Elevated Levels of Chemokine Receptor CXCR4 in HER-2 Negative ...

Journal of Surgical Research 141, 53–59 (2007) doi:10.1016/j.jss.2007.03.015

Elevated Levels of Chemokine Receptor CXCR4 in HER-2 Negative Breast Cancer Specimens Predict Recurrence Neal T. Holm, M.D.,* Kerry Byrnes, M.D.,* Benjamin D. L. Li, M.D.,* Richard H. Turnage, M.D.,* Fleurette Abreo, M.D.,† James M. Mathis, Ph.D.,‡ and Quyen D. Chu, M.D.*,1 *Department of Surgery, †Department of Pathology, and ‡Department of Biochemistry and Molecular Biology, Louisiana State University Health Sciences Center in Shreveport, and Feist-Weiller Cancer Center; Shreveport, Louisiana Submitted for publication January 8, 2007

overexpression between HER-2 (ⴙ) [5.6] and HER-2 (ⴚ) [6.6] cancers (P ⴝ 0.3; independent samples t-test). Recurrences occurred in 18 of 103 patients (17%); 10 occurred in HER-2 (ⴙ) tumors, and 8 occurred in HER-2 (ⴚ) patients. CXCR4 expression level was not predictive of cancer recurrence (P ⴝ 0.80) or overall survival (P ⴝ 0.70) in the HER-2 (ⴙ) group. However, among HER-2 negative tumors, 7 of 8 recurrences occurred in the high CXCR4 group (P ⴝ 0.037). There was no correlation between the degree of CXCR4 overexpression with tumor size (r ⴝ 0.13, P ⴝ 0.22), nodal status (r ⴝ 0.019, P ⴝ 0.4), ER/PR status (r ⴝ 0.12, P ⴝ 0.29), and HER-2 status (r ⴝ 0.091, P ⴝ 0.36). Conclusions. CXCR4 overexpression was observed in all 103 breast cancer specimens but was undetectable in benign breast tissues. CXCR4 overexpression does not correlate with tumor size, nodal status, ER/PR status, and HER-2 status. High CXCR4 overexpression had a significant impact on disease-free survival in HER-2 negative breast cancer patients and may help identify a subset of HER-2 negative breast cancers that have a more aggressive biological behavior. © 2007 Elsevier Inc. All rights reserved. Key Words: breast cancer; CXCR4; HER-2; cancer recurrence.

Introduction. CXCR4 is a chemokine receptor that has recently been implicated to play a pivotal role in breast cancer growth and metastasis. In animal models, reduction of CXCR4 expression significantly abrogated metastatic disease and prolonged survival. In human breast cancers, CXCR4 overexpression may portend a worse clinical course. Recent data suggest that HER-2 up-regulates CXCR4, but whether this is applicable in the clinical setting is not known. In this study, we evaluated the role of CXCR4 overexpression in breast cancer and determined whether it can serve as a potential marker of tumor recurrence in HER-2 negative tumors. Methods. One hundred three patients with stages I to III breast cancers and 6 benign breast tissues were prospectively accrued and analyzed. Study homogeneity was maintained by standardized treatment, surveillance, and compliance protocols. CXCR4 levels were detected using Western blots and results were quantified against 1 ␮g of HeLa cells (positive controls). HER-2 expression was evaluated using the Hercep program, (Dako Corp., Carpinteria, CA) with a positive result defined as >2. CXCR4 expression was defined as low (6.6-fold). Primary endpoints were cancer recurrence and death. Statistical analysis performed included Spearman’s correlation, independent samples t-test, Kaplan-Meier survival analysis, and log-rank test. Results. All 103 cancer specimens had CXCR4 overexpression (mean 6.6 ⴞ 4.7), while none of the 6 benign breast tissues had detectable level of CXCR4. There were 36 HER-2 (ⴙ) tumors and 67 HER-2 (ⴚ) tumors. There was no statistical significance in mean CXCR4

INTRODUCTION

In 2005, 211,240 women in United States were diagnosed with breast cancer. It is the most common malignancy and the second leading cause of cancer death in women [1]. Our understanding of crucial molecular defects that give rise to cancers has ushered in a new era of molecularly targeted therapy. In operable breast cancer, trastuzumab (Herceptin), a humanized monoclonal antibody directed against the extracellular domain of HER-2, significantly reduces the relative risk

1

To whom correspondence and reprint requests should be addressed at Louisiana State University Health Sciences Center and the Feist-Weiller Cancer Center, 1501 Kings Highway, Shreveport, LA 71130-3932. E-mail: [email protected].

53

0022-4804/07 $32.00 © 2007 Elsevier Inc. All rights reserved.

54

JOURNAL OF SURGICAL RESEARCH: VOL. 141, NO. 1, JULY 2007

of cancer recurrence by approximately 50%, a degree that has not been observed since the introduction of tamoxifen for hormone-receptor-positive tumors [2, 3]. However, this effective therapy is available only for those whose tumors are HER-2 positive, and this represents only 20% to 30% of all breast cancers. Although HER-2 negative tumors are perceived to carry a better prognosis than HER-2 positive tumors, there is a subset of patients whose tumors behave aggressively. The molecular mechanism(s) to account for this is poorly understood. The vast majority of breast cancer related deaths are due to the development and subsequent progression of systemic disease. The development of metastases is a complex process involving multiple molecular factors [4, 5]. Many of these factors have characteristics common to lymphocyte homing and organogenesis, one of which is the chemokine receptor CXCR4 [6 – 8]. CXCR4 is a G protein-coupled seven-transmembrane receptor that has been linked to invasion and metastasis in a variety of cancers, including breast cancer [9]. Although multiple studies on CXCR4 have recently yielded promising results, CXR4’s clinical significance has not been well elucidated. In this study, we used a large prospective database of invasive breast cancer patients to evaluate the role of CXCR4 overexpression in cancer specimens to predict outcome. We believe that CXCR4 plays a pivotal role in predicting a worse outcome in a subset of patients with HER-2 negative tumors. A better understanding of its role may help identify a group of high risk patients who may be candidates for more intensive treatment and/or for novel targeted therapy. MATERIALS AND METHODS Approval to use our tissue bank and database was obtained from our institutional Internal Review Board (IRB). Complete clinicopathologic data and tissue specimens from 103 patients with stages I to III breast cancers were examined. Treatment and surveillance protocols were standardized to ensure study homogeneity. Compliance with treatment and surveillance protocol was 95% and 99%, respectively. Surgical treatment consisted of either a modified radical mastectomy or breast conservation therapy (BCT, lumpectomy with tumor-free margin, axillary lymph node dissection, and breast irradiation; a subset of patients who had T1 lesion underwent sentinel node biopsy, followed by a complete axillary lymph node dissection for those who had positive sentinel nodes). Adjuvant axillary irradiation, systemic chemotherapy, and antiestrogen therapy were offered and administered as indicated per current standard of care. Surveillance protocol consisted of a history and physical examination every 3 mo for 3 y, every 6 mo in years 4 and 5, and annually thereafter. Annual chest X-ray, mammogram, complete blood count, and liver function test were obtained. Any additional radiological and/or histological evaluation was performed based on the patient’s examination and history. Clinical data were accrued and recorded prospectively and included age at diagnosis, comorbid conditions, stage of disease, treatment protocol, surveillance protocol compliance, and study endpoints. Study primary endpoints were cancer recurrence and cancer-related death.

Tissue Procurement A cancer specimen of at least 100 mg was obtained from the tumor core at the time of surgery from each patient. The specimen was verified by the study pathologist to be an invasive mammary carcinoma. It was then immediately frozen in liquid nitrogen and stored at ⫺70°C.

Assay for CXCR4 Specimen assay for CXCR4 expression was performed using Western blot analysis. In brief, a protein lysate from each breast specimen was prepared using a 10 mg portion of tumor tissue cut into tiny pieces, suspended in 0.5 mL RIPA buffer (150 mmol/L NaCl; 1% NP-40; 0.5% DOC; 0.1% SDS; 50 mmol/L Tris [pH 8.0]; 0.1 mmol/L PMSF), and mechanically homogenized using a Savant Bio 101 Fastprep FP120 system (Savant Instruments, Inc., Holbrook, NY). The lysate was then centrifuged at 10,000 g for 10 min (at 4°C), and total protein content was determined using a standard BCA (bicinchoninic acid) copper reduction assay kit (Pierce, Rockford, IL). An equal amount of protein lysate from each specimen as well as benign control breast tissue (20 ␮g diluted in 1:10 RIPA) was loaded onto and separated by using 4% to 20% denaturing gel Tris HCL polyacrylamide gel electrophoresis. Electroblotting onto a nylon membrane (Immobilon PVDF; Millipore, Bedford, MA) was performed, and the membranes were blocked with 5% nonfat milk for 1 h. Primary incubation of the membrane was performed using polyclonal goat x human anti-CXCR4 antibody (Fusin SC6190; Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Secondary incubation of the membrane was performed using a bovine/goat horseradish peroxidase conjugate. Blot development was accomplished using Opti 4CN (4-chloro-1-naphthol; Bio-Rad Laboratories, Hercules, CA). Quantification of CXCR4 protein expression was accomplished using the Biophotonics system (Biophotonics Corp., Ann Arbor, MI). The blots were scanned and the band intensity was evaluated using the Intelligent Quantifier software (Bio Image, Ann Arbor, MI). Since benign breast tissues have undetectable level of CXCR4, 1 ␮g total protein from cell lysate of HeLa (ATCC no.CCL-2.2) cell lines (known to overexpress CXCR4) was used, to which all cancer specimens were compared. Band intensity from tumor samples was compared against 1 ␮g HeLa cells. Quantification of CXCR4 level in each cancer specimen was expressed as x-fold elevated over known concentration of HeLa cells. This process was repeated three times for each specimen and the results were averaged. Protein extracts from the following cells obtained from American Type Culture Collection (ATCC) were used as controls: (1) CRL6509 cells (normal rat kidney) served as a negative control, (2) HeLa (human cervical cancer) and HL60 (human acute myeloid leukemia) cells served as positive controls.

Assay for HER-2 Expression The tumor’s HER-2 status is defined as two or more, using the Hercep program (Dako Corp., Carpinteria, CA).

Estrogen and Progesterone Receptor Status Estrogen receptor (ER) and progesterone receptor (PR) status was determined using immunohistochemical methods. Slides were stained and evaluated using the Dako Autostriker and the automated cellular imaging system. Activity greater than ten percent was considered positive.

Statistical Analysis Statistical analyses were performed using MedCalc software (Microsoft, Inc.). Level of CXCR4 overexpression, tumor size, tumor grade, nodal, HER-2, ER and PR statuses were correlated using the samples t-test and Spearman rank correlation. Survival analysis was

55

HOLM ET AL.: ELEVATED LEVELS OF CHEMOKINE RECEPTOR CXCR4

TABLE 1 T Stage, N Stage, and CXCR4 Expression Levels for All Studied Patients CXCR4 Groups

FIG. 1. CXCR4 expression profile in HER-2 positive and HER-2 negative breast cancer cells: Western blot analysis demonstrated that CXCR4 overexpression was observed in both HER-2 positive (lanes 1, 2) and HER-2 negative tumors (lanes 3, 4). Benign breast tissue failed to demonstrate any expression of CXCR4 (lane 5). CRL6509 cells, which are known to lack CXCR4 expression, was used as a negative control (lane 6), while HeLa and HL60 cells, which are known to overexpress CXCR4, were used as positive controls (lanes 7–9). Variable CXCR4 expression was demonstrated with different concentrations of HeLa cells (lanes 7, 8).

performed using the Kaplan-Meier method and Log-rank test. A P value ⱕ0.05 was considered statistically significant.

RESULTS

One hundred and three patients were accrued for this study. The mean age at diagnosis was 53 y, and the mean follow-up was 26 mo. There were 36 patients who had HER-2 positive tumors and 67 patients who had HER-2 negative tumors. Fig. 1 is a representative Western blot analysis for CXCR4 expression. Note that CXCR4 overexpression was observed in both HER-2 positive (lanes 1, 2) and HER-2 negative tumors (lanes 3, 4). Benign breast tissue had undetectable level of CXCR4 (lane 5). CRL6509 cells, known to

T stage T1 T2 T3 T4 N stage N0 N1 N2 N3

No. of patients (%)

High

Low

30 (29.2) 59 (57.3) 12 (11.6) 2 (1.9)

7 27 5 2

23 32 7 0

47 (45.6) 33 (32.1) 14 (13.6) 9 (8.7)

17 11 7 5

30 22 7 4

lack CXCR4, were used as a negative control (lane 6), while HeLa and HL60 cells, known to overexpress CXCR4, were used as positive controls (lanes 7–9). Variable CXCR4 expression was demonstrated with different concentrations of HeLa cells (lanes 7, 8). The mean level of CXCR4 overexpression was 6.6 ⫾ 4.7 and using this as our cutoff, patients were distributed into two groups: (1) low CXCR4 group (⬍6.6fold elevation, n ⫽ 62 patients), and (2) high CXCR4 group (ⱖ6.6-fold elevation, n ⫽ 41 patients) (Fig. 2). The breakdown of patients as grouped by tumor size (T stage) and nodal status (N) is shown in Table 1. The T stage distribution was as follows: T1 lesions (n ⫽ 30 patients), T2 (n ⫽ 59 patients), T3 (n ⫽ 12 patients), and T4 (n ⫽ 2 patients). The majority (57%) of patients had T2 lesions. There were 56 node-positive patients and 47 node negative patients. The N stage distribution was as follows: (1) N0 ⫽ 47 patients, (2) N1 ⫽ 33 patients, (3) N2 ⫽ 14 patients, and (4) N3 ⫽ 9 patients. Over 50% of patients were node positive with 32% having N1 disease. Table 2 demonstrates the correlation test results between CXCR4 overexpression with tumor size, nodal status, estrogen and progesterone receptors status, and HER-2 receptor status. We found no correlation between the degree of CXCR4 overexpression with tumor size (r ⫽ 0.13; P ⫽ 0.22), nodal status (r ⫽ 0.019; TABLE 2 Correlation of CXCR4 Overexpression with Known Clinicopathologic Parameters

FIG. 2. Distribution of breast cancer patients based on CXCR4 overexpression levels: this is a bar chart representation of CXCR4 distribution grouped by high and low groups relative to the mean CXCR4 overexpression. Sixty percent of patients had tumors that had low CXCR4 overexpression (⬍6.6-fold), and 40% of patients had tumors that had high CXCR4 overexpression (ⱖ6.6-fold).

Factors

P value

T-stage N-stage ER/PR status HER-2 status

0.22 0.40 0.29 0.36

CXCR4: low group: ⬍6.6; high group: ⱖ6.6.

56


TABLE 3 Incidence of Cancer Recurrence in 67 Patients with HER-2 Negative Tumors and CXCR4 Overexpression Levels HER-2 negative patients CXCR4 ⬍6.6

Stage I Stage II Stage III Total

FIG. 3. Disease-free survival for HER-2 positive breast cancer patients as stratified by CXCR4 overexpression levels: this KaplanMeier graph compares the disease-free survival for high and low groups of CXCR4 overexpression in HER-2 positive patients. Patients whose tumors were in the low CXCR4 group did not have a statistically significant difference in cancer recurrence when compared with the high CXCR4 overexpression group (P ⫽ 0.80, log-rank test).

P ⫽ 0.4), ER/PR status (r ⫽ 0.12; P ⫽ 0.29), or HER-2 status (r ⫽ 0.091; P ⫽ 0.36), thus suggesting that CXCR4 may be an independent factor that predicts outcome. In addition, contrary to the proposed link between CXCR4 expression level and HER-2 expression level submitted by others [10], not only did we not find a correlation between CXCR4 level and HER-2 status, but the mean CXCR4 overexpression level between HER-2 positive (5.6-fold) and HER-2 negative (6.6-fold) was not statistically different (P ⫽ 0.3).

FIG. 4. Overall survival for HER-2 positive breast cancer patients as stratified by CXCR4 overexpression levels: this KaplanMeier graph compares the disease-free survival for high and low groups of CXCR4 overexpression in HER-2 negative patients. Patients whose tumors were in the low CXCR4 group had a lower rate of cancer recurrence when compared with patients whose tumors were in the high CXCR4 group (P ⫽ 0.037, log-rank test).

CXCR4 ⱖ6.6

No. pts.

Recur

No. pts.

Recur

11 17 6 34

0 0 1 1

4 22 7 33

0 4 3 7

Recurrences occurred in 18 of 103 patients (17%); 10 occurred in HER-2 positive tumors, and 8 occurred in HER-2 negative tumors. Figures 3 and 4 show the Kaplan-Meier survival curves for the HER-2 positive group. Note that CXCR4 overexpression level was not predictive of cancer recurrence (P ⫽ 0.80) (Fig. 3) or overall survival (P ⫽ 0.70) (Fig. 4) for this group. However, among the HER-2 negative tumors, 7 of 8 recurrences occurred in the high CXCR4 overexpression group (Table 3), and all were systemic recurrences. Figures 5 and 6 show the Kaplan-Meier survival curves for the HER-2 negative group; high CXCR4 overexpression level significantly predicted cancer recurrence in HER-2 negative breast cancers (P ⫽ 0.037) (Fig. 5), although it had a borderline significance in predicting overall survival (P ⫽ 0.072) (Fig. 6). Figure 7 demonstrates the disease-free survival of patients with HER-2 negative tumors with elevated CXCR4 overexpression versus patients with HER-2

FIG. 5. Disease-free survival for HER-2 negative breast cancer patients as stratified by CXCR4 overexpression levels: this KaplanMeier graph compares overall survival for high and low CXCR4 overexpression in HER-2 positive patients. There was no statistical difference in cancer-related death between high and low groups of CXCR4 overexpression (P ⫽ 0.70 log-rank test).

HOLM ET AL.: ELEVATED LEVELS OF CHEMOKINE RECEPTOR CXCR4

FIG. 6. Overall survival for HER-2 negative breast cancer patients as stratified by CXCR4 overexpression levels: this KaplanMeier graph compares overall survival for high and low CXCR4 overexpression in HER-2 negative patients. Patients in the high CXCR4 group had a higher rate of death that approached statistical significance (P ⫽ 0.072 log-rank test) when compared with patients in the low CXCR4 group.

positive tumors. Note that patients with HER-2 negative, elevated CXCR4 tumors do just as poorly as those with HER-2 positive tumors, suggesting that this is indeed a group of patients at risk for worse outcome. As an internal control to assess the reliability of our sample size, we evaluated disease-free survival and overall survival based on known prognostic factors such as tumor size and nodal status. Our sample size appears to be accurate since disease-free survival and overall survival were statistically influenced by tumor size (P ⬍ 0.05) and nodal status (P ⬍ 0.05). DISCUSSION

Understanding the “molecular signatures” of cancer is essential for improved understanding of tumor biology and for the development of target-specific cancer therapies. Evidence suggests that the chemokine receptor CXCR4 plays a pivotal role in invasion and metastasis of various cancers including breast [11], colon [12], and melanoma [13]. This makes it an ideal candidate as a diagnostic marker and/or as a target for novel therapy. Cancer metastasis is a complex process involving cell dissociation, migration, matrix invasion, and transport to remote sites via adhesion [14]. In breast cancer, metastasis is generally an organ-selective process whereby cancer cells metastasize to regional lymph nodes, bone marrow, lungs, and liver. This directional, site-specific “homing” mechanism is the result of complex interactions between the chemokines and their respective receptors [11]. Target organs produce high level of CXCL12 chemokine (also called stromal cellderived factor or SDF-1) that attract nearby or distant breast cancer cells that possess high levels of the che-

57

mokine receptor CXCR4. Consequently, there is compelling evidence implicating CXCR4 to play a pivotal role in breast cancer metastasis [11, 15–18]. Müller et al. found that breast cancer cells’ motility and migration can be induced when they were exposed to their ligand, SDF-1 [11]. They have also shown that neutralizing CXCR4/SDF-1 interactions with anti-CXCR4 antibodies will abrogate the development of lung metastasis in SCID mice [11]. Breast cancer metastasis can be inhibited by silencing CXCR4 [16 –18]. Using a rodent model, Smith et al. demonstrated a substantial growth delay of metastatic breast cancer cells to the lung with small interfering RNAs (siRNA) against CXCR4 or with a CXCR4 antagonist AMD3100 [18]. Similarly, using siRNAs against CXCR4 in an animal model, Ligna et al. successfully blocked breast cancer cells in vitro invasion and in vivo metastasis [16]. CXCR4 has been demonstrated to promote breast cancer cell growth since stably transduced MDA-MB-231 breast cancer cells fail to grow in SCID mice when CXCR4 was silenced with siRNA [17]. The mechanism that drives CXCR4 expression has not been well elucidated. Li et al. recently speculated that CXCR4 expression and HER-2 expression are linked, which accounts for their observation that most CXCR4 positive cells were also HER-2 positive [10]. They demonstrated that HER-2 up-regulates CXCR4 overexpression by (1) enhancing its translation through the phosphatidylinositol 3-kinase (PI3k)/Akt/mTor signaling pathway, and (2) protecting CXCR4 from degradation by preventing it from being ubiquinated, a process that sorts internalized G protein-coupled receptors to the degradative pathway rather than recycling them back to the plasma membrane [10, 19, 20].

FIG. 7. Comparison of disease-free survival between HER-2 negative, elevated CXCR4 expression versus HER-2 positive patients: this Kaplan-Meier graph compares disease free survival for HER-2 negative patients with elevated CXCR4 levels to HER-2 positive patients. Patients with HER-2 negative tumors with high CXCR4 overexpression had no statistical difference in disease free survival compared with HER-2 positive patients.

58


Although there are strong preclinical evidences to support CXCR4’s role in breast cancer growth and metastasis, its clinical application remains elusive. There is paucity of clinical data and results have been equivocal, since most studies reported small sample size and/or failed to characterize their patients’ population. Immunohistochemical staining technique (IHC), which was used to identify and quantify CXCR4 expressions in all of these studies, was highly variable, thereby making it difficult to compare results. Finally, most studies were retrospective and, consequently, contained biases and confounders. As a result, there were conflicting reports of the utility of using CXCR4 as a predictive marker for breast cancer outcome. Kato et al. evaluated 79 breast cancer specimens and found no differences in the staining intensity among the majority of node-positive cases [21]. Furthermore, they did not find a significant correlation between CXCR4 expression level with hematogenous metastasis or overall survival [21]. Similarly, Kang et al. evaluated 120 cancer specimens via IHC and found that although CXCR4 level significantly correlated with lymph node metastases, it did not correlate to distant metastases or overall survival [22]. Cabioglu et al. evaluated 18 breast cancer specimens and found that high CXCR4 expression was predictive of isolated tumor cells in bone marrow [23]. However, as acknowledged by these investigators, whether this has any prognostic significance remains to be determined. In another study of 197 patients, Cabioglu et al. observed a significant increase of high cytoplasmic CXCR4 coexpression with HER-2 in small tumors (T1 or tumor less than 2 cm) that had lymph node metastases [24]. Whether these results can be applied to larger tumors, however, is not known. Finally, Li et al. examined 219 breast cancer samples and found that patients who had high CXCR4 expression, based on IHC scoring system, had a significantly poorer overall survival than those who had low CXCR4 expression [10]. However, the lack of reporting complete clinicopathologic details in these cohorts has made it difficult for us to fully interpret the data. It is unknown whether high CXCR4 in their study was co-sorted with nodal disease and thus predicted a worse clinical outcome. Given these equivocal results, we examined our large prospective database to ascertain whether CXCR4 has a role in predicting breast cancer outcome in a subset of patients. This database was created as part of several prospective clinical trials that were conducted over the past decade by our group. Complete clinicopathologic data and tissue specimens were available on all patients, and because all had standardized treatment regimens, compliance rate and follow-up data were near uniform (over 94%) [25]. Our study found that CXCR4 was overexpressed in all 103 cancer specimens but not in benign tissues.

What we found interesting was that CXCR4 overexpression was not predictive of outcome in HER-2 positive tumors, but was predictive in HER-2 negative tumors. HER-2 negative tumors have not received a lot of attention since they were generally believed to have a less aggressive clinical course than their counterpart. However, despite such belief, there is a subset of patients who are at high-risk for developing recurrences. Because HER-2 negative tumors represent over 70% of all breast cancers, such subset of patients becomes significant. In our training set of 67 HER-2 negative tumors, there were eight recurrences (12%), seven of which occurred in patients whose tumors had high level of CXCR4 overexpression. Using the KaplanMeier survival analysis and the log-rank test, we found that high CXCR4 overexpression level in tumor specimen significantly predicted cancer recurrence in HER-2 negative breast cancers (P ⫽ 0.037). We also found no correlation between the degree of CXCR4 overexpression with tumor size (P ⫽ 0.22), nodal status (P ⫽ 0.4), or ER/PR status (P ⫽ 0.29), suggesting that CXCR4 may be an independent marker for outcome. In addition, contrary to what was proposed by Li et al. [10], we did not find a correlation between CXCR4 level and HER-2 status (r ⫽ 0.091; P ⫽ 0.36), since the mean CXCR4 overexpression between HER-2 positive (5.6fold) and HER-2 negative tumors (6.6-fold) was not statistically different (P ⫽ 0.3). This suggests that CXCR4 expression is regulated by other yet to be identified mechanism(s). In conclusion, we identified a “molecular signature” that identified a subset of HER-2 negative breast cancers that have a more aggressive biological behavior. The poor outcome among patients with HER-2 negative and high CXCR4 overexpression tumors was similar to those with HER-2 positive tumors, suggesting that this subset of patients is indeed a high risk group. CXCR4 appears to be an independent predictor of outcome since it did not correlate with tumor size, nodal status, and ER/PR status. Additionally, in contrast to what has been speculated, we did not find a clinical link between HER-2 positive tumors and CXCR4 overexpression. This finding opens new avenues to explore for other mechanisms that regulate CXCR4 expression. Although our findings are encouraging, we are cautious to not over-interpret the data, since the number of recurrences was small relative to the dataset. Validation of these observations on a larger and/or independent database will be necessary before translating them into the clinical setting. REFERENCES 1.

Jemal A, Murray T, Ward E, et al. Cancer statistics 2005. CA Cancer J Clin 2005;55:10. 2. Slamon DJ, Clark GM, Wong SG, et al. Human breast cancer: Correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987;235:177.

HOLM ET AL.: ELEVATED LEVELS OF CHEMOKINE RECEPTOR CXCR4 3.

4.

5.

6. 7. 8. 9. 10. 11.

12.

13.

14. 15.

Slamon DJ, Leyland-Jones B, Shak S, et al. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 2001; 344:783. Weidner N, Cady B, Goodson WH, III. Pathologic prognostic factors for patients with breast carcinoma. Which factors are important? Surg Oncol Clin N Am 1997;6:415. Jatoi I, Hilsenbeck S, Clark G, Osborne K. Significance of axillary lymph node metastasis in primary breast cancer. J Clin Oncol 1999;17:2334. Chambers A. Dissemination and growth of cancer cells in metastatic sites. Nat Rev Cancer 2002;2:563. Walser T. The role of chemokines in the biology and therapy of breast cancer. Breast Dis 2004;20:137. Ben-Baruch A. The multifaceted roles of chemokines in malignancy. Cancer Metast Rev 2006;25:357. Proudfoot A. Chemokine receptors: Multifaceted therapeutic targets. Nat Rev Immunol 2002;2:106. Li Y, Pan Y. Up-regulation of CXCR4 is essential for HER2mediated tumor metastasis. Cancer Cell 2004;6:459. Müller A, Homey B, Soto H, Ge N, et al. Involvement of chemokine receptors in breast cancer metastasis. Nature 2001; 410:50. Ottaiano A. Overexpression of both CXC chemokine receptor 4 and vascular endothelial growth factor proteins predicts early distant relapse in stages II-III colorectal cancer patients. Clin Cancer Res 2006;12:2795. Murakami T, Maki W. Expression of CXC chemokine receptor-4 enhances the pulmonary metastatic potential of murine B16 melanoma cells. Cancer Res 2002;62:7328. Woodhouse EC, Chuaqui RF, Liotta LA. General mechanisms of metastasis. Cancer 1997;80:1529. Lee B, Lee T, Avraham S, Avraham H. Involvement of the

16. 17.

18.

19. 20.

21.

22.

23.

24.

25.

59

chemokine receptor CXCR4 and its ligand stromal cell-derived factor 1␣ in breast cancer cell migration through human brain microvascular endothelial cells. Molec Cancer Res 2004;6:327. Liang Z, Yoon Y, Votaw J, et al. Silencing of CXCR4 blocks breast cancer metastasis. Cancer Res 2005;65:967. Lapteva N, Yang A, Sanders D, et al. CXCR4 knockdown by small interfering RNA abrogates breast tumor growth in vivo. Cancer Gene Ther 2005;12:84. Smith M, Luker K, Garbow J, et al. CXCR4 regulates growth of both primary and metastatic breast cancer. Cancer Res 2004; 64:8604. Benovic JL, Marchese A. A new key in breast cancer metastasis. Cancer Cell 2004;6:429. Marchese A, Railborg C, Santini F, et al. The E3 ubiquitin ligase AIP4 mediates ubiquitination and sorting of the G protein-coupled receptor CXCR4. Dev Cell 2003;5:709. Kato M, Kitayama J, Kazama S, et al. Expression pattern of CXC chemokine receptor-4 is correlated with lymph node metastasis in human invasive ductal carcinoma. Breast Cancer Res 2003;5:144. Kang H, Watkins G, Douglas-Jones A, et al. The elevated level of CXCR4 is correlated with nodal metastasis of human breast cancer. Breast 2005;14:360. Cabioglu N, Sahin A, Doucet M, et al. Chemokine receptor CXCR4 expression in breast cancer as a potential predictive marker of isolated tumor cells in bone marrow. Clin Exp Metastasis 2005;22:39. Cabioglu N, Yazici M, Arun B, et al. CCR7 and CXCR4 as novel biomarkers predicting axillary lymph node metastasis in T1 breast cancer. Clin Cancer Res 2005;11:5686. McClusky DR, Chu QD, Yu H, DeBenedetti A, et al. A prospective trial on initiation factor 4E (eIF4E) overexpression and cancer recurrence in node-positive breast cancer. Ann Surg 2005;242:584.