Annals of Oncology Advance Access published July 30, 2012

review

Annals of Oncology 00: 17, 2012 doi:10.1093/annonc/mds238

A critical review why assessment of steroid hormone receptors in breast cancer should be quantitative
O. Brouckaert1*, R. Paridaens1, G. Floris1, E. Rakha2, K. Osborne3 & P. Neven1
1 Department of Gynaecological Oncology, Multidisciplinary Breast Centre UZ, Leuven, Belgium; 2Department of Cellular Pathology, University of Nottingham and Nottingham University Hospitals NHS Trust, Nottingham, UK; 3Dan L. Duncan Cancer Center, Baylor College of Medicine, Houston, USA

Received 15 May 2012; revised 10 June 2012; accepted 12 June 2012

Steroid receptors have been around in the eld of breast cancer for decades now. Still, controversy remains on how best to report steroid receptors. In this review, we will convince the reader why benets outweigh pitfalls, when reporting steroid receptors in a quantitative rather than qualitative way. Summarizing decades of research in this eld, we will explore the evidence why quantitative reporting is superior in all settings (neoadjuvant, adjuvant and metastatic settings). Furthermore, we will also summarize different staining methods, denitions and pitfalls that have shown to be important points of discussion in earlier debates. Although the molecular unraveling of breast cancer in the past decade has revolutionized the way we think about breast cancer, we should not easily abandon the classical pathological variables such as steroid receptors in favor of molecular tools. Key words: breast cancer, ER, PR, qualitative, quantitative, review

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 31, 2013

Sex steroids together with growth factors drive the development, growth and differentiation of breast epithelial tissue following activation of the nuclear estrogen (ER) and progesterone receptors (PR) [16] and are also critical for breast cancer development and progression. ER and PR are expressed in breast cancer, respectively, in 80% and 60% of cases. Their activationmainly through ligand binding or growth factor-induced phosphorylationtriggers diverse gene networks, metabolic and cell regulatory pathways, sustaining cancerous breast epithelium [7]. Whilst the PR is induced by an active estrogenER pathway, the expression of hormone receptors depends on many host (i.e. age at diagnosis, BRCA1/2 mutation, body mass index, parity etc.) and tumor characteristics [i.e. histological type and grade, human epidermal growth factor 2 (HER-2) expression] [8]. The qualitative expression (absent/present) of ER and PR is predictive for treatment response and prognostic for outcome but to a different extent for each receptor. Women with an ERpositive breast cancer not receiving endocrine therapy have a better short-term prognosis than women with an ER-negative tumor [9]. This difference is further enhanced with endocrine therapy as ER expression predicts response to anti-estrogen therapeutic strategies [10]. The PR is more complicated: PR expression is predictive for response to tamoxifen in metastatic breast cancer, but prognostic rather than predictive in the

*Correspondence to: Dr. O. Brouckaert, Department of Gynaecological Oncology & Multidisciplinary Breast Centre UZ Leuven, Herestraat 49, Leuven 3000, Belgium. Tel: +32 16344634; Fax: +32 16344634; E-mail: olivier.brouckaert@uzleuven.be

The Author 2012. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: journals.permissions@oup.com.

review

introduction

adjuvant setting [1120]. Compared with ER/PR-positive cases, ER-positive/PR-negative cases constitute a more aggressive socalled luminal B molecular class of breast cancer [2125] with lower nuclear ER-activity, greater genomic instability, higher proliferation, more PI3K mutations and more crosstalk between the ER and growth factors, although this might be age- and estrogen level related [26]. With high estrogen levels, the PR can be abundantly present together with alternative growth factor signalling whilst with low estrogen levels, the PR can be absent with an intact estrogenER pathway [27]. Loss of PR has several explanations but may reect treatment resistance or activation of growth factor receptor pathways [28]. The ER and PR should be reported in combination with the HER-2 status, since these clinical phenotypes possess discriminative prognostic information, and are able to challenge gene expression proling, particularly when carried out in a central lab [2932]. Semi-quantitative rather than qualitative expression of hormone receptors ne-tunes their predictive and prognostic capabilities. Semi-quantitative subgroups are dened by arbitrarily dened cut-offs of a continuous variable [3234]. Cut-offs are not based on measurements of functional activity of the ER pathway but response rates to endocrine therapy increase with a higher expression for ER [14, 3436]. Tumors with strong ER and PR expression most often show an indolent behavior and the benet from adjuvant chemotherapy is limited [37]. Tumors with a low expression of ER usually lack high PR expression and exhibit features associated with poor prognosis (such as HER-2 expression) and may benet more from adjuvant chemotherapy. Semi-quantitative PR

review
information may rene the 2011 St Gallen classication of luminal breast cancers [37, 38]. We here review why semi-quantitative assessment of hormone receptors provides an additional benet over qualitative reporting in the light of daily clinical decisionmaking. We rst review different methods for measuring hormone receptors and highlight potential pitfalls.

Annals of Oncology

measuring hormone receptors in breast cancer

ER assessment was initially based on standardized ligand binding assays (LBA), but retrospective studies showed that semi-quantitative immunohistochemical (IHC) analysis of ER expression was superior for prognostic and predictive purposes [3942]. The LBA was therefore abandoned, but prospective studies comparing the predictive value of LBA with IHC are not available [32, 43]. IHC is easy, cheaper, specic, safe and usable for evaluating cytology, fresh frozen tissue and formalin-xed parafn-embedded specimens. IHC also allows to discriminate between malignant and benign cells. A lack of standardized protocols in the pre-analytical phase (e.g. choice of specimen, time of xation, processing of different specimens, type of xative etc.), staining procedure (e.g. type of antigen retrieval, detection system), assessment methods (scoring, threshold for positivity) and availability of diverse antibodies remain critical issues leading to discordant results in different laboratories [32, 34, 43, 44]. Quality control programs are essential to guarantee reproducibility and comparability between different centres. False-negative/positive results for ER exist, with error rates as high as 20% [32, 43, 45]. Several IHC semi-quantitative scoring systems for ER and PR have been developed. Amongst the rst was McCartys H score, calculated by multiplying the percentage of positive cells by a factor representing the intensity of immunoreactivity (1 for weak, 2 for moderate and 3 for strong), giving a maximum score of 300 (3+) [39]. Another score is the immunoreactive Remmele score (IRS), the product of the proportion score (0 4) and an intensity score (03) with a range of 012 [46]. Nowadays, the most established score is the Allred (modied quick score) [40]. It uses the same intensity score as the Hscore and IRS system (03), but adds the proportion of cells staining into six subgroups (0, no staining; 1, <1%; 2, between 1% and 10%; 3, between 11% and 33%; 4, between 34% and 66% and 5, between 67%100% of the cells staining).

dening hormone receptor positivity in breast cancer

Although the use of clinically validated, reproducible and standardized cut-offs for determining hormone receptor positivity is critical, prospective studies addressing the optimal IHC cut-off for dening positivity based on the efcacy of endocrine therapy are lacking. Retrospective studies previously carried out using the LBA suggested that ER > 3 fmol/mg protein should be considered positive but others suggested >10 fmol/mg and even 410 fmol/mg [21, 32, 47, 48]. Harvey et al. compared the LBA with IHC in patients treated with

adjuvant tamoxifen [41]. An Allred score of 3 is a good cutoff to predict benet from tamoxifen, but only when using identical standardized staining methods and primary antibody. ER-positivity is considered only when 1% ( proportion score 2) of the tumor cells stain (intensity score = 1) and the tumor is, as per denition, ER-negative, regardless of the intensity score, when <1% of the tumor cells stain ( proportion score 1). Endocrine therapy is not useful in these cases [32, 49]. Although data used to derive the 1% IHC cut-off may be suboptimal (derived from a previous study), there is still an agreement to use it to guide anti-estrogen therapy [32, 50]. In routine practice however, there seems to be a wide range of arbitrary cut-offs in the percentage of stained cells being used (i.e. >0% [51, 52], 1% [53], 5%10% [5457] and 20% irrespective of the scoring method used [51]). In view of the great advancement in the IHC technique and the introduction of new highly sensitive antibodies and the wide use of 10% as a cut-off of positivity in many centers, categorization of tumors showing 1%9% IHC staining remains controversial. Further research in this area is needed to address not only the response rates of these tumors but also their biological features with regard to ER-pathway activation. A recent report in this eld states that almost half of tumors staining 1%9% IHC have molecular characteristic of the basal-like phenotype [58]. Discordances between the H-score and the Allred score for dening ER-positivity exist. These are mainly observed when 1%10% of cells stain, but also in Allred scores 78. The two scoring methods remain in use and a modied conversion table, originally suggested by Shousha, making the two systems more equivalent is shown here (Table 1) [59]. Discordances may also result from using different antibodies with variable sensitivity, particularly when the hormone receptor expression is low. The two antibodies used in the rst generation IHC studies were mouse monoclonal antibodies (MMabs), MMab 6F11 and MMab 1D5. Recently, rabbit monoclonal antibodies (RabMabs) have become increasingly popular because of higher afnities for the desired epitope [60]. RabMab SP1 is such an antibody exhibiting 8-fold higher afnity than MMab 1D5 for the ER [61]. Both the 6F11 and 1D5 as well as SP1 are correlated with clinical outcomes and outperform LBA [62]. A previous retrospective study comparing the newer RabMab SP1 with the conventionally used MMab 1D5 using mostly previously frozen tissue in 4105 cases reported that 8% of breast carcinomas were SP1+/1D5 (correlating with good outcomes) and 2% were SP1/1D5+ (correlating with poorer outcomes) [63]. This study showed that MMabs have lower sensitivity and specicity when compared with RabMabs potentially misclassifying patients who may benet or not from endocrine therapy. Another prospective study comparing SP1 and 1D5 stainings in 508 cases identied only 2 SP1+/1D5 and no SP1/1D5+ cases [64]. Discordance was mainly present in those with <1% staining and therefore of no clinical importance. SP1 was consistently noted as having a greater intensity of staining compared with 1D5, and thus, it was much easier to read positive-low results and negative results. Also, the cytoplasmic staining seen with 1D5 in SP1 ER cases was never seen with SP1. The use of RabMab SP1 therefore seems to reduce the likelihood of false-positive results for ER. The primary antibodies for the ER supported by the American

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 31, 2013

| Brouckaert et al.

Annals of Oncology
Table 1. Allred/H-score conversion table (modied from Shousha [58]) ER status Negative Allred score 0 1 (does not exist) 2/3/4 3/4/5 4/5 5 6 6 6 7 7/8 Allred equation 0+0 1 + 1/2/3 2 + 1/2/3 3 + 1/2 4+1 5+1 4+2 3+3 4+3 5 + 2/3 Stained cells (%) 0 <1 110 1133 3466 67100 3466 1133 3466 67100 Intensity of immunoreactivity 0 1/2/3 1/2/3 1/2 1 1 2 3 3 2/3

review
Equivalent H-score 0 () 0 () 130 (+) 1166 (+) 3466 (+) 67100 (+) 68132 (+/++) 3399 (+) 102198 (++) 134300 (++/+++)

Poor

Intermediate

Rich

Society of Clinical Oncology guidelines published in 2010 are 6F11, 1D5, SP1, ER.2.123+1D5, 1294, 1A6 and 312 and different cut-off points for positive ER status apply for different antibodies [32]. Some authors suggest that using these very sensitive antibodies, hormone receptor expression has now become bimodal [65, 66]. Others do not agree with this concept [33] and a quantitative approach is favored by several facts: (i) the LBA demonstrated ER expression to be continuous rather than bimodal, (ii) not all ER-positive breast cancer respond to endocrine therapy and tumors with high ER expression benet little from adjuvant chemotherapy (iii) molecular classication segregates breast cancer into different intrinsic subtypes with distinct hormone receptor expression and further molecular data reclassied up to 22% of IHC ER-negative cases as ER positive by a 20 gene ER signature and (iv) certain IHC assays like that of Allred showed a variable level of ER expression that correlated with response to tamoxifen in the adjuvant and metastatic settings [41, 42, 62, 67]. Overviews of clinical trials of breast cancer indicate that a small proportion (20%) of tumors with an Allred score of 2 or 3 respond to endocrine treatments and this proportion increases to 50% in tumors with a score of 46 and up to 75%80% in tumors with a score of 7 or 8 [14, 35, 36, 68]. Newer molecular technologies such as qRT-PCR (real-time quantitative reverse transcription polymerase chain reaction) and gene expression microarrays may further rene the assessment of hormone receptor expression. In view of the complexity of available therapy regimens and because endocrine responsiveness depends on more than qualitative and quantitative ER testing, we expect these newer techniques to include other clinically relevant genes aside from hormone receptor-related genes only. For example, the pre-treatment Oncotype DX recurrence score uses qRT-PCR ER gene information together with information obtained from 20 other genes and is not only a highly signicant predictor of recurrence ( prognostic), but it is also being used as a predictive tool to estimate the benet of adjuvant chemotherapy [69]. However, the prognostic information provided by the recurrence score is heavily dependent on the proliferative activity of ER-positive tumors and classical combined histopathologic variables (i.e. ER, PR, HER-2 and mitotic

count/Ki67) provide equivalent information [7074]. Although the techniques used to assess ER/PR are totally different (gene level versus protein level), the concordance among local IHC, central IHC and central qRT-PCR by the proprietary gene assay for ER and PR status is high [67]. Another exciting eld of research in cancer is represented by the epigenetic alterations (so-called epigenomics), including DNA methylation, chromatin remodeling and noncoding interference RNA (miRNA). These epigenetic mechanisms inuence substantially gene expression and protein expression (including ER/PR). Recently, a miRNA signature predictive for ER/PR status has been proposed in breast cancer [75]. Potential future integration of these novel molecular techniques with more classical morphological approaches will pave the way for a fully personalized treatment of breast cancer patients.

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 31, 2013

prognostic value of quantitative ER and/or PR-expression in ER-positive breast cancer

A prognostic factor is a clinical, pathological, or biological feature associated with the aggressiveness of (un)treated invasive breast cancer, and if adverse usually results in the use of (more) adjuvant therapies following surgery. The prognostic value of the qualitative ER and PR expression is well demonstrated both in patients treated with endocrine therapy and, but to a lesser extent, in endocrine therapy naive patients [41]. However, the prognostic effect is time-dependent, disappearing after 58 years, and survival curves between ERpositive and ER-negative disease cross somewhere between 5 and 10 years, also illustrative for late recurrences typical for ER-positive breast cancer [9, 7680]. This is probably explained by the strong correlation between ( positive) ER status and (lower) rates of tumor cell proliferation, as noticeable also in gene expression studies where lower quantitative ER correlates with higher recurrence score (which is mainly driven by proliferation) [81]. It takes ER+ tumors longer to recur and ER status may thus not be a marker for metastatic potential.

doi:10.1093/annonc/mds238 |

review
The prognostic value of quantitative hormone receptor expression in ER-positive endocrine treatment nave cases might be more important for PR than ER. Data from a placebo-controlled trial have shown that quantitative ER expression in patients with an ER-positive breast cancer is not strongly prognostic [35, 68]. ER expression was quantied by the 21 gene Oncotype DX assay using quantitative qRT-PCR in archived parafn blocks. ER and PR expressions were measured on a log based 2 scale from 0 to 15 (relative to reference genes) and a one unit increment is associated with a 2-fold change in expression. The KaplanMeier plot showed that untreated patients with an ER-positive breast cancer have signicant differences in time to distant recurrence by quantitative PR expression (P = 0.002). The highest PR tertile had the longest time to distant recurrence and the lowest PR tertile the shortest. Most other studies have conrmed these results in postmenopausal women [8284]. Interestingly, a small retrospective study including 138 ER-positive, lymphnode negative and stage I and II breast carcinomas recently showed that lack of PR expression associated with mitotic count greater than score 1 is a strong predictor of high Oncotype DX recurrence score [72]. In patients treated with endocrine agents, high ER expression is associated with better outcome than lower ER expression, which is better than no ER expression. This is mainly due to a better therapy response in patients with higher ER expression reecting its predictive effect. The same applies to the PR within the ER-positive treated breast cancer patients, but here, the quantitative PR is prognostic rather than predictive [15, 30, 68, 81, 85, 86]. Furthermore, the overall good prognosis of patients with tumors staining strongly for ER and PR is in line with their negative association with other unfavorable prognostic markers like HER-2 receptor expression and high histological grade, although this might be age related [8789].

Annals of Oncology

predictive value of quantitative ER and/or PR-expression in ER-positive breast cancer

the adjuvant setting
A factor is predictive if it predicts the likelihood of responding to specic types of therapies in both the adjuvant and metastatic settings. In the adjuvant setting, placebo-controlled tamoxifen studies showed that the benet of tamoxifen increases with increasing ER expression, independent of menopausal status [14, 35, 36, 68]. Tumors with low ER expression are less sensitive to tamoxifen. Although there is a gradient of increasing response to endocrine agents with increasing levels of ER, the gradient is skewed such that tumors expressing even very low numbers (e.g. between 1% and 10%) of positive cells show a signicant benet far above that of ER-negative tumors, which are essentially unresponsive. The predictive value of quantitative PR expression for response to tamoxifen is less clear (in the adjuvant setting). As indicated by the recently updated EBCTCG study, the benet of endocrine adjuvant therapy seems not to be affected by quantitative PR expression in ER-

positive breast cancers [36]. A Swedish study randomizing between tamoxifen and placebo in women under age 50 reported the benet of tamoxifen to be restricted to ERpositive breast cancers with a strong expression of PR (>75%) [90]. Strong PR expression can indeed be interpreted as a surrogate marker of stronger ER expression. There is little information on the prognostic value of quantitative expression of ER or PR from placebo-controlled trials using aromatase inhibitors (AIs) as such trials are almost nonexisting. A subgroup analysis of the MA 17 study showed that the clinical benet of extended letrozole versus placebo following 5 years of tamoxifen therapy was limited to the subgroup expressing PR but quantitative levels were not studied. The benet of tamoxifen in ER-negative breast cancer is conned to PR-positive cases and these cases may need to be restained to rule out a false-negative ER staining [32, 91]. As might be predicted from biology, ER-negative/PR-positive tumors are rare and may well represent a false-negative ER assay [31]. PR expression was originally thought to discriminate the benet of tamoxifen compared with an AI but these data did not hold up over time [15, 86, 92, 93]. A prospectively planned pathology substudy in the TEAM trial reported that patients with strong ER expressing tumors may derive an additional benet from an oral AI over tamoxifen although this was not statistically signicant [15]. HER-2-positive breast cancers expressing ER and/or PR belong to the so-called luminal B tumors [24]. Quantitative expression of ER and/or PR is lower than seen in so-called luminal A tumors [24]. The lower levels of ER in HER-2positive tumors may contribute to the lower benet from endocrine therapy seen in patients with these tumors. Such tumors may also be less responsive to trastuzumab. Tumors expressing both ER/PR and HER-2 may be driven by both pathways and may thus be resistant to a drug targeting only one of them. Combination therapy may be more benecial in these tumors as suggested by Neosphere trial [94] and the TBCRC 006 trial [95]. To our knowledge however, no data are available for a predictive role of quantitative expression of hormone receptors in response to trastuzumab or endocrine therapy in HER-2 overexpressing breast cancers. The statistical signicance of the benet from adjuvant trastuzumab is dependent on PRs presence in ER-positive breast cancers in HERA [96]. There is a direct correlation between the likelihood of clinical response to hormonal therapies and the level of ER expression [41]. Tumors with high levels of ER and PR are highly sensitive to endocrine therapy. The 2011 St-Gallen consensus was that benet of adjuvant chemotherapy is small here, irrespective of menopausal status [37, 97, 98]. It remains unclear whether the lack of benet from chemotherapy in these patients is due to an excellent outcome with endocrine therapy or due to genuine lack of biological effect. ER-positive tumors have lower rates of tumor cell proliferation and numerous prior studies have shown a relationship between the measures of cell proliferation and the benet from chemotherapy [99, 100]. Most postmenopausal patients with tumors expressing very high levels of ER can be optimally treated with adjuvant hormonal therapy alone, and can safely

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 31, 2013

| Brouckaert et al.

Annals of Oncology

review
tumors, higher PR expression is associated with increased time to treatment failure and this is not dependent only on the higher proliferation rate seen in PR-negative tumors [124, 125]. This parallels the decreased benet from tamoxifen in the metastatic setting with the absence of PR, although in the adjuvant setting PR expression does not predict clinical benet from AIs over tamoxifen or vice versa. For ER expression, there was only a nonsignicant trend for clinical benet with higher ER expression, consistent with the previously described ndings of Ravdin with tamoxifen. Even in the lower quartile of ER expression, clinical benet (complete/partial response/ stable disease for 6 months) is 50% but no effect on overall survival for different quantitative hormone receptor expression was found. Reports in the early 1980s suggested that lack of ER expression was associated with an increased response to chemotherapy in metastatic breast cancer but others found a high ER expression to be associated with higher response rates [126, 127]. Until today, the impact of quantitative and certainly qualitative hormone receptor expression on chemotherapy response rates in patients with ER-positive metastatic breast cancer remains unclear. The level of ER expression could be predictive of lower response to chemotherapy in the (neo) adjuvant setting [102] and some guidelines extend this knowledge into the metastatic setting [124, 128]. A major problem in metastatic breast cancer is that few studies looked at the level of ER on a metastatic biopsy but only at expression on the primary tumor. Quantitative changes in ER/PR/HER-2 between primary tumor and metastatic lesion occur in up to 16%/40%/10% of patients, respectively, leading to a change in management in 14% [129]. Qualitative ER expression remains concordant between primary tumor and metastatic lesion in 77% and the qualitative changes in the remaining 23% are small [130]. Single-agent rst-line trastuzumab achieves response rates of up to 41% in patients with a HER-2 FISH-positive tumor [131, 132]. Benet from trastuzumab with chemotherapy is similar when hormone receptor expression is assessed qualitatively [133], but decreases with (quantitatively) higher levels [8].

forego the rigors of chemotherapy [99, 101105]. A recent retrospective study conrms this [106]. This stresses the need for robust quantitative methods for analyzing hormone receptor expression. The IBCSG study IX investigated the role of CMF chemotherapy before tamoxifen therapy in postmenopausal women with a lymph node-negative ERpositive breast cancer. There was a survival benet but only in the tumors with low ER expression [107]. IBCSG study VIII in premenopausal women demonstrated that little benet was to be expected from chemotherapy in patients with tumors expressing PR [108]. The fact that AdjuvantOnline! does not incorporate quantitative ER receptor expression and no PR expression at all suggests that the benet of chemotherapy is therefore overestimated [109]. Many other studies describe the link between hormone receptor levels and breast cancer survival with or without chemotherapy. The studies are in line with the consensus that patients with strong hormone receptor expression have a good prognosis and that there is little benet from adjuvant chemotherapy [99,110112].

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 31, 2013

the neoadjuvant setting

In menopausal women, there is a linear relationship between the quantitative expression of ER and the chance of responding to endocrine therapy, both for tamoxifen and AI. Patients with a lower expression for ER had little benet from tamoxifen but more benet from an AI. This effect decreased with increasing expression for ER [113]. Quantitative expression of ER and PR can be used to determine the benet from preoperative chemotherapy [114, 115]. Pathological complete remission in patients with ERpositive breast cancer treated with neoadjuvant chemotherapy decreases with increasing ER expression [116]. This relation is also true for quantitative PR expression in ER-positive breast cancer, but apparently only in HER-2-negative cases [115]. Pathological complete response increases progressively in HER-2-positive patients treated with combined chemotherapy trastuzumab with decreasing semi-quantitative ER and PR expression. The benet from adding trastuzumab is highest in ER- and PR-negative cases [94, 117]. The prognostic value of pathological complete response also depends on the ER and PR status and is only signicant in luminal B/HER-2-negative, HER-2-positive (nonluminal) and triple-negative disease [118].

conclusion
Strong IHC expression for ER/PR correlates with better prognosis and higher chance of responding to endocrine therapy. The (semi)quantitative assessment of hormone receptor expression rather than qualitative assessment estimates the benet of (neo)adjuvant chemotherapy, since this benet decreases with increasing hormone receptor expression. This is known for over >30 years now and we daily implement this knowledge when we consider (neo)chemotherapy in patients of certain age with certain hormone receptor expression. Receptor assays showing a range of ER and PR expression levels will provide the most help to clinicians and their patients.

the metastatic setting

Metastatic postmenopausal breast cancer patients have a clinical benet from rst-line tamoxifen or AIs of 38%59% [119] depending on the quantitative ER expression [120122]. In ER low/intermediate/high expression tumors, the response rates of tamoxifen are 25%/46%/66% respectively [41]. For PR, there is a linear response ratio with increasing PR expression [20], but endocrine therapy can be useful even in patients with absent PR expression [123]. Surprisingly, until recently, there were no studies assessing the relationship between the use of the recent third generation AIs in the metastatic setting and quantitative hormone receptor expression [124]. Response rates for third generation AIs in rst-line metastatic setting are 30%50% [125]. However, in ER-positive

disclosures
KO is a member of the advisory boards of Genentech, Astra Zeneca and Novartis (compensated). All other authors have declared no conicts of interest.
doi:10.1093/annonc/mds238 |

review
references
References 61 to 134 are available as Supplementary data in Annals of Oncology online. 1. Jones B, Russo J. Inuence of steroid hormones on the growth fraction of human breast carcinomas. Am J Clin Pathol 1987; 88(2): 132138. 2. Hankinson SE, Colditz GA, Willett WC. Towards an integrated model for breast cancer etiology: the lifelong interplay of genes, lifestyle, and hormones. Breast Cancer Res 2004; 6(5): 213218. 3. Pike MC, Ross RK. Estrogen-progestin replacement therapy: regulatory action needed. Breast Cancer Res 2002; 4(6): 222223. 4. Anderson E, Clarke RB. Steroid receptors and cell cycle in normal mammary epithelium. J Mammary Gland Biol Neoplasia 2004; 9(1): 313. 5. Joshi PA, Jackson HW, Beristain AG et al Progesterone induces adult mammary stem cell expansion. Nature 2010; 465(7299): 803807. 6. Asselin-Labat ML, Vaillant F, Sheridan JM et al Control of mammary stem cell function by steroid hormone signalling. Nature 2010; 465(7299): 798802. 7. Frasor J, Danes JM, Komm B et al Proling of estrogen up- and down-regulated gene expression in human breast cancer cells: insights into gene networks and pathways underlying estrogenic control of proliferation and cell phenotype. Endocrinology 2003; 144(10): 45624574. 8. Montemurro F, Rossi V, Cossu Rocca M et al Hormone-receptor expression and activity of trastuzumab with chemotherapy in HER2-positive advanced breast cancer patients. Cancer 2012; 118(1): 1726. 9. Knight WA, Livingston RB, Gregory EJ, McGuire WL. Estrogen receptor as an independent prognostic factor for early recurrence in breast cancer. Cancer Res 1977; 37(12): 46694671. 10. Jensen EV. Mechanism of estrogen action in relation to carcinogenesis. Proc Can Cancer Conf 1966; 6: 143165. 11. Osborne CK, Yochmowitz MG, Knight WA et al The value of estrogen and progesterone receptors in the treatment of breast cancer. Cancer 1980; 46(12 Suppl): 28842888. 12. Singh L, Wilson AJ, Baum M, Whimster WF et al The relationship between histological grade, oestrogen receptor status, events and survival at 8 years in the NATO (Nolvadex) trial. Br J Cancer 1988; 57(6): 612614. 13. Kaufmann M, Pusztai L, BEPM. Expert panel members: use of standard markers and incorporation of molecular markers into breast cancer therapy: consensus recommendations from an international expert panel. Cancer 2011; 117(8): 15751582. 14. (EBCTCG). EBCTCG. Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet 2005; 365(9472): 16871717. 15. Bartlett JM, Brookes CL, Robson T, van de Velde CJ et al Estrogen receptor and progesterone receptor as predictive biomarkers of response to endocrine therapy: a prospectively powered pathology study in the tamoxifen and exemestane adjuvant multinational trial. J Clin Oncol 2011; 29(12): 15311538. 16. van de Velde CJ, Rea D, Seynaeve C, Putter H et al Adjuvant tamoxifen and exemestane in early breast cancer (TEAM): a randomised phase 3 trial. Lancet 2011; 377(9762): 321331. 17. Regan MM, Price KN, Giobbie-Hurder A, Thrlimann B et al Interpreting breast international group (BIG) 198: a randomized, double-blind, phase III trial comparing letrozole and tamoxifen as adjuvant endocrine therapy for postmenopausal women with hormone receptor-positive, early breast cancer. Breast Cancer Res 2011; 13(3): 209. 18. Cuzick J, Sestak I, Baum M, Buzdar A et al ATAC/LATTE investigators effect of anastrozole and tamoxifen as adjuvant treatment for early-stage breast cancer: 10-year analysis of the ATAC trial. Lancet Oncol 2010; 11(12): 11351141. 19. Hhnel R, Woodings T, Vivian AB. Prognostic value of estrogen receptors in primary breast cancer. Cancer 1979; 44(2): 671675. 20. Ravdin PM, Green S, Dorr TM et al. Prognostic signicance of progesterone receptor levels in estrogen receptor-positive patients with metastatic breast cancer treated with tamoxifen: results of a prospective southwest oncology group study. J Clin Oncol 1992; 10(8): 1284.

Annals of Oncology
21. Clark GM, Osborne CK, McGuire WL. Correlations between estrogen receptor, progesterone receptor, and patient characteristics in human breast cancer. J Clin Oncol 1984; 2(10): 11021109. 22. Arpino G, Weiss H, Lee AV et al Estrogen receptor-positive, progesterone receptor-negative breast cancer: association with growth factor receptor expression and tamoxifen resistance. J Natl Cancer Inst 2005; 97(17): 12541261. 23. Rakha EA, El-Sayed ME, Green AR et al Biologic and clinical characteristics of breast cancer with single hormone receptor positive phenotype. J Clin Oncol 2007; 25(30): 47724778. 24. Perou CM, Sorlie T, Eisen MB et al Molecular portraits of human breast tumours. Nature 2000; 406(6797): 747752. 25. Sorlie T, Tibshirani R, Parker J et al Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc Natl Acad Sci USA 2003; 100(14): 84188423. 26. Thakkar JP, Mehta DG. A review of an unfavorable subset of breast cancer: estrogen receptor positive progesterone receptor negative. Oncologist 2011; 16(3): 276285. 27. Huang HJ, Neven P, Drijkoningen M, Paridaens R et al Association between HER2/neu and the progesterone receptor in oestrogen-dependent breast cancer is age-related. Breast Cancer Res 2005; 91(1): 8187. 28. Cui X, Schiff R, Arpino G et al Biology of progesterone receptor loss in breast cancer and its implications for endocrine therapy. J Clin Oncol 2005; 23(30): 77217735. 29. Brouckaert O, Pintens S, Van Belle V et al Short-term outcome of primary operated early breast cancer by hormone and HER-2 receptors. Breast Cancer Res Treat 2009; 115(2): 349358. 30. Van Belle V, Van Calster B, Brouckaert O et al Qualitative assessment of the progesterone receptor and HER2 improves the Nottingham prognostic index up to 5 years after breast cancer diagnosis. J Clin Oncol 2010; 28(27): 41294134. 31. Rakha EA, Reis-Filho JS, Ellis IO. Combinatorial biomarker expression in breast cancer. Breast Cancer Res Treat 2010; 120(2): 293308. 32. Hammond ME, Hayes DF, Wolff AC et al American society of clinical oncology/ college of American pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Oncol Pract 2010; 6(4): 195197. 33. Allred DC, Mohsin SK. ER expression is not bimodal in breast cancer. Am J Clin Pathol 2005; 124(3): 474475. 34. Bartlett JM, Rea D, Rimm DL. Quantication of hormone receptors to guide adjuvant therapy choice in early breast cancer: better methods required for improved utility. J Clin Oncol 2011; 29(27): 37153716. 35. Paik S, Shak S, Tang G et al Expression of the 21 Genes in the Recurrence Score Assay and Tamoxifen Clinical Benet in the NSABP Study B-14 of NodeNegative, Estrogen Receptor-Positive Breast Cancer. In Presented at the ASCO Annual Meeting. (Abstr 510), 2005. Orlando, FL. 36. Davies C, Godwin J, Gray R et al Relevance of breast cancer hormone receptors and other factors to the efcacy of adjuvant tamoxifen: patient-level metaanalysis of randomised trials. Lancet 2011; 378(9793): 771784. 37. Goldhirsch A, Wood WC, Coates AS, Gelber RD et al Strategies for subtypes dealing with the diversity of breast cancer: highlights of the St Gallen international expert consensus on the primary therapy of early breast cancer 2011. Ann Oncol 2011; 22(8): 17361747. 38. Prat A, Cheang MC, Martin M et al Prognostic signicance of progesterone receptor-positive tumor cells within immunohistochemically-dened luminal A breast cancer. Ann Oncol 2012; 23(Suppl 2) (Abstr 10O_PR). 39. McCarty KS, Jr, Szabo E, Flowers JL, Cox EB et al Use of a monoclonal antiestrogen receptor antibody in the immunohistochemical evaluation of human tumors. Cancer Res 1986; 48(8 Suppl): 4244s4248s. 40. Ferno M, Andersson C, Fallenius G et al Oestrogen receptor analysis of parafn sections and cytosol samples of primary breast cancer in relation to outcome after adjuvant tamoxifen treatment. The south Sweden breast cancer group. Acta Oncol 1996; 35(1): 1722. 41. Harvey JM, Clark GM, Osborne CK, Allred DC. Estrogen receptor status by immunohistochemistry is superior to the ligand-binding assay for predicting

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 31, 2013

| Brouckaert et al.

Annals of Oncology
response to adjuvant endocrine therapy in breast cancer. J Clin Oncol 1999; 17(5): 14741481. Elledge RM, Green S, Pugh R, Allred DC et al Estrogen receptor (ER) and progesterone receptor (PgR), by ligand-binding assay compared with ER, PgR and pS2, by immunohistochemistry in predicting response to tamoxifen in metastatic breast cancer: a southwest oncology group study. Int J Cancer 2000; 89(2): 111117. Allred DC. Issues and updates: evaluating estrogen receptor-alpha, progesterone receptor, and HER2 in breast cancer. Mod Pathol 2010; 23(Suppl 2): S52S59. Mote PA, Johnston JF, Manninen T et al Detection of progesterone receptor forms A and B by immunohistochemical analysis. J Clin Pathol 2001; 54(8): 624630. Craig Allred D. Commentary: hormone receptor testing in breast cancer: a distress signal from Canada. Oncologist 2008; 13(11): 11341136. Remmele W, Stegner HE. Recommendation for uniform denition of an immunoreactive score (IRS) for immunohistochemical estrogen receptor detection (ER-ICA) in breast cancer tissue. Pathologe 1987; 8(3): 138140. Clark GM. Do we really need prognostic factors in breast cancer? Breast Cancer Res Treat 1993; 30: 117126. Clark GM, Wenger CR, Beardslee S et al How to integrate steroid hormone receptor, ow cytometric, and other prognostic information in regard to primary breast cancer. Cancer 1982; 71: 21572162. Yamashita H, Yando Y, Nishio M et al Immunohistochemical evaluation of hormone receptor status for predicting response to endocrine therapy in metastatic breast cancer. Breast Cancer 2006; 13(1): 7483. Pathology Reporting of Breast Disease. A joint document incorporating the third edition of the NHS breast screening programmes guidelines for pathology reporting in breast cancer screening and the second edition of the royal college of pathologists minimum dataset for breast cancer histopathology. NHSBSP Pub No 58, 2005.

review
51. Layeld LJ, Gupta D, Mooney EE. Assessment of tissue estrogen and progesterone receptor levels: a survey of current practice, techniques, and quantitation methods. Breast J 2000; 6(3): 189196. 52. Putti TC, Pinder SE, Elston CW, Lee AH et al Breast pathology practice: most common problems in a consultation service. Histopathology 2005; 47(5): 445457. 53. Dowsett M, Cuzick J, Wale C, Howell T et al Retrospective analysis of time to recurrence in the ATAC trial according to hormone receptor status: an hypothesis-generating study. J Clin Oncol. 2005; 23(30): 75127517. 54. Regitnig P, Reiner A, Dinges HP, Her G et al Quality assurance for detection of estrogen and progesterone receptors by immunohistochemistry in Austrian pathology laboratories. Virchows Arch 2002; 441(4): 328334. 55. Eiermann W, Paepke S, Appfelstaedt J, Llombart-Cussac A et al Letrozole NeoAdjuvant Breast Cancer Study Group. Preoperative treatment of postmenopausal breast cancer patients with letrozole: a randomized double-blind multicenter study. Ann Oncol 2001; 12(11): 15271532. 56. Stendahl M, Rydn L, Nordenskjld B, Jnsson PE et al High progesterone receptor expression correlates to the effect of adjuvant tamoxifen in premenopausal breast cancer patients. Clin Cancer Res 2006; 12(15): 46144618. 57. Yip CH, Pathy NB, Uiterwaal CS, Taib NA et al Factors affecting estrogen receptor status in a multiracial Asian country: an analysis of 3557 cases. Breast 2011; 20(Suppl 2): S60S64. 58. Iwamoto T, Booser D, Valero V et al Estrogen receptor (ER) mRNA and ERRelated gene expression in breast cancers that are 1% to 10% ER-Positive by Immunohistochemistry. J Clin Oncol 2012; 30(7): 729734. 59. Shousha S. Oestrogen receptor status of breast carcinoma: Allred/H score conversion table. Histopathology 2008; 53(3): 346347. 60. Rossi S, Laurino L, Furlanetto A, Chinellato S et al Rabbit monoclonal antibodies: a comparative study between a novel category of immunoreagents and the corresponding mouse monoclonal antibodies. Am J Clin Pathol 2005; 124(2): 295302.

42.

43. 44.

45. 46.

47. 48.

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 31, 2013

49.

50.

doi:10.1093/annonc/mds238 |