|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2000 American Society for Clinical Oncology Correlation of Vascular Endothelial Growth Factor Content With Recurrences, Survival, and First Relapse Site in Primary Node-Positive Breast Carcinoma After Adjuvant TreatmentFrom the Departments of Oncology and Clinical Chemistry, Umeå University, Umeå, Sweden. Address reprint requests to Barbro Linderholm, MD, Department of Oncology, Umeå University, SE-901 85 Umeå, Sweden; email barbro.linderholm{at}onkologi.umu.se
PURPOSE: To determine the predictive value of vascular endothelial growth factor (VEGF) for relapse-free survival (RFS) and overall survival (OS) in primary node-positive breast cancer (NPBC) after adjuvant endocrine treatment or adjuvant chemotherapy. MATERIALS AND METHODS: VEGF was quantitatively measured in tumor cytosols from 362 consecutive patients with primary NPBC using an enzyme immunoassay for human VEGF165. Adjuvant treatment was given to all patients, either as endocrine therapy (n = 250) or chemotherapy (n = 112). The median follow-up time was 56 months. RESULTS: Univariate analysis showed VEGF to be a significant predictor of RFS (P = .0289) and OS (P = .0004) in the total patient population and in patients who received adjuvant endocrine treatment (RFS, P = .0238; OS, P = .0121). In the group of patients who received adjuvant chemotherapy, no significant difference was seen in RFS, but a difference was seen in OS (P = .0235). Patients with bone recurrences tended to have lower VEGF expression (median, 2.17 pg/µg DNA) than patients with visceral metastasis (4.41 pg/µg), brain metastasis (8.29 pg/µg), or soft tissue recurrences (3.16 pg/µg). Multivariate analysis showed nodal status (P = .0004), estrogen receptor (ER) status (P < .0001), and tumor size (P = .0085) to be independent predictors of RFS. VEGF was found to be an independent predictor of OS (P = .0170; relative risk [RR] = 1.82), as were ER (P < .0001; RR = 5.19) and nodal status (P = .0002; RR = 2.58). For patients receiving adjuvant endocrine treatment, multivariate analysis showed VEGF content to be an independent predictor of OS (P = .0420; RR = 1.90) but not of RFS. CONCLUSION: The results suggest that VEGF165 content in tumor cytosols is a predictor of RFS and OS in primary NPBC. VEGF content might also predict outcome after adjuvant endocrine treatment, but further studies in a prospective setting with homologous treatments are required.
IN BREAST CARCINOMA, tumor involvement of the axillary lymph nodes is considered to be the most important predictor for relapse and thereby the strongest indicator for delivering systemic adjuvant therapy.1,2 However, the clinical outcome of node-positive breast cancer is heterogeneous despite adjuvant systemic treatment.3,4 In addition, lymph node metastasis neither predicts the complete biologic features of the tumor nor provides information concerning responsiveness to different types of systemic treatment given.5,6 Steroid receptor status is, at present, the only well-accepted predictor of responsiveness to adjuvant endocrine treatment.3,7 Extensive research has been carried out to identify novel biologic markers with the capacity to predict recurrence of the disease and responsiveness to systemic treatment. Although promising results have been demonstrated concerning overexpression of c-erbB-28 and p53 status,9-14 there are at present no conclusive markers recommended for use in the routine, clinical setting. Angiogenesis is stimulated by various peptides that induce proliferation of cells and is shown to be an essential factor for tumor growth and development of metastases.15 In breast carcinoma, several studies have suggested that the degree of vascularization of the primary tumor is an independent predictor of survival, regardless of nodal status.16-25 Vascular endothelial growth factor (VEGF), also known as vascular permeability factor, is a heparin-binding glycoprotein that has several important effects on vascular endothelial cells. Currently, VEGF is considered to be the most selective mitogen for endothelial cells.26 VEGF also increases vascular permeability27,28 and induces alterations in ion flow, cell proliferation, and migration27,29 and release of proteinases that are involved in tumor invasiveness.30,31 We recently reported data on VEGF165 as an independent predictive factor for overall survival (OS) in a series of 525 node-negative breast cancer patients.32 In this study, the aim was to determine the value of VEGF165 in predicting relapse-free survival (RFS) and OS in 362 node-positive patients compared with established prognostic indicators. Moreover, the outcome after adjuvant endocrine or cytotoxic treatment as well as the first predominant recurrence site in relation to the cytosolic VEGF content were studied.
Patient Data Clinical information and tumor samples were collected from 362 consecutive, unselected women with invasive breast carcinoma, stage T1-3N1-2M0, diagnosed and primarily treated between 1990 and 1995 in the northern health care region of Sweden. Tumor classification and staging were conducted in accordance with the International Union Against Cancer tumor-node-metastasis (UICC-TNM) classification. The patients’ records at the Department of Oncology, Umeå University (Umeå, Sweden), or at the departments of surgery in the countries of Norrbotten, Västerbotten, Västernorrland, and Jämtland, Sweden, were checked manually. Primary treatment was administered according to the guidelines of the North Swedish Breast Cancer Group. Two hundred sixty-one patients underwent modified radical mastectomy with axillary dissection, and radiotherapy was given to a total dose of 39.2 Gy (14 x 2.8 Gy). One hundred one patients were treated with breast conservation surgery and axillary dissection followed by radiotherapy to a total dose of 56 Gy (28 x 2.0 Gy). The axillary lymph nodes were not included in the target volume. Adjuvant systemic treatment was administered to all patients; premenopausal estrogen receptor (ER)-positive patients were randomized in a clinical trial to receive either chemotherapy with nine cycles of cyclophosphamide, methotrexate, and fluorouracil (CMF) or ovarian irradiation (7 x 2.5 Gy). Premenopausal ER-negative patients received chemotherapy with nine cycles of fluorouracil, epirubicin, and cyclophosphamide (FEC) or CMF within a randomized trial. Postmenopausal patients were randomized to receive 40 mg of tamoxifen daily for either 2 (n = 35) or 5 (n = 37) years. The majority of patients who received adjuvant tamoxifen in this patient population were not included in this trial. An overview of the patients and the adjuvant systemic treatment they received is included in Table 1. In total, 112 patients received adjuvant chemotherapy with FEC (n = 25) or CMF (n = 87); 11 of these patients also received adjuvant tamoxifen. Endocrine therapy was given to 250 patients, of whom 41 (16.4%) were found to be steroid receptor-negative. Tamoxifen was administered to 208 patients and 42 patients were treated with ovarian irradiation; in 12 cases, this was followed by treatment with tamoxifen. The number of patients for whom data were available varied with respect to different prognostic factors studied. In all cases, data regarding tumor size, number of axillary lymph node metastases, histologic type, ER status, progesterone receptor status, and VEGF protein content were collected. Histopathologic grade was evaluated in 301 cases, and the majority of those patients presented with a ductal carcinoma. S-phase fraction was, during this period, analyzed in few patients (n = 83). The median age in the total study population was 57 years and the median follow-up time for survivors was 56 months.
Follow-Up After primary and adjuvant treatment, the patients were, in most cases, followed-up by routine check-ups at the Department of Oncology, Umeå University, or at the departments of surgery in the northern health care region of Sweden for 5 to 10 years. The patients were invited to screening programs in 1990 in the counties of Norrbotten and Västernorrland, in 1995 in Västerbotten, and in 1996 in Jämtland. Before this, mammograms were recommended to be performed yearly or every other year. Radiographic studies were performed when indicated. In the randomized trials with chemotherapy, in addition to clinical examination and collection of blood samples, chest roentgenograms and bone scintigrams or skeleton roentgenograms were performed according to the protocol, every 6 months or yearly. Recurrences were defined as the first documented evidence of new disease manifestations in the locoregional area, in the contralateral breast, at distant sites, or in a combination of these. The RFS and the OS were calculated as the time from diagnosis to the date of first recurrence or death. The follow-up time for patients without documented recurrences or death has been calculated as the time from diagnosis to the last clinical examination (last follow-up date, August 31, 1998).
Tumor Tissue Preparation
VEGF Analysis
Steroid Receptor Analysis
Statistical Methods
Distribution of VEGF A wide range of VEGF protein content was found. The median value was 2.33 pg/µg DNA (range, 0.04 to 134.29 pg/µg DNA (Fig 1). This did not significantly differ from the findings in node-negative patients reported by us32 (median, 2.44 pg/µg DNA; range, 0.11 to 144.79; P = .722). No difference was found in VEGF content (median, range) between samples with the longest storage time (breast cancer diagnosis in 1990) as compared with samples with the shortest storage time (diagnosis in 1995; not shown).
Clinical Outcome of All Patients At 56 months, the probability of RFS and OS in all patients was 63.9% and 71.8%, respectively. Actual recurrences, sites of first recurrence, and deaths in the patient population are listed in Table 2. A total of 130 recurrences and 102 deaths was recorded; of those, six deaths were unrelated to breast cancer but were included in overall survival analysis. The recurrences were divided into four groups: soft tissue (including local failures), visceral (including lung and liver metastasis), brain metastasis, and bone metastasis. The first predominant recurrence site was determined as the clinically unfavorable one; ie, patients with liver and bone metastasis or local recurrences were placed in the visceral group.
The predominant relapse site was soft tissue in 36 cases, including 18 local in breast failure and one contralateral breast carcinoma. Fifty-four patients had visceral metastasis, five had brain metastasis, and 35 had bone metastasis as predominant sites. Of those, 15 patients (37.5%) with bone metastasis, 16 (40.0%) with soft tissue recurrences (including seven with local in-breast failures), and nine (22.5%) with visceral metastasis were alive at last follow-up. No patients with brain metastasis were alive at last follow-up.
Association Between VEGF Expression and Site of First Recurrence
Association Between VEGF and Other Variables In the total material, statistically significant inverse associations were seen between VEGF content and ER status (positive v negative; P < .001) or PgR status (positive v negative; P < .001). Moreover, a significant association was observed between VEGF and histologic type (ductal v lobular and others; P = .042). No significant correlation was found between VEGF and clinical stage (I v IIA v IIB v IIIA v IIIB; P = .071), histologic grade (P = .867), number of axillary lymph node metastases (one to five v six to 10 v > 10; P = .570), tumor size (P = .091), or age (  56 years v > 56 years; P = .742).
Univariate Analysis in Node-Positive Patients
Multivariate Analysis in Node-Positive Patients Multivariate analysis of the joint effect of combining VEGF determination with other prognostic factors showed number of axillary node metastases (one to five v > five; P = .0004), ER status (positive v negative; P < .0001), and tumor size (T1 v T2-3; P = .0085) to be independent predictors of RFS. VEGF content (P = .2222) was not an independent predictor of RFS. However, for OS, ER status (P < .0001), number of involved lymph nodes (P = .0002), and VEGF content ( 1.75 pg/µg DNA v > 1.75 pg/µg DNA; P = .0170) were all independent predictors of OS, with an increased risk for death of 5.19, 2.58, and 1.82, respectively. Tumor size was not found to be a independent predictor of OS. Histologic grade (1 and 2 v 3) was not found to be an independent predictor of RFS or OS (Table 4).
Univariate Analysis for RFS and OS After Adjuvant Treatment VEGF content was a statistically significant predictor of RFS and OS for patients who received adjuvant endocrine treatment (n = 250). Patients with higher expression had reduced survival times compared with those with lower VEGF content, both when 1.75 pg/µg DNA was used as the cut point (RFS, P = .0238; OS, P = .0121; Fig 3A and 3B; Table 3) and when the median level of 2.33 pg/µg DNA was used as the cut point (RFS, P = .0046; OS, P = .0051). However, in this group, 41 patients were found to have ER-negative disease; the results were similar when they were excluded from analysis (RFS, P = .0346; OS, P = .0263). Other factors of statistical significance for survival were ER (RFS, P < .0001; OS, P < .0001), PgR (RFS, P = .0003; OS, P < .0001), tumor size (RFS, P = .0002; OS, P = .0074), number of involved lymph nodes (RFS, P = .0005; OS, P = .0027), S-phase fraction (RFS, P = .0235; OS, P = .0329), and age (RFS, P = .0341; OS, P = .0024). Histopathologic grade and type were not significant predictors of RFS or OS (Table 3).
Among patients who received adjuvant chemotherapy, univariate analysis showed no significant difference in RFS with regard to VEGF (cut point of 1.75 pg/µg DNA, P = .6129; or 2.33 pg/µg DNA, P = .5369). Concerning OS, statistically significant reductions in survival times were seen for patients with higher VEGF content with the cut point of 1.75 pg/µg DNA (P = .0235). Other factors that were found to be statistically significant for prediction of RFS were ER (P = .0005), PgR (P < .0001), and tumor size (P = .0067); factors that were found to be statistically significant for prediction of OS included ER (P < .0001), PgR (P < .0001), S-phase fraction (P = .0246), and tumor size (P = .0463) (Table 3).
Multivariate Analysis for Patients Treated With Adjuvant Endocrine Therapy or Chemotherapy
The results from this study suggest that the tumor cytosolic content of VEGF165 is a possible predictor of survival in primary node-positive breast cancer and might also be useful to predict outcome after adjuvant endocrine therapy. This finding is consistent when using VEGF as a dichotomous variable as well as a continuous variable. The present results find support in earlier studies, which showed VEGF as the strongest predictor of survival in node-negative patients.25,32 Angiogenesis, measured by counting vessel density after immunocytochemical staining, has demonstrated a worse prognosis for patients with an increased vessel density in the primary tumor in several independent studies.16-23 A correlation has also been shown between vessel density and the content of VEGF protein in cytosols from the primary tumor.33 Thus quantitative measurement of VEGF might be one reproducible way to estimate the angiogenic activity in tumor samples. The isoform measured in this study, VEGF165, has been reported as the only detectable isoform in malignant breast tissue, despite the fact that mRNA transcripts were detected from three isoforms (namely, 121, 165, and 189).34 One interesting observation in this study is that the VEGF content seems to predict responsiveness to adjuvant endocrine therapy. Patients with ER-positive tumors and a high VEGF expression in the primary tumor had a significantly reduced RFS and OS despite administration of endocrine treatment. In accordance with the present results, a previous study showed that patients with higher vessel density had significantly reduced survival times compared with those with lower vessel density, regardless of treatment with tamoxifen.35 From the experimental level, it is of interest to note that in hormone-sensitive breast cancer cell lines, estrogen regulates the production of VEGF.36 Moreover, tamoxifen has been shown to directly inhibit angiogenesis.37 Tamoxifen has also been shown to have an antiangiogenic effect by decreasing transforming growth factor alpha, which is a stimulator of angiogenesis in ER-positive tumors, and by increasing transforming growth factor beta, which inhibits angiogenesis in ER-negative tumors.38,39 A subsequent suppression of neovascularization was also seen in these experiments. Nevertheless, the pertinent information from the present study suggests that endocrine treatment alone is insufficient as systemic treatment for patients with a high VEGF expression and suggests that VEGF content adds predictive information to ER status. This could be of value to delineate the optimal treatment. Concerning adjuvant polychemotherapy, VEGF content failed to predict RFS but was found to predict OS. Earlier results have shown reduced RFS and OS times for patients with higher microvessel density despite administration of adjuvant chemotherapy.35 Interestingly, P-glycoprotein and glutathione-S-transferase, which are at least to some degree responsible for resistance to chemotherapy, are expressed not only in tumor cells but also in endothelial cells.40,41 In tumors with high microvascular density, this could contribute to an increased risk of failure to conventional treatment. However, this study consists of too small a number of patients who received heterogenous types of adjuvant chemotherapy (in some cases, also followed by tamoxifen). A controlled prospective study with homologous treatment is, therefore, desirable. In this study, VEGF content of node-positive patients did not statistically differ from that of node-negative patients, as earlier reported.32 An increased microvessel density has, in some studies, been significantly correlated with an increased number of involved lymph nodes and/or distant metastases.16,17,19 Others have found a significant increase in microvessel density in node-positive compared with node-negative tumors but no difference within the node-positive group.16 In a xenograft of human breast carcinoma in nude mice, VEGF was critical during initial growth but was not necessary for continued growth after the tumor had reached a certain size. On the other hand, suppression of VEGF had no effect on the expansion of the larger tumors. Instead, upregulation of other angiogenic peptides, ie, basic fibroblast growth factor and transforming growth factor alpha, was found after suppression of VEGF.42 Therefore, it could be of interest to evaluate whether this shift in the angiogenic growth factors is also relevant in the clinical setting, in an effort to at least partially explain the present findings. The results from this study show a discrepancy between the predictive value of VEGF concerning RFS and OS. The predictive value of VEGF seems to be of higher importance with shorter follow-up time. Interestingly, patients with bone recurrences, a group demonstrating, in some cases, less aggressive clinical behavior, were found to have lower VEGF expression than patients with soft tissue and/or visceral metastasis. Patients with solely bone recurrences had VEGF levels equal to those of patients without documented recurrences. Somewhat in contrast to our results, Gasparini et al22 did not find any correlation between the microvessel density in the primary breast carcinoma and site of first recurrence.22 Also, different types of local or systemic treatment delivered at the time of first recurrence and responsiveness to those treatments must be considered. In fact, patients with a higher microvessel density or higher VEGF expression were reported to have shorter RFS and OS, both after adjuvant endocrine treatment and adjuvant chemotherapy with CMF.35,43 In our material, VEGF was not associated with nodal status and could not be significantly correlated with an increased tumor size, although a trend was found. The highest association was seen between increased VEGF expression and ER/PgR negativity, high proliferation rate, and poorly differentiated tumors. This could imply that VEGF might represent a biologic "tumor-dependent" factor, not a "time-dependent" factor, such as tumor size and lymph node status. Although VEGF has been suggested as the major angiogenic factor in tumor angiogenesis, it is obvious that determination of other factors and their association with clinical outcome would be of value. Moreover, one must consider the complexity of this process, which involves degradation of basement membranes, endothelial proliferation and migration, tube formation, and initiation of blood flow, a result that reflects the net balance between stimulating and inhibiting factors. Methodologic problems must also be considered. In this study, VEGF content has been measured in cytosols prepared for determination of steroid receptors. However, in a comparative study, it has been shown that for all VEGF isoforms, a higher level was found in the membrane fraction than in the cytosolic fraction.34 This suggests that difficulties in identifying patients with lower VEGF expression could be overcome by determination of VEGF in the membrane fraction. In summary, the results suggest that tumor cytosolic content of VEGF165 is a possible predictor of survival in patients with node-positive breast cancer. Moreover, assay for VEGF content might aid in identification of ER-positive patients who are unlikely to have an optimum outcome after administration of adjuvant hormonal therapy alone. The results from this study also suggest an association between an increased VEGF content and a higher risk of development of brain and visceral recurrences. This finding might partly explain why VEGF seems to be of higher predictive value for OS than for RFS. We conclude that further investigations are justified to elucidate the role of angiogenic factors as predictors of responsiveness to treatment and survival in primary breast cancer, alone and in comparison with other biologic markers, such as p53 and c-erbB-2.
Supported by grants from the Cancer Research Foundation and Lions Cancer Research Foundation, Umeå, Sweden.
1. Veronesi U, Luini A, Galimberti V, et al: Extent of metastatic axillary involvement in 1446 cases of breast cancer. Oncol 16:127-133, 1990 2. Fentiman IS, Mansel RE: The axilla: Not a no-go zone. Lancet 1:221-223, 1991 3. Fisher B, Redmond C, Brown A, et al: Influence of tumor estrogen and progesterone levels on the response to tamoxifen and chemotherapy in primary breast cancer. J Clin Oncol 1:227-241, 1983[Abstract] 4. Bonnadonna G, Valagussa P: Adjuvant systemic therapy for resectable breast cancer. J Clin Oncol 3:259-275, 1985[Medline] 5. Hendersson IC, Harris JR, Kinne DW, et al: Cancer of the breast, in DeVita VT Jr, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice of Oncology. Philadelphia, PA,Lippincott, 1989, pp 1230-1235
6.
Gasparini G, Pozza F, Harris AL: Evaluating usefulness of new prognostic and predictive indicators in node-negative breast cancer patients. J Natl Cancer Inst 85:1206-1219, 1993 7. Mason BH, Holdaway IM, Mullins PR, et al: Progesterone and estrogen receptors as prognostic variables in breast cancer. Res 43:2985-2990, 1983
8.
Clark GM: Should selection of adjuvant chemotherapy for patients with breast cancer be based on erb-2 status? J Natl Cancer Inst 90:1320-21, 1998 9. Cattoretti G, Rilke F, Andreola S, et al: P53 expression in breast cancer. Int J Cancer 41:178-183, 1988[Medline] 10. Silvestrini R, Veneroni S, Benini E, et al: Expression of p53 in node-negative breast carcinoma. J Natl Cancer Inst 16:965-970, 1993 11. Thor A, Moore DH II, Edgerton SM, et al: Accumulation of p53 tumor suppressor gene protein: An independent marker of prognosis in breast cancer. J Natl Cancer Inst 11:845-855, 1992
12.
Allred DC, Clark GM, Elledge R, et al: Association of p53 protein expression with tumor cell proliferation rate and outcome in node-negative breast cancer. J Natl Cancer Inst 85:200-206, 1993 13. Bergh J, Norberg T, Sjögren S, et al: Complete sequencing of the p53 gene provides prognostic information in breast cancer patients, particularly in relation to adjuvant systemic therapy and radiotherapy. Nat Med 10:1029-1034, 1995 14. Borg Å, Lennerstrand J, Stenmark-Askmalm M, et al: Prognostic significance of p53 overexpression in primary breast cancer: A novel luminometric immunoassay applicable on steroid receptor cytosols. Br J Cancer 71:1013-1017, 1995[Medline] 15. Folkman J: Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1:27-31, 1995[Medline] 16. Weidner N, Semple JP, Welch WR, et al: Tumor angiogenesis and metastasis-correlation in invasive breast carcinoma. N Engl J Med 324:1-8, 1991[Abstract] 17. Horak ER, Leek R, Klenk N, et al: Angiogenesis, assessed by platelet/endothelial cell adhesion molecule antibodies, as indicator of node metastasis and survival in breast cancer. Lancet 340:1120-1124, 1992[Medline] 18. Bosari S, Lee AKC, De Lellis RA, et al: Microvessel quantification and prognosis in invasive breast carcinoma. Hum Pathol 23:755-761, 1992[Medline] 19. Toi M, Kashitani J, Tominaga T: Tumor angiogenesis is an independent prognostic indicator in primary breast carcinoma. Int J Cancer 55:371-374, 1993[Medline] 20. Fox SB, Leek RD, Smith K, et al: Tumor angiogenesis in node-negative breast carcinomas: Relationship with epidermal growth factor receptor and survival. Res Treat 29:109-116, 1994 21. Obermair A, Czerzwenka K, Kurz C, et al: Influence of tumoral microvessel density on the recurrence-free survival in human breast cancer: Preliminary results. Onkologie 17:44-49, 1994 22. Gasparini G, Weidner N, Bevilacqua P, et al: Tumor microvessel density, p53 expression, tumor size and peritumoral lymphatic vessel invasion are relevant prognostic markers in node-negative breast carcinoma. Clin Oncol 12:454-466, 1994 23. Toi M, Inada K, Suzuki H, et al: Tumor angiogenesis in breast cancer: Its importance as a prognostic indicator and the association with vascular endothelial growth factor expression. Cancer Res Treat 36:193-204, 1995 24. Heimann R, Ferguson D, Powers C, et al: Angiogenesis as a predictor of long-term survival for patients with node-negative breast cancer. J Natl Cancer Inst 23:1764-1769, 1996
25.
Gasparini G, Toi M, Gion M, et al: Prognostic significance of vascular endothelial growth factor protein in node-negative breast carcinoma. J Natl Cancer Inst 89:139-147, 1997
26.
Leung DW, Cachianes G, Kuang WJ, et al: Vascular endothelial growth factor is a secreted angiogenic mitogen. Science 246:1306-1309, 1989
27.
Senger DR, Galli SJ, Dvorak AM, et al: Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid. Science 219:983-985, 1983 28. Dvorak HF: Tumours: Wounds that do not heal–Similarity between tumour stroma generation and wound healing. N Engl J Med 315:1650-1658, 1983[Medline] 29. Ferrara N, Henzel WJ: Pituitary follicular cells secrete a novel heparin-binding growth factor specific for vascular endothelial cells. Biochem Biophys Res Commun 161:851-859, 1989[Medline] 30. Unemori EN, Ferrara N, Bauer EA, et al: Vascular endothelial growth factor induces interstitial collagenase expression in human endothelial cells. J Cell Physiol 153:557-562, 1992[Medline] 31. Lindgren M, Johansson M, Sandström J, et al: VEGF and tPA co-expressed in malignant glioma. Acta Oncologica 6:615-618, 1997
32.
Linderholm B, Tavelin B, Grankvist K, et al: Vascular endothelial growth factor is of high prognostic value in node-negative breast carcinoma. J Clin Oncol 16:3121-3128, 1998 33. Toi M, Kondo S, Suzuhi H, et al: Quantitative analysis of vascular endothelial growth factor in primary breast cancer. Cancer 77:1101-1106, 1996[Medline] 34. Scott PAE, Smith K, Poulsom R, et al: Differential expression of vascular endothelial growth factor mRNA vs protein isoform expression in human breast cancer and relationship to eIF-4E. Br J Cancer 12:2120-2128, 1998 35. Gasparini G, Barbareshi M, Boracchi P, et al: Tumor angiogenesis predicts clinical outcome of node-positive breast cancer patients treated with adjuvant hormone therapy or chemotherapy. Cancer J 1:131-141, 1995 36. Kolch W, Martiny-Baron G, Kieser A, et al: Regulation of the expression of the VEGF/VPS and its receptors: Role in angiogenesis. Breast Cancer Res Treat 36:139-155, 1995[Medline]
37.
Gagliardi A, Collins DC: Inhibition of angiogenesis by antiestrogens. Cancer Res 53:533-535, 1993 38. Noguchi S, Motomura K, Inaji H, et al: Down regulation of transforming growth factor-alpha by tamoxifen in human breast cancer. Cancer 72:131-136, 1993[Medline]
39.
Antonelli-Orlidge A, Saunders KB, Smith SR, et al: An activated form of transforming growth factor beta is produced by cocultures of endothelial cells and pericytes. Proc Natl Acad Sci U S A 86:4544-4548, 1989 40. Cordon-Cardo C, O’Brian JP, Boccia J, et al: Expression of the multidrug resistence gene product P-glycoprotein in human normal and tumor tissues. J Histochem Cytochem 9:1277-1287, 1990 41. Terrier P, Townsend AJ, Coindre JM, et al: An immunohistochemical study of Pi class gluthatione S-transferase expression in normal human tissues. Am J Pathol 137:845-853, 1990[Abstract]
42.
Yoshiji H, Harris SR, Thorgeirsson Vascular endothelial growth factor is essential for initial but not continued in vivo growth of human breast carcinoma cells. Cancer Res 57:3924-3928, 1997 43. Gasparini G, Toi M, Miceli R, et al: Clinical relevance of vascular endothelial growth factor and thymidine phosphorylase in patients with node-positive breast cancer treated with either adjuvant chemotherapy or hormone therapy. Cancer J Sci Am 5:101-111, 1999[Medline] Submitted May 10, 1999; accepted November 29, 1999. This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
0.1 fmol ER or PgR/µg DNA were considered to be receptor-positive. 
2 test. Association between VEGF content and site of first relapse were tested with the Mann-Whitney U test. For the purpose of this study, VEGF was tested both as a dichotomous variable and as a continuous variable. Survival was estimated using the Kaplan-Meier method, and comparison between study groups was performed with the log-rank test. The optimal cut-off point for VEGF content, according to the lowest P values and the highest relative risk (RR), was found at 1.75 pg/µg DNA for OS, and this was used as the cut point in univariate and multivariate analysis. The median value of VEGF was also used as cut point in order to avoid statistical manipulation. VEGF was also tested as a continuous variable in a univariate Cox regression analysis. To evaluate the simultaneous effect of different factors on survival, the Cox proportional hazards model was used. For each variable, the patients were classified in two prognostic categories, and the putative best prognosis category was used as the reference. In all tests, the significance level was set to .05, and all tests were two-sided. 
