The assessment of angiogenesis in breast cancer is of importance as a key indicator of survival and response to therapy. Circulating vascular endothelial growth factor (VEGF) measurements may provide a less subjective analysis than microvessel density (MVD) or immunohistochemical analysis of VEGF expression; however, most studies have used serum, which is now known to largely reflect platelet-derived VEGF concentrations. This study examined for the first time both plasma(VEGFP) and serum (VEGFS) VEGF concentrations in 201 blood samples from pre- and postmenopausal healthy controls and from patients with benign breast disease, localized breast cancer,breast cancer in remission, or metastatic breast cancer and related these to other clinicopathological markers. VEGFP but not VEGFS concentrations of patients with localized disease were significantly elevated compared with normal controls(P = 0.016). Patients with metastatic disease had higher VEGFP and VEGFS levels than normal controls (P < 0.001, P = 0.044 respectively), and higher VEGFP, but not VEGFS, than patients with benign disease (P = 0.009) and patients with localized disease (P = 0.004). However,the highest VEGFP and VEGFS concentrations were seen in patients in remission compared with normal controls(P < 0.001 and P = 0.008, respectively). VEGFP concentrations in patients in remission were also higher than in patients with benign disease (P = 0.01) or patients with localized disease(P = 0.005). Tamoxifen treatment was significantly associated with higher circulating and platelet-derived VEGF levels. Circulating VEGF did not correlate with any clinicopathological factor, including MVD or VEGF expression. VEGF expression was significantly correlated with estrogen receptor status and inversely correlated with tumor grade. MVD correlated with tumor size. Tamoxifen-induced increases in VEGF may be important in clinical prognosis or associated pathologies.

Angiogenesis, the formation of new blood vessels from the existing vascular network, is essential for continued tumor growth and metastasis (1). One of the most potent and specific angiogenic factors is VEGF,3also known as vascular permeability factor and vasculotropin(reviewed in Ref. 2). Evidence for the pivotal role of this cytokine in tumor angiogenesis includes the observations of increased expression in tumor cells of numerous human cancers together with up-regulation of the receptors on the associated endothelial cells and the inhibitory effect of anti-VEGF antibodies on tumor growth in vivo (reviewed in Ref. 2).

Since the pivotal findings in breast cancer of a correlation between tumor angiogenesis, as assessed by microvessel count and density, and metastasis (3), many studies have confirmed the clinical value of this parameter (4), and IMD is now established as one of the standard prognostic factors for predicting metastasis and relapse-free or overall survival. Longer-term survival is also predicted by IMD as an independent prognostic factor (5),and in addition, IMD has been found to add prognostic information to ER status in predicting response to tamoxifen treatment (6). Higher VEGF mRNA levels have been found in invasive breast carcinoma or DCIS, compared with benign or normal breast tissue(7, 8, 9, 10), and in association with higher vessel counts (8, 10, 11, 12, 13). Assessment of VEGF expression by immunohistochemistry or immunoassay of tissue extracts has shown significant correlations with microvessel counts or density (10, 14, 15, 16). In node-negative breast cancer patients, VEGF is a strong independent predictor of relapse-free (17, 18) and overall (18, 19) survival.

Clearly, measurement of circulating soluble markers of angiogenesis would be of considerable benefit over more subjective approaches such as immunohistochemical assessments, or immunoassays which involve laborious tissue extraction procedures. Since the original finding of elevated VEGFS concentrations of in patients with cancer (20), many studies have reported similar findings in patients with breast cancer and many other types of cancer, with higher levels often found in metastatic disease than in localized disease or in progressive disease during treatment(21, 22, 23, 24). Correlations with prognosis have also been reported for several cancers, e.g., ovarian cancer(25, 26, 27). However, the interpretation of such studies has been complicated by the demonstration that much of the VEGF measured in serum samples is released from platelets upon activation after venipuncture (28, 29, 30), with mRNA for VEGF being found in all blood cell types examined, including megakaryocytes, platelets,lymphocytes, CD34+ cells, granulocytes, and monocytes (31, 32, 33, 34), and VEGF protein being shown to be released by neutrophils (35) and platelets (28, 29, 30, 33, 34) after activation in vitro. To measure true basal circulating levels of VEGF, rapidly processed citrated plasma samples are the only suitable material (28); they may,therefore, better reflect any circulating VEGF released by the tumor. However, this has yet to be investigated, and the biological significance of platelet- or leukocyte-derived VEGF has yet to be clarified.

This study examined for the first time both VEGFPand VEGFS in blood samples from pre- and postmenopausal healthy controls and patients with benign breast disease, localized breast cancer, breast cancer in remission, or metastatic breast cancer to establish the clinical validity of the different measurements. These results were related to the immunohistochemical expression of VEGF by the tumor, as determined by a new antibody specifically validated for use with paraffin-embedded material (36), and to the other established clinicopathological prognostic markers, namely MVD, tumor size and grade, lymph node status, and ER and PR status. In particular, the impact of therapy on circulating VEGF concentrations was investigated.

Patients.

The study included 138 patients with benign breast disease, localized breast cancer (prior to surgery), metastatic breast cancer, or in remission from breast cancer, together with 63 healthy controls (Table 1). Six patients with localized disease, 30 in remission, and 7 with metastatic disease were receiving tamoxifen at the time of sampling. Other hormone-based medication included arimidex (one with localized disease, one in remission, and two with metastatic disease), Megace(one metastatic), and Danazol (one with localized disease). Two patients with localized disease, eight in remission, and five with metastatic disease were receiving chemotherapy. Nine of the 62 patients with localized disease had DCIS alone; the clinicopathological details of the remaining 53 patients in this group shown in Table 2. Patients with metastatic disease included 1 at presentation, 16 at relapse, and 5 during treatment for relapse. The duration of remission for the patients in the remission group ranged from <1 to 123 months. In seven healthy females, changes in circulating VEGF concentrations during the menstrual cycle were examined with blood samples taken on a minimum of four occasions during one cycle and concomitant samples being assayed for circulating estradiol, progesterone, luteinizing hormone, and follicle-stimulating hormone. The study had been approved by the local ethics committee, and informed consent was obtained.

Blood Sampling.

Venous blood was collected in tubes containing either trisodium citrate[final concentration, 0.31% (w/v)] or no anticoagulant. Within 30 min, samples were centrifuged (2000 × g, 10 min), and the plasma and serum were removed, aliquoted, and stored at−80°C until analysis. Where possible, parallel samples were used for routine hematological counts in a Technicon H2 system. Serum samples from patients in remission were analyzed for circulating estradiol concentrations, using the Bayer Immuno 1 system.

Measurement of Circulating VEGF.

Plasma and serum samples were analyzed for VEGF using a commercially available sandwich ELISA obtained from R & D Systems Europe(Abingdon, United Kingdom). The sensitivity of the assay was 9.0 pg/ml as quoted by the manufacturer. All samples were assayed in duplicate. A preliminary evaluation of the assay included the examination of recovery of recombinant VEGF added to samples, the parallelism of diluted samples, and the inter- and intra-assay CVs.

Both serum and citrated plasma samples were assayed for VEGF with the concentration in the citrated plasma sample being corrected for the dilution effect of the anticoagulant present as a liquid form. Because our previous study had indicated that the majority of the difference in VEGF concentrations between serum and plasma samples was accounted for by release of VEGF from platelets during the clotting process(28), a theoretical platelet-derived VEGF was calculated and expressed per platelet number, taking into account the hematocrit,as follows:

\[\mathrm{VEGF\ (pg)/10}^{\mathrm{6}}\mathrm{\ platelets\ {=}\ }\ \frac{\mathrm{(VEGF}_{\mathrm{S}}\mathrm{\ {-}\ VEGF}_{\mathrm{P}}\mathrm{)\ {\times}\ (1\ {-}\ PCV)}}{\mathrm{platelet\ number}}\]

where PCV is the packed cell volume, and the units for VEGF measurement were in pg/ml of serum or plasma and the platelet number was × 109/liter of blood. These data were calculated for only 49 of the local disease group, 23 of the remission group, 10 of the metastatic group, 9 of the benign group, and 20 of the control group because of a lack of hematological data for many samples.

Histopathology and Immunohistochemistry.

Classification of tumors was according to the International Union against Cancer TNM system (37), with grade determined by the modified Bloom-Richardson criteria according to Elston and Ellis(38). Immunohistochemical staining for CD31, ER, PR, and VEGF was performed using a standard streptavidin-biotin-peroxidase detection system with the following primary antibodies (tissue culture supernatants): for ER, clone 1D5 (DAKO, Cambridge, United Kingdom)diluted 1:50; for PR, clone PgR 636 (DAKO) diluted 1:50; for CD31,clone JC/70A (DAKO) diluted 1:30; and for VEGF, clone VG1 (36, 39) used undiluted. Tissue sections for ER, PR, and VEGF immunolabeling were subject to antigen retrieval by immersion in boiling citrate buffer (pH 6.0) in a microwave three times for 5 min each, twice for 5 min each, and twice for 5 min each, respectively,before being either left to cool in the buffer for 20 min (ER and PR)or cooled on ice for 20 min (VEGF). For CD31 immunolabeling, antigen retrieval was carried out by immersing sections in boiling citrate buffer (pH 6.0) for 1 min in a domestic pressure cooker before cooling in water. A positive control was included in all runs, and for each case omission of the primary antibody was used as a negative control.

At least 10% of tumor cell nuclei were required to be positive for the tumor to be accepted as ER- or PR-positive, with cytoplasmic staining being disregarded. Staining intensity was variable, and positive staining of adjacent normal parenchymal elements was used as an internal control. VEGF staining was assessed semiquantitatively on a scale from 0 (negative) to 4 (strongly positive). IMD in invasive tumors was assessed after immunostaining for CD31 according to the recent international consensus (40), using a 25-point Chalkley graticule at ×200 magnification (field area, 0.67 mm2). The three most vascular fields, or“hotspots,” were initially identified at low power, and the maximum number of points on the graticule was counted for each area. Areas of sclerosis or dense inflammation were avoided. The mean Chalkley counts for the three areas were used in the statistical analysis. For cases of DCIS, “stromal” vascularity was assessed using a Chalkley graticule, and “periductal” vascularity was assessed by a semiquantitative, subjective grading system (40).

Statistical Analysis.

All statistical analyses were performed using SPSS for Windows (Version 8.0; SPSS, Chicago). The primary analysis was the comparison of the five patient groups and the circulating VEGF variables. The Kruskal-Wallis test was used to compare the five groups overall, and for two-group comparisons we used the Mann-Whitney test. (Corrections for multiple significance tests were made using the Bonferroni correction.) For subgroup analyses and to investigate the relationships between the circulating VEGF variables and other variables, Spearman’s correlation, χ2,Kruskal-Wallis, ANOVA, Mann-Whitney, and Student’s t tests were used as appropriate. P < 0.05 was considered statistically significant.

A limited evaluation of the VEGF immunoassay found VEGF recoveries of 85.1% (SD, 9.7%) for samples from patients with breast cancer(n = 10) compared with 86.3% (SD, 13.0%)for samples from healthy controls (n = 10) to which 130 pg/ml VEGF had been added. In general, intra-assay CVs were <10%, although they increased to 20% or more at lower concentrations of VEGF. Similarly, the interassay CVs were variable depending on the sample VEGF concentration, with values of 28% for a 18.2 pg/ml sample (n = 15) and 4.4% for a 186.0 pg/ml sample (n = 10).

Significantly higher VEGF concentrations were found in the serum samples (VEGFS) compared with the corresponding citrated plasma samples (VEGFP; P < 0.0001). When we used the 95th percentile as the cutoff points, the upper limits of normal values for VEGF in the female control group were 62.5 pg/ml plasma for plasma,590.2 pg/ml for serum, 546.9 pg/ml for the difference between serum and plasma (VEGFdiff), and 1.20 pg/106 platelets; there were no significant differences when compared with a male control group, but the numbers were small (n = 14, data not shown). There were significant differences in age between the groups(P < 0.001), but no significant effects of age or menopausal status were found on VEGF parameters. Analysis of the menstrual cycle data for seven healthy controls showed no definite trend for VEGFS, although within an individual cycle, variation was apparent with most values falling within 25% of the mean cycle value. No cyclical variation was detected in VEGFP because of the low and often undetectable circulating concentrations present.

The circulating VEGF concentrations for each study group are shown in Fig. 1. Significant differences between groups were seen in VEGFP (P < 0.001) and VEGFS (P = 0.048), but not in VEGFdiff or platelet VEGF concentrations. Further analysis revealed that VEGFPconcentrations of patients with localized disease were significantly elevated compared with normal controls (P = 0.016); patients with metastatic disease had higher levels than normal controls (P < 0.001), patients with benign disease (P = 0.009), and patients with localized disease (P = 0.004). However, the highest concentrations were seen in patients in remission compared with normal controls (P < 0.001), patients with benign disease (P = 0.01), or patients with localized disease (P = 0.005). VEGFS concentrations were significantly higher in patients in remission or those with metastatic disease compared with normal controls (P = 0.008 and 0.044,respectively). VEGFdiff was significantly higher only in the remission group compared with the control group(P = 0.023).

When the data from all of the groups were considered,VEGFP correlated strongly with VEGFS (ρ = 0.682; P < 0.001), VEGFdiff(ρ = 0.601; P < 0.001), and platelet VEGF concentration (ρ = 0.530; P < 0.001). Not surprisingly given the derivation and the relative proportions, VEGFScorrelated very strongly with VEGFdiff(ρ = 0.992; P < 0.001), but in addition, it correlated with platelet VEGF concentration(ρ = 0.923; P < 0.001). For the hematological variables, a significant but weak correlation was found between platelet count and VEGFS(ρ = 0.172; P = 0.05) and VEGFdiff (ρ = 0.204; P = 0.021), and a negative correlation was found between VEGFP and hemoglobin (ρ = −0.227; P = 0.01). When the groups were examined individually, correlations similar to those above were found for the VEGF parameters, but no significant correlations were seen between the VEGF parameters and hematological variables, with the exception of a significant correlation between white cell count and VEGFP in patients with metastatic disease(P < 0.047; ρ = 0.489).

Further analysis of the group of patients in remission revealed a significant effect of tamoxifen therapy on measured VEGF (Fig. 2). Significantly higher VEGFP levels were seen in those patients in remission receiving tamoxifen(P = 0.004), as were VEGFS, VEGFdiff, and platelet VEGF concentrations (P = 0.01,0.016, and 0.012, respectively). No significant differences were seen in estradiol levels when subdivided on the basis of tamoxifen therapy. Multivariate analysis failed to reveal any further influencing variables. The median time since diagnosis tended to be greater in patients not receiving tamoxifen (60 months; range, <1–85 months)compared with those receiving tamoxifen (25 months; range, 2–108 months), but the difference did not reach statistical significance. Although the numbers of patients were small (14 non-tamoxifen versus 7 tamoxifen), similar significant effects for VEGFP (P = 0.004) were seen in the group of metastatic patients, but the effects on other measured VEGF parameters failed to reach statistical significance(P = 0.064, 0.172, and 0.200, respectively). The plasma and serum VEGF concentrations of patients in remission or metastatic groups not receiving tamoxifen were not significantly different from the normal female control group or the patients with localized breast cancer.

The histological types, grades, and other clinicopathological variables examined together with the immunolabeling results in the group of patients with localized invasive breast cancer are summarized in Table 2. Grade and immunohistological variables were unavailable for two patients because neoadjuvant therapy made assessment unreliable. The majority of patients had ductal carcinoma of no special type. CD31 immunolabeling gave a highly specific and intense staining of vascular endothelium both within and outside the areas of tumor. Stromal cells were negative, and epithelial cytoplasmic positivity was seen in only one case. Areas of intense inflammation were strongly CD31-positive and were avoided when selecting areas for assessment. VEGF immunohistochemical staining was heterogeneous across different tumors and within tumors, with areas of very intense and very pale staining. There was staining of normal breast tissue, especially of apocrine epithelium, but no stromal staining was seen. Labeling of intratumoral macrophages was also apparent in one case. Two patterns of increased vascularity were observed in the cases of pure DCIS: stromal and periductal. All cases demonstrated the former, whereas the typical “necklace” distribution of the latter was observed in only one case, which was an unusual case of apocrine DCIS exhibiting the most intense VEGF positivity of any sample.

In patients with localized invasive cancer, no significant relationships were found between the circulating VEGF variables and the recognized immunohistochemical or immunopathological prognostic indicators examined, with the exception of a significant correlation between VEGFdiff and ER status (ρ = 0.325; P = 0.023). Interestingly, an association between platelets and grade was found (ρ = −0.359; P = 0.017) and between VEGF immunostaining and white cell count (ρ = 0.458; P = 0.006), with that between VEGF immunostaining and platelet number just failing to reach statistical significance (ρ=0.331; P = 0.052). The relationships between VEGF immunostaining or MVD and other recognized clinicopathological variables are shown in Table 3.

Of the six DCIS tumors for whom vascularity data were available, five had stromal vascular patterns of MVD staining, with one having both this and the necklaces of periductal vessels discussed above. The three most vascular (Chalkey assessment and semiquantitative) were all ER-and PR-negative, whereas the remaining three were positive. Similarly,the VEGF staining was more pronounced in the three with the highest MVD assessment.

In breast cancer, IMD is now established as one of the standard prognostic factors for predicting metastasis and relapse-free or overall survival (4, 5). The assessment of angiogenesis is also of potential relevance in identifying those who may benefit from antiangiogenic therapies. IMD is assessed primarily by quantification of MVD, and the techniques are laborious, require experience, and tend to observer variability (40). The measurement of circulating concentrations of specific angiogenic factors such as VEGF may provide less subjective measurements.

Initially interpreted as reflecting an “overspill” of tumor-derived VEGF into the systemic circulation, VEGFS is now known to be largely platelet-derived (28, 29, 30). A more accurate basal circulating level, which should reflect normal physiological angiogenesis and any disease-related overspill, should therefore be determined with minimal platelet activation, using citrated plasma (28). No study has as yet made these comparisons directly, and we report for the first time that VEGFP is a more sensitive measurement than VEGFS in terms of being significantly higher in patients with localized disease compared with controls or patients with benign disease, unlike VEGFS. However, it is also apparent that this measurement would not be an accurate discriminator between the groups because of considerable overlap. The lack of a significant difference in VEGFS between patients with localized breast cancer preoperatively and normal healthy controls confirms previous studies (22, 24, 41). Only two previous studies measured VEGFP in cancer, with concentrations being inversely related to survival and response to chemotherapy in gastrointestinal cancer (42) and with significant differences in patients with prostatic cancer depending on disease state, with a sensitivity of 81% and specificity of 71% for metastatic disease (43). Because the CVs of immunoassays at the low VEGF concentrations more often found in plasma samples are relatively high, interpretations of changes or differences should take into account their relative magnitude. Although decreases in VEGFS postoperatively (22, 23) have been described, the extent to which these measurements relate to tumor-derived VEGF is unclear. An alternative explanation may be the contribution of other tumor-derived cytokines, such as interleukin-6(44), which is known to induce VEGF expression(45); has been shown to correlate with platelet number,VEGFS, and platelet VEGF content(46); and thus may conceivably play a role in acting systemically to regulate platelet VEGF content. Changes in VEGFS during chemotherapy coincide with chemotherapy-induced thrombocytopenia and a subsequent rebound increase in platelet number rather than the sustained decrease in VEGFS that would be expected if it was tumor-derived (30). Similarly, the finding of lower VEGFS in breast cancer patients with disseminated cancer undergoing therapy compared with those untreated (21, 24) may reflect chemotherapy-induced effects on platelets,effects which may not only be on platelet number but also on platelet volume (47).

Several studies have now shown that VEGFScorrelates with platelet number (30, 46, 48), although significant interindividual variation exists for estimates of platelet VEGF content (28, 46, 49). The platelet VEGF concentration has been reported to be significantly higher in patients with cancer compared with normals, with medians of 1.6 pg/106platelets (range, 0.3–2.6 pg/106 platelets) and 0.5 pg/106 platelets (0.1–1.6 pg/106 platelets), respectively(49), but this was not confirmed in the present study, in which we found corresponding medians of 0.47 pg/106 platelets (range, 0.13–1.86 pg/106 platelets) in invasive breast cancer and 0.54 pg/106 platelets (range, 0.08–1.21 pg/106 platelets) in controls. The variability in estimates of platelet-derived VEGF may reflect the use of isolated platelets rather than theoretical derivations from plasma and serum levels, sample sizes, and possible confounding factors such as therapy or the omission of a correction for hematocrit in derived concentrations. For peripheral blood mononuclear cells, median VEGF concentrations of 10.6 pg/106 cells in cancer patients and 0.9 pg/106 cells in normals have been reported (49), values that clearly make their potential contribution to serum VEGF much smaller than that of platelets given the differential in numbers, but with more marked differences between diseased and normal states. The role of platelets in tumor biology is not clear, but cancer patients frequently have higher platelet counts and an increased platelet consumption compared with healthy individuals; in addition, thrombocytosis is a negative prognostic factor in some cancers, and antitumor effects and improvements in survival have been seen with anticoagulation therapy(50).

In the all-groups analysis but not in the subgroups,VEGFP was negatively correlated with hemoglobin. A significant negative correlation between VEGFSand hemoglobin has been reported previously in patients with a variety of localized cancers (51). Such correlations may reflect the disease state, with low hemoglobin often reflecting more advanced disease; alternatively, they may be related to hypoxia-induced VEGF expression, with the decrease in hemoglobin reflecting the hypoxia.

There is no consensus regarding VEGF concentrations during the menstrual cycle; studies have found variable correlations with hormones and have variably reported higher levels of VEGFSin the follicular phase (52), the luteal phase(53), the early proliferative (follicular) and secretory(luteal) phases (54), or no consistent changes (23, 41). Interestingly, the patient in our study with the highest circulating VEGF concentrations was receiving Danazol at the time of sampling. Significantly higher levels have been reported in postmenopausal compared with premenopausal women (41, 52), which we could not confirm, possibly because of sample sizes.

The inference, after subgroup analysis, that tamoxifen produces significant elevations in circulating VEGF in some patients may have prognostic implications. This was also reflected by an increase in the amount of theoretical platelet-derived VEGF, which indicates that tamoxifen is likely to be acting systemically in terms of its actions on VEGF production. To our knowledge, no previous studies have demonstrated this effect. The antitumor activity of tamoxifen is largely attributed to antiestrogenic properties, although it has mixed agonist/antagonist activity depending on the tissue. However, the effects of tamoxifen are not entirely explained by its action on the ER: some receptor-positive patients fail to respond. Other known potentially therapeutic actions include regulation of circulating concentrations of the breast epithelial mitogen insulin-like growth factor-I and insulin-like growth factor-binding protein I(55, 56). Hormonal regulation of VEGF has been demonstrated, with progesterone and progestins such as megestrol acetate stimulating VEGF production by some breast cancer cell lines in vitro(57) and estradiol being shown to stimulate VEGF production in several systems, including endometrial cultures and the uterus in vivo(58, 59, 60, 61). Tamoxifen itself has been shown to induce VEGF expression in the breast cancer cell line MCF7 by a mechanism that is not thought to be mediated by the ER (62).

The induction of VEGF by tamoxifen is paradoxical but may be implicated in tamoxifen-associated endometrial hyperplasia, with the endometrium being a possible source of some of the increase in VEGFP seen in this study. Tamoxifen has been shown to increase VEGF expression in uterine tissue (63)and, more recently, to induce production of the novel angiogenic factor adrenomedullin by endometrial macrophages via a nonestrogenic mechanism(64). In addition, tamoxifen exerts a direct antiproliferative activity on VEGF-stimulated endothelial cell growth,an effect not mediated by the ER (65). Angiogenesis has been reported to add information to ER status in predicting response to tamoxifen, with the patients having the best prognosis being characterized by low vascularity, ER positivity, few nodes involved, and small low-grade tumors (6). The determination of the magnitude of the VEGF increase while on tamoxifen may add further prognostic information both with regard to breast cancer and the incidence of endometrial hyperplasia (66).

In the present study, 80% of tumors were found to be VEGF-positive,which is in agreement with previous reports. However no significant correlation between VEGF staining intensity and Chalkley count was seen, unlike in earlier reports (10, 14, 15, 16), although one study found this relationship to be true only for invasive ductal carcinomas or pure DCIS, with VEGF being lower and negatively correlated with angiogenesis in invasive lobular carcinoma(10). Our findings may reflect the relatively small sample size, the relatively uniform patient populations examined in many earlier studies, antibody specificity, the predominant use of tissue homogenates for assessment of intratumoral VEGF in previous studies, or variability in the techniques used to assess immunohistochemical staining or angiogenesis. This may also reflect differing relative dependencies on other specific angiogenic factors such as thymidine phosphorylase.

IMD has also been reported to correlate with menopausal status, age,and tumor grade (67), although this is unconfirmed(6). In the present study, a significant association with tumor size was found that may be expected given the dependence of tumor expansion on angiogenesis, confirming previous reports (67, 68). The association of intratumoral VEGF with other clinical prognostic factors is unresolved, with no correlation being found with age, tumor type or size, or hormone receptor status in several studies(14, 15, 18); however, correlations with tumor size(17, 19) and grade (19) and an inverse correlation with ER in node-negative breast cancer patients have also been reported (19). In contrast, we found intratumoral VEGF to be correlated with ER status, which may reflect an estrogenic regulation of VEGF in some patients. Similarly,VEGFdiff was correlated with ER status. Unexpectedly, VEGF was inversely correlated with tumor grade. The prognostic use of intratumoral VEGF assessment is still not clear(15, 16), but in node-negative breast cancer, several studies have found VEGF to be a strong independent predictor of relapse-free (17, 18) and overall (18, 19)survival. This is being examined prospectively in our study cohort at present. No association between circulating levels of VEGF and intratumoral VEGF assessed immunohistochemically was seen, confirming previous results (69), although an additional study has reported that the serum-positive rate is associated with MVD and intratumoral VEGF expression when determined by immunoassay(23).

This study illustrates that VEGFP shows better discrimination than VEGFS in patients with breast cancer and highlights the tamoxifen-associated increase in VEGF concentrations occurring in some patients. The significance of this in relation to disease course or subsequent endometrial pathology,together with the relevance of platelet-tumor cell interactions in tumor angiogenesis and progression, warrants further prospective studies.

We thank Lynne Hill, Christine Gascoigne, Julie Booth, Sue Buckle, and Amanda Brazier for assistance.

Fig. 1.

VEGFP (a), VEGFS(b), VEGFS minus VEGFP(c), and theoretical platelet VEGF concentration(d) in normal healthy controls(Controls), patients with benign breast disease(Benign), or breast cancer patients with localized breast cancer (Local disease), in remission(Remission), or with metastatic disease(Metastatic). The dashed line indicates the limit of detection of the immunoassay, the dotted line indicates the upper limit of normal. The bars and numbers illustrate the median for each group.

Fig. 1.

VEGFP (a), VEGFS(b), VEGFS minus VEGFP(c), and theoretical platelet VEGF concentration(d) in normal healthy controls(Controls), patients with benign breast disease(Benign), or breast cancer patients with localized breast cancer (Local disease), in remission(Remission), or with metastatic disease(Metastatic). The dashed line indicates the limit of detection of the immunoassay, the dotted line indicates the upper limit of normal. The bars and numbers illustrate the median for each group.

Close modal
Fig. 2.

VEGFP (a), VEGFS(b), VEGFS minus VEGFP(c), and theoretical platelet VEGF concentration(d) in patients with breast cancer in remission,subdivided on the basis of whether they were receiving tamoxifen at the time of sampling. Arrows indicate those patients also receiving adjuvant chemotherapy. The dashed lineindicates the limit of detection of the immunoassay, and the dotted line indicates the upper limit of normal. The bars and numbers illustrate the median for each group. plts, platelets.

Fig. 2.

VEGFP (a), VEGFS(b), VEGFS minus VEGFP(c), and theoretical platelet VEGF concentration(d) in patients with breast cancer in remission,subdivided on the basis of whether they were receiving tamoxifen at the time of sampling. Arrows indicate those patients also receiving adjuvant chemotherapy. The dashed lineindicates the limit of detection of the immunoassay, and the dotted line indicates the upper limit of normal. The bars and numbers illustrate the median for each group. plts, platelets.

Close modal

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.

1

Financial support provided by the Imperial Cancer Research Fund and the Liz Dawn St James’s Hospital Breast Cancer Appeal.

3

The abbreviations used are: VEGF, vascular endothelial growth factor; IMD, intratumoral microvessel density; ER,estrogen receptor; DCIS, ductal carcinoma in situ;VEGFS and VEGFP, serum and plasma VEGF; MVD,microvessel density; PR, progesterone receptor; CV, coefficient of variation; VEGFdiff, serum minus plasma VEGF.

Table 1

Characteristics of the 201 patients and healthy controls included in the study

CategorynMedian age (range), yearsMenopausal status
PeriPrePost
Benign breast disease 12 47 (32–63) 
Localized breast cancer (presurgery) 62 56 (29–85) 14 44 
Remission 42 54 (38–85) 32 
Metastatic disease 22 52 (32–82) 15 
Healthy female controls 63 37 (20–72) 49 14 
CategorynMedian age (range), yearsMenopausal status
PeriPrePost
Benign breast disease 12 47 (32–63) 
Localized breast cancer (presurgery) 62 56 (29–85) 14 44 
Remission 42 54 (38–85) 32 
Metastatic disease 22 52 (32–82) 15 
Healthy female controls 63 37 (20–72) 49 14 
Table 2

Tumor characteristics of invasive breast cancer cases (n = 53) in the group of patients examined with locoregional disease prior to surgery

Nine additional patients with pure DCIS were included in the study.

Chalkley scoreVEGF immunohistochemistry
MeanRangeNDa0123ND
Histology         
Ductal (n = 47) 5.8 2.6–8.6 10 10 16 11 
Lobular (n = 2) 5.5 5.0–6.0 
Tubular (n = 2) 5.3 4.6–6.0 
Mucinous (n = 1) 8.3  
Medullary (n = 1) 11.3  
Grade         
I (n = 9) 5.8 3.3–8.3 
II (n = 22) 5.7 2.6–8.6 
III (n = 19) 6.3 3.0–11.3 11 
Unknown (n = 3)   
Tumor size (UICC/TNM)         
pT1 (n = 32) 5.5 2.6–8.6 
pT2 (n = 17) 6.8 3.6–11.3 
pT3–4 (n = 2) 7.0 7.0 
Unknown (n = 2)   
Node status         
Negative (n = 32) 6.2 3.3–11.3 
Positive (n = 21) 5.5 2.6–8.6 
ER status         
Negative (n = 17) 6.4 3.0–11.3 
Positive (n = 33) 5.8 2.6–8.6 14 
ND (n = 3) 6.0 6.0 
PR status         
Negative (n = 29) 6.2 3.0–11.3 
Positive (n = 21) 5.6 2.6–8.6 
ND (n = 3) 6.0 6.0 
Chalkley scoreVEGF immunohistochemistry
MeanRangeNDa0123ND
Histology         
Ductal (n = 47) 5.8 2.6–8.6 10 10 16 11 
Lobular (n = 2) 5.5 5.0–6.0 
Tubular (n = 2) 5.3 4.6–6.0 
Mucinous (n = 1) 8.3  
Medullary (n = 1) 11.3  
Grade         
I (n = 9) 5.8 3.3–8.3 
II (n = 22) 5.7 2.6–8.6 
III (n = 19) 6.3 3.0–11.3 11 
Unknown (n = 3)   
Tumor size (UICC/TNM)         
pT1 (n = 32) 5.5 2.6–8.6 
pT2 (n = 17) 6.8 3.6–11.3 
pT3–4 (n = 2) 7.0 7.0 
Unknown (n = 2)   
Node status         
Negative (n = 32) 6.2 3.3–11.3 
Positive (n = 21) 5.5 2.6–8.6 
ER status         
Negative (n = 17) 6.4 3.0–11.3 
Positive (n = 33) 5.8 2.6–8.6 14 
ND (n = 3) 6.0 6.0 
PR status         
Negative (n = 29) 6.2 3.0–11.3 
Positive (n = 21) 5.6 2.6–8.6 
ND (n = 3) 6.0 6.0 
a

ND, not done; UICC, International Union against Cancer.

Table 3

Correlations between clinicopathological variables and VEGF immunostaining or microvessel density (Chalkley count) in patients with invasive breast cancer with locoregional disease prior to surgery

GradeStageTaNERPRChalkley countVEGF
Grade         
Stage 0.326        
 P = 0.021        
 (50)        
0.452 0.733       
 P = 0.001 P < 0.001       
 (50) (51)       
NS 0.716 NS      
  P < 0.001       
  (52)       
ER −0.664 −0.314 −0.324 NS     
 P < 0.001 P = 0.028 P = 0.025      
 (48) (49) (48)      
PR −0.484 NS NS NS 0.611    
 P < 0.001    P < 0.001    
 (48)    (50)    
Chalkley count NS NS 0.339 NS NS NS   
   P = 0.026      
   (43)      
VEGF −0.597 NS NS NS 0.366 NS NS  
 P < 0.001    P = 0.019    
 (41)    (41)    
GradeStageTaNERPRChalkley countVEGF
Grade         
Stage 0.326        
 P = 0.021        
 (50)        
0.452 0.733       
 P = 0.001 P < 0.001       
 (50) (51)       
NS 0.716 NS      
  P < 0.001       
  (52)       
ER −0.664 −0.314 −0.324 NS     
 P < 0.001 P = 0.028 P = 0.025      
 (48) (49) (48)      
PR −0.484 NS NS NS 0.611    
 P < 0.001    P < 0.001    
 (48)    (50)    
Chalkley count NS NS 0.339 NS NS NS   
   P = 0.026      
   (43)      
VEGF −0.597 NS NS NS 0.366 NS NS  
 P < 0.001    P = 0.019    
 (41)    (41)    
a

T and N, tumor size and nodal status, respectively,according to the TNM classification system; NS, not significant.

1
Folkman J. The role of angiogenesis in tumor growth.
Semin. Cancer Biol.
,
3
:
65
-71,  
1992
.
2
Ferrara N. Vascular endothelial growth factor.
Eur. J. Cancer
,
32A
:
2413
-2422,  
1996
.
3
Weidner N., Semple J. P., Welch W. R., Folkman J. Tumor angiogenesis and metastasis—correlation in invasive breast carcinoma.
N. Engl. J. Med.
,
324
:
1
-8,  
1991
.
4
Gasparini G., Harris A. L. Clinical importance of the determination of tumor angiogenesis in breast carcinoma: much more than a new prognostic tool.
J. Clin. Oncol.
,
13
:
765
-782,  
1995
.
5
Heimann R., Ferguson D., Powers C., Recant W. M., Weichselbaum R. R., Hellman S. Angiogenesis as a predictor of long-term survival for patients with node-negative breast cancer.
J. Natl. Cancer Inst.
,
88
:
1764
-1769,  
1996
.
6
Gasparini G., Fox S. B., Verderio P., Bonoldi E., Bevilacqua P., Boracchi P., Dante S., Marubini E., Harris A. L. Determination of angiogenesis adds information to estrogen receptor status in predicting the efficacy of adjuvant tamoxifen in node-positive breast cancer patients.
Clin. Cancer Res.
,
2
:
1191
-1198,  
1996
.
7
Yoshiji H., Gomez D. E., Shibuya M., Thorgeirsson U. P. Expression of vascular endothelial growth factor, its receptor, and other angiogenic factors in human breast cancer.
Cancer Res.
,
56
:
2013
-2016,  
1996
.
8
Anan K., Morisaki T., Katano M., Ikubo A., Kitsuki H., Uchiyama A., Kuroki S., Tanaka M., Torisu M. Vascular endothelial growth factor and platelet-derived growth factor are potential angiogenic and metastatic factors in human breast cancer.
Surgery
,
119
:
333
-339,  
1996
.
9
Brown L. F., Berse B., Jackman R. W., Tognazzi K., Guidi A. J., Dvorak H. F., Senger D. R., Connolly J. L., Schnitt S. J. Expression of vascular permeability factor (vascular endothelial growth factor) and its receptors in breast cancer.
Hum. Pathol.
,
26
:
86
-91,  
1995
.
10
Lee A. H. S., Dublin E. A., Bobrow L. G., Poulsom R. Invasive lobular and invasive ductal carcinoma of the breast show distinct patterns of vascular endothelial growth factor expression and angiogenesis.
J. Pathol.
,
185
:
394
-401,  
1998
.
11
Anan K., Morisaki T., Katano M., Ikubo A., Tsukahara Y., Uchiyama A., Kuroki S., Tanaka M., Torisu M. Preoperative assessment of tumor angiogenesis by vascular endothelial growth factor mRNA expression in homogenate samples of breast carcinoma: fine-needle aspirates vs. resection samples.
J. Surg. Oncol.
,
66
:
257
-263,  
1997
.
12
Anan K., Morisaki T., Katano M., Ikubo M., Tsukahara Y., Kojima M., Uchiyama A., Kuroki S., Torisu M., Tanaka M. Assessment of c-erbB2 and vascular endothelial growth factor mRNA expression in fine-needle aspirates from early breast carcinomas: pre-operative determination of malignant potential.
Eur. J. Surg. Oncol.
,
24
:
28
-33,  
1998
.
13
Guidi A. J., Schnitt S. J., Fischer L., Tognazzi K., Harris J. R., Dvorak H. F., Brown L. F. Vascular permeability factor (vascular endothelial growth factor) expression and angiogenesis in patients with ductal carcinoma in situ of the breast.
Cancer (Phila.)
,
80
:
1945
-1953,  
1997
.
14
Toi M., Kondo S., Suzuki H., Yamamoto Y., Inada K., Imazawa T., Taniguchi T., Tominaga T. Quantitative analysis of vascular endothelial growth factor in primary breast cancer.
Cancer (Phila.)
,
77
:
1101
-1106,  
1996
.
15
Obermair A., Kucera E., Mayerhofer K., Speiser P., Seifert M., Czerwenka K., Kaider A., Leodolter S., Kainz C., Zeillinger R. Vascular endothelial growth factor (VEGF) in human breast cancer: correlation with disease-free survival.
Int. J. Cancer
,
74
:
455
-458,  
1997
.
16
Toi M., Hoshina S., Takayanagi T., Tominaga T. Association of vascular endothelial growth factor expression with tumor angiogenesis and with early relapse in primary breast cancer.
Jpn. J. Cancer Res.
,
85
:
1045
-1049,  
1994
.
17
Eppenberger U., Kueng W., Schlaeppi J-M., Roesel J. L., Benz C., Mueller H., Matter A., Zuber M., Luescher K., Litschgi M., Schmitt M., Foekens J. A., Eppenberger-Castori S. Markers of tumor angiogenesis and proteolysis independently define high- and low-risk subsets of node-negative breast cancer patients.
J. Clin. Oncol.
,
16
:
3129
-3136,  
1998
.
18
Gasparini G., Toi M., Gion M., Verderio P., Dittadi R., Hanatani M., Matsubara I., Vinante O., Bonoldi E., Boracchi P., Gatti C., Suzuki H., Tominaga T. Prognostic significance of vascular endothelial growth factor protein in node-negative breast carcinoma.
J. Natl. Cancer Inst.
,
89
:
139
-147,  
1997
.
19
Linderholm B., Tavelin B., Grankvist K., Henriksson R. Vascular endothelial growth factor is of high prognostic value in node-negative breast carcinoma.
J. Clin. Oncol.
,
16
:
3121
-3128,  
1998
.
20
Kondo S., Asano M., Matsuo K., Ohmori I., Suzuki H. Vascular endothelial growth factor/vascular permeability factor is detectable in the sera of tumor-bearing mice and cancer patients.
Biochim. Biophys. Acta
,
1221
:
211
-214,  
1994
.
21
Salven P., Maenpaa H., Orpana A., Alitalo K., Joensuu H. Serum vascular endothelial growth factor is often elevated in disseminated cancer.
Clin. Cancer Res.
,
3
:
647
-651,  
1997
.
22
Kraft A., Weindel K., Ochs A., Marth C., Zmija J., Schumacher P., Unger C., Marme D., Gastl G. Vascular endothelial growth factor in the sera and effusions of patients with malignant and nonmalignant disease.
Cancer (Phila.)
,
85
:
178
-187,  
1999
.
23
Yamamoto Y., Toi M., Kondo S., Matsumoto T., Suzuki H., Kitamura M., Tsuruta K., Taniguchi T., Okamoto A., Mori T., Yoshida M., Ikeda T., Tominaga T. Concentrations of vascular endothelial growth factor in the sera of normal controls and cancer patients.
Clin. Cancer Res.
,
2
:
821
-826,  
1996
.
24
Salven P., Perhoniemi V., Tykka H., Maenpaa H., Joensuu H. Serum VEGF levels in women with a benign breast tumor or breast cancer.
Breast Cancer Res. Treat.
,
53
:
161
-166,  
1999
.
25
Tempfer C., Obermair A., Hefler L., Haeusler G., Gitsch G., Kainz C. Vascular endothelial growth factor serum concentrations in ovarian cancer.
Obstet. Gynecol.
,
92
:
360
-363,  
1998
.
26
Yamamoto S., Konishi I., Mandai M., Kuroda H., Komatsu T., Nanbu K., Sakahara H., Mori T. Expression of vascular endothelial growth factor (VEGF) in epithelial ovarian neoplasms: correlation with clinicopathology and patient survival, and analysis of serum VEGF levels.
Br. J. Cancer
,
76
:
1221
-1227,  
1997
.
27
Chen C. A., Cheng W. F., Lee C. N., Chen T. M., Kung C. C., Hsieh F. J., Hsieh C. Y. Serum vascular endothelial growth factor in epithelial ovarian neoplasms: correlation with patient survival.
Gynecol. Oncol.
,
74
:
235
-240,  
1999
.
28
Banks R. E., Forbes M. A., Kinsey S. E., Stanley A., Ingham E., Walters C., Selby P. J. Release of the angiogenic cytokine vascular endothelial growth factor (VEGF) from platelets: significance for VEGF measurements and cancer biology.
Br. J. Cancer
,
77
:
956
-964,  
1998
.
29
Webb N. J. A., Bottomley M. J., Watson C. J., Brenchley P. E. C. Vascular endothelial growth factor (VEGF) is released from platelets during blood clotting: implications for measurement of circulating VEGF levels in clinical disease.
Clin. Sci.
,
94
:
395
-404,  
1998
.
30
Verheul H. M. W., Hoekman K., Luykx-de Bakker S., Eekman C. A., Folman C. C., Broxterman H. J., Pinedo H. M. Platelet: transporter of vascular endothelial growth factor.
Clin. Cancer Res.
,
3
:
2187
-2190,  
1997
.
31
Katoh O., Tauchi H., Kawaishi K., Kimura A., Satow Y. Expression of the vascular endothelial growth factor (VEGF) receptor gene, KDR, in hematopoietic cells and inhibitory effect of VEGF on apoptotic cell death caused by ionizing radiation.
Cancer Res.
,
55
:
5687
-5692,  
1995
.
32
Freeman M. R., Schneck F. X., Gagnon M. L., Corless C., Soker S., Niknejad K., Peoples G. E., Klagsbrun M. Peripheral blood T lymphocytes and lymphocytes infiltrating human cancers express vascular endothelial growth factor: a potential role for T cells in angiogenesis.
Cancer Res.
,
55
:
4140
-4145,  
1995
.
33
Wartiovaara U., Salven P., Mikkola H., Lassila R., Kaukonen J., Joukov V., Orpana A., Ristimaki A., Heikinheimo M., Joensuu H., Alitalo K., Palotie A. Peripheral blood platelets express VEGF-C and VEGF which are released during platelet activation.
Thromb. Haemost.
,
80
:
171
-175,  
1998
.
34
Möhle R., Green D., Moore M. A. S., Nachman R. L., Rafii S. Constitutive production and thrombin-induced release of vascular endothelial growth factor by human megakaryocytes and platelets.
Proc. Natl. Acad. Sci. USA
,
94
:
663
-668,  
1997
.
35
Gaudry M., Brgerie O., Andrieu V., El Benna J., Pocidalo M. A., Hakim J. Intracellular pool of vascular endothelial growth factor in human neutrophils.
Blood
,
90
:
4153
-4161,  
1997
.
36
Turley H., Scott P. A. E., Watts V. M., Bicknell R., Harris A. L., Gatter K. C. Expression of VEGF in routinely fixed material using a new monoclonal antibody VG1.
J. Pathol.
,
186
:
313
-318,  
1998
.
37
Sobin, L. H., and Wittekind, C., (eds.) UICC-TNM Classification of Malignant Tumours, 5th ed. New York: Wiley-Liss Inc., 1997.
38
Elston C. W., Ellis I. O. Pathological prognostic factors in breast cancer. : I. The value of histological grade in breast cancer: experience from a large study with long-term follow-up..
Histopathology
,
19
:
403
-410,  
1991
.
39
Zhang L., Scott P. A. E., Turley H., Leek R., Lewis C. E., Gatter K. C., Harris A. L., Mackenzie I. Z., Rees M. C. P., Bicknell R. Validation of anti-vascular endothelial growth factor (anti-VEGF) antibodies for immunohistochemical localization of VEGF in tissue sections: expression of VEGF in the human endometrium.
J. Pathol.
,
185
:
402
-408,  
1998
.
40
Vermeulen P. B., Gasparini G., Fox S. B., Toi M., Martin L., McCulloch P., Pezzella F., Viale G., Weidner N., Harris A. L., Dirix L. Y. Quantification of angiogenesis in solid human tumours: an International consensus on the methodology and criteria of evaluation.
Eur. J. Cancer
,
32A
:
2474
-2484,  
1996
.
41
Chung H. C., Rha S. Y., Ahn J. B., Shim K. Y., Yoo N. C., Kim J. H., Roh J. K., Lee K. S., Min J. S., Kim B. S., Kim J. J. Menstrual state should be considered in determining sero-positivity of soluble angiogenic factors in breast cancer.
Int. J. Mol. Med.
,
2
:
465
-470,  
1998
.
42
Hyodo I., Doi T., Endo H., Hosokawa Y., Nishikawa Y., Tanimizu M., Jinno K., Kotani Y. Clinical significance of plasma vascular endothelial growth factor in gastrointestinal cancer.
Eur. J. Cancer
,
13
:
2041
-2045,  
1998
.
43
Duque J. L., Loughlin K. R., Adam R. M., Kantoff P. W., Zurakowski D., Freeman M. R. Plasma levels of vascular endothelial growth factor are increased in patients with metastatic prostate cancer.
Urology
,
54
:
523
-527,  
1999
.
44
Crichton M. B., Nichols J. E., Zhao Y., Bulun S. E., Simpson E. R. Expression of transcripts of interleukin-6 and related cytokines by human breast tumors, breast cancer cells, and adipose stromal cells.
Mol. Cell. Endocrinol.
,
118
:
215
-220,  
1996
.
45
Hsich G., Kenney K., Gibbs C. J., Lee K. H., Harrington M. G. The 14-3-3 brain protein in cerebrospinal fluid as a marker for transmissible spongiform encephalopathies [see comments].
N. Engl. J. Med.
,
335
:
924
-930,  
1996
.
46
Salgado R., Vermeulen P. B., Benoy I., Weytjens R., Huget P., Van Marck E., Dirix L. Y. Platelet number and interleukin-6 correlate with VEGF but not with bFGF serum levels of advanced cancer patients.
Br. J. Cancer
,
80
:
892
-897,  
1999
.
47
Gunsilius E., Gastl G. Platelets and VEGF blood levels in cancer patients.
Br. J. Cancer
,
81
:
184
-186,  
1999
.
48
O’Byrne K. J., Dobbs N., Propper D., Smith K., Harris A. L. Vascular endothelial growth factor, platelet counts, and prognosis in renal cancer.
Lancet
,
353
:
1494
-1495,  
1999
.
49
Salven P., Orpana A., Joensuu H. Leukocytes and platelets of patients with cancer contain high levels of vascular endothelial growth factor.
Clin. Cancer Res.
,
5
:
487
-491,  
1999
.
50
Pinedo H. M., Verheul H. M. W., D’Amato R. J., Folkman J. Involvement of platelets in tumour angiogenesis?.
Lancet
,
352
:
1775
-1777,  
1998
.
51
Dunst J., Pigorsch S., Hansgen G., Hintner I., Lautenschlager C., Becker A. Low hemoglobin is associated with increased serum levels of vascular endothelial growth factor (VEGF) in cancer patients. : Does anemia stimulate angiogenesis?.
Strahlenther. Onkol.
,
175
:
93
-96,  
1999
.
52
Heer K., Kumar H., Speirs V., Greenman J., Drew P. J., Fox J. N., Carleton P. J., Monson J. R. T., Kerin M. J. Vascular endothelial growth factor in premenopausal women—indicator of the best time for breast cancer surgery?.
Br. J. Cancer
,
78
:
1203
-1207,  
1998
.
53
Agrawal R., Conway G. S., Sladkevicius P., Payne N. N., Bekir J., Campbell S., Tan S. L., Jacobs H. S. Serum vascular endothelial growth factor (VEGF) in the normal menstrual cycle: association with changes in ovarian and uterine Doppler blood flow.
Clin. Endocrinol.
,
50
:
101
-106,  
1999
.
54
Benoy I., Vermeulen P., Wuyts H., Dirix L. Vascular endothelial cell growth factor (VEGF) serum concentrations change according to the phase of the menstrual cycle.
Eur. J. Cancer
,
34
:
1298
-1299,  
1998
.
55
Lonning P. E., Hall K., Aakvaag A., Lien E. A. Influence of tamoxifen on plasma levels of insulin-like growth factor I and insulin-like growth factor binding protein I in breast cancer patients.
Cancer Res.
,
52
:
4719
-4723,  
1992
.
56
Pollak M., Costantino J., Polychronakos C., Blauer S-A., Guyda H., Redmond C., Fisher B., Margolese R. Effect of tamoxifen on serum insulinlike growth factor I levels in stage I breast cancer patients.
J. Natl. Cancer Inst.
,
82
:
1693
-1697,  
1999
.
57
Hyder S. M., Murthy L., Stancel G. M. Progestin regulation of vascular endothelial growth factor in human breast cancer cells.
Cancer Res.
,
58
:
392
-395,  
1998
.
58
Huang J. C., Liu D. Y., Dawood M. Y. The expression of vascular endothelial growth factor isoforms in cultured human endometrial stromal cells and its regulation by 17β-oestradiol.
Mol. Hum. Reprod.
,
4
:
603
-607,  
1998
.
59
Cullinan-Bove K., Koos R. D. Vascular endothelial growth factor/vascular permeability factor expression in the rat uterus: rapid stimulation by estrogen correlates with estrogen-induced increases in uterine capillary permeability and growth.
Endocrinology
,
133
:
829
-837,  
1993
.
60
Shifren J. L., Tseng J. F., Zaloudek C. J., Ryan I. P., Meng Y. G., Ferrara N., Jaffe R. B., Taylor R. N. Ovarian steroid regulation of vascular endothelial growth factor in the human endometrium: implications for angiogenesis during the menstrual cycle and in the pathogenesis of endometriosis.
J. Clin. Endocrinol. Metab.
,
81
:
3112
-3118,  
1996
.
61
Charnock-Jones D. S., Sharkey A. M., Rajput-Williams J., Burch D., Schofield J. P., Fountain S. A., Boocock C. A., Smith S. K. Identification and localization of alternatively spliced mRNAs for vascular endothelial growth factor in human uterus and estrogen regulation in endometrial carcinoma cell lines.
Biol. Reprod.
,
48
:
1120
-1128,  
1993
.
62
Ruohola J. K., Valve E. M., Karkkainen M. J., Joukov V., Alitalo K., Harkonen P. L. Vascular endothelial growth factors are differentially regulated by steroid hormones and antiestrogens in breast cancer cells.
Mol. Cell. Endocrinol.
,
149
:
29
-40,  
1999
.
63
Hyder S. M., Stancel G. M., Chiappetta C., Murthy L., Boettger-Tong H. L., Makela S. Uterine expression of vascular endothelial growth factor is increased by estradiol and tamoxifen.
Cancer Res.
,
56
:
3954
-3960,  
1996
.
64
Zhao Y., Hague S., Manek S., Zhang L., Bicknell R., Rees M. C. PCR display identifies tamoxifen induction of the novel angiogenic factor adrenomedullin by a non estrogenic mechanism in the human endometrium.
Oncogene
,
16
:
409
-415,  
1999
.
65
Gagliardi A. R., Hennig B., Collins D. C. Antiestrogens inhibit endothelial cell growth stimulated by angiogenic-growth factors.
Anticancer Res.
,
16
:
1101
-1106,  
1996
.
66
Kedar R. P., Bourne T. H., Powles T. J., Collins W. P., Ashley S. E., Cosgrove D. O., Campbell S. Effects of tamoxifen on uterus and ovaries of postmenopausal women in a randomised breast cancer prevention trial.
Lancet
,
343
:
1318
-1321,  
1994
.
67
Gasparini G., Weidner N., Bevilacqua P., Maluta S., Dalla Palma P., Caffo O., Barbareschi M., Boracchi P., Marubini E., Pozza F. Tumor microvessel density, p53 expression, tumor size, and peritumoral lymphatic vessel invasion are relevant prognostic markers in node-negative breast carcinoma.
J. Clin. Oncol.
,
12
:
454
-466,  
1994
.
68
Horak E. R., Leek R., Klenk N., Le Jeune S., Smith K., Stuart N., Greenall M., Stepniewska K., Harris A. L. Angiogenesis, assessed by platelet/endothelial cell adhesion molecule antibodies, as indicator of node metastases and survival in breast cancer.
Lancet
,
340
:
1120
-1124,  
1992
.
69
Balsari A., Maier J. A. M., Colnaghi M. I., Menard S. Correlation between tumor vascularity, vascular endothelial growth factor production by tumor cells, serum vascular endothelial growth factor levels, and serum angiogenic activity in patients with breast carcinoma.
Lab. Investig.
,
79
:
897
-902,  
1999
.