SEARCH

SEARCH BY CITATION

Keywords:

  • breast neoplasms;
  • estrogen receptors;
  • epithelial hyperplasia lacking atypia;
  • risk factors;
  • premalignant breast disease

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Estrogen is associated with many epidemiologic risk factors for invasive breast cancer. Cells that express estrogen receptors (ERs) in epithelial hyperplasia lacking atypia (EHLA) may influence breast cancer progression. We conducted a nested case-control study of 268 women with biopsy-confirmed EHLA to determine whether immunohistochemical expression of ERα in EHLA affects subsequent breast cancer risk. Study subjects could not have a prior or current history of breast cancer or atypical hyperplasia. Immunohistochemical stains in individual lesions and adjacent normal lobules were considered positive if ≥ 10% of epithelial cells stained for ERα. The risk of invasive breast cancer in EHLA patients with ERα-positive normal lobules was twice that of other EHLA patients (95% CI = 1.0–3.8). This risk was not affected by the ERα status of EHLA lesions. ERα expression in adjacent normal lobules increases the moderate breast cancer risk previously associated with EHLA. © 2004 Wiley-Liss, Inc.

Knowledge of molecular markers of breast neoplastic progression may help identify women at high risk for breast cancer. Such markers would facilitate the development of prophylactic therapies for premalignant breast disease.1 Estrogen has long been associated with stimulating the normal growth of breast epithelial cells and may also be associated with many epidemiologic risk factors for invasive breast cancer. Estrogen acts on cells via the estrogen receptors (ERs) to stimulate their mitogenic activity.2 Comparison of normal and precancerous breast biopsies has shown that ER expression is relatively low in normal epithelium and higher in atypical ductal hyperplasia and carcinomas in situ.3, 4 The earliest lesion associated with breast cancer is epithelial hyperplasia lacking atypia (EHLA), which has a 1.5- to 2-fold increased risk of development of breast cancer.5, 6, 7, 8 A slight increase in ER-positive cells has been described in these earliest lesions.3 In our study, we investigated the relationship between ERα expression in EHLA and subsequent risk of invasive breast cancer.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

We conducted a retrospective nested case-control study of women from the Nashville Breast Cohort Study.5, 9, 10, 11 This cohort consists of women who underwent benign breast biopsy at 1 of 3 hospitals in Nashville, Tennessee. Successful follow-up in this cohort has been achieved for 90% of study subjects. Women with prior or concurrent invasive breast cancer were excluded from this cohort. Cohort members were eligible for inclusion in this study if their entry biopsy contained EHLA and were performed between 1965 and 1982. This study subcohort consisted of 1,680 women. The average length of follow-up for these patients was 15.4 years. Cases consisted of all eligible subjects who subsequently developed invasive breast cancer. Two controls were selected for each case matched by age at biopsy and year of biopsy. To be selected as a control for a specific case, a patient had to be in the risk set for that case at the time when the case patient developed breast cancer,12 that is, her breast cancer-free follow-up had to be at least as long as that of the case under consideration. Selected controls could subsequently become cases, which in fact happened for 8 women. Each of these subjects was included as a control in one set of matched case-control triplets and as a case in another. There were 92 cases who developed breast cancer on follow-up who were matched to 184 controls. Since 8 of these controls subsequently became cases, this study consisted of 268 distinct women.

We cut 2 consecutive sections from formalin-fixed, paraffin-embedded tissue of the diagnostic biopsies of 268 study subjects. One section was stained with hematoxylin and eosin (H&E) to verify the presence of EHLA, and the other was used for ERα immunohistochemistry. The diagnostic criteria for EHLA are those previously given for moderate or florid hyperplasia.5, 13, 14, 15, 16 We performed immunostaining using an affinity-purified mouse monoclonal antibody against ERα (6F-11; Zymed Laboratories, South San Francisco, CA) and the streptoavidin-biotin immunoperoxidase method. Immunohistochemical stains were evaluated in individual EHLA lesions and in at least 3 adjacent histologically normal ducts and lobular units without knowledge of subsequent cancer outcome. Only nuclear staining in epithelial cells was evaluated; slides were classified as ER-positive when more than 10% of cells in at least one evaluated lobular unit exhibited positive nuclear staining. The cutoff of 10% staining used to define positivity was chosen because it was easy to apply consistently. Although there was considerable variability in the degree of staining, most ER-positive lobular units stained far more than 10% of cells and there were only a few biopsies with a staining frequency near our 10% cutoff. This definition of positivity was established before any data analysis was performed. It dichotomizes our biopsies into those that were virtually free of staining in their adjacent lobular units and those with extensive staining in at least one such unit. There were 6 cases and 7 controls who were excluded because of a complete absence of ERα staining. This left 86 cases and 177 matched controls for analysis.

We used conditional logistic regression analysis to assess the association between ERα expression and breast cancer risk (Stata Statistical Software Release 8.0; Stata, College Station, TX).17 Analyses are adjusted for both age at biopsy and year of biopsy by the matching criteria used to select controls and by including these variables in the regression models. All p-values are derived with respect to 2-sided alternative hypotheses.

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Examination of the stained slides revealed that expression of ERα was detectable in epithelial nuclei of normal breast ducts and lobular units (Fig. 1a and b). There was variation in the staining pattern and intensity of ERα expression in hyperplastic lesions and in normal epithelium of ducts and lobules (Fig. 1bd). Usually, epithelial cells at the periphery of positive-staining hyperplastic lesions stained stronger than the inner cells (Fig. 1b and d). Figure 1(a) shows a typical degree of staining in a positive lobular unit. Although the degree of staining varied, the overwhelming majority of subjects had either substantially less or substantially more than 10% of ERα-positive cells. Patients with more than 10% of ERα-positive cells of lobular units and ducts showed a 2-fold increase in the risk for breast cancer development (Table I). The presence of positive staining within the EHLA lesions as opposed to the adjacent lobular units did not indicate a further increase in breast cancer risk (Table I). The pattern of ERα immunoexpression in EHLA was similar between cases and controls (Fig. 1). There was a positive correlation between staining in EHLA and in the adjacent lobular units (Pearson's correlation coefficient = 0.41). This degree of correlation was similar for cases and controls. Seventy-one percent of controls and 83% of cases were concordant with respect to the presence or absence of staining in EHLA and adjacent lobules. Hence, the small number of discordant women makes it impossible to estimate accurately the cancer risk of either positive lobules with negative EHLA or positive EHLA with negative lobules.

thumbnail image

Figure 1. Immunohistochemical staining for ERα. (a) Nuclei of epithelial cells of lobular units are positive for estrogen receptor. (b and d) EHLA showing more than 10% of cells positive for ERα. (c) The majority of the hyperplastic cells (> 90%) are negative. Note the marked staining of most of the cells at the periphery of the spaces (b, d). Streptavidin-biotin peroxidase: 100× (a, b) and 200× (c, d).

Download figure to PowerPoint

Table I. Estrogen Receptor Expression in Normal Lobular Units and EHLA in the Entry Biopsies of Cases and Controls
ERα expressionControlsCasesOdds ratio95% confidence intervalp-value
  • Samples from 7 controls and 6 cases were excluded because of complete absence of ERα staining.

  • 1

    Denominator of odds ratio.

  • 2

    At least 10% of epithelial cells showed positive nuclear staining for ERα.

  • 3

    EHLA lesions not present on deeper cuts used for ERα staining.

In normal lobules     
 Negative63201.01  
 Positive2114661.991.0–3.80.04
  Total17786   
In EHLA     
 Negative79251.01  
 Positive292571.870.95–3.70.07
 Missing364   
  Total17786   

There was no evidence that a history of hormonal replacement therapy increased breast cancer risk among study subjects, regardless of ERα expression in their EHLA lesions or adjacent lobular units.

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

All of our study subjects had EHLA without a current or prior diagnosis of breast cancer or atypical hyperplasia. Hence, this study evaluates future breast cancer risk in women who have unequivocally benign lesions. In contrast, Khan et al.18, 19 performed a cross-sectional study in which benign epithelium from cases with concurrent breast cancer was contrasted with that of cancer-free controls. These investigators found higher levels of ER expression in the benign epithelium of cases compared to controls. However, their study does not demonstrate that this ER expression preceded the development of breast cancer. Shaaban et al.20 compared ERα expression in EHLA lesions of women who subsequently developed breast cancer (cases) compared to women who remained breast cancer-free (controls). They found that cases had significantly higher ERα expression levels than controls; expression levels in normal lobules were also elevated, but not significantly. Although these findings are consistent with ours, these investigators did not estimate the risk of subsequent breast cancer associated with their findings. It is possible that ERβ expression and other factors may also be important determinants of breast cancer risk in these early proliferative lesions.

It should be noted that EHLAs are unproven as deterministic precursor lesions. Rather, they have been shown to be markers of elevated breast cancer risk, that is, women who have them are at elevated risk, and it is plausible that, undisturbed, an EHLA lesion is more likely to progress to cancer than a normal lobular unit in a women without EHLA. Also, their indicator status is thought to depend on the presence of other EHLA within the breast at the time of biopsy. The presence of positive ERα staining increases the breast cancer risk but does not affect the stochastic nature of this progression. Presumably, the likelihood that EHLA will progress to cancer is greater in women with ERα staining than in women without. However, we make no claim that this progression is inevitable regardless of the patient's ERα status.

In this study, we evaluated ERα expression in normal lobular units and ducts regardless of age or menopausal status. Our results show that patients with increased numbers of ERα-positive epithelial cells are at increased risk of subsequent invasive breast cancer. They are consistent with the hypothesis that the overexpression of ERs in benign breast epithelium may augment estrogen sensitivity, thereby creating a permissive state for other initiating events that can lead to malignancy.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The authors thank Marcia Freudenthal for her assistance with this research, the staff of the 3 Nashville hospitals for their cooperation and aid in patient follow-up and the thousands of women who have participated in their studies.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References