Original Study
Prognostic Signi
fi
cance of Progesterone Receptor
Expression in Estrogen-Receptor Positive,
HER2-Negative, Node-Negative Invasive Breast Cancer
With a Low Ki-67 Labeling Index
Makiko Ono,
1,2Hitoshi Tsuda,
3,4Masayuki Yoshida,
4Chikako Shimizu,
2Takayuki Kinoshita,
5Kenji Tamura
2Abstract
The Ki-67 labeling index has been reported to discriminate luminal A (LA)-like and luminal B-like subtypes in patients with hormone-receptor positive, HER2-negative breast cancer. We analyzed prognostic factors including progesterone receptor (PR) expression. PR expression and tumor size were independent prognostic factors in the LA-like subtype, and the LA/PR-negative subtype had the higher risk of recurrence, especially late recurrence.
Introduction:We showed the clinical implications of the Ki-67 labeling index (LI) in hormone receptor (HR)-positive, HER2-negative (HER2) breast cancer patients with lower risk (luminal A [LA]-like) and higher risk (luminal B [LB]-like)
subtypes.Patients and Methods:Three hundred sixty-nine patients with HRþ, HER2cancers were eligible and we
dichotomized these patients into LA-like and LB-like according to Ki-67 LI (14% cutoff) and analyzed prognostic
significance of progesterone receptor (PR) expression. Results:Of 205 LA-like and 163 LB-like subtypes, PR was
positive in 149 (73%) and 103 (63%). PR expression was a prognostic factor in the LA-like subtype but not in the
LB-like subtype. In LA/PRþ, LA/PR, and LB-like subtypes, the 12-year disease-free survival (DFS) rates were 94.8%,
81.6%, and 79.7% (P ¼ .03), and breast cancer-specific survival (BCSS) rates were 98.4%, 97.4%, and 92.0%,
respectively (P¼.05). Late recurrence occurred in LA/PRsubtype, and differences in prognosis between LA/PRþand
LA/PRsubtypes emerged>5 years after surgery. Twelve-year DFS rates of the LA/PRsubtype were almost equal to
those of the LB-like subtype, whereas 12-year BCSS of the LA/PR subtype was superior to that of the LB-like
subtype. In multivariate analysis, PR expression and tumor size were significant or nearly significant prognostic
factors.Conclusion:PR expression and tumor size were independent prognostic factors in the LA-like subtype, and
the LA/PRsubtype had the higher risk of recurrence, especially late recurrence, than the LA/PRþsubtype. In LA-like
breast cancers, stratification of prognosis according to PR expression and tumor size is important.
Clinical Breast Cancer,Vol.-, No.-,---ª2016 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: Cutoff, Luminal subtype, PR, Prognostic factor, Tumor size
Introduction
Breast cancer is the most common cancer in women. The mor-tality in Western countries is decreasing because of early detection and effective systemic adjuvant therapy. In Japan, however, the
incidence is still increasing, and more than 70,000 cases were newly diagnosed in 2010, with more than 12,000 deaths in 2012.1More than half of newly diagnosed Japanese breast cancer patients are estimated to have node-negative, invasive disease, resulting in local
1
Division of Molecular and Cellular Medicine, National Cancer Center Research Institute, Tokyo, Japan
2
Department of Breast and Medical Oncology, National Cancer Center Hospital, Tokyo, Japan
3Department of Basic Pathology, National Defense Medical College, Saitama, Japan
4Department of Pathology and Clinical Laboratories, National Cancer Center Hospital,
Tokyo, Japan
5Department of Breast Surgery, National Cancer Center Hospital, Tokyo, Japan
Submitted: Feb 22, 2016; Accepted: Jun 16, 2016
Address for correspondence: Hitoshi Tsuda, MD, Department of Basic Pathology, National Defense Medical College, Namiki, Tokorozawa, Saitama 359-8513, Japan Fax:þ81-4-2996-5193; e-mail contact:htsuda@ndmc.ac.jp
1526-8209/ª2016 The Authors. Published by Elsevier Inc. This is an open access article
-or distant recurrence within 10 years of the diagnosis in 10% to 15% of these patients.2,3 Therefore, it has been important to establish criteria for identifying the high-risk patient groups for whom adjuvant chemotherapy might be beneficial.
Breast cancer has been reported to be a heterogeneous disease in previous studies in which a molecular subtype classification of invasive breast cancers using gene expression analysis was correlated with chemotherapeutic response and clinical outcome.4In clinical practice, simplified intrinsic subtype classification using immuno-histochemistry (IHC) is widely used for treatment choice in the neoadjuvant or adjuvant settings.5-7 Hormone receptor
(HR)-positive breast cancers are in practice classified into luminal A (LA)-like subtype and luminal B (LB)-(LA)-like subtype with or without HER2 overexpression. The LA-like subtype is shown to express high levels of HR and a favorable prognosis. In contrast, the LB-like subtype presents with a worse prognosis compared with the LA-like subtype.8In discriminating between LA-like and LB-like subtypes without HER2 overexpression, Ki-67 labeling index (LI) has been shown to be useful.7 We previously reported that Ki-67 LI was useful as a prognosticator to identify high-risk node-negative
luminal/HER2-negative (HER2) tumors in Japanese patients
regardless of 3 different cutoff values.9In these cases, it was shown that any of the Ki-67 LI cutoff values (10%, 14%, and 20%), could be applicable in making determinations between high-risk and low-risk luminal/HER2node-negative invasive breast cancers.
A recent study showed that progesterone receptor (PR) expression as well as Ki-67 LI using IHC were important prognostic factors in luminal/HER2breast cancer, which suggests the utility of PR as a marker for differentiation between LA-like and LB-like subtypes.6 Actually, previous reports revealed that PR expression was a prog-nostic factor in HRþbreast cancer.10In this study, we evaluated the prognostic significance of PR expression using the cohort in whom Ki-67 LI was previously examined9and investigated optimal cutoff levels of Ki-67 LI and PR expression.
Patients and Methods
Patient Selection and Data Collection
Details of patient selection were shown in our previous study.9 Briefly, of 530 consecutive patients with pathologically node-negative invasive primary breast cancer, 369 (70%) with HR-positive and HER2 (luminal/HER2) tumors were identified. We reviewed the medical charts and extracted patient data including age, regimens of adjuvant chemotherapy and endocrine therapies, and dates of recurrence, death, or the last follow-up examination.
Methods for formalin-fixed paraffin blocks, routine pathological evaluation, IHC for estrogen receptor (ER), PR, HER2, and Ki-67, fluorescence in situ hybridization (FISH) for HER2, and tissue microarray construction were described in our previous report.9The present study was approved by the institutional review board of the National Cancer Center, Tokyo, Japan.
Evaluation of IHC
The methods for evaluation of IHC have been shown in our previous report.9Briefly, ER was judged as positive when the Allred
score was3 and as negative when the score was211and PR was
evaluated with 3 cutoff levels of 10%, 20%, and 30%. HER2 was judged as positive when the IHC score was 3þor when it was 2þ
Table 1 Clinicopathological Features of 368 Patients With Luminal/HER2LNode-Negative Invasive Breast Cancers (n[368) Characteristic Patient n (%) Age, Years 50 145 (40) >50 223 (60) Menopause Premenopausal 151 (41) Postmenopausal 217 (59)
Invasive Tumor Size, cm
2.0 275 (75) >2.0 to5.0 89 (24) >5.0 4 (1) Histology IDC 302 (82) Mucinous carcinoma 25 (7) Invasive lobular carcinoma 22 (6) IDC with predominantly intraductal component 17 (4) Other (medullary, tubular) 2 (1)
Ki-67 (‡10%) Positive 221 (60) Negative 147 (40) Ki-67 (‡14%) Positive 163 (44) Negative 205 (56) Ki-67 (‡20%) Positive 87 (24) Negative 281 (76) Histological Grade 1 103 (28) 2 175 (47) 3 88 (24) Unknown 2 (1) Nuclear Grade 1 155 (42) 2 103 (28) 3 109 (30) Unknown 1 (0) Lymphovascular Invasion Positive 138 (37) Negative 230 (62) Adjuvant Chemotherapy Yes 152 (41) CMF 100 (65)a UFT 52 (34)a CEF 1 (1) No 216 (59)
Hormone Therapy (Tamoxifen)
Yes 187 (51)
No 181 (49)
Abbreviations: CEF¼cyclophosphamide, epirubicin, and 5-fluorouracil; CMF¼ cyclophospha-mide, methotrexate, and 5-fluorouracil; IDC¼invasive ductal carcinoma; UFT¼oral tegafur-uracil.
aOne patient received CMF and UFT.
-and the HER2/CEP17 (centromere enumeration probe 17) ratio was2.0 in FISH analysis.12For Ki-67 LI, we selected 2 to 3 areas per tissue core including hot spots of Ki-67þcells, and took mul-tiple photomicrographs of differentfields.13The photomicrographs were printed, and the proportion of Ki-67þ cancer cells with moderate to strong immunoreaction per 1000 cancer cells was calculated independently by 2 observers (M.O. and H.T.), with the average value adopted as Ki-67 LI of each case. LA-like and LB-like subtypes were defined as Ki-67 LI<14% and14%, respectively. We evaluated the prognostic significances of Ki-67 LI on the basis of 3 cutoff values, 10%, 14%, and 20%, and of PR expression on the basis of the cutoff values, 10%, 20%, and 30%.
Statistical Analyses
Statistical analyses were performed using IBM SPSS (SPSS Inc, Chicago, IL). Correlation analyses were performed using the
c
2test for categorical variables. The interval from the date of initial surgery to disease recurrence was defined as disease-free survival (DFS), and the interval from the date of surgery to death from breast cancer or the last follow-up examination was defined as breast cancer-specific survival (BCSS). Death from another disease was regarded as censored. DFS and BCSS curves were drawn using the KaplaneMeier method and compared using the log rank test. Cox univariate and multivariate regression analyses were performed to evaluate prognostic significance of each parameter in patients with luminal-subtype breast cancers. In all analyses, differences were considered significant atP<.05.Results
Correlation Between Ki-67 LI and PR Expression Of 369 tumors, we excluded 1 case because of insufficient ma-terial for evaluation of PR expression. The details of the 368 luminal/HER2breast cancer cases are shown inTable 1. There were 187 patients (51%) who had received adjuvant hormonal therapy with tamoxifen and 153 (41%) who had received adjuvant chemotherapy. The median duration of follow-up for all 368 pa-tients was 118 (range, 0.6-160) months. Overall, 38 recurrences and 14 deaths from breast cancer occurred.
The correlation between the Ki-67 LI and PR expression is
shown in Table 2. Although the Ki-67 LI was 10% in 221
(60%), 14% in 163 (44%), and 20% in 87 (24%) of the
tumors, PR expression was<10% in 66 (18%), <20% in 115 (31%), and<30% in 138 (37%). With every cutoff level of Ki-67 LI, there were substantial PRbreast cancers in patients with low Ki-67 LI breast cancer. For example, 22 (15%) patients with Ki-67 LI<10% had PRtumors when the cutoff level of PR was 10%, whereas 39 (27%) patients with Ki-67 LI <10% had PR tumors when the cutoff level of PR expression was 20%. As ex-pected, the percentage of PRtumors in high and low Ki-67 LI breast cancers increased as the cutoff level for PR expression went higher. Simultaneously, the higher the cutoff level of Ki-67 LI, the more low Ki-67 LI breast cancers were negative for PR. Accordingly, in case of the 20% Ki-67 LI cutoff level, PR expression with any cutoff level and Ki-67 LI were inversely associated (PR 10%, P < .001; PR 20%, P < .001; PR 30%,
P ¼ .003), indicating that 14% to 34% of low Ki-67 LI breast cancers were negative for PR.
When Ki-67 LI was divided into 6 groups with <5, 5 to 10, 10 to 14, 14 to 20, 20 to 30, and>30%, as shown in Figure 1, in breast cancers with high Ki-67 LI the percentage of PR-negative breast cancers increased, showing that Ki-67 LI and PR expression had a significant inverse correlation. Interestingly, when Ki-67
LI was 20%, the number of PR breast cancers greatly
increased. The percentages of PR breast cancers with <20%
Ki-67 LI and 20% Ki-67 LI were 26% and 49%, respectively
(P <.001).
Ki-67 LI and PR Expression as a Prognostic Marker for Luminal-Like Subtypes
To investigate the prognostic significance of PR expression, we examined the survival analyses for patients stratified according to 3 subtypes (ie, LA-like/PRþ[LA/PRþ], LA-like/PR[LA/PR], and LB-like subtypes). When the cutoff value of PR expression was set at 10% or 30%, there was no significant difference in DFS between
LA/PRþ and LA/PR breast cancers. However, as shown in
Figure 2A, with a PR expression cutoff level of 20%, DFS was
Table 2 Correlation Between Ki-67 LI and PR Expression
Ki-67 LI
Luminal/HER2LBreast Cancer (n[368), n (%)
PR 10% Cutoff PR 20% Cutoff PR 30% Cutoff
PRL P PRL P PRL P 10% Cutoff High (10%; n¼221) 44 (20) .23 76 (35) .11 88 (40) .24 Low (<10%; n¼147) 22 (15) 39 (27) 50 (34) 14% Cutoff High (14%; n¼163) 37 (23) .03 59 (36) .07 69 (42) .10 Low (<14%; n¼205) 29 (14) 56 (27) 69 (34) 20% Cutoff High (20%; n¼87) 26 (30) <.001 41 (47) <.001 44 (51) .003 Low (<20%; n¼281) 40 (14) 74 (26) 94 (33) Total 66 (18) 115 (31) 138 (37)
Abbreviations: LI¼labeling index; PR¼progesterone receptor.
-significantly different among 3 groups, and 12-year DFS was
94.8%, 81.6%, and 79.7% in LA/PRþ, LA/PR, and LB-like
groups, respectively (P ¼ .03). Strikingly, DFS for patients with LA/PRþsubtype was quite better than those of the other groups. Of note, the DFS curve for patients with the LA/PRsubtype was close to that for the LB-like subtype, indicating that the LA/PRsubtype has more aggressive characteristics similar to the LB-like subtype compared with the LA/PRþsubtype.
Interestingly, recurrence of the LA/PRsubtype predominantly occurred later, and differences in DFS curves between the patient
groups with the LA/PRþ subtype and the LA/PR subtype
emerged 5 years after curative surgery. With regard to BCSS curves, there was a significant difference among patients with LA/PRþ, LA/PR, and LB-like subtypes, and the 12-year BCSS rates were 98.4%, 97.5%, and 91.9% in these 3 groups, respectively (P ¼.04). Unlike DFS, the BCSS for patients with the LA/PR subtype was better than that for the LB-like subtype and close to that for the LA/PRþsubtype.
Cox Univariate and Multivariate Analyses of DFS In Cox univariate analysis of DFS for patients with the LA-like subtype, it was shown that large tumor size was a significant indica-tor of worse clinical outcome (hazard ratio, 3.36; 95% confidence interval [CI], 1.18-9.58;P¼ .02). Although the significance was marginal, negative PR expression with a cutoff level of 20% also tended to be correlated with worse clinical outcomes (hazard ratio, 2.81; 95% CI, 0.98-8.00;P¼.05;Table 3). Likewise, a multivariate analysis including pT factor and PR showed that both of these factors remained significant or marginally significant prognostic factors (tu-mor size, hazard ratio, 3.36, 95% CI, 1.18-9.59,P¼.02; PR, hazard ratio, 2.83, 95% CI, 0.99-8.07,P¼.05;Table 4).
In univariate analysis of BCSS for patients with the LA-like subtype, no factor had any prognostic significance (Table 5). This might be because a few number of events occurred in BCSS. However, in univariate analysis of DFS for patients with the LB-like subtype, unlike the LA-like subtype, there was no significant prognostic factor (data not shown), indicating that clinical outcome in patients with LB-like subtype is influenced by Ki-67 LI but not PR expression.
Discussion
For decades, the predictive and prognostic significance of PR expression in breast cancer have been debated, and so far there have been equivocal data about the role of PR expression in HRþbreast cancer.10,14In the present study we showed that PR expression was
a marginally significant independent prognostic factor in the LA-like subtype and that differences in prognosis between LA/PRþ and LA/PRbecame evident5 years after surgery.
A recent meta-analysis has shown a similar reduction in the risk of recurrence in ERþ/PRþ and ERþ/PR early breast cancers treated with adjuvant tamoxifen.10Simultaneously, PR expression had no influence on the efficacy of an aromatase inhibitor over tamoxifen in ERþearly breast cancer patients,15,16whereas baseline risk of recurrence is higher in the PRsubgroup than the PRþ subgroup in ERþbreast cancer patients.10Thesefindings revealed that PR expression is a prognostic but not a predictive factor.
Figure 1 Correlation Between Ki-67 Labeling Index (LI) and Progesterone Receptor (PgR) Negativity (<20%). Error Bar Is Presented as SD
Figure 2 Survival Curves for 368 Patients Stratified According to Luminal A (LA) Progesterone Receptor (PgR)-Positive, LA/PgRL, and Luminal B Subtypes. (A) Disease-Free Survival (DFS); (B) Breast Cancer-Specific Survival (BCSS)
-There have been several molecular mechanisms with the loss of PR expression in breast cancer reported so far, in which PR is found to be the end product of estrogen action, and the absence of PR expression means nonfunctional ER.17In addition, recent reports have shown that aberrant growth factor signaling resulted in reduction of PR expression.18 Indeed, in our cohort including luminal/HER2þsubtype, luminal/HER2þbreast cancers are more likely to be negative for PR expression compared with luminal/
HER2 breast cancers (data not shown). All these suggested
mechanisms of the loss of PR expression could be associated with highly malignant characteristics leading to poor clinical outcomes.
Interestingly, the present study has shown that PR expression affects late recurrence in the LA-like subtype. Therefore, BCSS for LA/PRwas as good as that for the LA/PRþsubtype over 10 years after surgery. It is assumed that differences in BCSS between LA/ PRþand LA/PRwould emerge later, even 20 years after surgery. Because the extended use of hormonal therapy beyond 5 years could prevent late recurrences,19,20it is important to clarify the risk factors of late recurrence in patient selection for its extended use. Previous studies have reported that proliferation markers and Grade were important prognostic factors for early recurrence within 5 years after diagnosis in HRþbreast cancer, but these studies failed to predict later recurrences.21Simultaneously, several molecular tests that give weight to proliferation and cell cycle progression such as Oncotype DX (Genomic Health, Redwood City, CA) and Mammaprint (Agendia, Amsterdam) could not significantly predict late re-currences.22,23 Thesefindings suggest that the molecular
mecha-nism might be different between early and late recurrence. Indeed,
gene signatures related to recurrence risk have been shown to differ between early and late recurrence.24,25
To date, there have been no reports showing PR expression as a prognostic marker for late recurrence. However, in a gene expres-sion analysis, HR signaling was shown to be a risk factor of late recurrence. The EndoPredict, a gene signature that combines proliferation-related genes and ER signaling genes, has been re-ported to predict early and late recurrences.24Interestingly, the risk
of late recurrences was significantly higher in breast cancers with the higher risk score estimated from expression of 5 ER signaling genes,
RBBP8 (retinoblastoma-binding protein 8), IL6ST (interleukin 6
signal transducer),AZGP1(alpha-2-glycoprotein 1, zinc-binding),
MGP (matrix Gla protein), and STC2 (stanniocalcin 2), than in those with the lower risk score.24Some of these 5 genes are shown to be regulated by progesterone. Therefore, a low expression level of PR is implied to be associated with risk of late recurrence.26,27
Similarly with EndoPredict, Breast Cancer Index (BCI) and PAM50 are able to predict late recurrences as well as early re-currences in HRþ breast cancer.22,25,28-30 BCI consists of the Molecular Grade Index (MGI), which depends on proliferation-related genes, andHOXB13(homeobox B13):IL17BR(interleukin 17 receptor B) ratio (H/I). MGI is associated with early but not late recurrence, whereas H/I has a greater prognostic power in late re-currences than in early rere-currences.25
In a study that showed the prognostic significance of H/I, PR expression was not associated with late recurrence risk.31 How-ever, in that study, the cutoff level of PR expression was much lower than 20%. In the present study, a PR cutoff level of 10% did not show a significant difference in prognosis in LA-like breast cancers. In these points, the results of the present study are consistent with those of previous studies that showed the importance of a 20% cutoff level for PR in IHC according to the data of PAM50.6,28,29
We also showed tumor size as an important prognostic factor. Although nowadays we give weight to breast cancer biology with subtypes in risk estimation, tumor burden such as tumor size and nodal status remain important prognostic factors.32It has been
re-ported that tumor size and nodal status were independent prog-nostic factors for late recurrence in patients with ERþbreast cancer
Table 3 Cox Univariate Model Analyses for Breast Cancer Recurrence in 205 Patients With Luminal A (Ki-67 LI<14%) Cancers
Factor Hazard Ratio 95% CI P
Age (£50 vs.>50), Years 0.77 0.26-2.30 .64
Menopause (Pre- vs. Post-) 1.41 0.49-4.02 .52
Pathological T (pT2 or pT3 vs. pT1)
3.36 1.18-9.58 .02
Grade (3 vs. 1, 2) 2.46 0.69-8.83 .17
Nuclear Grade (3 vs. 1, 2) 2.55 0.80-8.12 .11
Histology (IDC vs. Others) 1.78 0.40-7.97 .45
PR<10% vs.‡10% 1.15 0.25-4.98 .89
PR<20% vs.‡20% 2.81 0.98-8.00 .05 PR<30% vs.‡30% 2.05 0.72-5.84 .18
ThePvalues which are significant or almost significant are shown in bold.
Abbreviations: IDC¼invasive ductal carcinoma; LI¼labeling index; PR¼progesterone receptor.
Table 4 Cox Multivariate Model Analyses for Breast Cancer Recurrence in 205 Patients With Luminal A (Ki-67 LI<14%) Cancers
Factor Hazard Ratio 95% CI P
Pathological T (pT2 or pT3 vs. pT1)
3.36 1.18-9.59 .02
PR<20% vs.‡20% 2.83 0.99-8.07 .05
Abbreviation: LI¼labeling index; PR¼progesterone receptor.
Table 5 Cox Univariate Model Analyses for Breast Cancer-Specific Survival in 205 Patients With Luminal A (Ki-67 LI<14%) Cancers
Factor Hazard Ratio 95% CI P
Age (£50 vs. 50), Years 0.67 0.06-7.44 .75
Menopause (Pre- vs. Post-) 2.72 0.25-30.1 .41
Pathological T (pT2 or pT3 vs. pT1)
6.48 0.59-71.5 .13
Grade (3 vs. 1, 2) 4.32 0.39-47.7 .23
Nuclear Grade (3 vs. 1, 2) 3.07 0.28-33.8 .36
Histology (IDC vs. Others) 30.45 0.001-2814 .59
PR<10% vs.‡10% 0.04 0.001-8227 .66
PR<20% vs.‡20% 1.32 0.12-14.52 .82
PR<30% vs.‡30% 0.96 0.09-10.60 .97
Abbreviations: IDC¼invasive ductal carcinoma; LI¼labeling index; PR¼progesterone receptor.
-treated with tamoxifen. PAM50 is calculated not only on the basis of breast cancer subtype but also tumor burden with tumor size.30 These facts might have partly caused the data that PAM50 was a better prognostic marker than Oncotype DX and IHC4 (the immunohistochemical 4) in HRþbreast cancers.
Conclusion
We showed that PR expression and tumor size were significant or marginally significant independent prognostic factors in the LA-like subtype and that differences in prognosis between LA/PRþ and LA/PRbecame evident5 years after surgery, resulting in better prognosis in LA/PRthan in the LB-like subtype. In LA-like breast cancer, stratification of prognosis according to PR expression and tumor size is important.
Clinical Practice Points
The node-negative invasive breast cancer of the LA-like subtype is defined as HR-positive, HER2-negative, and Ki-67-low sub-group and is correlated with good prognosis.
Nonetheless, a minority of patients with these characteristics suffer from relapse of and death from LA-like breast cancer. In the present study, we examined 205 consecutive patients who
received surgical therapy to node-negative LA-like primary invasive breast cancer, and 149 (73%) and 56 cases (27%) of these 205 were PRþ(20% or more cells) and PR(<20% of cells), respectively.
In this LA-like group, in addition to pathological tumor size, PR negativity was a significant indicator of a lower DFS rate: 12-year DFS rates were 94.8% in the LA-like group with positive PR and 81.6% in the LA-like group with negative PR status.
Therefore, the status of PR in node-negative LA-like invasive carcinoma could be useful in the consideration of adjuvant therapy.
Acknowledgments
This work was supported in part by the National Cancer Center Research and Development Fund (23-A-17, 26-A-4), the Grant-in-Aid for Scientific Research (C) from the Japan Society for the Promotion of Science (22501039, 25430144), and Grant-in-Aid for the Third-Term Comprehensive 10-Year Strategy for Cancer Control from the Ministry of Health, Labour, and Welfare, Japan (H22-023). The authors thank Sachiko Miura, MT, Chizu Kina, MT, Ms Chinami Onuma (National Cancer Center Hospital), and Ms Kozue Suzuki (National Defense Medical College) for technical support.
Disclosure
The authors have stated that they have no conflicts of interest.
References
1.National Cancer Center for Cancer Control and Information Services, Japan, Cancer mortality (1958-2012). Tokyo: Vital Statistics Japan (Ministry of Health, Labour and Welfare); 2013.
2.Kurebayashi J, Kanomata N, Shimo T, et al. Marked lymphovascular invasion, progesterone receptor negativity, and high Ki-67 labeling index predict poor
outcome in breast cancer patients treated with endocrine therapy alone.Breast Cancer2014; 21:214-22.
3.Watanabe T, Sano M, Takashima S, et al. Oral uracil and tegafur compared with classic cyclophosphamide, methotrexate, fluorouracil as postoperative chemo-therapy in patients with node-negative, high-risk breast cancer: National Surgical Adjuvant Study for Breast Cancer 01 Trial.J Clin Oncol2009; 27:1368-74. 4.Perou CM, Sorlie T, Eisen MB, et al. Molecular portraits of human breast
tu-mours.Nature2000; 406:747-52.
5.Carey LA, Perou CM, Livasy CA, et al. Race, breast cancer subtypes, and survival in the Carolina Breast Cancer Study.JAMA2006; 295:2492-502.
6.Prat A, Cheang MC, Martin M, et al. Prognostic significance of progesterone receptor-positive tumor cells within immunohistochemically defined luminal A breast cancer.J Clin Oncol2013; 31:203-9.
7.Cheang MC, Chia SK, Voduc D, et al. Ki-67 index, HER2 status, and prognosis of patients with luminal B breast cancer.J Natl Cancer Inst2009; 101:736-50. 8.Sorlie T, Perou CM, Tibshirani R, et al. Gene expression patterns of breast
car-cinomas distinguish tumor subclasses with clinical implications.Proc Natl Acad Sci U S A2001; 98:10869-74.
9.Ono M, Tsuda H, Yunokawa M, et al. Prognostic impact of Ki-67 labeling indices with 3 different cutoff values, histological grade, and nuclear grade in hormone-receptor-positive, HER2-negative, node-negative invasive breast cancers.Breast Cancer2015; 22:141-52.
10.Davies C, Godwin J, Gray R, et al. Relevance of breast cancer hormone receptors and other factors to the efficacy of adjuvant tamoxifen: patient-level meta-analysis of randomised trials.Lancet2011; 378:771-84.
11.Harvey JM, Clark GM, Osborne CK, et al. Estrogen receptor status by immu-nohistochemistry is superior to the ligand-binding assay for predicting response to adjuvant endocrine therapy in breast cancer.J Clin Oncol1999; 17:1474-81. 12.Wolff AC, Hammond ME, Schwartz JN, et al. American Society of Clinical
Oncology/College of American Pathologists guideline recommendations for hu-man epidermal growth factor receptor 2 testing in breast cancer.J Clin Oncol2007; 25:118-45.
13.Dowsett M, Nielsen TO, A’Hern R, et al. Assessment of Ki-67 in breast cancer: recommendations from the International Ki-67 in Breast Cancer working group. J Natl Cancer Inst2011; 103:1656-64.
14.Bardou VJ, Arpino G, Elledge RM, et al. Progesterone receptor status significantly improves outcome prediction over estrogen receptor status alone for adjuvant endocrine therapy in two large breast cancer databases.J Clin Oncol2003; 21: 1973-9.
15.Dowsett M, Allred C, Knox J, et al. Relationship between quantitative estrogen and progesterone receptor expression and human epidermal growth factor receptor 2 (HER-2) status with recurrence in the Arimidex, Tamoxifen, Alone or in Combination trial.J Clin Oncol2008; 26:1059-65.
16.Viale G, Regan MM, Dell’Orto P, et al. Which patients benefit most from adjuvant aromatase inhibitors? Results using a composite measure of prognostic risk in the BIG 1-98 randomized trial.Ann Oncol2011; 22:2201-7.
17.Horwitz KB, McGuire WL. Estrogen control of progesterone receptor in human breast cancer. Correlation with nuclear processing of estrogen receptor.J Biol Chem 1978; 253:2223-8.
18.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 Inst2005; 97:1254-61.
19.Jin H, Tu D, Zhao N, et al. Longer-term outcomes of letrozole versus placebo after 5 years of tamoxifen in the NCIC CTG MA.17 trial: analyses adjusting for treatment crossover.J Clin Oncol2012; 30:718-21.
20.Davies C, Pan H, Godwin J, et al. Long-term effects of continuing adjuvant tamoxifen to 10 years versus stopping at 5 years after diagnosis of oestrogen receptor-positive breast cancer: ATLAS, a randomised trial.Lancet2013; 381:805-16. 21.Kennecke HF, Olivotto IA, Speers C, et al. Late risk of relapse and mortality
among postmenopausal women with estrogen responsive early breast cancer after 5 years of tamoxifen.Ann Oncol2007; 18:45-51.
22.Sgroi DC, Sestak I, Cuzick J, et al. Prediction of late distant recurrence in patients with oestrogen-receptor-positive breast cancer: a prospective comparison of the breast-cancer index (BCI) assay, 21-gene recurrence score, and IHC4 in the TransATAC study population.Lancet Oncol2013; 14:1067-76.
23.van’t Veer LJ, Dai H, van de Vijver MJ, et al. Gene expression profiling predicts clinical outcome of breast cancer.Nature2002; 415:530-6.
24.Dubsky P, Brase JC, Jakesz R, et al. The EndoPredict score provides prognostic information on late distant metastases in ERþ/HER2breast cancer patients.Br J Cancer2013; 109:2959-64.
25.Zhang Y, Schnabel CA, Schroeder BE, et al. Breast cancer index identifies early-stage estrogen receptor-positive breast cancer patients at risk for early- and late-distant recurrence.Clin Cancer Res2013; 19:4196-205.
26.Raulic S, Ramos-Valdes Y, DiMattia GE. Stanniocalcin 2 expression is regulated by hormone signalling and negatively affects breast cancer cell viability in vitro. J Endocrinol2008; 197:517-29.
27.Song G, Satterfield MC, Kim J, et al. Progesterone and interferon tau regulate leukemia inhibitory factor receptor and IL6ST in the ovine uterus during early pregnancy.Reproduction2009; 137:553-65.
28.Filipits M, Nielsen TO, Rudas M, et al. The PAM50 risk-of-recurrence score predicts risk for late distant recurrence after endocrine therapy in postmenopausal women with endocrine-responsive early breast cancer.Clin Cancer Res2014; 20: 1298-305.
29.Sestak I, Cuzick J, Dowsett M, et al. Prediction of late distant recurrence after 5 years of endocrine treatment: a combined analysis of patients from the Austrian
-breast and colorectal cancer study group 8 and arimidex, tamoxifen alone or in combination randomized trials using the PAM50 risk of recurrence score.J Clin Oncol2015; 33:916-22.
30.Dowsett M, Sestak I, Lopez-Knowles E, et al. Comparison of PAM50 risk of recurrence score with oncotype DX and IHC4 for predicting risk of distant recurrence after endocrine therapy.J Clin Oncol2013; 31:2783-90.
31.Sgroi DC, Carney E, Zarrella E, et al. Prediction of late disease recurrence and extended adjuvant letrozole benefit by the HOXB13/IL17BR biomarker.J Natl Cancer Inst2013; 105:1036-42.
32.Coates AS, Winer EP, Goldhirsch A, et al. Tailoring therapies-improving the management of early breast cancer: St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2015.Ann Oncol2015; 26:1533-46.
