Tumor-Specific Expression of Vascular Endothelial Growth Factor Receptor 2 but Not Vascular Endothelial Growth Factor or Human Epidermal Gro
http://www.100md.com
《临床肿瘤学》
the Department of Surgery, Helsingborgs Lasarett, Helsingborg
Department of Laboratory Medicine, Pathology, University Hospital, Malm
Department of Oncology, University Hospital, Lund
Department of Oncology, University Hospital, Linkping
Department of Pathology and Cytology, Kalmar County Hospital, Kalmar
South Swedish and Southeast Swedish Breast Cancer Groups, Sweden
ABSTRACT
PURPOSE: Vascular endothelial growth factor A (VEGF-A) and vascular endothelial growth factor receptor 2 (VEGFR2) are often coexpressed in breast cancer, and potentially affect cellular pathways and key proteins such as the estrogen receptor (ER) targeted by endocrine treatment. We therefore explored the association between adjuvant tamoxifen treatment in breast cancer and expression of VEGF-A and VEGFR2, as well as human epidermal growth factor receptor 2 (HER2), which represents a candidate gene product involved in tamoxifen resistance.
PATIENTS AND METHODS: Immunohistochemical expression of tumor-specific VEGF-A, VEGFR2, and HER2 was evaluated in tumor specimens from premenopausal breast cancer patients randomly assigned to 2 years of tamoxifen or no treatment (n = 564), with 14 years of follow-up. Hormone receptor status was determined in 96% of the tumors.
RESULTS: VEGF-A, VEGFR2, and HER2 were assessable in 460, 472, and 428 of the tumors, respectively. In patients with ER–positive and VEGFR2-low tumors, adjuvant tamoxifen significantly increased recurrence-free survival (RFS; [HR] hazard ratio for RFS, 0.53; P = .001). In contrast, tamoxifen treatment had no effect in patients with VEGFR2-high tumors (HR for RFS, 2.44; P = .2). When multivariate interaction analyses were used, this difference in treatment efficacy relative to VEGFR2 expression status was statistically significant for both ER-positive (P = .04) plus ER-positive and progesterone receptor–positive tumors. We found no significant difference in tamoxifen treatment effects in relation to VEGF-A or HER2 status.
CONCLUSION: Tumor-specific expression of VEGFR2 was associated with an impaired tamoxifen effect in hormone receptor–positive premenopausal breast cancer. Tamoxifen in combination with VEGFR2 inhibitors might be a novel treatment approach for VEGFR2-expressing breast cancer, and such a treatment might restore the tamoxifen response.
INTRODUCTION
Adjuvant tamoxifen treatment for 5 years improves 10-year survival in hormone receptor–positive breast cancer and reduces the recurrence rate by 47% and mortality by 27%.1,2 Even though the beneficial effect of tamoxifen is restricted to patients with estrogen receptor (ER) –positive and/or progesterone receptor (PR) –positive tumors, a large fraction of hormone receptor–positive tumors do not respond as expected to the treatment.2 Resistance to endocrine therapy with tamoxifen is an immense clinical problem and researchers have therefore actively pursued mechanistic studies of tamoxifen resistance and tested predictive markers other than hormone receptor status.2-6 Growth factors with tyrosine kinase activity can activate ER without ligand binding2,4-6 and the cross-talk between ER and growth factor receptors seems to be an essential part of tamoxifen resistance. PR can serve as an alternative predictor of the effect of tamoxifen treatment7 because low expression of PR may be an indicator of activated growth factor receptors.8 Overexpression of the human epidermal growth factor receptor 2 (HER2) is also associated with resistance to tamoxifen in in vitro experiments,4,7,9 and clinical data suggest that overexpression of HER2 may be a detriment to the beneficial effect of tamoxifen, although evidence from randomized studies is still limited.3,5,10-13
Researchers have addressed vascular endothelial growth factor A (VEGF-A) as a potential predictor of response to tamoxifen treatment in early and advanced breast cancer.14-18 VEGF-A exerts its effect via receptors with tyrosine kinase activity (vascular endothelial growth factor receptor 1 [VEGFR1] and VEGFR2), activating intracellular pathways similar to those of other tyrosine kinase receptors.19 VEGFR2 has been linked to proliferative features via the mitogen-activated protein kinase pathway, whereas VEGFR1 does not affect proliferation.19 The VEGF receptors are mainly located on the surface of endothelial cells, but have also been identified on breast cancer cells, implicating autocrine and paracrine roles for VEGF-A in addition to its proangiogenic property.20-23
We recently reported a beneficial effect from 2 years of tamoxifen treatment compared with no adjuvant treatment in premenopausal breast cancer patients who were included in a randomized treatment trial in which fewer than 2% of patients received additional adjuvant chemotherapy.24 A tissue microarray of the breast cancer samples used in this study has now been constructed, allowing efficient studies of tamoxifen response in subgroups of breast cancer defined by putative predictive markers. Currently, standard adjuvant therapy for hormone-responsive breast cancer is 5 years of tamoxifen, administered either alone or combined with chemotherapy. Nevertheless, a cohort of breast cancer patients included in a randomized 2-year trial of adjuvant tamoxifen, which showed a clear beneficial effect of tamoxifen, offers a simple and favorable study design for scrutinizing predictors of tamoxifen response without interference from prognostic considerations and effects of chemotherapy.
The aim of this study was to analyze the association of tumor-specific expression of VEGF-A and VEGFR2 with tamoxifen response in hormone receptor–positive primary breast cancer in a controlled, randomized trial of tamoxifen versus control. The predictive information of another growth factor receptor that is commonly overexpressed in breast cancer, HER2, was further delineated to validate this candidate marker for tamoxifen response.
PATIENTS AND METHODS
Study Design
Premenopausal patients or patients younger than 50 years with stage II (pT2, N0, M0; pT1, N1, M0; and pT2, N1, M0) invasive breast cancer were enrolled onto a multicenter clinical trial between 1986 and 1991 and were randomly assigned to control (n = 288) or 2 years of tamoxifen (n = 276). Random assignment was performed by the Regional Oncological Center and informed consent was registered for all included patients. The study was approved by the ethical committees at Lund and Linkping Universities. The primary end point was recurrence-free survival (RFS) and the treatment effect should be examined in relation to hormone receptor status, as stated in the study protocol. All patients underwent curative surgery either by modified radical mastectomy or by breast-conserving surgery with axillary lymph node dissection (levels one and two). After breast-conserving surgery, radiotherapy (50 Gy) was administered to the breast, and in patients with axillary lymph node metastases, locoregional radiotherapy was delivered. Additional adjuvant therapy was administered as polychemotherapy (n = 8) and goserelin (n = 1) to nine patients (< 2%), who were evenly distributed between the two treatment arms. Patients were included irrespective of hormone receptor status, although ER and PR status were determined in 80% of the patients at the time of surgery by cytosol-based methods. Patients' and tumor characteristics in relation to treatment arm in the original study are presented in Table 1.
The patients were observed for 5 years with annual mammogram and physical examination and then within the national program by screening mammogram every 18 months. The median duration of follow-up for patients without a breast cancer event was 14 years (range, 0 to 17 years). The median follow-up time was comparable in the two treatment arms. The study has been included in the Oxford meta-analysis1 and the main results have been presented in detail elsewhere.24 Briefly, two years of adjuvant tamoxifen increased RFS in all included patients (hazard ratio [HR], 0.77; 95% CI, 0.60 to 098; P = .03). Hormone receptor status could be determined in 541 patients (96%) prospectively by using cytosol-based methods and retrospectively by immunohistochemistry as described.24 The beneficial effect by 2 years of adjuvant tamoxifen in terms of RFS was restricted to 385 patients with ER-positive and/or PR-positive tumors, whereas no effect was recorded in patients with ER-negative and PR-negative tumors. The interaction between tamoxifen treatment and hormone receptor status in terms of RFS was explored with Cox multivariate analyses; a significant interaction between PR status and tamoxifen treatment (P = .015), as well as for the combination of ER and PR positivity (P = .031), was demonstrated. In contrast, ER status alone was not a significant predictor of tamoxifen response (P = .19), and the combination of ER and/or PR status was not a predictor of tamoxifen response (P = .11).
Hormone Receptor Status
Hormone receptor status was determined by cytosol-based methods prospectively and by immunohistochemical methods retrospectively, as described previously.24 Comparison between the two techniques of determining hormone receptor status resulted in 86% concordance for ER status (, 0.70) and 88% concordance for PR status (, 0.73), justifying that a combination of the methods could be used.
Tumor Tissue Microarray Preparation
All tumor specimens were obtained at the time of surgery before administration of adjuvant therapy. Formalin-fixed, paraffin-embedded tumor blocks were obtained from the primary breast tumors. Clearly defined tumor areas were selected and two biopsies, 0.6 mm in diameter, were obtained from each donor block and mounted in a recipient block using a tissue microarray (TMA) instrument according to the manufacturer's instructions (Beecher Instruments, MD).
Immunohistochemistry
VEGF-A and VEGFR2. Four-micrometer TMA sections were deparaffinized, rehydrated, and treated in a microwave for 5 + 5 minutes in a citrate buffer (pH 6.0) before they were processed in an automatic immunohistochemistry staining machine according to standard procedures (TechMate500; DakoCytomation, Glostrup, Denmark) using a polyclonal VEGF (A-20) antibody diluted 1:400 (Santa Cruz Biotechnology Inc, Santa Cruz, CA) and a monoclonal VEGFR2 antibody diluted 1:1000 (Santa Cruz Biotechnology Inc, Santa Cruz, CA). Normal human kidneys were used as positive controls for VEGF-A and human aortic endothelium was used as positive control for VEGFR2. VEGF-A has been validated previously using an enzyme-linked immunosorbent assay–based method.25 The cytoplasmic staining intensity was evaluated semiquantitatively using a classification from 0 to 3, representing lack of staining (grade 0), low staining intensity (grade 1), intermediate staining intensity (grade 2), and high staining intensity (grade 3). Only invasive tumor cells were evaluated. In survival analyses of patients with tumors where VEGF-A and VEGFR2 were determined, high staining intensity (grade 3) was compared with grades 0 to 2 because grade 0 to 2 produced overlapping survival curves, whereas grade 3 differed substantially. The TMA was examined by two investigators with blinded clinical data and the concordance between the investigators was 89% for VEGF-A and 93% for VEGFR2 when participants were divided into four groups. Divergent results were re-examined followed by a conclusive decision.
HER2. Protein expression of HER2 was determined by immunohistochemistry using the Ventana Benchmark System (Ventana Medical Systems Inc, Tucson, AZ) with a prediluted antibody (Pathway CB-11, 760-2694; Ventana Medical Systems Inc, Tucson, AZ). HER2 expression was evaluated semiquantitatively according to a standard protocol (HercepTest; DakoCytomation). The protocol categorizes tumors into four groups: grade 0, lack of staining in all tumor cells or membrane staining in less than 10% of the tumor cells; grade 1+, weak, not circumferential membrane staining in more than 10% of the tumor cells; grade 2+, intermediate, circumferential membrane staining in more than 10% of the tumor cells; and grade 3+, intense and circumferential staining in more than 10% of the tumor cells.
HER2 gene amplification. HER2 gene amplification was determined by fluorescent in situ hybridization using an automated staining procedure according to the manufacturer's recommendations (Ventana Medical Systems). Briefly, unstained, formalin-fixed, paraffin-embedded tumor sections were deparaffinized, pretreated, and denatured before incubation overnight with the hybridization probe. The sections were counterstained and evaluated for HER2 gene copy number using a fluorescence microscope at a magnification of x400. All tumor cells within the biopsies in the TMA were evaluated. Tumors were considered amplified when they displayed six or more signals per tumor cell.
Statistics
A 2 test was used to compare the distribution of categoric or categorized variables between the two treatment arms, whereas the Mann-Whitney U test was used to evaluate median differences for continuous variables. The relations among VEGF-A, VEGFR2, HER2, and clinicopathologic data were measured by Pearson correlation. The statistic was used as a measure of agreement.
For survival analyses, all calculations were done according to the intention-to-treat rule. RFS considered distant, local, and regional metastasis as a primary event as well as breast cancer death, whereas contralateral breast cancer was not included. The Kaplan-Meier method was used to estimate RFS and overall survival, and the log-rank test was used to compare survival in different strata. The Cox proportional hazards model was used for estimation of HRs in both univariate and multivariate analysis. All P values corresponded to two-sided tests and values less than .05 were considered significant. The statistics packages Stata 8.1 (STATA Corp, College Station, TX) and Statistical Package for the Social Sciences (SPSS version 11.0; SPSS Inc, Chicago, IL) were used for data analysis.
RESULTS
Patient and Tumor Characteristics
Tumor blocks were available in 500 of 564 patients (89%). Ten percent of the missing blocks were lost during a fire accident in one pathology department. Patient and tumor characteristics in relation to treatment arm in the original study and the present study were well balanced and the results of the immunohistochemical analysis for HER2, VEGF-A, and VEGFR2 as well as HER2 amplification by fluorescent in situ hybridization are presented in Table 1. In summary, the distribution of patient and tumor characteristics between tamoxifen-treated and -untreated patients was similar in the original and the present study, verifying the absence of selection bias in the randomization process as well as in the selection of the present study cohort.
Hormone receptor status was available for 97% of the tumors in this study and the distribution in relation to the patient treatment arm is listed in Table 1; 428 of 500 tumor samples were assessable for HER2 status by immunohistochemistry, and because of insufficient tumor material within the array biopsies and fixation artifacts, it was not possible to evaluate 127 of 500 tumors (26%) for HER2 amplification. The proportion of tumors with HER2 grade 3+ was 15%, and 13% of the tumors were HER2 amplified. Only 8% of the ER-positive tumors were HER2 grade 3+ or HER2 amplified. Among HER2 grade 2+ tumors, only one tumor was amplified, whereas there was a strong correlation between HER2 grade 3+ and HER2 gene amplification (r = 0.82; P < .001) with a value of 0.84.
Table 2 lists the relationship among VEGF-A, VEGFR2, and HER2, and clinical and pathologic characteristics; a strong correlation is identified between VEGF-A and VEGFR2. In addition, there was a link among VEGF-A and VEGFR2 and Nottingham histologic grade (NHG) 3, and VEGFR2 was also associated with ER negativity. HER2 grade 3 and HER2 gene amplification were not associated with VEGF-A or VEGFR2, but correlated strongly with NHG 3, ER negativity, PR negativity, and decreasing age.
Tamoxifen Treatment Prediction
In the survival analyses of tamoxifen response, patients with ER-positive tumors as well as ER-positive and PR-positive tumors were considered separately to optimally validate the predictive markers for tamoxifen response in two slightly different subgroups of hormone receptor–positive tumors. The group of patients with ER-positive and PR-negative tumors only constituted 24 patients and did not allow any additional subgroup analyses. VEGF-A, VEGFR2, and HER2 were dichotomized (grade 3 v 0 to 2) because the tamoxifen effects in terms of RFS were similar for patients with grade 0 to 2 tumors, but not for patients with grade 3 tumors.
VEGF-A. Tamoxifen significantly increased RFS in patients with ER-positive and VEGF-A grade 0 to 2 tumors (HR, 0.59; 95% CI, 0.39 to 0.90; P = .01), whereas there was no significant tamoxifen effect in VEGF grade 3 tumors (HR, 0.86; 95% CI, 0.39 to 1.88; P = .7), as illustrated in Figures 1A and 1B. When this potential difference in treatment effect was analyzed in relation to VEGF using a Cox proportional hazards model including VEGF status (±), tamoxifen treatment (±), and a treatment-interaction variable [VEGF status (±) x tamoxifen treatment (±)], there was no statistical difference in treatment effect in terms of RFS (P value for interaction variable, 0.8). Similar results were obtained when the model was adjusted for age, tumor size, node status, and NHG (data not shown). In addition, when ER-positive and PR-positive tumors were analyzed, there was no significant difference in tamoxifen response in relation to VEGF-A (Table 3).
VEGFR2. A clear beneficial effect of tamoxifen treatment was identified in patients with ER-positive and grade 0 to 2 VEGFR2 tumors in terms of RFS (HR, 0.53; 95% CI, 0.36 to 0.78; P = .001). This was in contrast to patients with VEGFR2 grade 3 tumors, in whom a positive effect of the tamoxifen treatment evidently is lacking (HR, 2.44; 95% CI, 0.61 to 9.77; P = .2), as illustrated by Kaplan-Meier estimates (Figs 2A and 2B), and instead illustrates a trend toward an adverse effect of the treatment. By using a Cox model as described in our study, the P value for the unadjusted interaction variable between treatment and VEGFR2 status was .04 for RFS, supporting a statistically significant difference in tamoxifen response in relation to VEGFR2. A similar and statistically significant difference in treatment response was also observed for patients with ER-positive and PR-positive tumors (Table 3). In addition, when the term of interaction was adjusted for conventional prognostic markers as listed in the preceding paragraph, the test was significant for both patients with ER-positive tumors (P = .049) and patients with ER-positive and PR-positive tumors (P = .042).
HER2. There was no difference in tamoxifen response in relation to HER2 status in patients with ER-positive tumors, as illustrated in Figures 3A and 3B and Table 3. Similar findings were observed for patients with ER-positive and PR-positive tumors (Table 3). When the treatment effect was analyzed using a multivariate interaction model, there was no statistically significant difference in tamoxifen treatment efficiency in relation to HER2 protein status.
HER2 gene amplification. Patients with ER-positive and HER2-nonamplified tumors did not show improved RFS (HR, 0.73; 95% CI, 0.47 to 1.12; P = .14) from 2 years of tamoxifen, nor did patients with HER2-amplified tumors (HR, 0.21; 95% CI, 0.03 to 1.67; P = .14). Furthermore, there was no significant difference in tamoxifen effect explored by a Cox model (P = .2). For patients with ER-positive and PR-positive tumors, similar results were achieved (data not shown).
Clinical Outcome in Untreated Patients in Relation to VEGF-A, VEGFR2, and HER2
The prognostic impact of the investigated markers was analyzed by both univariate and multivariate analyses in terms of RFS in the subgroup of untreated patients. Neither VEGF-A, VEGFR2, HER2, or HER2 amplification yielded any prognostic information, as listed in Table 4. However, when we stratified for ER status, VEGFR2 (3 v 0 to 2) was a significant prognostic factor (HR, 0.23; 95% CI, 0.07 to 0.74; P = .01) for patients with ER-positive tumors using univariate analysis, as indicated in Figures 2A and 2B. This finding was nevertheless not significant in multivariate analysis (data not shown).
DISCUSSION
The aim of this study was to explore tamoxifen response in premenopausal patients with hormone receptor–positive tumors in relation to tumor-specific expression of VEGF-A, VEGFR2, and HER2 in a controlled, randomized trial. In this trial we identified a beneficial effect from 2 years of tamoxifen treatment in terms of RFS in patients with hormone receptor–positive tumors, although no significant interaction was noted between ER status and tamoxifen treatment, in contrast to the significant interaction between PR status and treatment. This finding is in line with previous reports, indicating that PR status is especially useful as a predictor of tamoxifen treatment response.7
VEGFR2 has not been explored previously in relation to breast cancer prognosis or treatment prediction in a randomized study, but was, in addition to hormone receptor status, the only marker in this study with significant predictive value using multivariate analysis of RFS. Interestingly, this group of patients did not benefit from adjuvant tamoxifen and a significant interaction was noted between tamoxifen treatment and VEGFR2 status in patients with ER-positive tumors, as well as in patients with ER- positive and PR-positive tumors. Tamoxifen even seemed to have a slightly adverse effect in patients with ER-positive VEGFR2 grade 3 tumors (Fig 2); it is possible that it acted as an estrogen agonist in these tumors, although this descriptive and not statistically significant finding needs to be substantiated by experimental models and other clinical studies.
VEGFR2 is the most important receptor for VEGF-A; it stimulates proliferation and migration via its tyrosine kinase activity.19 VEGFR2 is mainly located on endothelial cells, but has also been identified on tumor cells both in vivo and in vitro, implicating autocrine activity for VEGF-A in addition to its proangiogenic properties.20-23 The intracellular signaling pathways for VEGFRs in malignant cells have been studied20,25,26 and seem to involve mitogen-activated protein kinase activity, extracellular regulated kinase 1/2, and PI3 in T47D breast cancer cell lines.20 These pathways may be similar, in part, to the signaling cascades identified for EGFR and HER2, resulting in a ligand-independent activation of the ER.2 VEGFR2 added no prognostic information in untreated patients in general, but was a prognostic factor for ER-positive tumors in univariate analysis of RFS. The finding of a biologic marker identified as a negative predictive factor for adjuvant tamoxifen treatment and a positive prognostic factor has been described for the coactivator A1B1.5 Regarding A1B1, no biologic explanation for this observation can be given without experimental studies, and additional results from clinical trials are needed to confirm the data.
This is the first report from a randomized trial in which the expression of VEGF-A was explored in relation to tamoxifen response in patients with ER-positive tumors. Experimental data have shown that both estrogen and tamoxifen can stimulate VEGF secretion, theoretically counteracting the tamoxifen effect.27 We observed a significant tamoxifen response in patients with grade 0 to 2 expression of tumor-specific VEGF-A, which was not observed in tumors with grade 3 expression of VEGF-A. Nevertheless, these results were not significant in multivariate interaction analyses, suggesting that VEGF status was not a predictive marker for tamoxifen response. This finding is in contrast to results from nonrandomized trials of adjuvant tamoxifen treatment, and could possibly be explained by the study design or the selection of patients, in addition to methodologic differences.14-16
Only a few randomized studies have examined HER2 status and tamoxifen response, with conflicting results.3,11,13 In a study of premenopausal patients treated with both oophorectomy and 5 years of tamoxifen, HER2 status was not associated with treatment response, and the beneficial effect of adjuvant treatment in HER2-overexpressing patients was even greater than in HER2-negative patients.11 However, the impact of HER2 status in relation to tamoxifen response could not be clarified, given that patients received estrogen-suppressing therapy as well. This study showed no significant difference in tamoxifen response in relation to HER2 protein or gene amplification status, supporting the view that HER2 overexpression is not a predictor for tamoxifen response in premenopausal breast cancer.
We found no significant difference in clinicopathologic variables between untreated and tamoxifen-treated patients in the initial study and in the subgroup for which additional survival analyses were performed within this study using the primary end point stated in the initial protocol. This observation is critical because even patients with hormone receptor–negative tumors entered the initial study with risk of a selection bias in subsequent subgroup analyses. By including patients with hormone receptor–negative tumors, we were also able to demonstrate that these patients did not benefit from tamoxifen treatment, further supporting the assertion that no selection bias exists between untreated and tamoxifen-treated patients. To assess and validate the difference in treatment response in the various subgroups of breast cancer defined by VEGF, VEGFR2, and HER2 expression, we used multivariate interaction analysis. This is purported to be the optimal statistical approach when testing for different treatment effects in subgroups of patients.28 A problem when validating treatment response in small subgroups of breast cancer patients is the relatively low statistical power obtained, potentially disqualifying certain candidate genes. This might have affected the analyses of HER2, where the number of patients at risk with hormone receptor–positive and HER2 3+ tumors was low, due to the inverse relationship between HER2 status and hormone receptor status.13 Nevertheless, our results clearly suggest that VEGFR2 status is a significant predictor of tamoxifen response, which is also consistent with a recent report on postmenopausal breast cancer using a similar approach as in this study.29 This finding needs to be reproduced in larger trials before we can make any definitive clinical conclusions and potential clinical recommendations.
In summary, we have demonstrated that expression of tumor-specific VEGFR2 is a new and specific predictive marker for tamoxifen response in breast cancer patients, in contrast to VEGF-A and HER2, which were not significantly linked to tamoxifen response. Although this study explored the effect of 2 years of tamoxifen treatment, there are no data to support that length of treatment affects the mechanisms involved in tamoxifen resistance. Despite the frequent use of tamoxifen in combination with other adjuvant therapies, a potentially impaired tamoxifen response in defined subgroups of hormone receptor–positive breast cancer will most certainly affect the treatment efficiency of chemoendocrine therapy, given that the tamoxifen effect is known to be additive to adjuvant chemotherapy.1 VEGFR2 inhibitors may further affect the VEGFR2 signaling, potentially restoring the tamoxifen response, and monoclonal antibodies directed toward the membranous portion of the receptor as well as tyrosine kinase inhibitors have been developed and are currently being tested in clinical trials. Combination strategies with tamoxifen and inhibitors of EGFR and/or HER2 have shown promise as a therapeutic alternative for combating endocrine therapy resistance. The same concept theoretically could be applied to patients with hormone receptor–and VEGFR2-expressing tumors. In addition, compounds initially developed for antiangiogenic therapy may serve as adjuncts to tamoxifen, representing a new and potentially much more effective combination therapy approach to hormone-resistant breast cancer.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
The authors thank Elise Nilsson and Monica Haglund for their excellent technical assistance.
NOTES
Supported by grants from the Swedish Cancer Society, Malm University Hospital funds, Gunnar Nilssons Cancerstiftelse, the Zoega Fund, the Gorthon Fund, Kristianstad School of Higher Education Foundation, and a project grant from Swegene/Wallenberg Consortium North.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Early Breast Cancer Trialists' Collaborative Group: Tamoxifen for early breast cancer: An overview of the randomised trials. Lancet 351:1451-1467, 1998
Ali S, Coombes RC: Endocrine-responsive breast cancer and strategies for combating resistance. Nat Rev Cancer 2:101-112, 2002
De Placido S, De Laurentiis M, Carlomagno C, et al: Twenty-year results of the Naples GUN Randomized Trial: Predictive factors of adjuvant tamoxifen efficacy in early breast cancer. Clin Cancer Res 9:1039-1046, 2003
Nicholson RI, Hutcheson IR, Knowlden JM, et al: Nonendocrine pathways and endocrine resistance: Observations with antiestrogens and signal transduction inhibitors in combination. Clin Cancer Res 10:346S-354S, 2004
Osborne CK, Bardou V, Hopp TA, et al: Role of the estrogen receptor coactivator AIB1 (SRC-3) and HER-2/neu in tamoxifen resistance in breast cancer. J Natl Cancer Inst 95:353-361, 2003
Schiff R, Massarweh SA, Shou J, et al: Cross-talk between estrogen receptor and growth factor pathways as a molecular target for overcoming endocrine resistance. Clin Cancer Res 10:331S-336S, 2000
Bardou VJ, Arpino G, Elledge RM, et al: Progesterone receptor status significantly improves outcome prediction over estrogen receptor status alone for adjuvant endocrine therapy in two large breast cancer databases. J Clin Oncol 21:1973-1979, 2003
Cui X, Zhang P, Deng W, et al: Insulin-like growth factor-I inhibits progesterone receptor expression in breast cancer cells via the phosphatidylinositol 3-kinase/Akt/mammalian target of rapamycin pathway: Progesterone receptor as a potential indicator of growth factor activity in breast cancer. Mol Endocrinol 17:575-588, 2003
Benz CC, Scott GK, Sarup JC, et al: Estrogen-dependent, tamoxifen-resistant tumorigenic growth of MCF-7 cells transfected with HER2/neu. Breast Cancer Res Treat 24:85-95, 1993
Borg A, Baldetorp B, Ferno M, et al: ERBB2 amplification is associated with tamoxifen resistance in steroid-receptor positive breast cancer. Cancer Lett 81:137-144, 1994
Love RR, Duc NB, Havighurst TC, et al: Her-2/neu overexpression and response to oophorectomy plus tamoxifen adjuvant therapy in estrogen receptor-positive premenopausal women with operable breast cancer. J Clin Oncol 21:453-457, 2003
Stal O, Borg A, Ferno M, et al: ErbB2 status and the benefit from two or five years of adjuvant tamoxifen in postmenopausal early stage breast cancer. Ann Oncol 11:1545-1550, 2000
Knoop AS, Bentzen SM, Nielsen MM, et al: Value of epidermal growth factor receptor, HER2, p53, and steroid receptors in predicting the efficacy of tamoxifen in high-risk postmenopausal breast cancer patients. J Clin Oncol 19:3376-3384, 2001
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
Linderholm B, Grankvist K, Wilking N, et al: Correlation of vascular endothelial growth factor content with recurrences, survival, and first relapse site in primary node-positive breast carcinoma after adjuvant treatment. J Clin Oncol 18:1423-1431, 2000
Coradini D, Biganzoli E, Pellizzaro C, et al: Vascular endothelial growth factor in node-positive breast cancer patients treated with adjuvant tamoxifen. Br J Cancer 89:268-270, 2003
Berns EM, Klijn JG, Look MP, et al: Combined vascular endothelial growth factor and TP53 status predicts poor response to tamoxifen therapy in estrogen receptor-positive advanced breast cancer. Clin Cancer Res 9:1253-1258, 2003
Foekens JA, Peters HA, Grebenchtchikov N, et al: High tumor levels of vascular endothelial growth factor predict poor response to systemic therapy in advanced breast cancer. Cancer Res 61:5407-5414, 2001
Veikkola T, Karkkainen M, Claesson-Welsh L, et al: Regulation of angiogenesis via vascular endothelial growth factor receptors. Cancer Res 60:203-212, 2000
Miralem T, Steinberg R, Price D, et al: VEGF(165) requires extracellular matrix components to induce mitogenic effects and migratory response in breast cancer cells. Oncogene 20:5511-5524, 2001
Speirs V, Atkin SL: Production of VEGF and expression of the VEGF receptors Flt-1 and KDR in primary cultures of epithelial and stromal cells derived from breast tumours. Br J Cancer 80:898-903, 1999
Xie B, Tam NN, Tsao SW, et al: Co-expression of vascular endothelial growth factor (VEGF) and its receptors (flk-1 and flt-1) in hormone-induced mammary cancer in the Noble rat. Br J Cancer 81:1335-1343, 1999
Ryden L, Linderholm B, Nielsen NH, et al: Tumor specific VEGF-A and VEGFR2/KDR protein are co-expressed in breast cancer. Breast Cancer Res Treat 82:147-154, 2003
Ryden L, Jonsson PE, Chebil G, et al: Two years of adjuvant tamoxifen in premenopausal patients with breast cancer: A randomised, controlled trial with long-term follow-up. Eur J Cancer 41:256-264, 2005
von Marschall Z, Cramer T, Hocker M, et al: De novo expression of vascular endothelial growth factor in human pancreatic cancer: Evidence for an autocrine mitogenic loop. Gastroenterology 119:1358-1372, 2000
Catalano A, Caprari P, Rodilossi S, et al: Cross-talk between vascular endothelial growth factor and semaphorin-3A pathway in the regulation of normal and malignant mesothelial cell proliferation. FASEB J 18:358-360, 2004
Ruohola JK, Valve EM, Karkkainen MJ, et al: Vascular endothelial growth factors are differentially regulated by steroid hormones and antiestrogens in breast cancer cells. Mol Cell Endocrinol 149:29-40, 1999
Rothwell PM: Treating individuals 2: Subgroup analysis in randomised controlled trials—Importance, indications, and interpretation. Lancet 365:176-186, 2005
Ryden L, Stendahl M, Jonsson H, et al: Tumor-specific VEGF-A and VEGFR2 in postmenopausal breast cancer patients with long-term follow-up: Implication of a link between VEGF pathway and tamoxifen response. Breast Cancer Res Treat 89:135-143, 2005(Lisa Ryden, Karin Jirstrm)
Department of Laboratory Medicine, Pathology, University Hospital, Malm
Department of Oncology, University Hospital, Lund
Department of Oncology, University Hospital, Linkping
Department of Pathology and Cytology, Kalmar County Hospital, Kalmar
South Swedish and Southeast Swedish Breast Cancer Groups, Sweden
ABSTRACT
PURPOSE: Vascular endothelial growth factor A (VEGF-A) and vascular endothelial growth factor receptor 2 (VEGFR2) are often coexpressed in breast cancer, and potentially affect cellular pathways and key proteins such as the estrogen receptor (ER) targeted by endocrine treatment. We therefore explored the association between adjuvant tamoxifen treatment in breast cancer and expression of VEGF-A and VEGFR2, as well as human epidermal growth factor receptor 2 (HER2), which represents a candidate gene product involved in tamoxifen resistance.
PATIENTS AND METHODS: Immunohistochemical expression of tumor-specific VEGF-A, VEGFR2, and HER2 was evaluated in tumor specimens from premenopausal breast cancer patients randomly assigned to 2 years of tamoxifen or no treatment (n = 564), with 14 years of follow-up. Hormone receptor status was determined in 96% of the tumors.
RESULTS: VEGF-A, VEGFR2, and HER2 were assessable in 460, 472, and 428 of the tumors, respectively. In patients with ER–positive and VEGFR2-low tumors, adjuvant tamoxifen significantly increased recurrence-free survival (RFS; [HR] hazard ratio for RFS, 0.53; P = .001). In contrast, tamoxifen treatment had no effect in patients with VEGFR2-high tumors (HR for RFS, 2.44; P = .2). When multivariate interaction analyses were used, this difference in treatment efficacy relative to VEGFR2 expression status was statistically significant for both ER-positive (P = .04) plus ER-positive and progesterone receptor–positive tumors. We found no significant difference in tamoxifen treatment effects in relation to VEGF-A or HER2 status.
CONCLUSION: Tumor-specific expression of VEGFR2 was associated with an impaired tamoxifen effect in hormone receptor–positive premenopausal breast cancer. Tamoxifen in combination with VEGFR2 inhibitors might be a novel treatment approach for VEGFR2-expressing breast cancer, and such a treatment might restore the tamoxifen response.
INTRODUCTION
Adjuvant tamoxifen treatment for 5 years improves 10-year survival in hormone receptor–positive breast cancer and reduces the recurrence rate by 47% and mortality by 27%.1,2 Even though the beneficial effect of tamoxifen is restricted to patients with estrogen receptor (ER) –positive and/or progesterone receptor (PR) –positive tumors, a large fraction of hormone receptor–positive tumors do not respond as expected to the treatment.2 Resistance to endocrine therapy with tamoxifen is an immense clinical problem and researchers have therefore actively pursued mechanistic studies of tamoxifen resistance and tested predictive markers other than hormone receptor status.2-6 Growth factors with tyrosine kinase activity can activate ER without ligand binding2,4-6 and the cross-talk between ER and growth factor receptors seems to be an essential part of tamoxifen resistance. PR can serve as an alternative predictor of the effect of tamoxifen treatment7 because low expression of PR may be an indicator of activated growth factor receptors.8 Overexpression of the human epidermal growth factor receptor 2 (HER2) is also associated with resistance to tamoxifen in in vitro experiments,4,7,9 and clinical data suggest that overexpression of HER2 may be a detriment to the beneficial effect of tamoxifen, although evidence from randomized studies is still limited.3,5,10-13
Researchers have addressed vascular endothelial growth factor A (VEGF-A) as a potential predictor of response to tamoxifen treatment in early and advanced breast cancer.14-18 VEGF-A exerts its effect via receptors with tyrosine kinase activity (vascular endothelial growth factor receptor 1 [VEGFR1] and VEGFR2), activating intracellular pathways similar to those of other tyrosine kinase receptors.19 VEGFR2 has been linked to proliferative features via the mitogen-activated protein kinase pathway, whereas VEGFR1 does not affect proliferation.19 The VEGF receptors are mainly located on the surface of endothelial cells, but have also been identified on breast cancer cells, implicating autocrine and paracrine roles for VEGF-A in addition to its proangiogenic property.20-23
We recently reported a beneficial effect from 2 years of tamoxifen treatment compared with no adjuvant treatment in premenopausal breast cancer patients who were included in a randomized treatment trial in which fewer than 2% of patients received additional adjuvant chemotherapy.24 A tissue microarray of the breast cancer samples used in this study has now been constructed, allowing efficient studies of tamoxifen response in subgroups of breast cancer defined by putative predictive markers. Currently, standard adjuvant therapy for hormone-responsive breast cancer is 5 years of tamoxifen, administered either alone or combined with chemotherapy. Nevertheless, a cohort of breast cancer patients included in a randomized 2-year trial of adjuvant tamoxifen, which showed a clear beneficial effect of tamoxifen, offers a simple and favorable study design for scrutinizing predictors of tamoxifen response without interference from prognostic considerations and effects of chemotherapy.
The aim of this study was to analyze the association of tumor-specific expression of VEGF-A and VEGFR2 with tamoxifen response in hormone receptor–positive primary breast cancer in a controlled, randomized trial of tamoxifen versus control. The predictive information of another growth factor receptor that is commonly overexpressed in breast cancer, HER2, was further delineated to validate this candidate marker for tamoxifen response.
PATIENTS AND METHODS
Study Design
Premenopausal patients or patients younger than 50 years with stage II (pT2, N0, M0; pT1, N1, M0; and pT2, N1, M0) invasive breast cancer were enrolled onto a multicenter clinical trial between 1986 and 1991 and were randomly assigned to control (n = 288) or 2 years of tamoxifen (n = 276). Random assignment was performed by the Regional Oncological Center and informed consent was registered for all included patients. The study was approved by the ethical committees at Lund and Linkping Universities. The primary end point was recurrence-free survival (RFS) and the treatment effect should be examined in relation to hormone receptor status, as stated in the study protocol. All patients underwent curative surgery either by modified radical mastectomy or by breast-conserving surgery with axillary lymph node dissection (levels one and two). After breast-conserving surgery, radiotherapy (50 Gy) was administered to the breast, and in patients with axillary lymph node metastases, locoregional radiotherapy was delivered. Additional adjuvant therapy was administered as polychemotherapy (n = 8) and goserelin (n = 1) to nine patients (< 2%), who were evenly distributed between the two treatment arms. Patients were included irrespective of hormone receptor status, although ER and PR status were determined in 80% of the patients at the time of surgery by cytosol-based methods. Patients' and tumor characteristics in relation to treatment arm in the original study are presented in Table 1.
The patients were observed for 5 years with annual mammogram and physical examination and then within the national program by screening mammogram every 18 months. The median duration of follow-up for patients without a breast cancer event was 14 years (range, 0 to 17 years). The median follow-up time was comparable in the two treatment arms. The study has been included in the Oxford meta-analysis1 and the main results have been presented in detail elsewhere.24 Briefly, two years of adjuvant tamoxifen increased RFS in all included patients (hazard ratio [HR], 0.77; 95% CI, 0.60 to 098; P = .03). Hormone receptor status could be determined in 541 patients (96%) prospectively by using cytosol-based methods and retrospectively by immunohistochemistry as described.24 The beneficial effect by 2 years of adjuvant tamoxifen in terms of RFS was restricted to 385 patients with ER-positive and/or PR-positive tumors, whereas no effect was recorded in patients with ER-negative and PR-negative tumors. The interaction between tamoxifen treatment and hormone receptor status in terms of RFS was explored with Cox multivariate analyses; a significant interaction between PR status and tamoxifen treatment (P = .015), as well as for the combination of ER and PR positivity (P = .031), was demonstrated. In contrast, ER status alone was not a significant predictor of tamoxifen response (P = .19), and the combination of ER and/or PR status was not a predictor of tamoxifen response (P = .11).
Hormone Receptor Status
Hormone receptor status was determined by cytosol-based methods prospectively and by immunohistochemical methods retrospectively, as described previously.24 Comparison between the two techniques of determining hormone receptor status resulted in 86% concordance for ER status (, 0.70) and 88% concordance for PR status (, 0.73), justifying that a combination of the methods could be used.
Tumor Tissue Microarray Preparation
All tumor specimens were obtained at the time of surgery before administration of adjuvant therapy. Formalin-fixed, paraffin-embedded tumor blocks were obtained from the primary breast tumors. Clearly defined tumor areas were selected and two biopsies, 0.6 mm in diameter, were obtained from each donor block and mounted in a recipient block using a tissue microarray (TMA) instrument according to the manufacturer's instructions (Beecher Instruments, MD).
Immunohistochemistry
VEGF-A and VEGFR2. Four-micrometer TMA sections were deparaffinized, rehydrated, and treated in a microwave for 5 + 5 minutes in a citrate buffer (pH 6.0) before they were processed in an automatic immunohistochemistry staining machine according to standard procedures (TechMate500; DakoCytomation, Glostrup, Denmark) using a polyclonal VEGF (A-20) antibody diluted 1:400 (Santa Cruz Biotechnology Inc, Santa Cruz, CA) and a monoclonal VEGFR2 antibody diluted 1:1000 (Santa Cruz Biotechnology Inc, Santa Cruz, CA). Normal human kidneys were used as positive controls for VEGF-A and human aortic endothelium was used as positive control for VEGFR2. VEGF-A has been validated previously using an enzyme-linked immunosorbent assay–based method.25 The cytoplasmic staining intensity was evaluated semiquantitatively using a classification from 0 to 3, representing lack of staining (grade 0), low staining intensity (grade 1), intermediate staining intensity (grade 2), and high staining intensity (grade 3). Only invasive tumor cells were evaluated. In survival analyses of patients with tumors where VEGF-A and VEGFR2 were determined, high staining intensity (grade 3) was compared with grades 0 to 2 because grade 0 to 2 produced overlapping survival curves, whereas grade 3 differed substantially. The TMA was examined by two investigators with blinded clinical data and the concordance between the investigators was 89% for VEGF-A and 93% for VEGFR2 when participants were divided into four groups. Divergent results were re-examined followed by a conclusive decision.
HER2. Protein expression of HER2 was determined by immunohistochemistry using the Ventana Benchmark System (Ventana Medical Systems Inc, Tucson, AZ) with a prediluted antibody (Pathway CB-11, 760-2694; Ventana Medical Systems Inc, Tucson, AZ). HER2 expression was evaluated semiquantitatively according to a standard protocol (HercepTest; DakoCytomation). The protocol categorizes tumors into four groups: grade 0, lack of staining in all tumor cells or membrane staining in less than 10% of the tumor cells; grade 1+, weak, not circumferential membrane staining in more than 10% of the tumor cells; grade 2+, intermediate, circumferential membrane staining in more than 10% of the tumor cells; and grade 3+, intense and circumferential staining in more than 10% of the tumor cells.
HER2 gene amplification. HER2 gene amplification was determined by fluorescent in situ hybridization using an automated staining procedure according to the manufacturer's recommendations (Ventana Medical Systems). Briefly, unstained, formalin-fixed, paraffin-embedded tumor sections were deparaffinized, pretreated, and denatured before incubation overnight with the hybridization probe. The sections were counterstained and evaluated for HER2 gene copy number using a fluorescence microscope at a magnification of x400. All tumor cells within the biopsies in the TMA were evaluated. Tumors were considered amplified when they displayed six or more signals per tumor cell.
Statistics
A 2 test was used to compare the distribution of categoric or categorized variables between the two treatment arms, whereas the Mann-Whitney U test was used to evaluate median differences for continuous variables. The relations among VEGF-A, VEGFR2, HER2, and clinicopathologic data were measured by Pearson correlation. The statistic was used as a measure of agreement.
For survival analyses, all calculations were done according to the intention-to-treat rule. RFS considered distant, local, and regional metastasis as a primary event as well as breast cancer death, whereas contralateral breast cancer was not included. The Kaplan-Meier method was used to estimate RFS and overall survival, and the log-rank test was used to compare survival in different strata. The Cox proportional hazards model was used for estimation of HRs in both univariate and multivariate analysis. All P values corresponded to two-sided tests and values less than .05 were considered significant. The statistics packages Stata 8.1 (STATA Corp, College Station, TX) and Statistical Package for the Social Sciences (SPSS version 11.0; SPSS Inc, Chicago, IL) were used for data analysis.
RESULTS
Patient and Tumor Characteristics
Tumor blocks were available in 500 of 564 patients (89%). Ten percent of the missing blocks were lost during a fire accident in one pathology department. Patient and tumor characteristics in relation to treatment arm in the original study and the present study were well balanced and the results of the immunohistochemical analysis for HER2, VEGF-A, and VEGFR2 as well as HER2 amplification by fluorescent in situ hybridization are presented in Table 1. In summary, the distribution of patient and tumor characteristics between tamoxifen-treated and -untreated patients was similar in the original and the present study, verifying the absence of selection bias in the randomization process as well as in the selection of the present study cohort.
Hormone receptor status was available for 97% of the tumors in this study and the distribution in relation to the patient treatment arm is listed in Table 1; 428 of 500 tumor samples were assessable for HER2 status by immunohistochemistry, and because of insufficient tumor material within the array biopsies and fixation artifacts, it was not possible to evaluate 127 of 500 tumors (26%) for HER2 amplification. The proportion of tumors with HER2 grade 3+ was 15%, and 13% of the tumors were HER2 amplified. Only 8% of the ER-positive tumors were HER2 grade 3+ or HER2 amplified. Among HER2 grade 2+ tumors, only one tumor was amplified, whereas there was a strong correlation between HER2 grade 3+ and HER2 gene amplification (r = 0.82; P < .001) with a value of 0.84.
Table 2 lists the relationship among VEGF-A, VEGFR2, and HER2, and clinical and pathologic characteristics; a strong correlation is identified between VEGF-A and VEGFR2. In addition, there was a link among VEGF-A and VEGFR2 and Nottingham histologic grade (NHG) 3, and VEGFR2 was also associated with ER negativity. HER2 grade 3 and HER2 gene amplification were not associated with VEGF-A or VEGFR2, but correlated strongly with NHG 3, ER negativity, PR negativity, and decreasing age.
Tamoxifen Treatment Prediction
In the survival analyses of tamoxifen response, patients with ER-positive tumors as well as ER-positive and PR-positive tumors were considered separately to optimally validate the predictive markers for tamoxifen response in two slightly different subgroups of hormone receptor–positive tumors. The group of patients with ER-positive and PR-negative tumors only constituted 24 patients and did not allow any additional subgroup analyses. VEGF-A, VEGFR2, and HER2 were dichotomized (grade 3 v 0 to 2) because the tamoxifen effects in terms of RFS were similar for patients with grade 0 to 2 tumors, but not for patients with grade 3 tumors.
VEGF-A. Tamoxifen significantly increased RFS in patients with ER-positive and VEGF-A grade 0 to 2 tumors (HR, 0.59; 95% CI, 0.39 to 0.90; P = .01), whereas there was no significant tamoxifen effect in VEGF grade 3 tumors (HR, 0.86; 95% CI, 0.39 to 1.88; P = .7), as illustrated in Figures 1A and 1B. When this potential difference in treatment effect was analyzed in relation to VEGF using a Cox proportional hazards model including VEGF status (±), tamoxifen treatment (±), and a treatment-interaction variable [VEGF status (±) x tamoxifen treatment (±)], there was no statistical difference in treatment effect in terms of RFS (P value for interaction variable, 0.8). Similar results were obtained when the model was adjusted for age, tumor size, node status, and NHG (data not shown). In addition, when ER-positive and PR-positive tumors were analyzed, there was no significant difference in tamoxifen response in relation to VEGF-A (Table 3).
VEGFR2. A clear beneficial effect of tamoxifen treatment was identified in patients with ER-positive and grade 0 to 2 VEGFR2 tumors in terms of RFS (HR, 0.53; 95% CI, 0.36 to 0.78; P = .001). This was in contrast to patients with VEGFR2 grade 3 tumors, in whom a positive effect of the tamoxifen treatment evidently is lacking (HR, 2.44; 95% CI, 0.61 to 9.77; P = .2), as illustrated by Kaplan-Meier estimates (Figs 2A and 2B), and instead illustrates a trend toward an adverse effect of the treatment. By using a Cox model as described in our study, the P value for the unadjusted interaction variable between treatment and VEGFR2 status was .04 for RFS, supporting a statistically significant difference in tamoxifen response in relation to VEGFR2. A similar and statistically significant difference in treatment response was also observed for patients with ER-positive and PR-positive tumors (Table 3). In addition, when the term of interaction was adjusted for conventional prognostic markers as listed in the preceding paragraph, the test was significant for both patients with ER-positive tumors (P = .049) and patients with ER-positive and PR-positive tumors (P = .042).
HER2. There was no difference in tamoxifen response in relation to HER2 status in patients with ER-positive tumors, as illustrated in Figures 3A and 3B and Table 3. Similar findings were observed for patients with ER-positive and PR-positive tumors (Table 3). When the treatment effect was analyzed using a multivariate interaction model, there was no statistically significant difference in tamoxifen treatment efficiency in relation to HER2 protein status.
HER2 gene amplification. Patients with ER-positive and HER2-nonamplified tumors did not show improved RFS (HR, 0.73; 95% CI, 0.47 to 1.12; P = .14) from 2 years of tamoxifen, nor did patients with HER2-amplified tumors (HR, 0.21; 95% CI, 0.03 to 1.67; P = .14). Furthermore, there was no significant difference in tamoxifen effect explored by a Cox model (P = .2). For patients with ER-positive and PR-positive tumors, similar results were achieved (data not shown).
Clinical Outcome in Untreated Patients in Relation to VEGF-A, VEGFR2, and HER2
The prognostic impact of the investigated markers was analyzed by both univariate and multivariate analyses in terms of RFS in the subgroup of untreated patients. Neither VEGF-A, VEGFR2, HER2, or HER2 amplification yielded any prognostic information, as listed in Table 4. However, when we stratified for ER status, VEGFR2 (3 v 0 to 2) was a significant prognostic factor (HR, 0.23; 95% CI, 0.07 to 0.74; P = .01) for patients with ER-positive tumors using univariate analysis, as indicated in Figures 2A and 2B. This finding was nevertheless not significant in multivariate analysis (data not shown).
DISCUSSION
The aim of this study was to explore tamoxifen response in premenopausal patients with hormone receptor–positive tumors in relation to tumor-specific expression of VEGF-A, VEGFR2, and HER2 in a controlled, randomized trial. In this trial we identified a beneficial effect from 2 years of tamoxifen treatment in terms of RFS in patients with hormone receptor–positive tumors, although no significant interaction was noted between ER status and tamoxifen treatment, in contrast to the significant interaction between PR status and treatment. This finding is in line with previous reports, indicating that PR status is especially useful as a predictor of tamoxifen treatment response.7
VEGFR2 has not been explored previously in relation to breast cancer prognosis or treatment prediction in a randomized study, but was, in addition to hormone receptor status, the only marker in this study with significant predictive value using multivariate analysis of RFS. Interestingly, this group of patients did not benefit from adjuvant tamoxifen and a significant interaction was noted between tamoxifen treatment and VEGFR2 status in patients with ER-positive tumors, as well as in patients with ER- positive and PR-positive tumors. Tamoxifen even seemed to have a slightly adverse effect in patients with ER-positive VEGFR2 grade 3 tumors (Fig 2); it is possible that it acted as an estrogen agonist in these tumors, although this descriptive and not statistically significant finding needs to be substantiated by experimental models and other clinical studies.
VEGFR2 is the most important receptor for VEGF-A; it stimulates proliferation and migration via its tyrosine kinase activity.19 VEGFR2 is mainly located on endothelial cells, but has also been identified on tumor cells both in vivo and in vitro, implicating autocrine activity for VEGF-A in addition to its proangiogenic properties.20-23 The intracellular signaling pathways for VEGFRs in malignant cells have been studied20,25,26 and seem to involve mitogen-activated protein kinase activity, extracellular regulated kinase 1/2, and PI3 in T47D breast cancer cell lines.20 These pathways may be similar, in part, to the signaling cascades identified for EGFR and HER2, resulting in a ligand-independent activation of the ER.2 VEGFR2 added no prognostic information in untreated patients in general, but was a prognostic factor for ER-positive tumors in univariate analysis of RFS. The finding of a biologic marker identified as a negative predictive factor for adjuvant tamoxifen treatment and a positive prognostic factor has been described for the coactivator A1B1.5 Regarding A1B1, no biologic explanation for this observation can be given without experimental studies, and additional results from clinical trials are needed to confirm the data.
This is the first report from a randomized trial in which the expression of VEGF-A was explored in relation to tamoxifen response in patients with ER-positive tumors. Experimental data have shown that both estrogen and tamoxifen can stimulate VEGF secretion, theoretically counteracting the tamoxifen effect.27 We observed a significant tamoxifen response in patients with grade 0 to 2 expression of tumor-specific VEGF-A, which was not observed in tumors with grade 3 expression of VEGF-A. Nevertheless, these results were not significant in multivariate interaction analyses, suggesting that VEGF status was not a predictive marker for tamoxifen response. This finding is in contrast to results from nonrandomized trials of adjuvant tamoxifen treatment, and could possibly be explained by the study design or the selection of patients, in addition to methodologic differences.14-16
Only a few randomized studies have examined HER2 status and tamoxifen response, with conflicting results.3,11,13 In a study of premenopausal patients treated with both oophorectomy and 5 years of tamoxifen, HER2 status was not associated with treatment response, and the beneficial effect of adjuvant treatment in HER2-overexpressing patients was even greater than in HER2-negative patients.11 However, the impact of HER2 status in relation to tamoxifen response could not be clarified, given that patients received estrogen-suppressing therapy as well. This study showed no significant difference in tamoxifen response in relation to HER2 protein or gene amplification status, supporting the view that HER2 overexpression is not a predictor for tamoxifen response in premenopausal breast cancer.
We found no significant difference in clinicopathologic variables between untreated and tamoxifen-treated patients in the initial study and in the subgroup for which additional survival analyses were performed within this study using the primary end point stated in the initial protocol. This observation is critical because even patients with hormone receptor–negative tumors entered the initial study with risk of a selection bias in subsequent subgroup analyses. By including patients with hormone receptor–negative tumors, we were also able to demonstrate that these patients did not benefit from tamoxifen treatment, further supporting the assertion that no selection bias exists between untreated and tamoxifen-treated patients. To assess and validate the difference in treatment response in the various subgroups of breast cancer defined by VEGF, VEGFR2, and HER2 expression, we used multivariate interaction analysis. This is purported to be the optimal statistical approach when testing for different treatment effects in subgroups of patients.28 A problem when validating treatment response in small subgroups of breast cancer patients is the relatively low statistical power obtained, potentially disqualifying certain candidate genes. This might have affected the analyses of HER2, where the number of patients at risk with hormone receptor–positive and HER2 3+ tumors was low, due to the inverse relationship between HER2 status and hormone receptor status.13 Nevertheless, our results clearly suggest that VEGFR2 status is a significant predictor of tamoxifen response, which is also consistent with a recent report on postmenopausal breast cancer using a similar approach as in this study.29 This finding needs to be reproduced in larger trials before we can make any definitive clinical conclusions and potential clinical recommendations.
In summary, we have demonstrated that expression of tumor-specific VEGFR2 is a new and specific predictive marker for tamoxifen response in breast cancer patients, in contrast to VEGF-A and HER2, which were not significantly linked to tamoxifen response. Although this study explored the effect of 2 years of tamoxifen treatment, there are no data to support that length of treatment affects the mechanisms involved in tamoxifen resistance. Despite the frequent use of tamoxifen in combination with other adjuvant therapies, a potentially impaired tamoxifen response in defined subgroups of hormone receptor–positive breast cancer will most certainly affect the treatment efficiency of chemoendocrine therapy, given that the tamoxifen effect is known to be additive to adjuvant chemotherapy.1 VEGFR2 inhibitors may further affect the VEGFR2 signaling, potentially restoring the tamoxifen response, and monoclonal antibodies directed toward the membranous portion of the receptor as well as tyrosine kinase inhibitors have been developed and are currently being tested in clinical trials. Combination strategies with tamoxifen and inhibitors of EGFR and/or HER2 have shown promise as a therapeutic alternative for combating endocrine therapy resistance. The same concept theoretically could be applied to patients with hormone receptor–and VEGFR2-expressing tumors. In addition, compounds initially developed for antiangiogenic therapy may serve as adjuncts to tamoxifen, representing a new and potentially much more effective combination therapy approach to hormone-resistant breast cancer.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
The authors thank Elise Nilsson and Monica Haglund for their excellent technical assistance.
NOTES
Supported by grants from the Swedish Cancer Society, Malm University Hospital funds, Gunnar Nilssons Cancerstiftelse, the Zoega Fund, the Gorthon Fund, Kristianstad School of Higher Education Foundation, and a project grant from Swegene/Wallenberg Consortium North.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Early Breast Cancer Trialists' Collaborative Group: Tamoxifen for early breast cancer: An overview of the randomised trials. Lancet 351:1451-1467, 1998
Ali S, Coombes RC: Endocrine-responsive breast cancer and strategies for combating resistance. Nat Rev Cancer 2:101-112, 2002
De Placido S, De Laurentiis M, Carlomagno C, et al: Twenty-year results of the Naples GUN Randomized Trial: Predictive factors of adjuvant tamoxifen efficacy in early breast cancer. Clin Cancer Res 9:1039-1046, 2003
Nicholson RI, Hutcheson IR, Knowlden JM, et al: Nonendocrine pathways and endocrine resistance: Observations with antiestrogens and signal transduction inhibitors in combination. Clin Cancer Res 10:346S-354S, 2004
Osborne CK, Bardou V, Hopp TA, et al: Role of the estrogen receptor coactivator AIB1 (SRC-3) and HER-2/neu in tamoxifen resistance in breast cancer. J Natl Cancer Inst 95:353-361, 2003
Schiff R, Massarweh SA, Shou J, et al: Cross-talk between estrogen receptor and growth factor pathways as a molecular target for overcoming endocrine resistance. Clin Cancer Res 10:331S-336S, 2000
Bardou VJ, Arpino G, Elledge RM, et al: Progesterone receptor status significantly improves outcome prediction over estrogen receptor status alone for adjuvant endocrine therapy in two large breast cancer databases. J Clin Oncol 21:1973-1979, 2003
Cui X, Zhang P, Deng W, et al: Insulin-like growth factor-I inhibits progesterone receptor expression in breast cancer cells via the phosphatidylinositol 3-kinase/Akt/mammalian target of rapamycin pathway: Progesterone receptor as a potential indicator of growth factor activity in breast cancer. Mol Endocrinol 17:575-588, 2003
Benz CC, Scott GK, Sarup JC, et al: Estrogen-dependent, tamoxifen-resistant tumorigenic growth of MCF-7 cells transfected with HER2/neu. Breast Cancer Res Treat 24:85-95, 1993
Borg A, Baldetorp B, Ferno M, et al: ERBB2 amplification is associated with tamoxifen resistance in steroid-receptor positive breast cancer. Cancer Lett 81:137-144, 1994
Love RR, Duc NB, Havighurst TC, et al: Her-2/neu overexpression and response to oophorectomy plus tamoxifen adjuvant therapy in estrogen receptor-positive premenopausal women with operable breast cancer. J Clin Oncol 21:453-457, 2003
Stal O, Borg A, Ferno M, et al: ErbB2 status and the benefit from two or five years of adjuvant tamoxifen in postmenopausal early stage breast cancer. Ann Oncol 11:1545-1550, 2000
Knoop AS, Bentzen SM, Nielsen MM, et al: Value of epidermal growth factor receptor, HER2, p53, and steroid receptors in predicting the efficacy of tamoxifen in high-risk postmenopausal breast cancer patients. J Clin Oncol 19:3376-3384, 2001
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
Linderholm B, Grankvist K, Wilking N, et al: Correlation of vascular endothelial growth factor content with recurrences, survival, and first relapse site in primary node-positive breast carcinoma after adjuvant treatment. J Clin Oncol 18:1423-1431, 2000
Coradini D, Biganzoli E, Pellizzaro C, et al: Vascular endothelial growth factor in node-positive breast cancer patients treated with adjuvant tamoxifen. Br J Cancer 89:268-270, 2003
Berns EM, Klijn JG, Look MP, et al: Combined vascular endothelial growth factor and TP53 status predicts poor response to tamoxifen therapy in estrogen receptor-positive advanced breast cancer. Clin Cancer Res 9:1253-1258, 2003
Foekens JA, Peters HA, Grebenchtchikov N, et al: High tumor levels of vascular endothelial growth factor predict poor response to systemic therapy in advanced breast cancer. Cancer Res 61:5407-5414, 2001
Veikkola T, Karkkainen M, Claesson-Welsh L, et al: Regulation of angiogenesis via vascular endothelial growth factor receptors. Cancer Res 60:203-212, 2000
Miralem T, Steinberg R, Price D, et al: VEGF(165) requires extracellular matrix components to induce mitogenic effects and migratory response in breast cancer cells. Oncogene 20:5511-5524, 2001
Speirs V, Atkin SL: Production of VEGF and expression of the VEGF receptors Flt-1 and KDR in primary cultures of epithelial and stromal cells derived from breast tumours. Br J Cancer 80:898-903, 1999
Xie B, Tam NN, Tsao SW, et al: Co-expression of vascular endothelial growth factor (VEGF) and its receptors (flk-1 and flt-1) in hormone-induced mammary cancer in the Noble rat. Br J Cancer 81:1335-1343, 1999
Ryden L, Linderholm B, Nielsen NH, et al: Tumor specific VEGF-A and VEGFR2/KDR protein are co-expressed in breast cancer. Breast Cancer Res Treat 82:147-154, 2003
Ryden L, Jonsson PE, Chebil G, et al: Two years of adjuvant tamoxifen in premenopausal patients with breast cancer: A randomised, controlled trial with long-term follow-up. Eur J Cancer 41:256-264, 2005
von Marschall Z, Cramer T, Hocker M, et al: De novo expression of vascular endothelial growth factor in human pancreatic cancer: Evidence for an autocrine mitogenic loop. Gastroenterology 119:1358-1372, 2000
Catalano A, Caprari P, Rodilossi S, et al: Cross-talk between vascular endothelial growth factor and semaphorin-3A pathway in the regulation of normal and malignant mesothelial cell proliferation. FASEB J 18:358-360, 2004
Ruohola JK, Valve EM, Karkkainen MJ, et al: Vascular endothelial growth factors are differentially regulated by steroid hormones and antiestrogens in breast cancer cells. Mol Cell Endocrinol 149:29-40, 1999
Rothwell PM: Treating individuals 2: Subgroup analysis in randomised controlled trials—Importance, indications, and interpretation. Lancet 365:176-186, 2005
Ryden L, Stendahl M, Jonsson H, et al: Tumor-specific VEGF-A and VEGFR2 in postmenopausal breast cancer patients with long-term follow-up: Implication of a link between VEGF pathway and tamoxifen response. Breast Cancer Res Treat 89:135-143, 2005(Lisa Ryden, Karin Jirstrm)