significantly correlated with lower tumor grade and longer survival. Multivariate analyses showed that undifferentiated MVD was an independent prognostic factor for patient survival. An inverse correlation between undifferentiated MVD and differentiated MVD was also identified in CCRCC.

Conclusions: This is the first report showing distinct types of vasculature in CCRCC correlated with contrasting prognoses. A refined classification of CCRCC based on vasculature is therefore important for evaluating prognosis, and it may also have therapeutic implications.

Angiogenesis, the generation of new blood vessels from preexisting microvasculature, is an essential process for tumor growth and is related to blood-borne metastasis (1). The quantification of various aspects of tumor vasculature might provide an indication of angiogenic activity. Microvascular density (MVD) is an often-quantified variable of tumor vasculature. Recent reports suggest that increased MVD is associated with poor outcome in several malignancies, including breast, prostate, lung, and nasopharyngeal cancers (2–7).

Clear cell renal cell carcinoma (CCRCC) is the most common subtype of malignant renal tumors, representing ~80% of renal cell carcinoma. Despite improvements in medical imaging for early diagnosis, >40% of the patients with clear cell metastatic cancers remain incurable (8). The underlying mechanism of CCRCC metastasis is unclear.

The value of MVD as a predictor of prognosis in CCRCC is controversial. Several reports have shown a positive correlation between MVD and survival or prognosis [e.g., a higher blood vessel density in CCRCC indicates a better prognosis or longer patient survival (9–13)]. However, some researchers have reported an inverse relationship (14–16), and others were unable to find a significant correlation between MVD and survival (17, 18).

The blood vessel markers used to identify tumor blood vessels are heterogeneous and are known to reveal different aspects and characteristics of tumor vasculature. For example, CD34 is expressed in differentiated endothelial cells, whereas CD31 is expressed in both differentiated and undifferentiated endothelia (19). Recently, CD105 was reported to be superior to CD34 and CD31 in evaluating tumor angiogenesis because of greater affinity for activated endothelial cells (6, 20). Moreover, some immunohistochemical studies have shown that increased MVD as determined by CD105 staining is an independent prognostic indicator of shorter survival in breast cancer and colorectal cancer (21, 22).

To date, the studies of blood vessel differentiation have not been correlated with prognosis of cancer, especially for CCRCC. The recruitment of pericytes is a crucial process in the formation and stabilization of mature blood vessels (23–25). We hypothesized that CCRCC has different characteristics in terms of vascular differentiation and maturation that correlate with different patient outcomes. In the present study, we characterized angiogenesis in CCRCC by quantitatively evaluating tumor vasculature in 78 patients that had sufficient follow-up information. Through immunohistochemical staining of CD31, CD34, CD105, and several pericyte markers on tissue microarrays (TMA) and conventional tissue sections, we were able to analyze the tumor vasculature in CCRCC and its relationship to patient survival.


   Materials and Methods
Top
 Abstract
 Materials and Methods
Results
 Discussion
 References
 
Human tissue specimens and patient information. Paraffin-embedded, formalin-fixed specimens were collected from 78 patients with histopathologically verified CCRCC (26) who underwent nephrectomy between 1983 and 2004 at Tianjin Cancer Institute and Hospital (Tianjin, China). All patients were treated with surgery alone; 53 patients relapsed after surgery and 25 were disease-free at the end of the follow-up. All patients had a performance status (Eastern Cooperative Oncology Group) of 0 before surgery. The median follow-up period for the 49 patients who died of a CCRCC-related cause was 37 months, and for the 29 survivors it was 111.5 months (range, 34-140 months). Survival times were calculated from the date of surgery. The clinical staging for each patient at the time of surgery was corrected by using the 1997 International Union Against Cancer tumor-node-metastasis classification of malignant tumors (27). All tissue samples were reevaluated by a pathologist (X.J.Y.) using the Fuhrman grading system (28). A set of 16 nonmalignant kidney cortex tissue samples was obtained from the Shared Pathology Informatics Network at the Van Andel Research Institute (Grand Rapids, MI; ref. 29). The University of California at Los Angeles Integrated Staging System has been reported to be a useful system for predicting outcome (30, 31). Among the patients with CCRCC in the present study, 74 patients without metastasis at diagnosis were divided into low-risk (n = 15), intermediate-risk (n = 49), and high-risk (n = 10) groups. The four cases with lymph node metastases at diagnosis, which were removed in the surgery, were combined into the high-risk group in the later survival analyses.

TMAs and tissue sections. TMAs were constructed as described previously (32) using Beecher instruments (Beecher Instruments, Silver Spring, MD). Three 1.0-mm cores were punched from each of the 78 cases and assembled in three TMA blocks. The three cored areas on each donor block were randomly selected from three different parts of the tumor tissue avoiding necrosis (based on a histologic characterization of the H&E-stained slide) and were entirely tumor cells. All of the 78 primary CCRCCs and 16 nonmalignant kidney tissues were included in the TMA blocks. After the development of the TMA blocks, the remaining tissues in the donor blocks were sufficient for conventional sectioning in 76 cases. Consecutive 4-µm-thick sections were cut from each of the TMA blocks, the 76 remaining CCRCC blocks, and the 16 nonmalignant kidney tissue blocks for routine H&E staining and immunohistochemical evaluation.

Immunohistochemistry and quantification of stained vessels. Immunohistochemical analyses of CD31, CD34, and CD105 were carried out using a sensitive streptavidin-biotinylated horseradish peroxidase complex system (Catalyzed Signal Amplification System, DAKO, Carpinteria, CA) according to the manufacturer's instructions. Briefly, after deparaffinization, slides were steam pretreated in a citrate buffer at pH 6.0 for 30 min. The endogenous peroxidase activity and endogenous biotin were blocked with 3% hydrogen peroxide and protein block buffer, respectively. The conventional tissue sections or the TMA sections were then incubated at room temperature for 30 min with one of the primary antibodies: mouse anti-human CD31 monoclonal antibody (mAb; working dilution 1:40, JC70A, DAKO), mouse anti-human CD34 mAb (1:50, QBEnd10, DAKO), or mouse anti-human CD105 mAb (1:100, clone SN6h, DAKO). Normal mouse IgG1 was used as a substitute for the primary antibody in the negative controls. After washes using TBS with 0.1% Tween 20, the sections were incubated at room temperature with biotinylated rabbit anti-mouse secondary antibody for 15 min followed by TBS washes. The sections were then incubated with streptavidin-biotin complex for 15 min. For the staining of smooth muscle actin (SMA), a slightly different immunohistochemical procedure was used than described previously (33). Briefly, after antigen retrieval, the sections were incubated with rabbit anti-human SMA (1:100, RB-9010-P0, Lab Vision, Fremont, CA) at 4°C overnight. Then, the sections were incubated at room temperature with the secondary antibody (biotinylated anti-mouse IgG raised in a horse) at 1:200 (Vector, Burlingame, CA) for 30 min. Staining was carried out using 3,3'-diaminobenzidine and hydrogen peroxide. All sections were counterstained with hematoxylin.

Intratumoral MVD was assessed according to the criteria described previously, with some modifications (34, 35). Briefly, the three most vascularized areas (hotspots) within each TMA core were selected for quantification of blood vessels at a magnification of x400 with a Nikon (Melville, NY) microscope. Therefore, a total of nine hotspots was evaluated from the three TMA cores for each patient. For the conventional CCRCC tissue sections, five hotspots were selected at a magnification of x200 for each patient. The five selected hotspots were randomly distributed in the center of tumor nests or in the peripheral areas of tumor nests. For evaluating the discrepant expression of CD31 and CD34 in two consecutive conventional tissue sections, the hotspots were defined by the most vascularized area in CD31 staining slides. Any brown-staining endothelial cell or endothelial cell cluster that clearly separated from adjacent microvessels, tumor cells, and connective elements was considered as a single, countable microvessel regardless of whether a vessel lumen was seen. Image analysis for vessel counting was done using interactive software developed by our laboratory. The software allows the application of standardized computational algorithms as well as review and refinement of the results of those algorithms by human operators. The mean value of the vessel counts in the selected spots was retained as the final MVD value. The count of undifferentiated microvessels in each case was obtained by subtracting the CD34+ vessel count from the CD31+ vessel count.

Immunofluorescent staining. Conventional tissue sections of CCRCC and normal kidney tissues were steam pretreated in a citrate buffer at pH 6.0 for 30 min after deparaffinization. Nonspecific antibody-binding sites in the tissue sections were blocked by incubation in 5% donkey serum (30 min at room temperature). Next, the sections were incubated overnight at 4°C in various combinations of two antibodies staining different pericyte markers and/or endothelial cell markers. The mouse anti-human CD31 mAb (working dilution 1:20) and mouse anti-human CD34 mAb (1:25) were the same as in immunohistochemical staining. To stain the pericyte markers, several antibodies were used, including rabbit anti-human SMA antibody (1:50), mouse anti-human SMA mAb (ready to use, clone CGA7, Chemicon, Temecula, CA), rabbit anti-human desmin antibody (1:40, Abcam, Cambridge, MA), and mouse anti-human NG2 mAb (1:20, clone LHM-2, R&D Systems, Minneapolis, MN). After rinses with TBS with 0.1% Tween 20, sections were incubated in rhodamine-conjugated donkey anti-mouse and fluorescein-conjugated donkey anti-rabbit secondary antibodies at 1:150 dilution (Jackson ImmunoResearch Laboratories, West Grove, PA) for 1 h at room temperature. 4',6-Diamidino-2-phenylindole was then applied to stain the nuclei. Confocal fluorescent microscopy was done to evaluate the slides.

Statistical methods. We used the median as the cutoff value for vessel counts of TMA tissues. The Cox proportional hazard regression models were fitted for both univariate and multivariate analyses. The censoring time distribution was estimated by Kaplan-Meier method. Pearson's product-moment correlations were calculated and then tested for significance between the TMA CD34+ vessel and TMA CD105+ vessel MVDs and between the differentiated blood vessel and undifferentiated blood vessel MVDs. Spearman rank correlation coefficients were calculated and then tested between Fuhrman grading and MVD as well as between University of California at Los Angeles Integrated Staging System stage and MVD.


   Results
Top
 Abstract
 Materials and Methods
 Results
Discussion
 References
 
Two distinct types of microvessels can be identified in CCRCC. Of all the blood vessels in CCRCC revealed by anti-CD31 antibody staining, most could also be stained by anti-CD34 antibody (Fig. 1A and B ); however, certain blood vessels were stained only by anti-CD31 antibody but not by anti-CD34 antibody (Fig. 1B, arrows). It has been reported that CD34 is only expressed in differentiated endothelial cells, whereas CD31 is expressed in both differentiated and undifferentiated endothelial cells (19). We further evaluated the pericyte coverage of the blood vessels by staining for the pericyte marker SMA (36). The results showed that the CD34+ vessels could be surrounded by pericytes, whereas CD31+/CD34– microvessels had no surrounding pericytes (Fig. 1C). Therefore, there were two types of blood vessels in CCRCC: CD34+ and CD31+/CD34– vessels. The morphologic characteristics of CD31+/CD34– blood vessels in our observation included no or small lumen, thicker vessel wall, and smaller size when compared with the CD34+ vessels. Based on these observations, we defined CD34+ vessels as differentiated vessels and CD31+/CD34– vessels as undifferentiated counterpart.


Figure 1
View larger version (83K):
[in this window]
[in a new window]

  		Fig. 1. A to C, serial sections of CCRCC tissue were used for immunohistochemical staining. A, staining for CD34 clearly revealed the differentiated blood vessels in CCRCC. B, staining for CD31 in the same area of (A) showed additional blood vessels. Arrows, undifferentiated vessels that were not stained by anti-CD34 antibody. C, staining of SMA proved those vessels revealed by CD34 staining were surrounded by the pericytes. Note that the undifferentiated vessels in (B) (arrows) were not surrounded by pericytes. D to F, triple immunofluorescent staining images of CCRCC tissues. Blue, 4',6-diamidino-2-phenylindole stained the nuclei. D, close association of pericytes [SMA staining (green)] to endothelial cells [CD34 staining (red)] was seen in most of the differentiated blood vessels. E, partial or complete absence of association of pericytes and endothelial cells was also found in some differentiated blood vessels. F, desmin (green) was only expressed in a part of a few pericytes [SMA staining (red)] in CCRCC.

 
The association of pericytes to endothelial cells varied in human CCRCC. For the undifferentiated vessels, no pericyte coverage was found. For those differentiated vessels, a close association of pericytes and CD34+ endothelial cells was usually found in the peripheral area of tumor mass (Fig. 1D). However, loose association, partial absence, or even complete absence of pericytes was also found for other CD34+ differentiated vessels (Fig. 1E). Therefore, in terms of blood vessel maturation defined by pericyte coverage, the undifferentiated vessels were more immature than the differentiated vessels in CCRCC. Although both SMA and desmin were reported to be effective tumor pericyte markers in animal models (37), we found that desmin was not expressed in most of the tumor pericytes in human CCRCC (Fig. 1F).

TMA analyses and patient prognoses. We prepared CCRCC TMAs from all 78 cases. From these, we determined the correlation(s) between the expression of different blood vessel markers as well as between the MVD and patient survival. The immunoactivity of CD31 covered the largest number of intratumor vessels, including undifferentiated vessels, which were characterized as small, scattered, and having no lumen, together with differentiated vessels that were stained similarly by CD34 and CD105 (Fig. 2 ). The patient follow-up curve is shown in Fig. 3A . A significant correlation between T staging or Fuhrman grading and survival time was found (Fig. 3B and C). The clinicopathologic characteristics of the patients grouped by high and low MVD (based on TMA results) are shown in Table 1 . The MVDs of CD31+, CD34+, and CD105+ vessels were separately evaluated from their respective hotspots. The MVD of undifferentiated blood vessels was derived by subtracting the CD34+ vessel MVD from the CD31+ vessel MVD. The median was used as the cutoff point to divide the patients into high- and low-MVD groups.


Figure 2
View larger version (139K):
[in this window]
[in a new window]

  		Fig. 2. Serial sections of a normal kidney tissue, a low-grade CCRCC, and a high-grade CCRCC stained by different endothelial markers. The staining for CD105 was not specific in normal renal vasculature, although it was as good as CD34 in identifying differentiated tumor blood vessels. The undifferentiated blood vessels in high-grade CCRCC, which were a portion of the CD31+ vessels, were not stained by anti-CD34 or anti-CD105 antibodies.

 

Figure 3
View larger version (17K):
[in this window]
[in a new window]

  		Fig. 3. A, Kaplan-Meier estimate of the censoring time distribution. Except for one patient with postoperational metastases to the lymph node and the lung who was lost 34 mos after nephrectomy, the follow-up duration of the censored cases was longer than 7 yrs. B, Kaplan-Meier overall survival curves in terms of T stages. C, Kaplan-Meier overall survival curves in terms of pathologic grades (Fuhrman grading system). D, scatter plots of MVD identified by CD34 staining versus CD105 staining in their respective hotspots on the TMA slides. A total of 77 CCRCC cases qualified for the comparisons. E and F, scatter plots of MVD identified by CD34 (E) or CD31 (F) staining in TMA slides versus conventional slides. Analyses of correlation efficiency (COR) revealed significant consistency in both markers. G and H, rank correlation analyses were used to test the relationship between different types of MVD and pathologic grades (G) or University of California at Los Angeles Integrated Staging System (UISS) groups (H) in 76 cases of CCRCC.

 

View this table:
[in this window]
[in a new window]

  		Table 1. Clinicopathologic characteristics and MVD in CCRCC analyzed by TMAs (number of patients)

 
Univariate survival analyses (Table 2 ) showed that the MVD of CD34+ vessels from TMA analyses had a hazard ratio (HR) of 0.673 for every 10 adding vessels within a tissue area of 0.065 mm2 (x400 magnification), meaning that higher MVD indicated longer survival. Similar results were also found in the analysis of the CD105+ vessels (HR, 0.809). In contrast, the MVD of undifferentiated vessels had a HR of 1.68, meaning that higher MVD indicated shorter survival.


View this table:
[in this window]
[in a new window]

  		Table 2. Univariate Cox proportional hazard analysis of the relations between MVD and overall survival

 
As shown in Fig. 2, CD105 was not specifically expressed in the endothelial cells of normal kidney tissue. In tumor vasculature, CD105 was only expressed in the CD34+ (differentiated) blood vessels. A high correlation coefficient of 0.89 was found between CD34+ MVD and CD105+ MVD (Fig. 3D), implying that these two markers identified almost the same quantity of differentiated tumor blood vessels and no additional information could be provided by CD105 immunoactivity. Therefore, we only used CD34 staining to evaluate differentiated blood vessels in further studies.

Comparisons of the MVDs determined by TMA and conventional sectioning approaches. The interesting results from our TMA analyses encouraged us to confirm the correlation between tumor MVD and tumor grade in CCRCC using conventional tissue sections. The TMAs and the conventional sectioning approaches showed significant linear correlation in identifying the MVDs of CD34+ and CD31+ vessels (Fig. 3E and F), implying the effectiveness of TMA analyses for preliminary study in MVD evaluation.

The correlation between MVD and tumor grade in CCRCC. Fuhrman grading has been proven to be a strong prognostic indicator for RCC (18, 28), with the higher grades associated with poorer prognosis, which is consistent with our data (Fig. 3C). We found that the density of undifferentiated blood vessels determined in conventional sections was positively associated with tumor grade (Fig. 3G), with significantly more undifferentiated blood vessels in higher-grade tumors than in lower-grade tumors. However, the number of differentiated blood vessels was negatively correlated with tumor grade; there were significantly fewer differentiated blood vessels in high-grade tumors. In terms of total blood vessel density stained by anti-CD31 antibody, a negative correlation was also found with the tumor grade, meaning that the total vessel MVD decreased as the grade increased. The histologic appearance of the vasculature in low- and high-grade tumors is shown in Fig. 2. The correlation between the densities of undifferentiated or differentiated blood vessels and University of California at Los Angeles Integrated Staging System was also significant (Fig. 3H).

The correlation between MVD and patient survival in CCRCC. Univariate Cox proportional hazard analysis was done to evaluate the relation between every 10-vessel increment in conventional sections and overall survival. The assumption of proportionality of hazards was tested using the method outlined by Grambsch and Therneau (38), which showed no significant correlation between the Schoenfeld residuals and transformed survival time (Supplementary Fig. S1). The HRs of both undifferentiated vessels and differentiated CD34+ vessels in conventional sections derived from an area of 0.26 mm2 (x200 magnification) were consistent with those from TMA (Table 2). These results implied that higher undifferentiated vessel MVD indicates poorer prognosis, whereas higher differentiated vessel MVD indicated better prognosis. In terms of total vessels stained with CD31, the HR of 0.926 was also found to be significant in conventional section analyses.

For multivariate analyses, we started from Cox proportional hazard models with covariates of age, gender, T stage, and tumor grade with either undifferentiated or differentiated MVD, or their interaction term, or any combinations of them, except undifferentiated MVD and differentiated MVD should not be in the same model because of their obvious correlationship. The variables of age and gender were eliminated because they were not significant, leaving the three models listed in Table 3 . These results showed that, besides tumor grade and T stage, undifferentiated MVD was also an independent prognostic factor of CCRCC in multivariate analyses (model II). However, the differentiated MVD was not an independent prognostic factor any more as long as the more important prognostic factors (i.e., tumor grade and T stage) presented (model I).


View this table:
[in this window]
[in a new window]

  		Table 3. Multivariate Cox proportional hazard analyses

 
Inverse correlation between differentiated MVD and undifferentiated MVD. The relationship between undifferentiated vessels and differentiated vessels was also of interest. We found an inverse correlation between undifferentiated MVD and differentiated MVD (Fig. 4A ). A correlation coefficient of –0.54 was calculated, with a P value of <0.0001, meaning more undifferentiated blood vessels in the tumor were usually accompanied by fewer differentiated blood vessels, and conversely, fewer undifferentiated blood vessels in the tumor were usually accompanied by more differentiated blood vessels.


Figure 4
View larger version (9K):
[in this window]
[in a new window]

  		Fig. 4. A, scatter plots of differentiated MVD versus undifferentiated MVD in the same hotspots of the conventional tissue sections of CCRCCs. The differentiated vessels were identified by CD34 staining. The undifferentiated vessel count was derived from the subtraction of the CD34+ vessel count from the CD31+ vessel count. A total of 76 cases qualified for the comparisons. B, Kaplan-Meier overall survival analysis. L-U, low undifferentiated MVD; H-U, high undifferentiated MVD; L-D, low differentiated MVD; H-D, high differentiated MVD. The survival was compared among the four recombination groups: low undifferentiated plus high differentiated MVD, low undifferentiated plus low differentiated MVD, high undifferentiated plus low differentiated MVD, high undifferentiated plus high differentiated MVD. The survival time was significant longer in low undifferentiated plus high differentiated MVD group than in high undifferentiated plus low differentiated MVD group (P = 0.0029).

 
To clarify the contrasting prognostic implications of differentiated and undifferentiated MVD, the patients were regrouped by the recombination of high or low undifferentiated MVD plus high or low differentiated MVD (Fig. 4B). The survival time of the patients with low undifferentiated plus high differentiated MVD was significantly longer than that of patients with high undifferentiated plus low differentiated MVD (P = 0.0029). Interestingly, the patients with higher differentiated plus high undifferentiated MVD seemed to have better prognosis than those with low differentiated plus low undifferentiated MVD, but this tendency was not statistically significant (P = 0.18).

The variable of interaction between undifferentiated and differentiated MVD was derived by multiplying undifferentiated MVD by differentiated MVD. When multivariate Cox proportional survival analysis was used to evaluate the prognosis prediction of the interaction, a significant association with survival was also found (model III; Table 3), implying that the interaction between undifferentiated and differentiated MVD was also an independent prognostic factor in CCRCC. The variable of interaction could not be involved in a multivariate model together with undifferentiated MVD because it had a significantly positive correlation with undifferentiated MVD (Supplementary Fig. S2).


   Discussion
Top
 Abstract
 Materials and Methods
 Results
 Discussion
References
 
The stability and the repeatability of immunohistochemical staining are critical in the evaluation of MVD when using different antibodies. For example, the affinity of most of the anti-CD31 mAbs is less than that of most anti-CD34 mAbs, and sometimes, this could result in insufficient labeling of the tumor vasculature (39). The result of this situation would be an underestimation of MVD as determined by CD31 staining (2). In our study, the immunohistochemical staining quality was highly controlled; therefore, a critical difference between CD31 and CD34 staining in CCRCC was discovered. We found two distinct types of blood vessels [i.e., undifferentiated (CD31+/CD34–) and differentiated (CD34+) vessels] by direct immunohistochemical staining of the vasculature. Although we assumed no vessel was CD31–/CD34+, there were 5 of 76 conventional sections showing a slightly higher (0.01-4.4% higher) MVD of CD34+ vessels versus CD31+ vessels, leading to a few negative values of undifferentiated MVD instead of zeros in Fig. 4A. The mild discrepancy of MVD counted by CD34 staining versus CD31 staining could result from the systemic error of continuous sectioning of the tissues for the two separately staining procedures. For example, on very rare occasions, one particular vessel in a tissue section could disappear or become two separate vessels in the next section.

Our results are consistent with the finding of Eberhard et al. (23) that there is a significant proportion of blood vessels in RCC without pericyte coverage. By analyzing pericyte recruitment to tumor microvasculature, they found that the microvessel pericyte coverage index in RCC was only 17.9 ± 7.8% versus 65.4 ± 10.5% in colon carcinoma, 67.3 ± 14.2% in mammary carcinoma, 40.8 ± 14.5% in lung carcinoma, and 29.6 ± 9.5% in prostate carcinoma. Only glioblastoma has a lower percentage of pericyte coverage index than RCC. Their study implies that, compared with most of the common cancers, CCRCC possesses unique vasculature characteristics with a significantly higher proportion of immature microvessels. Because the kidney is essentially a blood purification system, it is not unexpected that the vasculature will be unique and significant in the characterization and analysis of CCRCC.

It has been reported that pericyte coverage is a correct functional reflection of microvessel maturation despite the fact that it is not the only indicator for maturation (23). Several molecular markers, such as SMA, desmin, platelet-derived growth factor receptor-ß, and NG2, have been used to identify pericytes. We compared the immunoactivity of NG2, desmin, and SMA in pericytes within CCRCC tissues. We found that there were overwhelmingly more SMA-positive pericytes than desmin-positive or NG2-positive pericytes, implying that desmin and NG2 were not effective pericyte markers in CCRCC. In our observation, the coverage of pericytes on CD34+ vessels varied in CCRCC, consistent with the finding from animal models showing multiple structural abnormalities in tumor pericytes (37). The significance of the association between pericytes and endothelial cells in terms of patient prognosis is unclear.

In this study, we were able for the first time to quantitatively evaluate the MVD of undifferentiated microvessels in CCRCC. We found that an increased undifferentiated vessel MVD significantly correlated with higher pathologic grades and shorter patient survival. Moreover, it was an independent prognostic factor of survival in multivariate analyses. In contrast, a higher differentiated vessel MVD significantly correlated with lower pathologic grades and longer survival. The MVD of undifferentiated vessels itself was negatively correlated with the MVD of differentiated vessels.

When undifferentiated MVD and differentiated MVD were considered together, we found that the patients with low counts of undifferentiated vessels plus high counts of differentiated vessels had significantly better prognosis than the patients with high counts of undifferentiated plus low counts of differentiated, further indicating the contrasting prognostic implications of these two types of blood vessels. Because the interaction of these two types of blood vessel had a linear correlation with undifferentiated MVD, it was not surprising that the interaction was significantly correlated with poorer prognosis.

Interestingly, when the whole vasculature was considered by staining with the broad-spectrum vasculature marker CD31, a higher MVD significantly correlated with lower pathologic grades and longer patient survival. Our results support the observation that the number of differentiated microvessels in CCRCC is a significant prognostic factor. The critical effect of undifferentiated vessels might not be noticed when the analysis was focused on the general profile of tumor vasculature in CCRCC. It is believed that rapid tumor growth does not imply high vascular density, and the MVD of a tumor need not be higher, and is often lower, than that of its corresponding normal tissue (40). RCC is just one of the malignancies with a lower MVD in tumor tissue than in normal tissue (23).

We believe that simultaneous comparison of the immunoactivities of CD31 and CD34 in the CCRCC vasculature can not only display the whole profile of tumor vasculature but also reveal a unique vascular category of CD31+/CD34– vessels that is significantly correlated with patient prognosis. The therapeutic implication of this unique vascular category in CCRCC is unclear, although immature vasculature is believed to be more vulnerable for therapeutic targeting (41). Antiangiogenic therapies are thought to induce the pruning of immature and inefficient vessels and to promote the maturation of the remaining vessels (42). In recent clinical trials, a mild benefit in survival of kidney cancer patients has been achieved by administration of Raf kinase inhibitor sorafenib, which inhibits both tumor cell proliferation and tumor angiogenesis (43, 44). Another multitargeted tyrosine kinase inhibitor, sunitinib, which specifically inhibits vascular endothelial growth factor receptor and platelet-derived growth factor receptor as well as other tyrosine protein kinases, is also effective in improving the survival of patients with metastatic CCRCC (45, 46). Evaluation of the MVD of CD31+/CD34– vessels may help to identify the tumors susceptible for antiangiogenic targeted therapy. Better understanding of the detailed mechanism of angiogenesis in CCRCC will undoubtedly facilitate the development of more effective antiangiogenic therapy for this deadly disease.

In conclusion, we quantitatively evaluated two distinct types of microvessels in CCRCC, analysis of which resulting in contrasting prognostic implications. A higher density of undifferentiated microvessels was an independent prognostic factor of shorter survival. In contrast, a higher differentiated microvessel density was significantly correlated with longer survival. Our finding suggests that the tumor vasculature in CCRCC is heterogeneous, and a refined classification of tumor vasculature in CCRCC is necessary for further exploration of the role of angiogenesis in this particular tumor. Our study also has therapeutic implications: the undifferentiated blood vessels are potentially therapeutic targets and therefore it is imperative to study if current or future antiangiogenic drugs act on differentiated or undifferentiated blood vessels or on both.


   Acknowledgments
 
We thank Bree Berghuis, Eric Hudson, and J.C. Goolsby (Laboratory of Analytical, Cellular, and Molecular Microscopy, Van Andel Research Institute) for technical support in the immunohistochemical staining; Brandon Leeser (Laboratory of Analytical, Cellular, and Molecular Microscopy, Van Andel Research Institute) for technical support in the development of the TMA blocks; David Nadziejka (Van Andel Research Institute) for proofreading the manuscript; and Sabrina Antio for assisting in preparing and submitting the manuscript.


   Footnotes
 
Grant support: Hauenstein Foundation, Schregardus Family Foundation, Fischer Family Trust, and The Gerber Foundation.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).

Received 3/29/06; revised 8/29/06; accepted 9/15/06.


   References
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 References
 

  1. Folkman J, Kalluri R. Tumor angiogenesis. In: Kufe DW, Pollock RE, Weichselbaum RR, Bast RCJ, Gansler TS, Holland JF, Frei EI, editors. Cancer medicine. Hamilton (Ontario, Canada): B.C. Decker, Inc.; 2003. p. 161–94.
  2. de la Taille A, Katz AE, Bagiella E, et al. Microvessel density as a predictor of PSA recurrence after radical prostatectomy. A comparison of CD34 and CD31. Am J Clin Pathol 2000;113:555–62.[CrossRef][Medline]
  3. Kato T, Kameoka S, Kimura T, Nishikawa T, Kobayashi M. The combination of angiogenesis and blood vessel invasion as a prognostic indicator in primary breast cancer. Br J Cancer 2003;88:1900–8.[CrossRef][Medline]
  4. Mineo TC, Ambrogi V, Baldi A, et al. Prognostic impact of VEGF, CD31, CD34, and CD105 expression and tumour vessel invasion after radical surgery for IB-IIA non-small cell lung cancer. J Clin Pathol 2004;57:591–7.[Abstract/Free Full Text]
  5. Qian CN, Min HQ, Liang XM, Zheng SS, Lin HL. Primary study of neovasculature correlating with metastatic nasopharyngeal carcinoma using computer image analysis. J Cancer Res Clin Oncol 1997;123:645–51.[CrossRef][Medline]
  6. Tanaka F, Otake Y, Yanagihara K, et al. Evaluation of angiogenesis in non-small cell lung cancer: comparison between anti-CD34 antibody and anti-CD105 antibody. Clin Cancer Res 2001;7:3410–5.[Abstract/Free Full Text]
  7. Weidner N, Folkman J, Pozza F, et al. Tumor angiogenesis: a new significant and independent prognostic indicator in early-stage breast carcinoma. J Natl Cancer Inst 1992;84:1875–87.[Abstract/Free Full Text]
  8. Edwards BK, Brown ML, Wingo PA, et al. Annual report to the nation on the status of cancer, 1975-2002, featuring population-based trends in cancer treatment. J Natl Cancer Inst 2005;97:1407–27.[Abstract/Free Full Text]
  9. Anastassiou G, Duensing S, Steinhoff G, et al. Platelet endothelial cell adhesion molecule-1 (PECAM-1): a potential prognostic marker involved in leukocyte infiltration of renal cell carcinoma. Oncology 1996;53:127–32.[Medline]
  10. Imao T, Egawa M, Takashima H, Koshida K, Namiki M. Inverse correlation of microvessel density with metastasis and prognosis in renal cell carcinoma. Int J Urol 2004;11:948–53.[CrossRef][Medline]
  11. Rioux-Leclercq N, Epstein JI, Bansard JY, et al. Clinical significance of cell proliferation, microvessel density, and CD44 adhesion molecule expression in renal cell carcinoma. Hum Pathol 2001;32:1209–15.[CrossRef][Medline]
  12. Sabo E, Boltenko A, Sova Y, et al. Microscopic analysis and significance of vascular architectural complexity in renal cell carcinoma. Clin Cancer Res 2001;7:533–7.[Abstract/Free Full Text]
  13. Schraml P, Struckmann K, Hatz F, et al. VHL mutations and their correlation with tumour cell proliferation, microvessel density, and patient prognosis in clear cell renal cell carcinoma. J Pathol 2002;196:186–93.[CrossRef][Medline]
  14. Joo HJ, Oh DK, Kim YS, Lee KB, Kim SJ. Increased expression of caveolin-1 and microvessel density correlates with metastasis and poor prognosis in clear cell renal cell carcinoma. BJU Int 2004;93:291–6.[CrossRef][Medline]
  15. Nativ O, Sabo E, Reiss A, et al. Clinical significance of tumor angiogenesis in patients with localized renal cell carcinoma. Urology 1998;51:693–6.[CrossRef][Medline]
  16. Yoshino S, Kato M, Okada K. Prognostic significance of microvessel count in low stage renal cell carcinoma. Int J Urol 1995;2:156–60.[Medline]
  17. MacLennan GT, Bostwick DG. Microvessel density in renal cell carcinoma: lack of prognostic significance. Urology 1995;46:27–30.[CrossRef][Medline]
  18. Minardi D, Lucarini G, Mazzucchelli R, et al. Prognostic role of Fuhrman grade and vascular endothelial growth factor in pT1a clear cell carcinoma in partial nephrectomy specimens. J Urol 2005;174:1208–12.[CrossRef][Medline]
  19. Poblet E, Gonzalez-Palacios F, Jimenez FJ. Different immunoreactivity of endothelial markers in well and poorly differentiated areas of angiosarcomas. Virchows Arch 1996;428:217–21.[Medline]
  20. Saad RS, El-Gohary Y, Memari E, Liu YL, Silverman JF. Endoglin (CD105) and vascular endothelial growth factor as prognostic markers in esophageal adenocarcinoma. Hum Pathol 2005;36:955–61.[CrossRef][Medline]
  21. Charpin C, Dales JP, Garcia S, et al. Tumor neoangiogenesis by CD31 and CD105 expression evaluation in breast carcinoma tissue microarrays. Clin Cancer Res 2004;10:5815–9.[Abstract/Free Full Text]
  22. Li C, Gardy R, Seon BK, et al. Both high intratumoral microvessel density determined using CD105 antibody and elevated plasma levels of CD105 in colorectal cancer patients correlate with poor prognosis. Br J Cancer 2003;88:1424–31.[CrossRef][Medline]
  23. Eberhard A, Kahlert S, Goede V, et al. Heterogeneity of angiogenesis and blood vessel maturation in human tumors: implications for antiangiogenic tumor therapies. Cancer Res 2000;60:1388–93.[Abstract/Free Full Text]
  24. Jain RK. Molecular regulation of vessel maturation. Nat Med 2003;9:685–93.[CrossRef][Medline]
  25. Lindahl P, Johansson BR, Levâeen P, Betsholtz C. Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science 1997;277:242–5.[Abstract/Free Full Text]
  26. Kovacs G, Akhtar M, Beckwith BJ, et al. The Heidelberg classification of renal cell tumours. J Pathol 1997;183:131–3.[CrossRef][Medline]
  27. Guinan P, Sobin LH, Algaba F, et al. TNM staging of renal cell carcinoma: Workgroup No. 3. Union International Contre le Cancer (UICC) and the American Joint Committee on Cancer (AJCC). Cancer 1997;80:992–3.[CrossRef][Medline]
  28. Fuhrman SA, Lasky LC, Limas C. Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol 1982;6:655–63.[Medline]
  29. Kort EJ, Campbell B, Resau JH. A human tissue and data resource: an overview of opportunities, challenges, and development of a provider/researcher partnership model. Comput Methods Programs Biomed 2003;70:137–50.[CrossRef][Medline]
  30. Zisman A, Pantuck AJ, Figlin RA, Belldegrun AS. Validation of the ucla integrated staging system for patients with renal cell carcinoma. J Clin Oncol 2001;19:3792–3.[Free Full Text]
  31. Zisman A, Pantuck AJ, Wieder J, et al. Risk group assessment and clinical outcome algorithm to predict the natural history of patients with surgically resected renal cell carcinoma. J Clin Oncol 2002;20:4559–66.[Abstract/Free Full Text]
  32. Sauter G, Simon R, Hillan K. Tissue microarrays in drug discovery. Nat Rev Drug Discov 2003;2:962–72.[CrossRef][Medline]
  33. Qian CN, Knol J, Igarashi P, et al. Cystic renal neoplasia following conditional inactivation of apc in mouse renal tubular epithelium. J Biol Chem 2005;280:3938–45.[Abstract/Free Full Text]
  34. Folkman J. Seminars in Medicine of the Beth Israel Hospital, Boston. Clinical applications of research on angiogenesis. N Engl J Med 1995;333:1757–63.[Free Full Text]
  35. Weidner N, Semple JP, Welch WR, Folkman J. Tumor angiogenesis and metastasis—correlation in invasive breast carcinoma. N Engl J Med 1991;324:1–8.[Abstract]
  36. Yonenaga Y, Mori A, Onodera H, et al. Absence of smooth muscle actin-positive pericyte coverage of tumor vessels correlates with hematogenous metastasis and prognosis of colorectal cancer patients. Oncology 2005;69:159–66.[CrossRef][Medline]
  37. Morikawa S, Baluk P, Kaidoh T, et al. Abnormalities in pericytes on blood vessels and endothelial sprouts in tumors. Am J Pathol 2002;160:985–1000.[Abstract/Free Full Text]
  38. Grambsch PM, Therneau TM. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 1994;81:515–26.[Abstract/Free Full Text]
  39. Beenken SW, Bland KI. Biomarkers for breast cancer. Minerva Chir 2002;57:437–48.[Medline]
  40. Hlatky L, Hahnfeldt P, Folkman J. Clinical application of antiangiogenic therapy: microvessel density, what it does and doesn't tell us. J Natl Cancer Inst 2002;94:883–93.[Free Full Text]
  41. Feron O. Targeting the tumor vascular compartment to improve conventional cancer therapy. Trends Pharmacol Sci 2004;25:536–42.[CrossRef][Medline]
  42. Jain RK. Molecular regulation of vessel maturation. Nat Med 2003;9:685–93.[CrossRef][Medline]
  43. Gore ME, Escudier B. Emerging efficacy endpoints for targeted therapies in advanced renal cell carcinoma. Oncology (Williston Park) 2006;20:19–24.
  44. Ratain MJ, Eisen T, Stadler WM, et al. Phase II placebo-controlled randomized discontinuation trial of sorafenib in patients with metastatic renal cell carcinoma. J Clin Oncol 2006;24:2505–12.[Abstract/Free Full Text]
  45. Motzer RJ, Michaelson MD, Redman BG, et al. Activity of SU11248, a multitargeted inhibitor of vascular endothelial growth factor receptor and platelet-derived growth factor receptor, in patients with metastatic renal cell carcinoma. J Clin Oncol 2006;24:16–24.[Abstract/Free Full Text]
  46. Motzer RJ, Rini BI, Bukowski RM, et al. Sunitinib in patients with metastatic renal cell carcinoma. JAMA 2006;295:2516–24.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Clin. Cancer Res.Home page
R. Riesenberg, C. Weiler, O. Spring, M. Eder, A. Buchner, T. Popp, M. Castro, R. Kammerer, O. Takikawa, R. A. Hatz, et al.
Expression of Indoleamine 2,3-Dioxygenase in Tumor Endothelial Cells Correlates with Long-term Survival of Patients with Renal Cell Carcinoma
Clin. Cancer Res., December 1, 2007; 13(23): 6993 - 7002.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yao, X.
Right arrow Articles by Teh, B. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yao, X.
Right arrow Articles by Teh, B. T.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online