The utility of CD44 and D2-40 as a prognostic predictor in invasive carcinomas of the breast
Article information
Trans Abstract
Purpose:
Local recurrence and distant metastasis are major prognostic factors associated with breast cancer. Lymphovascular invasion is an important pathway of metastatic spread. CD44 and D2-40 are assumed to be related with invasion of malignant cells to the adjacent lymphatic or vascular structures by different mechanisms. This study was conducted to examine whether CD44 and D2-40 expression together have prognostic value in invasive carcinoma of the breast.
Methods:
A total of 46 surgically resected tissue samples of invasive carcinoma of the breast were analyzed for immunohistochemical expression of CD44 and D2-40, along with other clinicopathologic factors. Association between patient related factors and disease free survival was evaluated by statistical analysis.
Results:
Disease free survival was shorter in CD44-positive patients (median, 66.5 months) than in CD44-negative patients (median, 112.6 months; P<0.001). All of 17 patients with recurrence showed D2-40 positivity. When both markers were positive, disease free survival was considerably shorter (63.3 months) than when at least one marker was negative (112.9 months, P<0.001).
Conclusion:
We concluded that expression of CD44 and D2-40 is associated with unfavorable prognosis. Concurrent positivity of both markers may be useful in predicting malignant relapse following surgical resection with shorter disease free survival in patients with invasive carcinoma of the breast.
INTRODUCTION
Development of local recurrence or distant metastasis is one of the most powerful prognostic factors for breast cancer and a leading cause of cancer related death in women in Korea. In spite of surgery, cytotoxic chemotherapy and hormonal therapy, a substantial proportion of breast cancer patients develop recurrence within several years.
CD44 is a transmembrane adhesion receptor which mediates cellular interaction with its microenvironment [1]. Previous literature have shown that CD44 promotes metastatic signaling and acts as a putative tumor suppressor in breast cancer [2-4]. In some studies, CD44 positive cells in primary breast tumors manifest and increase invasiveness and progenitor ability [5,6]. CD44 is associated with cellular invasion and metastasis of cancer cells by underpinning the reorganization of the cytoskeleton to facilitate active migration. CD44 enhances invasiveness of cancer cells by upregulating protease and collagen degrading enzymatic expression and activity [7]. The capacity of CD44 to induce protease activity is consistent with aggressive clinical characteristics of basal-like breast cancer cells and increased propensity to invade locally through breast tissue [7,8].
On the contrary, other studies reported that CD44 expression was an independent predictor of good prognosis and was associated positively with prolonged metastasis free survival [9,10]. CD44 negative phenotype was related with poorer clinical outcome [9,11].
One of the most common pathways of initial tumor dissemination is via regional lymphatics, following routes of natural lymphatic drainage [12]. Lymphovascular invasion of breast cancer is an independent predictor of regional lymph node metastasis that has critical prognostic influence on patients. Thus, accurate detection of lymphovascular invasion constitutes an essential part in predicting prognosis. As a marker of lymphatic endothelium, D2-40 plays a role in detecting tumor emboli in lymphatic vessels. D2-40 is a recently available monoclonal antibody specific for human podoplanin and has been used in identifying lymphovascular invasion of tumors [13]. Various studies have evaluated D2-40 expression in several malignant neoplasms. Positivity of D2-40 was helpful in detecting lymphatic tumor emboli and was associated with poorer prognosis in early colorectal cancer, esophageal squamous cell carcinoma, and small cell lung cancer [14-16]. Immunostaining with D2-40 significantly increased the accuracy of detection of lymphatic invasion compared to conventional hematoxylin and eosin (H&E) staining in early breast cancer [17]. Because lymphovascular invasion is a strong prognostic marker, D2-40 has potential as a key prognostic predictor for breast cancer and its effectiveness is actively investigated [18].
Local recurrence and distant metastasis are major prognostic factors associated with breast cancer. Lymphovascular invasion is an important pathway of metastatic spread. CD44 and D2-40 are assumed to be related with invasion of malignant cells to the adjacent lymphatic or vascular structures by different mechanisms. This study was conducted to examine that CD44 and D2-40 expression together may have prognostic value in invasive carcinoma of the breast.
METHODS
Patients
Charts and pathology reports of 62 patients with invasive ductal carcinoma of the breast from January 2003 to December 2006 were retrospectively reviewed. The patients had been treated from initial diagnosis to the completion of follow-up sessions. Patients without surgical treatment, those diagnosed with ductal carcinoma in situ or lobular carcinoma in situ, presence of metastasis, and insufficient number of pathologic slides were excluded. A total of 46 patients were included in the study and were managed with cytotoxic chemotherapy, hormonal therapy or radiation therapy according to pathologic stage and hormone receptor status. Patients were staged based on the TNM (tumor, nodes, metastasis) classification of the American Joint Committee on Cancer (AJCC 7th ed., 2010).
In current study, we used representative formalin fixed, paraffin embedded tissue sections. Each sample was routinely stained with H&E and was histologically classified according to the World Health Organization (WHO) classification.
Immunohistochemistry
After reviewing tumor specimens, one of the representative specimens was chosen. Three paraffin-embedded tissue samples were cut into 4 micrometer thick sections. One section was routinely stained with H&E, whereas others were stained with CD44 (Novo-Castra, NewCastle, UK) and D2-40 (DakoCytomation, Glostrup, Denmark). Sections were deparaffinized in xylene and rehydrated in a graded series of ethanol to water. Peroxidase activity was inhibited by incubating the tissues in methanol with 3% hydrogen peroxide for 5minutes. After rinsing with tris-buffered saline, sections were immersed in 0.05 M sodium citrate buffer and were heated in a microwave oven at 100°C for 10 minutes for antigen retrieval, followed by incubating with the respective primary antibody at room temperature for 30 minutes. After washing, the sections were incubated with a secondary antibody (DakoCytomation) at room temperature for 30 minutes. Subsequently, they were subjected to DAB+ substrate buffer and DAB+ chromogen (DakoCytomation) for 5 minutes. Counterstaining was performed with hematoxylin to identify nuclei and then specimens were mounted.
For each staining of D2-40, the positive and negative controls were used during pathologic evaluation of slides. Typically, normal non-neoplastic lymphatic channels and veins are scattered in the breast tumor tissue so that both positive and negative controls are comparable in a single slide simultaneously. Pathologic ‘internal control’ method was used in determining D2-40 positivity. D2-40 was determined to be positive when the tumor thrombi exists inside the tumor containing lymphatic channel which shows diffuse cytoplasmic reactivity in endothelial cells.
For each staining of CD44, the human tonsil was used as positive control. Negative controls were included for all sections by omission of primary antibody. CD44-positivity was defined when more than 20% of the tumor cells demonstrate cytoplasmic membrane staining [19,20]. Several related studies determined CD44 positivity when 10% or more of tumor cells demonstrated cytoplasmic membrane staining [21-23]. But regarding small sized study, detailed grading of raw data would provide inappropriate patient number for statistical analysis.
Lymphatic invasion was considered positive when tumor emboli were detected in D2-40 positive, lymphatic vessels. One pathologist evaluated the samples with a light microscope. Staining results were correlated with clinicopathologic features such as tumor size, lymph node involvement of tumor, TNM stage, estrogen receptor (ER) status, progesterone receptor (PR) status, HER2 expression, and tumor recurrence.
Statistical analysis
Statistical analysis was conducted. Correlations between clinico pathological factors, CD44 and D2-40 was analyzed with chisquare tests. Kaplan-Meier survival curves and the log rank test and Cox-regression test were utilized to evaluate factors associated with tumor recurrence. A P<0.05 was considered statistically significant.
RESULTS
Clinical and immunohistochemical expression data of patients
Baseline patient characteristics are presented in Table 1. All patients were female and aged between 30 and 77 years (mean age, 51.0 years). Forty-four patients had undergone a modified radical mastectomy and wide excision was performed for 2 patients. Median follow-up period was 70.4 months, ranging from 7.1 to 120.3 months, during which 17 patients (37%) experienced local recurrence or developed metastasis.
Correlation between clinicopathological factors and CD44, D2-40
Expression of CD44 was positively correlated with lymph node involvement (P=0.004), lymphovascular invasion detected by H&E stain (P=0.002), and recurrence (P<0.001) but not with tumor size (P=0.855), ER (P=0.246), PR (P=0.895), and HER2 (P=0.478).
Variables significantly associated with D2-40, positivity were lymph node involvement (P=0.002), lymphovascular invasion detected with H&E stain (P<0.001), and recurrence (P<0.001). CD44 and D2-40 showed similar patterns of correlation with clinicopathological factors (Table 2).
Correlation between disease free survival and CD44, D2-40
Univariate analysis revealed statistically significant factors involving lymph node status, stage, lymphovascular invasion by H&E, CD44, and D2-40 (Table 3). Kaplan-Meier analysis showed difference in disease free survival between patients subgroups classified by the presence of CD44 or D2-40. Median disease free survival was 66.5 months in CD44-positive patients and 112.6 months in CD44-negative patients (Fig. 1A). D2-40 also showed a similar disease free survival pattern as CD44. All of 17 patients with recurrences were positive to D2-40 and mean disease free survival was 60.9 months (Fig. 1B). When both markers were positive, disease free survival was shorter (63.3 months) with statistical significance than when at least one marker was negative (112.9 months, P<0.001) (Fig. 1C).

Univariate analysis of prognostic factors associated with disease free survival in breast cancer patients

Kaplan–Meier survival curves for analyses of times from breast cancer surgery to recurrence. (A) Disease free survival of CD44 positive or negative group. (B) Disease free survival of D2-40 positive or negative group. (C) Disease free survival of patient groups with CD44/D-40 both positive or not.
Statistical result of CD44 and D2-40 using Cox regression test is shown in Table 4. CD44-positive patients had 13.76 times increased risk of tumor recurrence than CD44-negative patients (P=0.001). And when the data was adjusted for lymph node involvement, TNM stage, and lymphovascular invasion by H&E, the hazard ratio was 10.82 (P=0.014). Patients who were positive to both CD44 and D2-40 had 16.22 times increased risk of developing tumor relapse than patients who were negative for at least one marker (P<0.001). The hazard ratio was increased to 22.72 when the data was adjusted (P<0.006).
However, multivariate analysis with clinicopathological factors revealed no significant prognostic effect on disease free survival except lymphovascular invasion. Cox regression test included factors of lymph node involvement (P=0.420), TNM stage (P=0.500), CD44 (P=0.552), D2-40 (P=0.877), and lymphovascular invasion by H&E (P=0.009). The limitation of small sized study might resulted in these consequences. And because of the fact that the majority of included patients were in stage I or II, relatively favorable prognosis might be reflected to such results.
DISCUSSION
Mechanisms underlying ability to metastasize are poorly understood in the microscopic environment surrounding tumor cells. Migration of carcinoma cells into regional lymphatic vessels at the primary site initially occurs during metastatic progression. Lymphovascular invasion is recognized as tumor cell nests floating within empty spaces, which are surrounded by thin, spindle shaped endothelial cells. D2-40 has been utilized recently to detect lymphovascular invasion more objectively [24]. Because typical breast cancer specimens have dense stromal tissue, artifacts can mimic lymphatic invasion of tumor cells. D2-40 immunohistochemistry has shown to improve accuracy in detecting lymphovascular invasion of breast carcinoma [25].
In breast cancer cell lines, CD44 activation has been shown to promote cytoskeletal remodeling, increase metastasis and invasion [26]. In contrast, several study results showed CD44 as a tumor progression suppressor. Its overexpression correlated with increased disease free survival and overall survival in human breast cancer [10,28]. Previous studies reported that expression of CD44 was related with both favorable and unfavorable prognosis of breast cancer with no clear consensus.
In the present study, we hypothesized that D2-40 and CD44 represent different mechanisms of microinvasive property respectively and thus positivity of both markers would be related with poorer prognosis due to presumed increased metastatic preponderance. We observed shorter disease free survival in the CD44-positive tumor group (P<0.001) and the D2-40 positive tumor group (P<0.001). Patients who were positive to both markers had 46.6 months shorter disease free survival period than those who were negative to at least one marker (P<0.001). Especially, when adjusted with other factors, the hazard ratio was 10.82 (P=0.014) in the CD44 positive alone group and was further increased to 22.72 (P=0.006) in the group with both markers positive.
In conclusion, CD44 may be a prognostic marker associated with unfavorable prognosis. Positivity of both CD44 and D2-40 may be a useful factor in predicting malignant relapse with shorter disease free survival.
These data lend additional informations on the role of CD44 as a tumor suppressor or promoter. As CD44-positive breast cancer cells have been shown to spread early in the course of disease, inhibitors of CD44-induced protease may contribute to counteract invasion and dissemination of tumor cells.
Although hazard ratio was increased with statistical significance in the CD44-positive group and the CD44, D2-40 positive group, the range of the confidence interval was rather wide. In addition, some factors were statistically insignificant in multivariate analysis, presumably due to the relatively small sample size. Further investigation with more sample size would provide better understanding with statistical significance.
We concluded that expression of CD44 and D2-40 is associated with unfavorable prognosis. Concurrent positivity of both markers may be more useful in predicting malignant relapse following surgical resection with shorter disease free survival in patients with invasive carcinoma of the breast.
Notes
This work was supported by the 2013 Inje University research grant.