|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2001 American Society for Clinical Oncology Prognostic and Predictive Significance of ErbB-2 Breast Tumor Levels Measured by Enzyme ImmunoassayFrom the Stiftung Tumorbank Basel, Departments of Research, Gynecology, and Surgery, University Clinics, Kantonsspital, Basel; Department of Gynecology and Institute for Clinical Pathology, University Hospital, Zurich; Pathology Institute Triemli, Zurich, and Kantonsspital, Baden, Switzerland; Department of Medicine, University of California at San Francisco, San Francisco, CA; and Breast Cancer Center, Rheinfelden, Germany. Address reprint requests to Urs Eppenberger, PhD, MD, Stiftung Tumorbank Basel, University Women’s Clinic, Kantonsspital Basel, Schanzenstrasse 46, CH-4031 Basel, Switzerland; email: urs-sere.eppenberger{at}unibas.ch
PURPOSE: A retrospective analysis to assess the prognostic and predictive clinical value of breast tumor ErbB-2 receptor expression quantified by enzyme immunoassay (EIA), to compare levels measured by EIA with ErbB-2 status determined by immunohistochemistry (IHC), and to correlate receptor content with levels of phosphorylated (Y1248-P) ErbB-2, a measure of functional tyrosine kinase activity. MATERIALS AND METHODS: EIA quantification of ErbB-2 was performed on membrane extracts from 3,208 well-characterized primary breast cancers. Overall, relapse-free, distant disease-free, and local/regional-free patient survival data were available on 1,123 of these tumors. IHC scoring for ErbB-2 status (HercepTest; DAKO, Glostrup, Denmark) was performed on adjacent sections of 151 cases, and receptor functionality was measured in 230 tumors by an antibody specific for phosphorylated (Y1248-P) ErbB-2. RESULTS: Unlike nonmalignant breast tissues, breast tumors showed increased ErbB-2 levels in a bimodal distribution, with 12% constituting a distinct set of ErbB-2–overexpressing tumors. The intermodal threshold value for ErbB-2 overexpression distinguished tumors with reduced estrogen and progesterone receptor content, high IHC score for ErbB-2, and significantly increased levels of phosphorylated (Y1248-P) ErbB-2 receptor. By multivariate analysis, EIA-determined ErbB-2 overexpression predicted significantly reduced patient survival that was unaffected by tamoxifen or cyclophosphamide, methotrexate, and fluorouracil adjuvant therapy. CONCLUSION: Determination of ErbB-2 receptor expression by EIA offers a clinically valuable alternative to semiquantitative IHC assessment of breast tumor ErbB-2 overexpression and affords the opportunity to evaluate ErbB-2 phosphorylation, which may represent an important predictive parameter of receptor functionality.
OVEREXPRESSION OF the 185-kd ErbB-2 (HER2/neu) membrane receptor tyrosine kinase occurs by transcriptional upregulation of the genomically amplified c-ErbB-2 oncogene early during breast tumorigenesis and almost exclusively in transformed mammary ductal (v lobular) epithelium.1-4 As a breast tumor marker, ErbB-2 overexpression identifies more aggressive cancers that are characterized by early metastatic spread and reduced patient survival.2 More important than its prognostic value is its predictive utility, because overexpression of ErbB-2 is associated with altered clinical responsiveness to some systemic breast cancer treatments and is the tumor-specific target for the recently approved antibody trastuzumab (Herceptin; Genentech, South San Francisco, CA, and F. Hoffmann-La Roche Ltd, Basel, Switzerland).3-5 Despite more than a decade of clinical research focused on this oncogenic receptor system, we still lack full understanding of its prognostic and predictive mechanisms as exemplified by the fact that approximately 80% of patients with ErbB- 2–overexpressing breast cancer do not respond to trastuzumab.3,4 Contributing to uncertainty in the clinical evaluation of this breast tumor marker has been the increasing variety of assays available to assess ErbB-2 receptor overexpression (mRNA, protein) and c-erbB-2 gene amplification.1 For routine clinical assessment of breast tumor ErbB-2 status, early solid matrix blotting techniques (eg, Southern, Northern, and Western blots that detect DNA, RNA, or protein levels from tumor extracts) and polymerase chain reaction (PCR) methods have been largely replaced by slide-based ErbB-2 immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) assays yielding semiquantitative scores for membrane receptor expression and relative DNA copy number, respectively.1 The tumors of all patients entered onto the phase II and III trastuzumab clinical trials had to express either 2+ or 3+ levels of ErbB-2 membrane immunoreactivity (on a scale of 0 to 3+, assayed in a single reference laboratory).3-5 As reported,5-7 patients with 3+ overexpressing tumors, as compared with those with 2+ overexpressing tumors, had significantly higher objective response rates to trastuzumab monotherapy and combination therapy with either doxorubicin-cyclophosphamide or paclitaxel (17%, 53%, and 44% v 4%, 40%, and 21%, respectively). Accordingly, the Food and Drug Administration (FDA) approved trastuzumab for use against 2+ and 3+ ErbB-2–overexpressing metastatic breast cancers and then also approved a commercial IHC assay (HercepTest; DAKO, Glostrap, Denmark) for routine clinical determination of breast tumor ErbB-2 status and trastuzumab eligibility.1,4 The performance of this commercial IHC assay was reasonably concordant with the reference laboratory results used during the clinical trials; however, there was significant discordance between these IHC assays in distinguishing tumors with a 2+ level of ErbB-2 overexpression.1,4,5 Commercially available FISH assays quantitate c-ErbB-2 gene signals per nucleus (Ventana Medical Systems, Inc, Tucson, AZ) or gene signals per chromosome 17 (Vysis, Inc, Downers Grove, IL) but are currently FDA approved only for predicting disease prognosis and response to taxane chemotherapy.1 Compared with more commonly used semiquantitative IHC assays for ErbB-2 assessment, FISH is more expensive and requires special instrumentation not generally available to pathologists. FISH can more specifically identify tumors with more than two-fold c-ErbB-2 gene amplification but cannot detect single-copy ErbB-2 overexpressors and underestimates trastuzumab-eligible patients whose tumors show 2+ ErbB-2 overexpression by IHC. A recent review comparing all gene- and protein-based assays for ErbB-2 highlighted the conclusion that assays performed on frozen or fresh tumor tissue produce more consistent results than those subject to variable types of tissue fixation and archival storage of paraffin-embedded sections.1 In particular, different anti-ErbB-2 IHC antibodies are known to give different staining patterns and intensities depending on the type of tissue fixative and the length of slide storage.8 Slide-based IHC results are routinely reported using semiquantitative scoring systems, which are subject to interobserver interpretation error and which prevent statistical analysis of the ErbB-2 value as a continuous variable function. Although limited by the need for milligram quantities of fresh or frozen tumor sample, enzyme immunoassay (EIA) measurement of tumor ErbB-2 receptor level averts the potential antigen damage associated with fixation, embedding, and uncontrolled storage and yields a highly reproducible continuous value instrument readout (units per milligram of tumor protein). As summarized in a comprehensive review of 52 breast cancer prognosis studies published between 1987 and 1998 analyzing 16,975 tumors for ErbB-2 status,1 only three previous studies have used EIA measurement of ErbB-2 levels. All three studies, which analyzed a total of 315 breast tumors, identified threshold levels for ErbB-2 overexpression that by multivariate analysis proved to have independent prognostic significance for early breast cancer relapse.9-11 The present study was designed to validate and determine the prognostic and predictive clinical significance of EIA-measured ErbB-2 receptor levels using a cryobanked collection of 3,208 primary human breast tumors. EIA assay validation was performed on different subsets of this tumor collective by comparing (1) tumor and normal breast levels of ErbB-2 expression (n = 334), (2) ErbB-2 tumor levels measured by two different commercially available immunoassays (n = 120), and (3) EIA-measured and IHC-scored levels of ErbB-2 expression (n = 151). The prognostic and predictive clinical significance of EIA-measured levels of total ErbB-2 expression was determined from a subset of 1,123 tumors for which patient outcome data were available and from which two distinct clinical populations could be further analyzed for therapeutic response: those with estrogen receptor (ER)–positive tumors who received tamoxifen as their only adjuvant therapy (n = 520) and those who received only adjuvant cyclophosphamide, methotrexate, and fluorouracil (CMF) chemotherapy (n = 154). Lastly, a novel immunoassay of functional ErbB-2 receptor was performed on primary breast tumors of known ErbB-2 receptor content (n = 230) to identify those with increased levels of activated ErbB-2 receptor tyrosine (Y) kinase product, Y-1248 phosphorylated ErbB-2, a potentially more specific prognostic and predictive ErbB-2 marker.
Tumor and Patient Characteristics All (n = 3,208) primary breast tumors were cryopreserved (-70°C) immediately after surgical resection. In 334 cases, adjacent nonmalignant breast epithelium was first carefully resected away from the grossly apparent tumor mass and the paired sets of tumor and nonmalignant adjacent breast tissue (NAT) were cryopreserved for later analysis. Patient and tumor characteristics, including tumor-node-metastasis stage, histologic grade, and ER and progesterone receptor (PgR) content for the entire tumor collection are listed in Table 1. Patient outcome data, including overall survival (OS), relapse-free survival (RFS), distant disease-free survival (DDFS), and local- and regional-free survival (LRFS), were available for 1,123 of these cases diagnosed between 1992 and 1996 (Table 1). Median follow-up duration of this subset was 41.7 months (range, 12 to 84 months), and from this subset, 120 (10.6%) died with disease and 271 (24%) developed metastatic recurrence within the follow-up interval. Of the recurrences, 120 were local and all occurred after primary treatment by either modified radical mastectomy (72 of 120) or radiation therapy in conjunction with breast-conserving surgery (48 of 120). Of the 533 (47.5%) node-negative patients in this cohort, 272 (51%) received adjuvant tamoxifen, 46 (9%) received adjuvant chemotherapy, 158 (30%) received no systemic therapy, and the remainder (10%) received combination therapy or unknown additional therapy. Of the 437 (38.9%; in 13.6% of cases the nodal status was unknown) node-positive patients, 110 (25%) received adjuvant tamoxifen, 37 (8%) received adjuvant chemotherapy, eight (2%) received no systemic therapy, and the remainder (65%) received combination therapy. Of all patients, 950 (84.6%) had ER-positive tumors (> 10 fmol/mg protein); 520 patients with ER-positive tumors received tamoxifen as their only adjuvant therapy. CMF was used as the only adjuvant therapy in a subset of 130 patients.
Tumor Extract Preparation for Immunoassays, Determination of ER and PgR Levels Tissue homogenates were prepared in accordance with standard procedures for determination of receptor and tumor marker EIA analyses, as previously described.12 In brief, the frozen tissues were finely pulverized in liquid nitrogen using a Micro-Dismembrator U (B. Braun Melsungen AG, Melsungen, Germany). The powders were homogenized with a tissue homogenizer (Ultra-Turrax; Janke & Kunkel, IKA-Werke, Staufen, Germany) for 20 seconds in three volumes of ice-cold extraction buffer containing Tris 10 mmol/L, EDTA 1.5 mmol/L, 10% glycerol, disodiummolybdate 5 mmol/L, and monothioglycerol 1 mmol/L. The homogenate was centrifuged for 3 minutes at 4°C, and the supernatant was recentrifuged in an ultracentrifuge (Beckman Instruments, Fullerton, CA) at 100,000 x g for 40 minutes at 4°C. The resulting supernatants (cytosols) were used for determination of the hormone receptors (ER, PgR), and the pellets were used for EIA determination of membrane-associated ErbB-2. As previously described,12 ER and PgR concentrations were assessed from tumor cytosolic extracts by commercial quantitative ER and PgR EIA kits (Abbott Laboratories, Abbott Park, IL) using a Quantum II photometer. Quality of ER and PgR determination was regularly assessed in trial studies organized by the Receptor Study Group of the European Organization for Research and Treatment of Cancer.
Quantitative Immunoassay of Total ErbB-2 Receptor Level
IHC Determination of ErbB-2 Expression
Quantitative Immunoassay of Phosphorylated ErbB-2 Receptor Level Tumor extracts for determination of phosphorylated ErbB-2 were prepared from pellet suspensions using Triton X-100, activated orthovanadate, and sodium fluoride at final concentrations of 1%, 1 mmol/L, and 1 mmol/L, respectively. After centrifugation at 13,000 rpm at 4°C for 15 minutes, the supernatants were diluted 1:20 with Tris-buffered saline (TBS) 20 mmol/L containing 0.05% Tween 20 (Pierce Surfact Amps), activated orthovanadate 1 mmol/L, and NaF 1 mmol/L. The protein concentrations of these pellet extracts were measured with the microprotein assay from Pierce. Before application of the samples, the coated microtiters were washed five times with Tris-buffered saline (300 µL/well) and then blocked for 2 hours at room temperature with 200 µL of ChemiBLOCKER (Chemicon International, Temecula, CA) diluted 1:8 with TBS. The blocked wells were washed five times with 300 µL of TBS and then 100 µL of the diluted tumor membrane extracts or reference material was added to the wells and incubated overnight at 4°C. As a reference for each assay, a cell extract of A431 cells was used. A431 tumor cells were treated for 2 days with serum-free culture medium and then stimulated with epidermal growth factor (100 ng/mL) for 6 minutes at 37°C. After the cells were washed twice with cold phosphate-buffered saline, the cells were scraped from the culture flasks, collected by gentle centrifugation, and extracted for 10 minutes at room temperature with a HEPES buffer (50 mmol/L) containing 0.5% Triton X-100, NaCl 150 mmol/L, activated orthovanadate 1 mmol/L, and protease inhibitors. The extract was centrifuged at 13,000 rpm for 15 minutes in a cooled microfuge, and the supernatants were aliquoted and stored at -70°C. For use in ELISAs, the A431 cell membrane extract was sequentially diluted with sample dilution buffer at ratios of 1:5, 1:10, 1:20, 1:40, 1:80, 1:160, and 1:320, and then 100-µL aliquots were incubated on each microtiter plate together with the tumor tissue extracts and blanks (containing only dilution buffer). After incubation of the samples and references, the wells were washed five times with 300 µL of TBS containing 0.05% Tween 20 at room temperature to eliminate unbound particles. Subsequent steps included the use of the detection system of a commercially available ELISA assay for the quantitative measurement of erbB-2 (product no. OSDI-10M; Oncogene Science Diagnostics). The response data of the series of diluted reference material were fitted and the curve was used for quantitation of the tumor extracts. The value of the undiluted A431 extract was denominated 1.0 U/mL.
Statistical Analyses
Distribution of ErbB-2 Levels in Malignant and Normal Mammary Epithelial Tissue For the entire collective of 3,208 tumors (whose clinical and pathologic features are described in Table 1), EIA-quantified ErbB-2 levels ranged from 0 to 3,333 U/mg and showed a bimodal distribution (88% in peak A, 12% in peak B) about an intermodal value of approximately 500 U/mg, as shown in Fig 1A. A sufficient number of tumor cases for each of five different patient age groups (22 to 40, 40 to 50, 50 to 60, 60 to 70, and 70 to 80 years) permitted us to address the controversial question of age-specific incidence for tumor ErbB-2 overexpression.21-23 All five patient age groups demonstrated similar distributions of tumor ErbB-2 levels, with median and mean values not significantly different from those of the entire collective (141 and 233 U/mg, respectively, for n = 3,208). However, the percentage of peak B was significantly higher in patients younger than 40 (20%) when compared with all other age groups (12%) (data not shown).
Nonmalignant breast epithelial samples (from 334 matched sets of tumor and NAT) displayed a distribution of low ErbB-2 receptor levels, with a median value of 37 U/mg and with 75% of NAT samples expressing ErbB-2 levels less than 80 U/mg tissue protein, as shown in Fig 1B. When compared with the ErbB-2 values determined in the adjacent set of tumor tissue samples (or those in the low-ErbB-2–expressing peak A from the entire 3,208 tumor collective), the entire distribution of NAT ErbB-2 values was significantly lower than that of the malignant tissues (P < .001, Mann-Whitney test). Of interest, NAT levels ranged from 0 to 1,000 U/mg, with nine (2.7%) outlying cases with ErbB-2 levels higher than 500 U/mg. This low-frequency detection of ErbB-2 overexpression in NAT may reflect undetected contamination of these samples with ErbB-2–overexpressing ductal carcinoma-in-situ,1,24 or morphologically normal-appearing epithelium-bearing activated ErbB-2 and at high risk for malignant transformation.25
Prognostic Significance of EIA-Determined Threshold Value of Tumor ErbB-2 Overexpression
Correlation Between EIA- and IHC-Determined ErbB-2 Overexpression For 151 tumor cases, ErbB-2 status was assessed by IHC scoring (HercepTest kit) of membrane immunoreactivity from formalin-fixed and paraffin-embedded tumor sections obtained adjacent to tumor tissues that were extracted for EIA analysis of ErbB-2 levels. As shown in Fig 3A, EIA-quantified ErbB-2 levels correlated generally with the FDA-approved IHC scoring for ErbB-2 because 99% of tumors scoring 0 and 1+ expressed less than 500 U/mg ErbB-2, whereas only 20% of 2+ tumors and 80% of 3+ tumors expressed more than 500 U/mg ErbB-2. Reanalysis by EIA of the five tumors with low (< 500 U/mg) ErbB-2 protein levels and 3+ IHC scores did not significantly change the EIA results, and histologic re-evaluation did not reveal any obvious explanation for this discrepancy (all samples showed primary membrane immunoreactivity and low background staining), leading to an overall 80% concordance between the two assays. The Kaplan-Meier curves in Fig 3B show that patient survival outcome was known for 98 of these 151 cases, and as with the larger group of 1,123 cases analyzed for outcome, the small subset of ErbB-2–overexpressing tumor patients (> 500 U/mg) had a significantly reduced (P <0.05) probability of DFS compared with those with low-ErbB-2–expressing tumors. When analyzed by IHC scores for ErbB-2 overexpression, however, 3+ but not 2+ ErbB-2 immunoreactivity correlated with significantly reduced DFS, consistent with the observation that 80% of 2+ scoring tumors but only 20% of 3+ scoring tumors actually express low levels (< 500 U/mg) of ErbB-2 receptor measured by EIA.
EIA-Determined ErbB-2 Overexpression Is Associated With Reduced Tumor ER and PgR Content Only 36% of tumors in peak A (Fig 1A) were ER- and/or PgR-negative (< 10 fmol/mg), and those that were ER- or PgR-positive showed no correlation between receptor content and level of ErbB-2 expression (data not shown). In contrast, 48% of tumors in peak B (Fig 1A) were ER- and/or PgR-negative (P < .0001 for peak B v peak A, Mann-Whitney test). The differences in distribution of ER and PgR levels between peak A and peak B tumors are shown in Fig 4, revealing the near six-fold reduction in median ER content and near seven-fold reduction in median PgR content associated with EIA-determined overexpression of ErbB-2. These data suggest that patients with ER- and/or PgR-positive breast tumors that also overexpressed ErbB-2 might be less responsive to the antiestrogen tamoxifen.
EIA-Determined ErbB-2 Overexpression Predicts Patient Outcome After Tamoxifen or CMF Adjuvant Therapy Of the 1,123 cases with known patient follow-up, 950 (84.6%) had ER-positive tumors (Table 1). One hundred sixty-two of these cases received no adjuvant therapy; of these, 18 (12%) of 149 (confidence interval [CI], 7.5% to 18.7%) with low tumor ErbB-2 expression relapsed at a median of 37 months, whereas three (23%) of 13 (CI, 6.2% to 54%) with ErbB-2 overexpression relapsed at a median of 28.5 months. Because 520 of the 1,123 cases with known patient follow-up were ER-positive and were treated only with systemic adjuvant tamoxifen therapy, survival of tamoxifen-treated patients with peak B tumors was compared with survival of tamoxifen-treated patients with peak A tumors to test the hypothesis that ErbB-2 overexpression is associated with reduced responsiveness to tamoxifen therapy.2-4,26,27 As shown in Fig 5A, tamoxifen-treated patients with ER-positive breast tumors that overexpressed ErbB-2 had significantly reduced OS (P = .03) and DDFS (P = .05), with reduction in both relapse rates and RFS as compared with similarly treated cases with low ErbB-2 expression. Relapse rates and median times to relapse for those with low- versus high-ErbB-2–expressing tumors were 84 (17.5%) of 480 (CI, 14.3% to 21.3%) and 24 months versus 10 (25%) of 40 (CI, 13.2% to 41.5%) and 16.8 months, respectively. Because total receptor content represented the only significant difference between the peak A and peak B (ErbB-2–overexpressing) tamoxifen-treated tumors, these outcome data are consistent with other clinical data showing lack of patient benefit when ErbB-2–overexpressing tumors are treated with adjuvant tamoxifen28,29 and suggest that part of this endocrine resistance mechanism may be due to the net reduction in tumor ER and PR content induced by ErbB-2 overexpression.
Previous clinical studies have also suggested that ErbB-2 overexpression is associated with resistance to CMF-based adjuvant chemotherapy.2,3,30-33 Because 154 of the 1,123 tumor cases with known patient follow-up were treated only with systemic adjuvant CMF chemotherapy, survival of CMF-treated patients with peak B tumors was compared with survival of CMF-treated patients with peak A tumors. As shown in Fig 5B, CMF-treated patients with breast tumors that overexpressed ErbB-2 had significantly reduced OS (P = .006), RFS (P = .05), and DDFS (P < .0001) as compared with similarly treated patients whose comparably staged and graded tumors expressed low ErbB-2 levels. Unfortunately, these limited sample sizes lead to underpowered subset analyses and the consequent risk of type II statistical errors.
Overexpression of Total ErbB-2 Is Associated With Phosphorylated ErbB-2
Determination of ErbB-2 receptor expression by EIA offers a clinically valuable alternative to the FDA-approved semiquantitative IHC assessment of breast tumor ErbB-2 status. Unlike nonmalignant breast tissues, which display a unimodal distribution of low ErbB-2 receptor levels (median 37 U/mg), breast tumors show at least an order of magnitude increase in the level of ErbB-2 expression that occurs in a bimodal distribution (peaks A and B), with 12% comprising a biologically distinct set of ErbB-2–overexpressing tumors (peak B). Of interest, this study observed a 2.7% incidence of EIA-determined ErbB-2 overexpression in the nonmalignant mammary epithelium (NAT) adjacent to primary invasive breast tumors. This finding may reflect contamination of this surgically resected tissue by occult premalignant disease such as ductal carcinoma-in-situ, which is known to overexpress ErbB-2 even more frequently than invasive ductal cancers;1,24 it is consistent with previously reported 4.5% and 9.5% incidence rates for c-ErbB-2 gene amplification in pathologically confirmed benign breast tissue from control and high-risk patients, respectively, who have undergone breast biopsies.25 Tumors constituting peak A of the bimodal distribution of EIA-determined ErbB-2 levels showed no significant characteristics or patient outcome differences dependent on their variable levels of ErbB-2 receptor expression. In contrast, peak B tumors containing ErbB-2 levels above the intermodal threshold for overexpression (> 500 U/mg) showed significant biologic and clinical associations with this threshold level of receptor overexpression. Tumor characteristics significantly associated with ErbB-2 overexpression at or above this level included reduced content of ER and PgR, increased tumor size and histologic grade, and significantly enhanced levels of phosphorylated (Y1248-P) ErbB-2 receptor. As well, patient outcome differences distinguished low-ErbB-2–expressing (peak A) tumors from ErbB-2–overexpressing (peak B) tumors because the latter were associated with significantly reduced OS, RFS, DDFS, and LRFS by univariate analysis. Multivariate analysis indicated that EIA-determined ErbB-2 overexpression is an independent prognostic factor associated with a 1.7-fold increased relative risk for breast cancer relapse; this risk magnitude is also consistent with the 1.6-fold relative risk for RFS previously reported from a multivariate analysis of IHC-determined ErbB-2 overexpression in a mixed collection of node-negative and node-positive primary breast tumors.23 Controversies abound regarding the predictive influence of ErbB-2 overexpression on clinical resistance to adjuvant tamoxifen or CMF chemotherapy, and these issues have been thoroughly reviewed.2-4 A reasonable mechanism has not been put forward to explain why ErbB-2–overexpressing breast tumors should be relatively unresponsive to adjuvant CMF. Preclinical studies do not support this hypothesis,32 although at least three clinical studies support it,31-33 and the suggestion has recently been made that this clinical resistance reflects a difference in tumor cell kinetics (as opposed to true drug resistance) wherein the ErbB-2–overexpressing tumors with their presumed higher growth fraction relapse earlier than slower growing tumors with low ErbB-2 expression.3 Our subset analysis performed on 130 comparably staged and graded tumors uniformly treated with adjuvant CMF showed significantly reduced OS (P = .006), RFS (P = .05), and DDFS (P < .0001) associated with ErbB-2 overexpression. Although they do not resolve the controversy about CMF resistance, our findings are in keeping with previous clinical reports in which ErbB-2 overexpression was determined by IHC,31-33 and they support the clinical relevance of our EIA-determined threshold for ErbB-2 overexpression. With regard to the question of ErbB-2 overexpression and clinical resistance to tamoxifen therapy, for which there is supporting preclinical26,27 but contradictory clinical evidence,2,3 the results of our subset analysis on 520 tamoxifen-treated patients with ER-positive tumors are consistent with a recent retrospective analysis showing lack of clinical benefit from adjuvant tamoxifen in patients with ErbB-2–overexpressing tumors.28,29 Apart from providing additional confirming evidence for the biologic relevance of our EIA-determined threshold for ErbB-2 overexpression, the lower ER and PgR content of ErbB-2–overexpressing tumors (Fig 4) supports the hypothesis that ErbB-2-induced downmodulation of ER might contribute to a clinical pattern of tamoxifen resistance in patients with ErbB-2–overexpressing breast tumors. Although limited by the need for milligram quantities (50 to 100 mg) of fresh/frozen tumor samples, EIA measurements of tumor ErbB-2 receptor offer several advantages over slide-based assays of tumor ErbB-2 expression status. Unlike the semiquantitative scores resulting from the FDA-approved IHC assay (HercepTest) for ErbB-2 receptor expression, EIA measurements of ErbB-2 expression avoid potential antigen damage associated with fixation, embedding, and length of slide storage. Furthermore, EIA yields a reproducible and instrument-automated continuous variable readout that is more suitable for multivariate statistical analysis. Correlation of EIA-determined ErbB-2 levels with clinical response to trastuzumab therapy is still needed, but it is expected that use of an EIA-determined threshold for ErbB-2 overexpression will provide a more reliable predictive indicator for trastuzumab eligibility and responsiveness. This study found that only 20% of 2+ HercepTest scoring tumors overexpress ErbB-2 as determined by EIA, and 2+ IHC tumor scores are known to be both unreliable and poorly predictive of trastuzumab responsiveness.1,4,5 Even with more reliable measurements of tumor ErbB-2 receptor content, more specific ErbB-2 assays are needed, as evidenced by clinical trials showing that fewer than 20% of patients with tumors expressing the highest levels of ErbB-2 (3+ IHC scores) experience clinical responses to trastuzumab therapy, and not all of these patients benefit from the combination of trastuzumab with chemotherapy.3,4 Studies with a new monoclonal antibody, PN2A, that recognizes only autophosphorylated (functionally active) ErbB-2 suggest that this may represent a marker for more aggressive (node-metastasizing) ErbB-2–overexpressing breast tumors.13,14 IHC assays using this antibody with special antigen retrieval methods that result in both membrane and cytoplasmic PN2A immunoreactivity indicate that fewer than 20% of ErbB-2–overexpressing breast cancers are PN2A-positive and that phosphorylated ErbB-2 expression is strongly associated with higher tumor levels of ErbB-2.14 Using a modified EIA assay and a commercial polyclonal antibody that recognizes the same Y1248-P ErbB-2 epitope as PN2A, we measured increased Y1248-P ErbB-2 content in 5% (10 of 189) of low-ErbB-2–expressing (peak A) breast tumors and 85% (35 of 41) of ErbB-2–overexpressing (peak B) tumors. Additional studies assessing the prognostic and predictive value of Y1248-P ErbB-2 levels are necessary, but preliminary data suggest that this parameter reflecting overexpression of functionally activated ErbB-2 receptor tyrosine kinase may be a more specific breast tumor marker than total ErbB-2 receptor levels alone.
Supported by Swiss Cancer League grant no. 390-9-1996 (to W.K.), Swiss National Science Foundation grant no. 3100-49505.96 (to U.E.), and the Tumorbank Foundation of Basel. C.B. was supported in part by grants no. P01-CA44768, P50-CA58207, and R01-CA36773 from the National Institutes of Health and by grants from the University of California at San Francisco Hazel P. Munroe fund and the Mt Zion Health Systems (Janet Landfear Memorial) fund. C.B. offers special thanks to Ernest H. Rosenbaum, MD, for his longstanding support and dedication to women’s health care research. We thank Christine Wullschleger, Katja Meyer, and A. Takahashi for technical assistance, data management, and tumor banking. We are indebted to A. Almendral, M. Anabitare, C. Braschler, B. von Castelberg, R. Flury, R. Gaudenz, F. Harder, T. Hardmeier, S. Heinzl, O. Köchli, K. Lüscher, M. Mihatsch, F. Mross, G. Sauter, J. Torhorst, and M. Zuber as clinicians and pathologists. In particular, we are grateful to W.P. Carney and D. Tenney of Oncogene Science Diagnostics, Cambridge, MA, for providing the OSDI immunoassay kits.
1. Ross JS, Fletcher JA: HER2/neu (c-ErbB-2) gene and protein in breast cancer. Am J Clin Pathol 112: S53–S67, 1999[Medline] 2. DiGiovanna MP: Clinical significance of HER2/neu overexpression: Part II, in Rosenberg SA (ed): Principles & Practice of Oncology Updates (vol 13, no 10). Philadelphia, PA, Lippincott Williams & Wilkins, 1999 3. Pegram MD, Konecny G, Slamon DJ: Use of HER2 for predicting response to breast cancer therapy. Dis Breast Updates 3: 1-9, 1999 4. Benz CC, Tripathy D: ErbB-2 overexpression in breast cancer: Biology and clinical translation. J Women’s Cancer 2: 33-40, 2000 5. Trastuzumab/Herceptin. South San Francisco, CA, Genentec, Inc, 1998 (package insert) 6. Slamon D, Leyland-Jones B, Shak S, et al: Addition of Herceptin (humanized anti-HER2 antibody) to first line chemotherapy for HER2 overexpressing metastatic breast cancer (HER2+/MBC) markedly increases anticancer activity: A randomized multinational controlled phase III trial. Proc Am Soc Clin Oncol 17: 98a, 1998 (abstr 377) 7. Norton L, Slamon D, Leyland-Jones B, et al: Overall survival (OS) advantage to simultaneous chemotherapy (CRx) plus the humanized anti-HER2 monoclonal antibody Herceptin (H) in HER2-overexpressing (HER2+) metastatic breast cancer (MBC). Proc Am Soc Clin Oncol 18: 127, 1999 (abstr 483)
8.
Press MF, Hung G, Godolphin W, et al: Sensitivity of HER2/neu antibodies in archival tissue samples: Potential source of error in immunohistochemical studies of oncogene expression. Cancer Res 54: 2771-2777, 1994 9. Hieken TJ, Mehta RR, Shilkaitis A, et al: Her2/neu and p53 expression in breast cancer: Valid prognostic markers when assessed by direct immunoassay, but not by immunochemistry. Proc Am Soc Clin Oncol 15: 112, 1996 (abstr 113) 10. Dittadi R, Brazzale A, Pappagallo G, et al: ErbB-2 assay in breast cancer: Possibly improved clinical information using a quantitative method. Anticancer Res 17: 1245-1247, 1997[Medline] 11. Eissa S, Khalifa A, el-Gharib A, et al: Multivariate analysis of DNA ploidy, p53, c-ErbB-2 proteins, EGFR, and steroid hormone receptors for short-term prognosis in breast cancer. Anticancer Res 17: 3091-3097, 1997[Medline]
12.
Eppenberger U, Kueng W, Schlaeppi J-M, et al: Markers of tumor angiogenesis and proteolysis independently define high- and low-risk subsets of node-negative breast cancer patients. J Clin Oncol 16: 3129-3136, 1998 13. DiGiovanna MP, Carter D, Flynn SD, et al: Functional assay for HER2/neu demonstrates active signalling in a minority of HER2/neu-overexpressing invasive human breast tumours. Br J Cancer 74: 802-806, 1996[Medline]
14.
Thor AD, Liu S, Edgerton S, et al: Activation (tyrosine phosphorylation) of ErbB-2 (HER2/neu): A study of incidence and correlation with outcome in breast cancer. J Clin Oncol 18: 3230-3239, 2000 15. Barlow RE, Bartholomew DJ, Bremmer JM, et al: Statistical Interference Under Order Restrictions. London, United Kingdom, Wiley, 1972 16. Breimann L, Friedman JH, Olsen RA, et al: Classification and Regression Trees. Monterey, CA, Wadsworth and Brooks/Cole, 1984 17. Mantel N: Evaluation of survival data and the two new rank order statistics arising in its consideration. Cancer Chemother Rep 50: 163-170, 1966[Medline] 18. Cox DR: Regression models and life-tables. J R Stat Soc 34: 187-220, 1972 19. Tibshirani R: A plain man’s guide to the proportional hazards model. Clin Invest Med 5: 63-68, 1982[Medline] 20. Levine MN, Browman G, Gent M, et al: When is a prognostic factor useful? A guide for the perplexed. J Clin Oncol 9: 348-356, 1991[Abstract] 21. Kurtz JM, Jacquemier J, Amalric R, et al: Why are local recurrences after breast-conserving therapy more frequent in younger patients? J Clin Oncol 8: 591-598, 1990[Abstract] 22. Xing WR, Gilchrist KW, Harris CP, et al: FISH detection of HER2/neu oncogene amplification in early onset breast cancer. Breast Cancer Res Treat 39: 203-212, 1996[Medline] 23. Sjogren S, Inganas M, Lindgren A, et al: Prognostic and predictive value of c-ErbB-2 overexpression in primary breast cancer, alone and in combination with other prognostic markers. J Clin Oncol 16: 462-469, 1998[Abstract] 24. Liu E, Thor A, He M, et al: The HER2 (c-erbB-2) oncogene is frequently amplified in in situ carcinomas of the breast. Oncogene 7: 1027-1032, 1992[Medline]
25.
Stark A, Hulka BS, Joens S, et al: HER2/neu amplification in benign breast disease and the risk of subsequent breast cancer. J Clin Oncol 18: 267-274, 2000 26. Benz CC, Scott GK, Sarup JC, et al: Estrogen-dependent tamoxifen-resistant tumorigenic growth of MCF-7 cells transfected with HER2/neu. Breast Cancer Res Treat 24: 85-95, 1993 27. Pietras RJ, Arboleda J, Reese DM, et al: HER2 tyrosine kinase pathway targets estrogen receptor and promotes hormone-independent growth in human breast cancer cells. Oncogene 10: 2435-2446, 1995[Medline]
28.
Carlomagno C, Perrone F, Gallo C, et al: c-ErbB-2 overexpression decreases the benefit of adjuvant tamoxifen in early-stage breast cancer without axillary lymph node metastases. J Clin Oncol 14: 2702-2708, 1996 29. Bianco AR, De Laurentiis M, Carlomagno C, et al: HER2 overexpression predicts adjuvant tamoxifen (TAM) failure for early breast cancer (EBC): Complete data at 20 yr of the Naples GUN randomized trial 289. Proc Am Soc Clin Oncol 19: 75a, 2000 (abstr 289)
30.
Allred DC, Clark GM, Tandon AK, et al: HER2/neu in node-negative breast cancer: Prognostic significance of overexpression influenced by the presence of in situ carcinoma. J Clin Oncol 10: 599-605, 1992 31. Gusterson BA, Gelber RD, Goldhirsch A, et al: Prognostic importance of c-ErbB-2 expression in breast cancer: International (Ludwig) Breast Cancer Study Group. J Clin Oncol 10: 1049-1056, 1992[Abstract] 32. Paik S, Bryant J, Wolmark N: ERBB-2 overexpression and response to chemotherapy: NSABP study. Breast Cancer Res Treat 50: 231, 1998 (abstr) 33. Pegram MD, Finn RS, Arzoo K, et al: The effect of HER2/neu overexpression on chemotherapeutic drug sensitivity in human breast and ovarian cancer cells. Oncogene 15: 537-547, 1997[Medline] Submitted February 23, 2000; accepted September 14, 2000.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2001 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 test. The continuous variable function of EIA-determined ErbB-2 values was first tested for prognostic significance by univariate regression analysis against patient relapse rate; logarithmic transformation was also evaluated for its ability to improve these results. A cutoff or prognostic threshold value was searched for by means of classification and regression trees analysis.
