Clinical Role of Multidrug Resistance Protein 1 Expression in Chemotherapy Resistance in Early-Stage Breast Cancer: The Austrian Breast and
http://www.100md.com
《临床肿瘤学》
the Departments of Internal Medicine I, Pathology, Surgery, and Gynecology, Medical University of Vienna, Vienna
Department of Pathology and Third Medical Department, Salzburg Hospital, Salzburg
Departments of Pathology and Internal Medicine, University of Graz, Graz
Department of Pathology-Histology, Hanusch Medical Center, Vienna, Austria
ABSTRACT
PATIENTS AND METHODS: Immunostaining for MRP1 was performed on tissue sections from 516 premenopausal, hormone receptor-positive breast cancer patients with stage I and II disease. Statistical analyses were performed to assess the effect of MRP1 expression on survival and to test for interaction between MRP1 expression and treatment.
RESULTS: MRP1 expression independently predicted shorter relapse-free survival (RFS) and overall survival (OS) in patients treated with cyclophosphamide, methotrexate, and fluorouracil (CMF; RFS: hazard ratio [HR] = 1.48; 95% CI, 1.16 to 1.88; P = .002; OS: HR = 1.82; 95% CI, 1.10 to 3.01; P = .02), but it did not predict shorter RFS and OS in patients who received tamoxifen plus goserelin (RFS: HR = 0.99; 95% CI, 0.74 to 1.31; P = .9; OS: HR = 0.68; 95% CI, 0.40 to 1.15; P = .1). Tests for interaction between MRP1 expression and treatment were statistically significant for both RFS (P = .04) and OS (P = .006).
CONCLUSION: Our data suggest that MRP1 expression plays an important role in the clinical resistance to adjuvant CMF chemotherapy but does not seem to affect response to adjuvant endocrine treatment with tamoxifen plus goserelin. Thus, MRP1 may be a useful marker for the selection of patients with early-stage breast cancer for the appropriate adjuvant therapy after prospective confirmatory evaluation.
INTRODUCTION
The multidrug resistance protein 1 (MRP1), also called ABCC1, is another member of the ABC transporter family.4 Overexpression of this protein in tumor cells confers resistance to various anticancer drugs, such as anthracyclines or methotrexate,4-7 which are drugs that are part of the chemotherapy regimens currently used in the treatment of breast cancer.8 Prior studies suggest an association between MRP1 protein expression and shorter disease-free survival and overall survival (OS) of breast cancer patients.9 However, the heterogeneity of these studies as a result of differences in the study populations, the number of patients examined, the length of follow-up, the use of different types of adjuvant endocrine treatment and/or chemotherapy regimens, and the detection methods used does not allow a definitive conclusion with regard to the impact of MRP1 expression on clinical outcome.
To further evaluate the clinical role of MRP1 in breast carcinomas, we studied the relationship between MRP1 expression of the primary tumors and survival of the patients who were enrolled onto the Austrian Breast and Colorectal Cancer Study Group (ABCSG) Trial 5, a prospective randomized trial comparing the efficacy of cyclophosphamide, methotrexate, and fluorouracil (CMF) chemotherapy with a combination endocrine treatment.
PATIENTS AND METHODS
Tumor Specimens
All tumor specimens were obtained at the time of surgery before adjuvant therapy. Formalin-fixed, paraffin-embedded tumor blocks containing primary breast cancer tissue from patients who participated in ABCSG Trial 5 were retrospectively collected from the major participating centers. Details with regard to the collection of the samples and the preparation of slides have been published as part of a study describing the role of p27Kip1 expression in this population.11
Immunohistochemistry
A hematoxylin and eosin-stained slide was prepared from each block and used for pathologic confirmation of present invasive breast cancer by an experienced breast pathologist. Immunohistochemical analysis reported in this study was carried out in a single laboratory and performed as described previously.12 Briefly, tissue sections of 4-μm thickness were prepared, mounted on poly-L-lysine-coated slides, deparaffinized, and rehydrated with distilled water. Endogenous peroxidase activity was blocked by incubation in 0.06% hydrogen peroxide for 10 minutes at room temperature. After washing in phosphate-buffered saline (pH 7.4), the tissues were preincubated for 20 minutes in normal serum (dilution 1:20; Dako, Glostrup, Denmark) before a 90-minute incubation with the anti-MRP1 monoclonal antibody MRPr1 (dilution 1:50; Alexis, Lufelfingen, Switzerland). Antibody binding was detected by the avidin-biotin-peroxidase method. Bound peroxidase was developed with 3,3'-diaminobenzidine (Dako). The slides were counterstained with Mayer's Hmalaun and mounted with Aquatex (Merck, Darmstadt, Germany). All of the washes were performed in phosphate-buffered saline (pH 7.4).
Normal human kidney tissue served as positive control for MRP1 expression. In addition, we used breast cancer specimens known to be positive or negative for MRP1 as positive and negative controls, respectively. Negative controls without the primary antibody were performed as described earlier. Staining of tumor cells was examined by an experienced breast pathologist who was blinded to the clinical outcome of the patients. All tumor cells on each slide were evaluated, and interpretation of the results was limited to the invasive portion of the tumor. MRP1 immunostaining was classified as previously described by us,12 and was as follows: negative, 0% of tumor cells showing staining reactivity; low, less than 10% of tumor cells showing staining reactivity; intermediate, 10% to 30% of tumor cells showing staining reactivity; and high, more than 30% of tumor cells showing staining reactivity. To assess interobserver reproducibility, 100 randomly selected patients were analyzed by a second investigator, and {kappa} statistics were used to assess interobserver reliability. The observed {kappa} was 0.72, and therefore, interobserver reliability was considered satisfactory (data not shown). The rare discrepant cases were reassessed together by both investigators using a double-headed microscope, and a consensus was reached.
Statistical Analysis
Associations of MRP1 expression with age, tumor size, lymph node status, tumor grade, estrogen receptor (ER) expression, and progesterone receptor (PgR) expression were assessed by the Spearman rank correlation. Survival probabilities were estimated with the product limit method according to Kaplan and Meier.13 Survival time was defined as the period between the time of randomization and death (OS) or the period between the time of randomization and documented relapse (relapse-free survival [RFS]). Relapse was defined as the first reappearance of breast cancer at any local, distant, or contralateral site. Patients who died because of reasons other than breast cancer were considered as censored with death. Survival times of patients still alive were censored with the date of the last follow-up. Differences between survival curves were analyzed using the log-rank test. To describe the unadjusted effects of covariates on RFS and OS, univariate Cox proportional hazards regression models were used. Multiple Cox proportional hazards regression models were used to assess the independent effects of MRP1 expression on RFS and OS.14 In addition, multiple Cox proportional hazards regression models with backward elimination of nonsignificant variables were applied. We used a backward elimination algorithm with a 0.1 significance level for removing explanatory variables. Variables were coded as follows: age, less than 35 years or ≥ 35 years; tumor size, pT1, pT2, or pT3; lymph node status, 0, 1 to 3, 4 to 10, or more than 10 axillary lymph node metastases; tumor grade, grade 1, grade 2, or grade 3; ER, negative, positive, or strongly positive; PgR, negative, positive, or strongly positive; treatment, combination endocrine therapy or chemotherapy; p27Kip1, less than 50% or ≥ 50%; MRP1, negative, low, intermediate, or high; and an interaction term, the product of MRP1 and treatment (MRP1 x treatment). Tumor size, lymph node status, tumor grade, ER status, PgR status, and MRP1 expression were included as ordered factors in the statistical analyses on the assumption that the relative risk between two successive categories would remain the same. The Cox proportional hazards regression model was also applied to assess interactions between treatment and the other covariates. All P values are results of two-sided tests. SPSS 11.5.1 statistical software (SPSS Inc, Chicago, IL) was used for calculations.
RESULTS
MRP1 immunostaining was both membranous and cytoplasmic and ranged from 0% to 100% (median, 10%) of the breast cancer cells. For comparisons with clinical parameters, MRP1 expression was categorized as negative, low, intermediate, and high. MRP1 expression was negative in 148 patients (29%), low in 90 patients (17%), intermediate in 130 patients (25%), and high in 148 patients (29%).
When we examined the correlations between MRP1 expression and clinical parameters, MRP1 expression showed a weak positive correlation with tumor size (r = 0.115; P = .009) and tumor grade (r = 0.096; P = .03; Table 3). In contrast, MRP1 expression was not significantly correlated with age, lymph node status, hormone receptor status, treatment, and p27Kip1 expression (Table 3). Similar results were obtained when MRP1 expression was analyzed as a continuous variable (data not shown).
At a median follow-up of 5.6 years, 105 patients (20%) had relapsed (60 CMF patients and 45 endocrine therapy patients), and 38 patients (7%) had died from cancer (22 CMF patients and 16 endocrine therapy patients). The 5-year RFS and OS rates were 80% and 93%, respectively. Univariate analyses demonstrated that younger age, larger tumor size, higher number of positive lymph nodes, lower levels of PgR expression, low p27Kip1 expression, and increasing levels of MRP1 expression were significantly associated with shorter RFS (Table 4). Larger tumor size, higher number of involved lymph nodes, higher tumor grade, low p27Kip1 expression, and higher MRP1 expression were also significantly associated with shorter OS (Table 4).
The independent effects of MRP1 expression on RFS and OS were assessed by multiple Cox proportional hazards regression models. All variables listed in Table 4 were included in the models. Because previous studies suggested that MRP1 expression may be associated with resistance to CMF chemotherapy, we also tested for the existence of interaction between MRP1 expression and treatment by incorporating an interaction term, the product of MRP1 expression and treatment (MRP1 x treatment), into the Cox model. Age, tumor size, lymph node status, PgR expression, p27Kip1 expression, and the interaction term (MRP1 x treatment) were identified as statistically significant risk factors for RFS (Table 4). Lymph node status, treatment, p27Kip1 expression, and the interaction term (MRP1 x treatment) were independently associated with OS (Table 4). Interestingly, the interaction term was statistically significant for both RFS (P = .04) and OS (P = .006) in these analyses (Table 4). Similar results were obtained when multiple Cox proportional hazards regression models with backward elimination of nonsignificant variables were applied. In particular, the interaction term remained significant after elimination of the nonsignificant variables (P = .04 and P = .004 for RFS and OS, respectively; data not shown).
Because the interaction term was statistically significant for both RFS and OS, we determined the association of MRP1 expression with survival in subgroups of patients treated with either chemotherapy or endocrine therapy. To assess whether the effect of MRP1 expression by treatment is independent of other variables, hazard ratios were estimated before and after adjusting for age, tumor size, lymph node status, tumor grade, ER status, PgR status, and p27Kip1 expression. The results of these analyses are listed in Table 5 and shown in Figure 1. In the cohort of CMF-treated patients, higher MRP1 expression was associated with shorter RFS and OS of the patients. The adjusted hazard ratios for relapse and death were 1.48 (95% CI, 1.16 to 1.88; P = .002) and 1.82 (95% CI, 1.10 to 3.01; P = .02), respectively (Table 5). In contrast, in patients who received combination endocrine therapy, MRP1 expression did not predict RFS and OS. The adjusted hazard ratios for relapse and death were 0.99 (95% CI, 0.74 to 1.31; P = .9) and 0.68 (95% CI, 0.40 to 1.15; P = .1), respectively (Table 5). These data were also shown in Figure 1. At 5 years, the RFS rates for CMF-treated patients were 92%, 84%, 76%, and 56% for patients with negative, low, intermediate, and high MRP1 expression, respectively (Fig 1A); the corresponding 5-year OS rates were 98%, 97%, 91%, and 79%, respectively (Fig 1C). In contrast, the 5-year RFS rates for patients assigned to endocrine therapy were 83%, 81%, 82%, and 84% for patients with negative, low, intermediate, and high MRP1 expression, respectively (Fig 1B); the corresponding 5-year OS rates for patients who received endocrine therapy were 95%, 88%, 98%, and 100%, respectively (Fig 1D). Similar results were obtained when we performed an analysis on patients with T1 and T2 tumors only (data not shown).
These results indicate that MRP1 expression independently predicted shorter RFS and OS in CMF-treated patients and suggest that MRP1 expression is associated with treatment failure in those patients. In contrast, no such association was found for MRP1 and outcome in patients receiving endocrine therapy with tamoxifen plus goserelin.
DISCUSSION
In the current study, we have examined MRP1 expression in a large, homogenous, and well-defined patient population consisting of premenopausal, hormone receptor-positive breast cancer patients with stage I and II disease who were enrolled onto a prospective randomized trial. Our results indicate that MRP1 expression is a prognostic factor for shorter RFS and OS in the total study population. These findings are consistent with previous reports from both our group12,16 and other investigators.17-19 More importantly, however, we found an interaction between MRP1 expression and adjuvant treatment. Patients who were treated with CMF chemotherapy and whose tumors were MRP1 negative experienced a significant reduction in relapse rate and a decrease in mortality compared with patients who had highly expressed MRP1. In contrast, in patients who received tamoxifen plus goserelin, MRP1 expression had no impact on the relapse rate or mortality. These results suggest that MRP1 expression is associated with clinical resistance to CMF chemotherapy but not with tamoxifen resistance and confirm the hypothesis derived from preclinical studies suggesting that methotrexate but not tamoxifen is transported by MRP1.
Several studies have investigated the expression of MRP1 in clinical breast cancer samples.12,16-27 MRP1 mRNA was detected by reverse transcriptase polymerase chain reaction (RT-PCR) in 488 (98%) of 496 breast cancer specimens,19-25 and MRP1 protein expression determined by immunohistochemistry was detectable in 291 (53%) of 544 samples.12,16-18,21,26,27 The higher frequency of MRP1 expression obtained with RT-PCR is not surprising because MRP1 is ubiquitously expressed at low levels. Thus, by using RT-PCR, MRP1 mRNA can be detected in most breast cancer samples.
Only a few previous studies have correlated MRP1 expression with patients' clinical outcome. One study evaluated the impact of MRP1 expression on the prognosis and response to chemotherapy of recurrent breast cancer.17 In 23 previously untreated patients who received first-line chemotherapy for recurrence, the presence of MRP1 protein was associated with a lower response rate (13% of MRP1-positive patients had an objective response compared with 40% of MRP1-negative patients) and with a significantly shorter progression-free survival, independent of the type of chemotherapy. No association between MRP1 expression and response rate or progression-free survival was observed in patients who received chemotherapy after treatment with endocrine therapy. The results of another study suggested that MRP1 may be associated with an increased relapse rate in subgroups of patients with a more favorable prognosis (eg, patients with T1 tumors or node-negative patients) and in node-positive patients who received adjuvant CMF chemotherapy.18 MRP1 expression was also associated with an increased risk of death in patients with T1 tumors and in node-positive patients who were treated with adjuvant CMF chemotherapy. In our previous report, MRP1 expression in primary breast tumors correlated with shorter disease-free survival and OS in a heterogeneously treated cohort of patients.16 In patients treated with preoperative chemotherapy, patients with MRP1 expression had a significantly shorter progression-free survival than patients without MRP1 expression, independent of the type of chemotherapy. In a multivariate Cox regression analysis, prechemotherapy MRP1 expression was identified as an independent prognostic factor for shorter progression-free survival.12 A recently published study has shown MRP1 expression to be related to clinical outcome only in the subgroup of patients treated with anthracycline-based chemotherapy and not in patients who received CMF chemotherapy.19 These data suggest that the clinical role of MRP1 expression in chemotherapy resistance of breast cancer may be restricted to anthracycline resistance. However, the results are based on low numbers of patients, particularly in treatment subgroups, and, therefore, have to be viewed with caution.
The results of these previous studies suffer from various shortcomings and are difficult to compare because of differences in the study populations, the number of patients examined, the length of follow-up, the use of different types of adjuvant endocrine treatment and/or chemotherapy regimens, and the detection methods used. However, the results suggest that MRP1 expression is of clinical relevance in breast cancer and support the results obtained in our present study.
The treatment protocol underlying the present study consists of reasonable treatment options with regard to the management of premenopausal women with early-stage hormone receptor-positive breast cancer. Various randomized trials have shown that ovarian ablation with or without tamoxifen and standard chemotherapy regimens like CMF have similar benefits for premenopausal women with early-stage, receptor-positive breast cancer.8 However, because adjuvant polychemotherapy improves survival, the National Institutes of Health Consensus Development Panel concluded that polychemotherapy should be recommended to the majority of women with localized breast cancer.28 The inclusion of anthracyclines in adjuvant chemotherapy regimens results in a small but statistically significant improvement in survival over nonanthracycline-containing regimens. Four cycles of doxorubicin plus cyclophosphamide have been shown to be equivalent to six cycles of classical CMF, and only six cycles of anthracycline-containing regimens yield superior results but at the cost of greater toxicity.8 Therefore, our present results may have important implications in the treatment of the majority of women with localized breast cancer by the more precise identification of those patients who benefit from adjuvant chemotherapy. Previously, we concluded that, "Overall, our data suggest that the goserelin-tamoxifen combination is significantly more effective than CMF in the adjuvant treatment of premenopausal patients with stage I and II breast cancer."10 In light of the strong predictive effect of MRP1 among patients who received CMF chemotherapy, the results of the current study suggest that patients with intermediate or high levels of MRP1 expression should not be treated with CMF chemotherapy but may benefit from treatment with endocrine therapy, whereas MRP1-negative patients may be considered for treatment with CMF chemotherapy or anthracycline-based chemotherapy. However, this proposed treatment strategy has to be confirmed in well-designed clinical trials.
The current data are encouraging but require validation in a prospective trial before being implemented in clinical routine. Our results confirm the hypotheses derived from preclinical studies that suggest that higher MRP1 expression levels may be associated with clinical drug resistance in breast cancer. Thus, MRP1 may be a useful marker for the selection of patients with early-stage, hormone receptor-positive breast cancer for appropriate adjuvant therapy.
Appendix
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank Ngoc-Anh Huynh for excellent technical assistance and Harald Heinzl, Andreas Gleiss, and Karl Thomanek for their biometrical and editorial expertise.
NOTES
This study was supported by a grant from the Austrian Science Fund (Project No. P15377) and the Center of Excellence in Clinical and Experimental Oncology of the Medical University of Vienna.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Gottesman MM, Fojo T, Bates SE: Multidrug resistance in cancer: Role of ATP-dependent transporters. Nat Rev Cancer 2:48-58, 2002
2. Borst P, Evers R, Kool M, et al: A family of drug transporters: The multidrug resistance-associated proteins. J Natl Cancer Inst 92:1295-1302, 2000
3. Trock BJ, Leonessa F, Clarke R: Multidrug resistance in breast cancer: A meta-analysis of MDR1/gp170 expression and its possible functional significance. J Natl Cancer Inst 89:917-931, 1997
4. Cole SPC, Bhardwaj G, Gerlach JH, et al: Overexpression of a transporter gene in a multidrug-resistant human lung cancer cell line. Science 258:1650-1654, 1992
5. Cole SPC, Sparks KE, Fraser K, et al: Pharmacological characterization of multidrug resistant MRP-transfected human tumor cells. Cancer Res 54:5902-5910, 1994
6. Hooijberg JH, Broxterman HJ, Kool M, et al: Antifolate resistance mediated by the multidrug resistance proteins MRP1 and MRP2. Cancer Res 59:2532-2535, 1999
7. Zeng H, Chen Z-S, Belinsky MG, et al: Transport of methotrexate (MTX) and folates by multidrug resistance protein (MRP) 3 and MRP1: Effect of polyglutamylation on MTX transport. Cancer Res 61:7225-7232, 2001
8. Goldhirsch A, Wood WC, Gelber RD, et al: Meeting highlights: Updated international expert consensus on the primary therapy of early breast cancer. J Clin Oncol 21:3357-3365, 2003
9. Leonessa F, Clarke R: ATP binding cassette transporters and drug resistance in breast cancer. Endocr Relat Cancer 10:43-73, 2003
10. Jakesz R, Hausmaninger H, Kubista E, et al: Randomized adjuvant trial of tamoxifen and goserelin versus cyclophosphamide, methotrexate, and fluorouracil: Evidence for the superiority of treatment with endocrine blockade in premenopausal patients with hormone-responsive breast cancer—Austrian Breast and Colorectal Cancer Study Group Trial 5. J Clin Oncol 20:4621-4627, 2002
11. Pohl G, Rudas M, Dietze O, et al: High p27Kip1 expression predicts superior relapse-free and overall survival for premenopausal women with early-stage breast cancer receiving adjuvant treatment with tamoxifen plus goserelin. J Clin Oncol 21:3594-3600, 2003
12. Rudas M, Filipits M, Taucher S, et al: Expression of MRP1, LRP and Pgp in breast carcinoma patients treated with preoperative chemotherapy. Breast Cancer Res Treat 81:149-157, 2003
13. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
14. Cox DR: Regression models and life tables. J R Stat Soc 34:187-220, 1972
15. Bast RC Jr, Ravdin P, Hayes DF, et al: 2000 update of recommendations for the use of tumor markers in breast and colorectal cancer: Clinical practice guidelines of the American Society of Clinical Oncology. J Clin Oncol 19:1865-1878, 2001
16. Filipits M, Malayeri R, Suchomel RW, et al: Expression of the multidrug resistance protein (MRP1) in breast cancer. Anticancer Res 19:5043-5050, 1999
17. Nooter K, Brutel de la Riviere G, Klijn J, et al: Multidrug resistance protein in recurrent breast cancer. Lancet 349:1885-1886, 1997
18. Nooter K, Brutel de la Riviere G, Look MP, et al: The prognostic significance of expression of the multidrug resistance-associated protein (MRP) in primary breast cancer. Br J Cancer 76:486-493, 1997
19. Burger H, Foekens JA, Look MP, et al: RNA expression of breast cancer resistance protein, lung resistance-related protein, multidrug resistance-associated proteins 1 and 2, and multidrug resistance gene 1 in breast cancer: Correlation with chemotherapeutic response. Clin Cancer Res 9:827-836, 2003
20. Filipits M, Suchomel RW, Dekan G, et al: MRP and MDR1 gene expression in primary breast carcinomas. Clin Cancer Res 2:1231-1237, 1996
21. Dexter DW, Reddy RK, Geles KG, et al: Quantitative reverse transcriptase-polymerase chain reaction measured expression of MDR1 and MRP in primary breast carcinoma. Clin Cancer Res 4:1533-1542, 1998
22. Ito K, Fujimori M, Nakata S, et al: Clinical significance of the increased multidrug resistance-associated protein (MRP) gene expression in patients with primary breast cancer. Oncol Res 10:99-109, 1998
23. Lacave R, Coulet F, Ricci S, et al: Comparative evaluation by semiquantitative reverse transcriptase polymerase chain reaction of MDR1, MRP and GSTp gene expression in breast carcinomas. Br J Cancer 77:694-702, 1998
24. Ferrero JM, Etienne MC, Formento JL, et al: Application of an original RT-PCR-ELISA multiplex assay for MDR1 and MRP, along with p53 determination in node-positive breast cancer patients. Br J Cancer 82:171-177, 2000
25. Kanzaki A, Toi M, Nakayama K, et al: Expression of multidrug resistance-related transporters in human breast carcinoma. Jpn J Cancer Res 92:452-458, 2001
26. Linn SC, Pinedo HM, van Ark-Otte J, et al: Expression of drug resistance proteins in breast cancer, in relation to chemotherapy. Int J Cancer 71:787-795, 1997
27. Zchbauer-Müller S, Filipits M, Rudas M, et al: P-glycoprotein and MRP1 expression in axillary lymph node metastases of breast cancer patients. Anticancer Res 21:119-124, 2001
28. National Institutes of Health Consensus Development Panel: National Institutes of Health Consensus Development Conference Statement: Adjuvant therapy for breast cancer, November 1-3, 2000. J Natl Cancer Inst Monogr 30:5-15, 2001(Martin Filipits, Gudrun P)
Department of Pathology and Third Medical Department, Salzburg Hospital, Salzburg
Departments of Pathology and Internal Medicine, University of Graz, Graz
Department of Pathology-Histology, Hanusch Medical Center, Vienna, Austria
ABSTRACT
PATIENTS AND METHODS: Immunostaining for MRP1 was performed on tissue sections from 516 premenopausal, hormone receptor-positive breast cancer patients with stage I and II disease. Statistical analyses were performed to assess the effect of MRP1 expression on survival and to test for interaction between MRP1 expression and treatment.
RESULTS: MRP1 expression independently predicted shorter relapse-free survival (RFS) and overall survival (OS) in patients treated with cyclophosphamide, methotrexate, and fluorouracil (CMF; RFS: hazard ratio [HR] = 1.48; 95% CI, 1.16 to 1.88; P = .002; OS: HR = 1.82; 95% CI, 1.10 to 3.01; P = .02), but it did not predict shorter RFS and OS in patients who received tamoxifen plus goserelin (RFS: HR = 0.99; 95% CI, 0.74 to 1.31; P = .9; OS: HR = 0.68; 95% CI, 0.40 to 1.15; P = .1). Tests for interaction between MRP1 expression and treatment were statistically significant for both RFS (P = .04) and OS (P = .006).
CONCLUSION: Our data suggest that MRP1 expression plays an important role in the clinical resistance to adjuvant CMF chemotherapy but does not seem to affect response to adjuvant endocrine treatment with tamoxifen plus goserelin. Thus, MRP1 may be a useful marker for the selection of patients with early-stage breast cancer for the appropriate adjuvant therapy after prospective confirmatory evaluation.
INTRODUCTION
The multidrug resistance protein 1 (MRP1), also called ABCC1, is another member of the ABC transporter family.4 Overexpression of this protein in tumor cells confers resistance to various anticancer drugs, such as anthracyclines or methotrexate,4-7 which are drugs that are part of the chemotherapy regimens currently used in the treatment of breast cancer.8 Prior studies suggest an association between MRP1 protein expression and shorter disease-free survival and overall survival (OS) of breast cancer patients.9 However, the heterogeneity of these studies as a result of differences in the study populations, the number of patients examined, the length of follow-up, the use of different types of adjuvant endocrine treatment and/or chemotherapy regimens, and the detection methods used does not allow a definitive conclusion with regard to the impact of MRP1 expression on clinical outcome.
To further evaluate the clinical role of MRP1 in breast carcinomas, we studied the relationship between MRP1 expression of the primary tumors and survival of the patients who were enrolled onto the Austrian Breast and Colorectal Cancer Study Group (ABCSG) Trial 5, a prospective randomized trial comparing the efficacy of cyclophosphamide, methotrexate, and fluorouracil (CMF) chemotherapy with a combination endocrine treatment.
PATIENTS AND METHODS
Tumor Specimens
All tumor specimens were obtained at the time of surgery before adjuvant therapy. Formalin-fixed, paraffin-embedded tumor blocks containing primary breast cancer tissue from patients who participated in ABCSG Trial 5 were retrospectively collected from the major participating centers. Details with regard to the collection of the samples and the preparation of slides have been published as part of a study describing the role of p27Kip1 expression in this population.11
Immunohistochemistry
A hematoxylin and eosin-stained slide was prepared from each block and used for pathologic confirmation of present invasive breast cancer by an experienced breast pathologist. Immunohistochemical analysis reported in this study was carried out in a single laboratory and performed as described previously.12 Briefly, tissue sections of 4-μm thickness were prepared, mounted on poly-L-lysine-coated slides, deparaffinized, and rehydrated with distilled water. Endogenous peroxidase activity was blocked by incubation in 0.06% hydrogen peroxide for 10 minutes at room temperature. After washing in phosphate-buffered saline (pH 7.4), the tissues were preincubated for 20 minutes in normal serum (dilution 1:20; Dako, Glostrup, Denmark) before a 90-minute incubation with the anti-MRP1 monoclonal antibody MRPr1 (dilution 1:50; Alexis, Lufelfingen, Switzerland). Antibody binding was detected by the avidin-biotin-peroxidase method. Bound peroxidase was developed with 3,3'-diaminobenzidine (Dako). The slides were counterstained with Mayer's Hmalaun and mounted with Aquatex (Merck, Darmstadt, Germany). All of the washes were performed in phosphate-buffered saline (pH 7.4).
Normal human kidney tissue served as positive control for MRP1 expression. In addition, we used breast cancer specimens known to be positive or negative for MRP1 as positive and negative controls, respectively. Negative controls without the primary antibody were performed as described earlier. Staining of tumor cells was examined by an experienced breast pathologist who was blinded to the clinical outcome of the patients. All tumor cells on each slide were evaluated, and interpretation of the results was limited to the invasive portion of the tumor. MRP1 immunostaining was classified as previously described by us,12 and was as follows: negative, 0% of tumor cells showing staining reactivity; low, less than 10% of tumor cells showing staining reactivity; intermediate, 10% to 30% of tumor cells showing staining reactivity; and high, more than 30% of tumor cells showing staining reactivity. To assess interobserver reproducibility, 100 randomly selected patients were analyzed by a second investigator, and {kappa} statistics were used to assess interobserver reliability. The observed {kappa} was 0.72, and therefore, interobserver reliability was considered satisfactory (data not shown). The rare discrepant cases were reassessed together by both investigators using a double-headed microscope, and a consensus was reached.
Statistical Analysis
Associations of MRP1 expression with age, tumor size, lymph node status, tumor grade, estrogen receptor (ER) expression, and progesterone receptor (PgR) expression were assessed by the Spearman rank correlation. Survival probabilities were estimated with the product limit method according to Kaplan and Meier.13 Survival time was defined as the period between the time of randomization and death (OS) or the period between the time of randomization and documented relapse (relapse-free survival [RFS]). Relapse was defined as the first reappearance of breast cancer at any local, distant, or contralateral site. Patients who died because of reasons other than breast cancer were considered as censored with death. Survival times of patients still alive were censored with the date of the last follow-up. Differences between survival curves were analyzed using the log-rank test. To describe the unadjusted effects of covariates on RFS and OS, univariate Cox proportional hazards regression models were used. Multiple Cox proportional hazards regression models were used to assess the independent effects of MRP1 expression on RFS and OS.14 In addition, multiple Cox proportional hazards regression models with backward elimination of nonsignificant variables were applied. We used a backward elimination algorithm with a 0.1 significance level for removing explanatory variables. Variables were coded as follows: age, less than 35 years or ≥ 35 years; tumor size, pT1, pT2, or pT3; lymph node status, 0, 1 to 3, 4 to 10, or more than 10 axillary lymph node metastases; tumor grade, grade 1, grade 2, or grade 3; ER, negative, positive, or strongly positive; PgR, negative, positive, or strongly positive; treatment, combination endocrine therapy or chemotherapy; p27Kip1, less than 50% or ≥ 50%; MRP1, negative, low, intermediate, or high; and an interaction term, the product of MRP1 and treatment (MRP1 x treatment). Tumor size, lymph node status, tumor grade, ER status, PgR status, and MRP1 expression were included as ordered factors in the statistical analyses on the assumption that the relative risk between two successive categories would remain the same. The Cox proportional hazards regression model was also applied to assess interactions between treatment and the other covariates. All P values are results of two-sided tests. SPSS 11.5.1 statistical software (SPSS Inc, Chicago, IL) was used for calculations.
RESULTS
MRP1 immunostaining was both membranous and cytoplasmic and ranged from 0% to 100% (median, 10%) of the breast cancer cells. For comparisons with clinical parameters, MRP1 expression was categorized as negative, low, intermediate, and high. MRP1 expression was negative in 148 patients (29%), low in 90 patients (17%), intermediate in 130 patients (25%), and high in 148 patients (29%).
When we examined the correlations between MRP1 expression and clinical parameters, MRP1 expression showed a weak positive correlation with tumor size (r = 0.115; P = .009) and tumor grade (r = 0.096; P = .03; Table 3). In contrast, MRP1 expression was not significantly correlated with age, lymph node status, hormone receptor status, treatment, and p27Kip1 expression (Table 3). Similar results were obtained when MRP1 expression was analyzed as a continuous variable (data not shown).
At a median follow-up of 5.6 years, 105 patients (20%) had relapsed (60 CMF patients and 45 endocrine therapy patients), and 38 patients (7%) had died from cancer (22 CMF patients and 16 endocrine therapy patients). The 5-year RFS and OS rates were 80% and 93%, respectively. Univariate analyses demonstrated that younger age, larger tumor size, higher number of positive lymph nodes, lower levels of PgR expression, low p27Kip1 expression, and increasing levels of MRP1 expression were significantly associated with shorter RFS (Table 4). Larger tumor size, higher number of involved lymph nodes, higher tumor grade, low p27Kip1 expression, and higher MRP1 expression were also significantly associated with shorter OS (Table 4).
The independent effects of MRP1 expression on RFS and OS were assessed by multiple Cox proportional hazards regression models. All variables listed in Table 4 were included in the models. Because previous studies suggested that MRP1 expression may be associated with resistance to CMF chemotherapy, we also tested for the existence of interaction between MRP1 expression and treatment by incorporating an interaction term, the product of MRP1 expression and treatment (MRP1 x treatment), into the Cox model. Age, tumor size, lymph node status, PgR expression, p27Kip1 expression, and the interaction term (MRP1 x treatment) were identified as statistically significant risk factors for RFS (Table 4). Lymph node status, treatment, p27Kip1 expression, and the interaction term (MRP1 x treatment) were independently associated with OS (Table 4). Interestingly, the interaction term was statistically significant for both RFS (P = .04) and OS (P = .006) in these analyses (Table 4). Similar results were obtained when multiple Cox proportional hazards regression models with backward elimination of nonsignificant variables were applied. In particular, the interaction term remained significant after elimination of the nonsignificant variables (P = .04 and P = .004 for RFS and OS, respectively; data not shown).
Because the interaction term was statistically significant for both RFS and OS, we determined the association of MRP1 expression with survival in subgroups of patients treated with either chemotherapy or endocrine therapy. To assess whether the effect of MRP1 expression by treatment is independent of other variables, hazard ratios were estimated before and after adjusting for age, tumor size, lymph node status, tumor grade, ER status, PgR status, and p27Kip1 expression. The results of these analyses are listed in Table 5 and shown in Figure 1. In the cohort of CMF-treated patients, higher MRP1 expression was associated with shorter RFS and OS of the patients. The adjusted hazard ratios for relapse and death were 1.48 (95% CI, 1.16 to 1.88; P = .002) and 1.82 (95% CI, 1.10 to 3.01; P = .02), respectively (Table 5). In contrast, in patients who received combination endocrine therapy, MRP1 expression did not predict RFS and OS. The adjusted hazard ratios for relapse and death were 0.99 (95% CI, 0.74 to 1.31; P = .9) and 0.68 (95% CI, 0.40 to 1.15; P = .1), respectively (Table 5). These data were also shown in Figure 1. At 5 years, the RFS rates for CMF-treated patients were 92%, 84%, 76%, and 56% for patients with negative, low, intermediate, and high MRP1 expression, respectively (Fig 1A); the corresponding 5-year OS rates were 98%, 97%, 91%, and 79%, respectively (Fig 1C). In contrast, the 5-year RFS rates for patients assigned to endocrine therapy were 83%, 81%, 82%, and 84% for patients with negative, low, intermediate, and high MRP1 expression, respectively (Fig 1B); the corresponding 5-year OS rates for patients who received endocrine therapy were 95%, 88%, 98%, and 100%, respectively (Fig 1D). Similar results were obtained when we performed an analysis on patients with T1 and T2 tumors only (data not shown).
These results indicate that MRP1 expression independently predicted shorter RFS and OS in CMF-treated patients and suggest that MRP1 expression is associated with treatment failure in those patients. In contrast, no such association was found for MRP1 and outcome in patients receiving endocrine therapy with tamoxifen plus goserelin.
DISCUSSION
In the current study, we have examined MRP1 expression in a large, homogenous, and well-defined patient population consisting of premenopausal, hormone receptor-positive breast cancer patients with stage I and II disease who were enrolled onto a prospective randomized trial. Our results indicate that MRP1 expression is a prognostic factor for shorter RFS and OS in the total study population. These findings are consistent with previous reports from both our group12,16 and other investigators.17-19 More importantly, however, we found an interaction between MRP1 expression and adjuvant treatment. Patients who were treated with CMF chemotherapy and whose tumors were MRP1 negative experienced a significant reduction in relapse rate and a decrease in mortality compared with patients who had highly expressed MRP1. In contrast, in patients who received tamoxifen plus goserelin, MRP1 expression had no impact on the relapse rate or mortality. These results suggest that MRP1 expression is associated with clinical resistance to CMF chemotherapy but not with tamoxifen resistance and confirm the hypothesis derived from preclinical studies suggesting that methotrexate but not tamoxifen is transported by MRP1.
Several studies have investigated the expression of MRP1 in clinical breast cancer samples.12,16-27 MRP1 mRNA was detected by reverse transcriptase polymerase chain reaction (RT-PCR) in 488 (98%) of 496 breast cancer specimens,19-25 and MRP1 protein expression determined by immunohistochemistry was detectable in 291 (53%) of 544 samples.12,16-18,21,26,27 The higher frequency of MRP1 expression obtained with RT-PCR is not surprising because MRP1 is ubiquitously expressed at low levels. Thus, by using RT-PCR, MRP1 mRNA can be detected in most breast cancer samples.
Only a few previous studies have correlated MRP1 expression with patients' clinical outcome. One study evaluated the impact of MRP1 expression on the prognosis and response to chemotherapy of recurrent breast cancer.17 In 23 previously untreated patients who received first-line chemotherapy for recurrence, the presence of MRP1 protein was associated with a lower response rate (13% of MRP1-positive patients had an objective response compared with 40% of MRP1-negative patients) and with a significantly shorter progression-free survival, independent of the type of chemotherapy. No association between MRP1 expression and response rate or progression-free survival was observed in patients who received chemotherapy after treatment with endocrine therapy. The results of another study suggested that MRP1 may be associated with an increased relapse rate in subgroups of patients with a more favorable prognosis (eg, patients with T1 tumors or node-negative patients) and in node-positive patients who received adjuvant CMF chemotherapy.18 MRP1 expression was also associated with an increased risk of death in patients with T1 tumors and in node-positive patients who were treated with adjuvant CMF chemotherapy. In our previous report, MRP1 expression in primary breast tumors correlated with shorter disease-free survival and OS in a heterogeneously treated cohort of patients.16 In patients treated with preoperative chemotherapy, patients with MRP1 expression had a significantly shorter progression-free survival than patients without MRP1 expression, independent of the type of chemotherapy. In a multivariate Cox regression analysis, prechemotherapy MRP1 expression was identified as an independent prognostic factor for shorter progression-free survival.12 A recently published study has shown MRP1 expression to be related to clinical outcome only in the subgroup of patients treated with anthracycline-based chemotherapy and not in patients who received CMF chemotherapy.19 These data suggest that the clinical role of MRP1 expression in chemotherapy resistance of breast cancer may be restricted to anthracycline resistance. However, the results are based on low numbers of patients, particularly in treatment subgroups, and, therefore, have to be viewed with caution.
The results of these previous studies suffer from various shortcomings and are difficult to compare because of differences in the study populations, the number of patients examined, the length of follow-up, the use of different types of adjuvant endocrine treatment and/or chemotherapy regimens, and the detection methods used. However, the results suggest that MRP1 expression is of clinical relevance in breast cancer and support the results obtained in our present study.
The treatment protocol underlying the present study consists of reasonable treatment options with regard to the management of premenopausal women with early-stage hormone receptor-positive breast cancer. Various randomized trials have shown that ovarian ablation with or without tamoxifen and standard chemotherapy regimens like CMF have similar benefits for premenopausal women with early-stage, receptor-positive breast cancer.8 However, because adjuvant polychemotherapy improves survival, the National Institutes of Health Consensus Development Panel concluded that polychemotherapy should be recommended to the majority of women with localized breast cancer.28 The inclusion of anthracyclines in adjuvant chemotherapy regimens results in a small but statistically significant improvement in survival over nonanthracycline-containing regimens. Four cycles of doxorubicin plus cyclophosphamide have been shown to be equivalent to six cycles of classical CMF, and only six cycles of anthracycline-containing regimens yield superior results but at the cost of greater toxicity.8 Therefore, our present results may have important implications in the treatment of the majority of women with localized breast cancer by the more precise identification of those patients who benefit from adjuvant chemotherapy. Previously, we concluded that, "Overall, our data suggest that the goserelin-tamoxifen combination is significantly more effective than CMF in the adjuvant treatment of premenopausal patients with stage I and II breast cancer."10 In light of the strong predictive effect of MRP1 among patients who received CMF chemotherapy, the results of the current study suggest that patients with intermediate or high levels of MRP1 expression should not be treated with CMF chemotherapy but may benefit from treatment with endocrine therapy, whereas MRP1-negative patients may be considered for treatment with CMF chemotherapy or anthracycline-based chemotherapy. However, this proposed treatment strategy has to be confirmed in well-designed clinical trials.
The current data are encouraging but require validation in a prospective trial before being implemented in clinical routine. Our results confirm the hypotheses derived from preclinical studies that suggest that higher MRP1 expression levels may be associated with clinical drug resistance in breast cancer. Thus, MRP1 may be a useful marker for the selection of patients with early-stage, hormone receptor-positive breast cancer for appropriate adjuvant therapy.
Appendix
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank Ngoc-Anh Huynh for excellent technical assistance and Harald Heinzl, Andreas Gleiss, and Karl Thomanek for their biometrical and editorial expertise.
NOTES
This study was supported by a grant from the Austrian Science Fund (Project No. P15377) and the Center of Excellence in Clinical and Experimental Oncology of the Medical University of Vienna.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Gottesman MM, Fojo T, Bates SE: Multidrug resistance in cancer: Role of ATP-dependent transporters. Nat Rev Cancer 2:48-58, 2002
2. Borst P, Evers R, Kool M, et al: A family of drug transporters: The multidrug resistance-associated proteins. J Natl Cancer Inst 92:1295-1302, 2000
3. Trock BJ, Leonessa F, Clarke R: Multidrug resistance in breast cancer: A meta-analysis of MDR1/gp170 expression and its possible functional significance. J Natl Cancer Inst 89:917-931, 1997
4. Cole SPC, Bhardwaj G, Gerlach JH, et al: Overexpression of a transporter gene in a multidrug-resistant human lung cancer cell line. Science 258:1650-1654, 1992
5. Cole SPC, Sparks KE, Fraser K, et al: Pharmacological characterization of multidrug resistant MRP-transfected human tumor cells. Cancer Res 54:5902-5910, 1994
6. Hooijberg JH, Broxterman HJ, Kool M, et al: Antifolate resistance mediated by the multidrug resistance proteins MRP1 and MRP2. Cancer Res 59:2532-2535, 1999
7. Zeng H, Chen Z-S, Belinsky MG, et al: Transport of methotrexate (MTX) and folates by multidrug resistance protein (MRP) 3 and MRP1: Effect of polyglutamylation on MTX transport. Cancer Res 61:7225-7232, 2001
8. Goldhirsch A, Wood WC, Gelber RD, et al: Meeting highlights: Updated international expert consensus on the primary therapy of early breast cancer. J Clin Oncol 21:3357-3365, 2003
9. Leonessa F, Clarke R: ATP binding cassette transporters and drug resistance in breast cancer. Endocr Relat Cancer 10:43-73, 2003
10. Jakesz R, Hausmaninger H, Kubista E, et al: Randomized adjuvant trial of tamoxifen and goserelin versus cyclophosphamide, methotrexate, and fluorouracil: Evidence for the superiority of treatment with endocrine blockade in premenopausal patients with hormone-responsive breast cancer—Austrian Breast and Colorectal Cancer Study Group Trial 5. J Clin Oncol 20:4621-4627, 2002
11. Pohl G, Rudas M, Dietze O, et al: High p27Kip1 expression predicts superior relapse-free and overall survival for premenopausal women with early-stage breast cancer receiving adjuvant treatment with tamoxifen plus goserelin. J Clin Oncol 21:3594-3600, 2003
12. Rudas M, Filipits M, Taucher S, et al: Expression of MRP1, LRP and Pgp in breast carcinoma patients treated with preoperative chemotherapy. Breast Cancer Res Treat 81:149-157, 2003
13. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
14. Cox DR: Regression models and life tables. J R Stat Soc 34:187-220, 1972
15. Bast RC Jr, Ravdin P, Hayes DF, et al: 2000 update of recommendations for the use of tumor markers in breast and colorectal cancer: Clinical practice guidelines of the American Society of Clinical Oncology. J Clin Oncol 19:1865-1878, 2001
16. Filipits M, Malayeri R, Suchomel RW, et al: Expression of the multidrug resistance protein (MRP1) in breast cancer. Anticancer Res 19:5043-5050, 1999
17. Nooter K, Brutel de la Riviere G, Klijn J, et al: Multidrug resistance protein in recurrent breast cancer. Lancet 349:1885-1886, 1997
18. Nooter K, Brutel de la Riviere G, Look MP, et al: The prognostic significance of expression of the multidrug resistance-associated protein (MRP) in primary breast cancer. Br J Cancer 76:486-493, 1997
19. Burger H, Foekens JA, Look MP, et al: RNA expression of breast cancer resistance protein, lung resistance-related protein, multidrug resistance-associated proteins 1 and 2, and multidrug resistance gene 1 in breast cancer: Correlation with chemotherapeutic response. Clin Cancer Res 9:827-836, 2003
20. Filipits M, Suchomel RW, Dekan G, et al: MRP and MDR1 gene expression in primary breast carcinomas. Clin Cancer Res 2:1231-1237, 1996
21. Dexter DW, Reddy RK, Geles KG, et al: Quantitative reverse transcriptase-polymerase chain reaction measured expression of MDR1 and MRP in primary breast carcinoma. Clin Cancer Res 4:1533-1542, 1998
22. Ito K, Fujimori M, Nakata S, et al: Clinical significance of the increased multidrug resistance-associated protein (MRP) gene expression in patients with primary breast cancer. Oncol Res 10:99-109, 1998
23. Lacave R, Coulet F, Ricci S, et al: Comparative evaluation by semiquantitative reverse transcriptase polymerase chain reaction of MDR1, MRP and GSTp gene expression in breast carcinomas. Br J Cancer 77:694-702, 1998
24. Ferrero JM, Etienne MC, Formento JL, et al: Application of an original RT-PCR-ELISA multiplex assay for MDR1 and MRP, along with p53 determination in node-positive breast cancer patients. Br J Cancer 82:171-177, 2000
25. Kanzaki A, Toi M, Nakayama K, et al: Expression of multidrug resistance-related transporters in human breast carcinoma. Jpn J Cancer Res 92:452-458, 2001
26. Linn SC, Pinedo HM, van Ark-Otte J, et al: Expression of drug resistance proteins in breast cancer, in relation to chemotherapy. Int J Cancer 71:787-795, 1997
27. Zchbauer-Müller S, Filipits M, Rudas M, et al: P-glycoprotein and MRP1 expression in axillary lymph node metastases of breast cancer patients. Anticancer Res 21:119-124, 2001
28. National Institutes of Health Consensus Development Panel: National Institutes of Health Consensus Development Conference Statement: Adjuvant therapy for breast cancer, November 1-3, 2000. J Natl Cancer Inst Monogr 30:5-15, 2001(Martin Filipits, Gudrun P)