This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Fritz, P. Articles by Zanger, U. M. Search for Related Content PubMed PubMed Citation Articles by Fritz, P. Articles by Zanger, U. M.

Microsomal Epoxide Hydrolase Expression as a Predictor of Tamoxifen Response in Primary Breast Cancer: A Retrospective Exploratory Study With Long-Term Follow-Up

From the Center of Diagnostic Pathology, Robert-Bosch-Krankenhaus, Stuttgart; and the Dr Margarete Fischer-Bosch Institute of Clinical Pharmacology, Stuttgart, Germany.

Address reprint requests to Ulrich M. Zanger, PhD, Dr Margarete Fischer-Bosch Institute of Clinical Pharmacology, Auerbachstr 112, D-70376, Stuttgart, Germany; email uli.zanger{at}ikp-stuttgart.de

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: It has been suggested that estrogen receptor–independent high-affinity binding sites for antiestrogens could limit their local bioavailability and response. Microsomal epoxide hydrolase (mEH) was recently shown to be a component of the antiestrogen binding site complex. We investigated whether mEH expression in primary breast tumors is related to disease outcome and to the efficacy of tamoxifen treatment.

PATIENTS AND METHODS: Expression of mEH was semiquantitatively assessed by immunohistochemistry in sections prepared from archival paraffin blocks of primary breast cancers from 179 patients with a mean follow-up time of 81 months.

RESULTS: Expression of mEH was correlated with poor disease outcome in all patients (P < .01; n = 179) and in patients receiving tamoxifen (P < .01; n = 78), but not in patients not treated with tamoxifen. Moreover, mEH was an independent prognostic factor by Cox regression analysis.

CONCLUSION: The results of this first exploratory study suggest that mEH expression in primary breast cancer could be of predictive value for response to tamoxifen treatment and/or may be a novel independent prognostic factor for survival. The results are in agreement with the model that mEH participates in an estrogen receptor–independent tamoxifen- binding complex.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
BREAST CANCER IS the most frequent cancer occurring in women from developed countries and one of the leading causes of tumor death in women. Therapy of breast cancer is widely guided by assessing the traditional tumor/nodal/metastasis (tumor-node-metastasis) staging variables of the tumor.¹ More recently, tumor type grading and steroid receptor status have gained wider acceptance to support further therapy decisions. Additional newer prognostic variables under current investigation include c-erbB-2, p53, and Ki67.² After surgery, patients are either not treated or receive combination chemotherapy, and a subset of patients, usually those with tumors expressing estrogen and/or progesterone receptors (ER/PR), are treated with the nonsteroidal antiestrogen tamoxifen for up to 5 years.³ Tamoxifen significantly reduces the risk of recurrence and deaths in pre- and postmenopausal women.⁴ However, treatment response is variable, and approximately one third of receptor-positive patients have no benefit from tamoxifen treatment, whereas occasionally response of receptor-negative patients is reported.^1,3 The growth-inhibiting effects of tamoxifen are generally thought to be mediated by competitive inhibition of estrogen binding to the ER, thereby inhibiting ER-dependent expression of estrogen-regulated growth factors and angiogenic factors that stimulate tumor growth by autocrine or paracrine mechanisms.⁵ However, almost 20 years ago a second, ER-independent saturable high-affinity binding site for tamoxifen and other nonsteroidal antiestrogens was identified by detailed binding studies.⁶ Whereas numerous studies were undertaken to elucidate putative agonistic and antagonistic functions of this so-called antiestrogen binding site (AEBS⁷), its molecular nature remained elusive. One of the protein components of AEBS was identified only recently as the microsomal epoxide hydrolase (mEH; International Enzyme Classification System 3.3.2.3), a xenobiotic drug metabolism and carcinogen-activating enzyme that is ubiquitously expressed but preferentially found in the liver.⁸ Sutherland et al⁶ estimated high-affinity AEBS to be in excess over ER sites in breast tumor cells, and they hypothesized that a significant proportion of antiestrogenic drug could be locally absorbed in vivo, thus limiting the amount available for binding to the ER. Thus the level of mEH in tumor cells could modulate the growth-inhibiting effect of tamoxifen. Provided this hypothesis is correct, variable expression of mEH in the tumor should be related to tamoxifen efficacy in such a way that high expression results in poor response. To test this hypothesis, we analyzed archival paraffin-embedded tissues from 179 breast cancer patients for expression of mEH in primary tumors and correlated these results with overall survival and response to treatment with tamoxifen.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patients
Files of patients with breast cancer (n = 260; surgery performed 1986 through 1988) were retrieved from the archives of surgical pathology and followed up with a questionnaire sent to the general practitioner. The drop-out rate was 81 patients (31.2%) because of unreturned questionnaires (n = 55), loss of paraffin material (n = 14), or unknown steroid receptor status (n = 12), without further selection or bias in the 179 residual cases. The mean follow-up time was 81.1 months (median, 91.0 months; range, 2 to 143 months). Tumor staging was performed according to the tumor-node-metastasis classification system and histologic typing was performed according to World Health Organization guidelines.⁹ Briefly, we discriminated between lobular invasive, ductal invasive breast cancer, and specified types of breast cancer, namely medullar, inflammatory, and mucinous carcinoma. The following data were recorded in each case: tumor-node-metastasis, differentiation of the tumor (high, G1; moderate, G2; low, G3), ER and PR status, menopausal status, age, and type of therapy. ER and PR were analyzed biochemically with charcoal and dextran using 20 fmol/mg protein as the cutoff point. All 179 patients had initially undergone either ablatio mammae surgery or a breast-conserving resection of their primary carcinomas. Of all included patients, 159 patients were being treated in the adjuvant setting and 20 received palliative treatment after relapse. Adjuvant treatment with tamoxifen was given to 72 patients, whereas 46 patients were not further treated, 37 received adjuvant treatment with either cyclophosphamide, methotrexate, and fluorouracil or anthracycline regimes (either mitoxantrone and cyclophosphamide; doxorubicin and cyclophosphamide; fluorouracil, doxorubicin, and cyclophosphamide; or belladonna), and 16 were treated with radiation (Table 1). Of all 78 tamoxifen-treated patients, 63 received tamoxifen alone, whereas 15 also received chemotherapy.

Antibody Characterization
The polyclonal rabbit antibody against human mEH and its use in immunohistochemistry have been described by us previously.¹⁰ Western blot analysis was performed with microsomes prepared from several fresh breast tumors. Proteins were separated in a 10% sodium dodecyl sulfate–polyacrylamide gel, transferred to nitrocellulose (Schleicher und Schüll, Einbeck, Germany) and probed with anti-mEH serum (dilution 1:2000). Detection was performed with alkaline phosphatase-conjugated antirabbit immunoglobulins (1:1000, Dako, Glostrup, Denmark) and 5-bromo-4-chloro-indolyl-phosphate/nitro blue tetrazolium as substrate.

Immunohistochemical Methods
Staining was performed by the avidin-biotin-peroxidase complex method.¹¹ The slides were incubated overnight with anti-mEH serum diluted 1:100. A biotinylated secondary goat antirabbit antibody (Vector, Burlingame, CA) was used at 1:200 dilution. Visualization of the avidin-biotin-peroxidase complex (Vector) was achieved with H₂O₂ and diaminobenzidine (Sigma, Deisenhofen, Germany) resulting in brown-colored final reaction products. The stained breast cancer sections were scored blindly with respect to clinical patient data. For nontumorous tissue only, presence or absence of mEH staining was recorded using negative controls for comparison. In tumor tissues, staining intensity was visually scored in four degrees: absent (0), weak (1), moderate (2), and strong (3). The percentage of mEH-positive tumor cells was graded as absent (0), 1% to 10% (1), 11% to 50% (2), 51% to 80% (3), and more than 80% (4). The immunoreactive score (IRS) index was calculated as the product of both values.

Statistical Methods
Data assessment was made using the statistics software program SPSS (SPSS Software GmbH, Munich, Germany). Survival curves were established according to the Kaplan-Meier method, and comparisons between survival curves were performed with the log-rank test. Overall survival was calculated from the date of surgery to death or to the date of the last patient contact. Disease-free survival was measured from the date of surgery until the time of relapse, cancer-related death, or last contact. Patients who died from unrelated causes were considered as censored by the time of their death. To define a cutoff point for mEH expression, the minimal P value approach was used and Bonferroni correction for multiple testing was applied.¹² The IRS 2 was used for all further analyses. Multivariate analyses were performed using Cox regression analysis in a model with T, N, M, G, and ER and PR status. Association between mEH expression and other parameters such as age, tumor size, nodal status, and hormonal status was assessed by the ² test.

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Specificity of the mEH Antibody
In liver microsomes, the polyclonal antibody used in this study reacted with a single 50-kd protein by Western blot analysis (Fig 1A). To examine its specificity in breast cancer, whole homogenates and microsomes prepared from several fresh breast tumor tissue samples were analyzed. A single 50-kd protein was recognized, which comigrated with mEH in control liver microsomes (Fig 1A). For immunohistochemical analyses, the antibody’s capability to detect mEH in tissue was confirmed with liver sections that were used as positive controls, whereas absence of mEH in tissue was determined by comparison with negative control sections obtained by omitting either the primary or the secondary antibody.

View larger version (147K): [in this window] [in a new window]   Fig 1. Immunological analysis of mEH expression. (A) Western blot of homogenate (BC-H) and microsomes (BC-M, 10 µg each) prepared from breast cancer tissue of two patients (nos. 1 and 2); L-M, liver microsomes (1 µg). (B-D) Breast cancer paraffin sections showing (B) negative, (C) moderate, and (D) strongly positive mEH staining (counterstained with hematoxylin).

Univariate Analysis
Kaplan-Meier survival curves were calculated to evaluate the prognostic value of established parameters for overall survival. As expected, the classical prognostic factors, ie, tumor size, nodal status, histologic grade, and steroid receptor status were all significantly associated with overall survival, whereas age and menopausal status were not (Table 1). Expression of mEH in the tumor was independent of all other parameters, as shown by ² testing (Table 1).

To evaluate a possible association between mEH expression and disease outcome, different IRS subgroups were initially defined and Kaplan-Meier analysis for overall survival was performed. In these analyses, it became apparent that subgroups with higher mEH expression levels consistently had shorter mean survival times than subgroups with lower expression (data not shown). This suggested the use of a single cutoff value to simplify further analyses. To select a value, the minimal P value approach was used, and cutoffs from IRS 1 to 6 were compared by Kaplan-Meier analysis. The statistical results, with and without application of the Bonferroni correction for multiple statistical testing, are listed in Table 2. The most discriminative value (IRS 3) was used consistently for further subgroup analyses and for multivariate Cox regression analysis. The mean survival time of the subgroup with mEH 3 (n = 139) was 111.6 months, compared with 82.5 months of the group with mEH 3 (n = 40; log-rank test; P = .005). At 5 years, 77.5% of patients with low mEH expression in their tumors lived compared with 58.1% in the positive subgroup; at 10 years, 64.6% of patients with low mEH expression in their tumors lived compared with 42.8% the positive subgroup (Fig 2A).

View larger version (21K): [in this window] [in a new window]   Fig 2. Relationship between mEH expression and overall survival. The Kaplan-Meier survival curves shown are for subgroups with low (immunoreactive score index IRS 2) and high (IRS 3) mEH expression among (A) all patients under study (n = 179), (B) patients treated with tamoxifen (n = 78), and (C) patients not treated with tamoxifen (n = 101).

Subgroup Analysis
Stratification for the different types of treatment revealed no significant association between mEH expression and survival in any of these treatment groups except for patients treated with tamoxifen, who lived significantly longer if they had low mEH expression (Table 1). There was also no significant association for the patient group in the palliative situation, which included six patients treated with tamoxifen (Table 1). In Fig 2, Kaplan-Meier survival curves are exemplarily shown for (A) all patients, (B) for all patients receiving tamoxifen, and (C) for patients receiving other types of therapy. Patients treated with tamoxifen who had low expression of mEH in their tumors lived significantly longer (mean, 115.9 months; n = 62) as compared with patients in the high-mEH group (n = 16; mean, 75.5 months; log-rank test, P = .009; Fig 2B). At 5 and 10 years, the differences in the survival rates between patients with mEH 3 and mEH 2 were 26.1% and 31.2%, respectively (Fig 2B). For patients not treated with tamoxifen, a difference in survival between high-mEH and low-mEH groups was also apparent but did not reach significance (n = 101; log-rank test, P = .15; Fig 2C).

Although disease-free survival could not be determined with certainty as it was not consequently recorded in this retrospective study, the effect of mEH expression on this end point was also analyzed. A qualitatively similar association was found, which however did not reach significance. The 5-year and 10-year disease-free survival rates were 53.3% and 33.3%, respectively, for all tamoxifen-treated patients with high mEH expression (n = 16) and 69.4% and 59.4%, respectively, for patients in the low mEH group (n = 62; log-rank test, P = .086).

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
To our knowledge, this retrospective, exploratory study represents the first investigation of mEH expression in tumor cells as a prognostic factor for overall survival of women with breast cancer and as a predictive factor for the response to tamoxifen treatment. We found that mEH is expressed at higher levels in approximately one fourth to one third of all primary tumors regardless of age, menopausal status, and ER/PR status, and that high expression is significantly associated with poor disease outcome. Furthermore, we observed that high mEH expression was associated with poor response to tamoxifen treatment, whereas other treatment groups showed less or no such association.

mEH is a highly conserved and widely expressed protein, yet its biologic function is still unclear.^13,14 It is involved in metabolic activation or detoxification of epoxides, which are generated from aromatic and olefinic structures by microsomal mono-oxygenases. Its role in carcinogenesis by polycyclic aromatic hydrocarbons is particularly well established.¹⁵ Because both arylhydrocarbon hydroxylase (AHH) and mEH were shown to be simultaneously induced by tobacco smoke in lung cancer,¹⁶ it may be hypothesized that mEH expression in the tumor largely reflects total body accumulation of tobacco-inhaled carcinogens. These highly lipophilic substances are stored in adipose tissue and lead to continuous activation of the AH receptor, which mediates enzyme induction and participates in growth regulation by blocking apoptosis.¹⁷ Thus expression of AHH and mEH may be useful markers for tumor aggressiveness and overall prognosis. Increased AHH activity was indeed related to poor prognosis in breast cancer,¹⁸ and both AHH and mEH were shown to have prognostic significance in tobacco-related lung cancer.¹⁹ Unfortunately, information on smoking history was not available for the patients of this retrospective study. An association between mEH expression, smoking, and overall prognosis may thus not be excluded.

There is, however, evidence for a more direct involvement of mEH expression in hormonal treatment of breast cancer. Recently, it was shown that mEH is one of the two protein components of the heterodimeric AEBS complex and that the capability of AEBS to bind antiestrogens with high affinity depends on the presence of mEH.⁸ Sutherland et al,⁶ who first described AEBS, showed that its concentration in breast cancer may even exceed that of the ER. Furthermore, because tamoxifen binds with higher affinity to AEBS (K_d approximately 1 nmol/L) than to the ER (K_d approximately 80 nmol/L), a significant proportion of antiestrogenic drug could be bound to AEBS, leading to reduced available concentrations for binding to the ER, thus allowing the tumor to grow in the presence of antiestrogen.^6,20 Provided that mEH in breast tumors indeed represents the high-affinity AEBS, our observation of mEH expression related to poor prognosis in the tamoxifen-treated but not in the tamoxifen-free subgroup (Fig 2) supports this hypothesis.

AEBS was found to be present at highest levels in the liver but it occurs in many other animal and human tissues, including breast tumors,^6,8,20 which is in agreement with the tissue distribution of mEH.^13,14,21 Initially, mEH was identified in purified AEBS preparations obtained from rat liver by using a photoaffinity analog of tamoxifen. It was shown further that mEH is necessary to reconstitute AEBS in human hepatoma cells, because high-affinity binding of tamoxifen decreased when mEH expression was blocked by an antisense strategy.⁸ It is, therefore, reasonable to assume that mEH expression reflects the presence of AEBS in breast cancer tumor cells.

Expression of mEH in breast tumor tissue was reported previously; however, there are discrepancies with this study. Whereas Murray et al²¹ detected mEH in immunohistochemical studies in 89% of the tumors, our study detected mEH in only 34% (IRS > 0, Table 2). This may be explained by the use of different antibodies. We used an antibody highly specific for the microsomal form of epoxide hydrolase, which is encoded by a single gene on chromosome 1 in humans.^13,14 It is possible that this antibody missed the immunohistochemical detection of mEH expressed at very low levels, whereas AEBS was usually detected in all tissue samples by more sensitive radioactive antiestrogen detection methods.

Finally, it is important to consider the metabolism of tamoxifen. There are two major metabolic routes in humans that lead to 4-OH-tamoxifen and to N-desmethyltamoxifen. Although 4-OH-tamoxifen is a minor metabolite, it is a potent antiestrogen that binds to the ER with approximately 100-fold higher affinity than tamoxifen and with similar affinity to 17-beta-estradiol.^3,7 On the other hand, 4-OH-tamoxifen is less stable than tamoxifen and is enzymatically isomerized from the trans-configuration to the cis-configuration, which displays less antiestrogenic potency. With respect to our study, it is important that 4-OH-tamoxifen was shown to bind to AEBS with similar affinity to tamoxifen.⁶ Additional minor metabolic pathways of tamoxifen lead to other metabolites, some of which were shown to have estrogenic properties.^3,7 Therefore, the enzymes catalyzing these various biotransformation steps and their interindividual variability must be considered as factors determining the intratumoral balance of metabolites with antiestrogenic and estrogenic properties.

In conclusion, we obtained first results indicating that high expression of mEH in breast tumors may adversely influence the efficacy of antihormonal treatment. To explore the clinical consequences, our findings need to be confirmed in more extensive larger follow-up studies. Furthermore, the detailed mechanism of the impact of mEH expression in breast carcinoma on response to tamoxifen, including a possible role in tamoxifen-binding and/or metabolism, remains to be clarified.

	ACKNOWLEDGMENTS

 
Supported by the Robert Bosch Foundation, Stuttgart, and by a grant from the Mildred Scheel Foundation, Bonn, Germany (grant no. 10-0952-Ei3).

We thank Kerstin Gawronski for excellent technical assistance, Nadja Gugeler for statistical advice, and Hiltrud Brauch, Stuttgart, for critically reading the manuscript. Furthermore, we thank the numerous unnamed participating general practitioners and patients.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. Hortobagyi GN: Treatment of breast cancer. N Engl J Med 339: 974-984, 1998[Free Full Text]

2. Porter-Jordan K, Lippman ME: Overview of the biologic markers of breast cancer. Hematol Oncol Clin North Am 8: 73-100, 1994[Medline]

3. Osborne CK: Tamoxifen in the treatment of breast cancer. N Engl J Med 339: 1609-1618, 1998[Free Full Text]

4. Early Breast Cancer Trialists’ Collaborative Group: Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy: 133 randomized trials involving 31,000 recurrences and 24,000 deaths among 75,000 women. Lancet 339:1-15, 1992

5. MacGregor JI, Jordan VC: Basic guide to the mechanism of antiestrogen action. Pharmacol Rev 50: 151-196, 1998[Abstract/Free Full Text]

6. Sutherland RL, Murphy LC, Foo MS, et al: High-affinity anti-oestrogen binding site distinct from the oestrogen receptor. Nature 288: 273-275, 1980[Medline]

7. Jordan VC, Murphy CS: Endocrine pharmacology of antiestrogens as antitumor agents. Endocrine Rev 11: 578-610, 1990[Medline]

8. Mesange F, Sebbar M, Kedjouar B, et al: Microsomal epoxide hydrolase of rat liver is a subunit of the anti-oestrogen-binding site. Biochem J 334: 107-112, 1998

9. World Health Organization: Histological Typing of Breast Tumors, ed 2. Geneva, Switzerland, World Health Organization, 1981

10. Fritz P, Behrle E, Zanger UM, et al: Immunohistochemical assessment of human microsomal epoxide hydrolase in primary and secondary liver neoplasm: A quantitative approach. Xenobiotica 26: 107-116, 1996[Medline]

11. Hsu SM, Raine L: Protein A, avidin, and biotin in immunohistochemistry. J Histochem Cytochem 29: 1349-1353, 1981[Abstract]

12. Altman DG, Lausen B, Sauerbrei W, et al: Dangers of using "optimal" cutpoints in the evaluation of prognostic factors. J Natl Cancer Inst 86: 829-835, 1994[Free Full Text]

13. Skoda RC, Demierre A, McBride OW, et al: Human microsomal xenobiotic epoxide hydrolase: Complementary DNA sequence, complementary DNA-directed expression in COS-1 cells, and chromosomal localization. J Biol Chem 263: 1549-1554, 1988[Abstract/Free Full Text]

14. Beetham JK, Grant D, Arand M, et al: Gene evolution of epoxide hydrolases and recommended nomenclature. DNA Cell Biol 14: 61-71, 1995[Medline]

15. Miyata M, Kudo G, Lee YH, et al: Targeted disruption of the microsomal epoxide hydrolase gene: Microsomal epoxide hydrolase is required for the carcinogenic activity of 7,12-dimethylbenz[a]anthracene. J Biol Chem 274: 23963-23968, 1999[Abstract/Free Full Text]

16. Petruzzelli S, De Flora S, Bagnasco M, et al: Carcinogen metabolism studies in human bronchial and lung parenchymal tissues. Am Rev Respir Dis 140: 417-422, 1989[Medline]

17. Zaher H, Fernandez-Salguero PM, Letterio J, et al: The involvement of aryl hydrocarbon receptor in the activation of transforming growth factor-beta and apoptosis. Mol Pharmacol 54: 313-321, 1998[Abstract/Free Full Text]

18. Pyykkö K, Tuimala R, Aalto L, et al: Is aryl hydrocarbon hydroxylase activity a new prognostic indicator for breast cancer? Br J Cancer 63: 596-600, 1991[Medline]

19. Bartsch H, Hietanen E, Petruzzelli S, et al: Possible prognostic value of pulmonary AH-locus-linked enzymes in patients with tobacco-related lung cancer. Int J Cancer 46: 185-188, 1990[Medline]

20. Jordan VC: Biochemical pharmacology of antiestrogen action. Pharmacol Rev 36: 245-276, 1984[Medline]

21. Murray GI, Weaver RJ, Paterson PJ, et al: Expression of xenobiotic metabolizing enzymes in breast cancer. J Pathol 169: 347-353, 1993[Medline]

Submitted December 20, 1999; accepted July 24, 2000.

This article has been cited by other articles:

				M. B. Buck, P. Fritz, J. Dippon, G. Zugmaier, and C. Knabbe Prognostic Significance of Transforming Growth Factor {beta} Receptor II in Estrogen Receptor-Negative Breast Cancer Patients Clin. Cancer Res., January 15, 2004; 10(2): 491 - 498. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Fritz, P. Articles by Zanger, U. M. Search for Related Content PubMed PubMed Citation Articles by Fritz, P. Articles by Zanger, U. M.