|Year : 2018 | Volume
| Issue : 1 | Page : 5-15
Risk stratification in early-stage estrogen receptor+/HER2-breast cancer patients: Comparative analysis of cost-effective methods
Chandra Prakash, Aparna Gunda, Arun Kumar Attuluri, Lekshmi Madhav, Charusheila Ramkumar, Chetana Basavaraj, Nirupama Naidu, Manjiri M Bakre
OncoStem Diagnostics Pvt. Ltd., Bengaluru, Karnataka, India
|Date of Web Publication||18-Jun-2018|
Dr. Manjiri M Bakre
OncoStem Diagnostics Pvt. Ltd., 4, Raja Ram Mohan Roy Road, Anand Towers, 2nd Floor, Bengaluru - 560 027, Karnataka
Source of Support: None, Conflict of Interest: None
Context: Treatment decisions in early-stage hormone receptor-positive breast cancer patients are dependent on the potential risk of cancer recurrence. Multiple expensive gene expression based or cost-effective methods are used to assess the risk in conjunction with traditional prognostic determinants – age, tumor parameters – size, node, grade, and gold standard biomarkers-estrogen receptor, progesterone receptor (PR), and Human epidermal growth factor receptor 2. Aim: The aim of this study is to compare the performance of multiple economic methods, namely, (1) Ki67; (2) immunohistochemistry 4 (IHC4)-multi-biomarker test; (3) Luminal A/B subtyping (4) PREDICT-an online tool. Settings and Design: IHC was performed as per standard protocol on a retrospective cohort of 401 patients. The Kaplan–Meier analysis and Cox proportional-hazards model were used. Results: The results confirmed that lymph node status is the most useful prognostic indicator among the traditional clinicopathological parameters. IHC4 had a hazard ratio (HR) of 2.847 and separated the low-, intermediate- and high-risk groups significantly (P = 0.0248). Luminal subtyping (HR = 2.530) also stratified the two risk groups significantly (P = 0.0321), but had HR lesser than IHC4. Ki67 and PREDICT could not separate the cohort into low- and high-risk groups with statistical significance. All tools compared separated the low-, intermediate- and high-risk groups with a maximum of 7% difference in metastasis-free survival significantly less compared to Oncotype Dx, which separates with 28% difference in survival. Conclusions: IHC4 is a significant predictor of prognosis among the four tools tested. However, multiple limitations of IHC4 tool about validation and lack of standardized protocols for IHC create a need for a robust, accurate, and cost-effective risk assessment tool.
Keywords: Breast cancer, cost-effective, prognostic tools, risk stratification
|How to cite this article:|
Prakash C, Gunda A, Attuluri AK, Madhav L, Ramkumar C, Basavaraj C, Naidu N, Bakre MM. Risk stratification in early-stage estrogen receptor+/HER2-breast cancer patients: Comparative analysis of cost-effective methods. J Curr Oncol 2018;1:5-15
|How to cite this URL:|
Prakash C, Gunda A, Attuluri AK, Madhav L, Ramkumar C, Basavaraj C, Naidu N, Bakre MM. Risk stratification in early-stage estrogen receptor+/HER2-breast cancer patients: Comparative analysis of cost-effective methods. J Curr Oncol [serial online] 2018 [cited 2020 Jul 7];1:5-15. Available from: http://www.journalofcurrentoncology.org/text.asp?2018/1/1/5/234540
| Introduction|| |
Age and clinicopathological parameters-tumor size, node status, and grade are routinely used by clinicians as prognostic tools in the developing countries.,,,, Immunohistochemistry (IHC) based expression of gold standard biomarkers – estrogen receptor (ER), progesterone receptor (PR), and HER2/neu and gene expression-based tests have improved the scope of prognosis., In this study, we compared four prognostic tools, for example, Ki67 expression, luminal A/B subtyping, IHC4, and PREDICT in a cohort of 401 cases. IHC4 tool provided prognostic information by segregating patients into low-, intermediate- and high-risk groups with statistical significance. However, IHC4 has certain limitations, for example, lack of standardized IHC protocols worldwide, major validation on postmenopausal women and a wide intermediate-risk zone which indicate an unmet need for a broadly applicable, accurate, extensively validated, and cost-effective risk of recurrence assessment test.
| Materials and Methods|| |
Ethics approval and consent to participate
All studies were performed with the approval of the Institutional Review Board and Ethics Committees of the Hospitals participating in the study. Informed consent was obtained according to the Indian Council of Medical Research guidelines since the study was retrospective, observational, noninterventional, and anonymized.
We selected women with Stage I and II invasive ductal carcinoma or invasive lobular carcinoma of the breast, ER + or PR±, HER2/neu-with minimum 5-year follow-up and known the clinical outcome. All patient samples were stripped of personal identifiers. Information was collected on age and calendar year of diagnosis, surgery, tumor size, tumor grade, histologic type, ER, PR, nodal status, radiation treatment, hormonal therapy or chemotherapy (CT), and clinical follow-up for a period of minimum of 5 years, including local, locoregional, or distant recurrences, second primary malignancies, death or date of the last visit. Paraffin-embedded blocks of primary breast tumor from lumpectomy/breast-conserving surgery/modified radical mastectomy which had been fixed and processed as per prescribed norms were used. A detailed protocol of processing of tumor blocks, IHC staining, and grading are given as supplementary information [Appendix [Additional file 1][Additional file 6]].
Intrinsic molecular subtyping
Luminal A and B molecular subtype grouping were done as described.
Immunohistochemistry 4 calculation
Cuzick et al. 2011 reported an equation entailing of four variables, i.e., ER, PR, HER2/neu, and Ki67 for measuring the risk of distant recurrence in breast cancer patients. We adopted the same equation to calculate the IHC4 scores using the below equation,
IHC4 Score = 94.7 × (−0.100 ER10-0.079 PR10 + 0.586 HER2/neu + 0.240 ln [1 + 10x Ki67])
Further details on IHC4 calculation are given in supplementary information.
Kaplan–Meier survival curves were plotted using GraphPad Prism. The hazard ratio (HR) was calculated using log-rank test. For calculation of HR for the grade, tumor size, IHC4; G2 and G3, T2 and T3, and IHC4 intermediate and high were combined. Cox proportional hazards regression analysis was performed using MedCalc software. Comparison between groups of patients was made using the two-sided t-test. Values of P < 0.05 were considered statistically significant.
| Results|| |
Demographics of the study cohort
About 81% of the cohort are Indian patients and the remaining from the USA. As shown in [Table 1], majority of the patients were aged <60 (67%) and the median age of the cohort was 55 years (range 26–75 years). All the patients were positive for ER expression, and 85% of the patients expressed PR. Approximately 50% of the patients had node negative tumors of T2 size (56%) and grade 2 (51%). Of the total patients, 33% of patients were N1. About 71.3% (n = 286) of total patients belonged to Stage II and 2.7% (n = 11) to stage IIIA and 39.15% of the patients (n = 157 received only hormone therapy (CT-naïve) and the rest (n = 244) received both hormone therapy and CT.
|Table 1: Demographics of patients: The table includes the demographic characteristics of patients included in the study|
Click here to view
Risk stratification by traditional prognostic tools
Age, tumor size, node status, and tumor grade are routinely used prognostic factors by many clinicians., We analyzed distant metastasis-free survival (DMFS) as a measure of prognosis in early-stage breast cancer patients for these prognostic factors described below. Many studies have shown that younger women have a more aggressive disease with a greater likelihood of distant recurrence and decreased lifespan., We analyzed the prognostic function of age/usefulness of age (cut off 60 years) to stratify patients into low- and high-risk groups. We did not find any significant difference in DMFS in these two age groups (P = 0.665) [Figure 1]a. Analysis of the cohort by tumor size has showed a higher DMFS for T3 tumors (100%) than T2 and T1 tumors, which can be attributed to a low number of T3 tumors (n = 24). T1 tumors had DMFS (96.35%) compared to T2 (92.11%) [Figure 1]b. However, this was not statistically significant (P = 0.1076). We examined whether tumor grade was a prognostic factor, and found no significant difference in DMFS based on grade [Figure 1]c. Next, we tested the performance of node status as a predictor of prognosis and found as expected node-negative patients had a significantly higher DMFS (95.8%) than patients with 1–3 lymph nodes involvement, N1 (90.37%) (P = 0.0239) [Figure 1]d. Univariate analysis also showed that node status is a useful predictor of prognosis among these traditional prognostic factors, with a hazard ratio of 2.675 (95% confidence interval [CI], 1.139–6.285; P = 0.0239) [Table 2]a.
|Figure 1: Kaplan-Meier survival analysis using traditional prognostic factors: Kaplan-Meier survival curves for patients below and above 60 years (a) for patients with T1, T2 and T3 tumors (b); for patients with grade 1, grade 2 and grade 3 of (c) and for patients with node-positive and negative disease. (d)|
Click here to view
Finally, we studied whether CT had any additional benefit regarding increased DMFS over patients who had received endocrine therapy alone. There was no significant difference in the DMFS of these two groups, suggesting that CT did not really benefit these early-stage hormone positive diseased patients [Supplementary Figure 1 [Additional file 2]].
Our analysis showed that among the traditional prognostic tools tested node status can be used to assess DMFS, this finding is concordant with other published studies about the role of node status in disease prognosis. Next, we tried to evaluate the biomarker-based prognostic tools.
Risk-stratification by biomarker-based prognostic tools in a mixed cohort
We first examined the prognostic performance of the mitotic marker Ki67. Patients with high Ki67 expression are known to have a poor prognosis and early recurrence., We used 14% as cut off to segregate patients into low/high-risk groups. Kaplan–Meier survival analysis [Figure 2]a showed that Ki67 segregated low- and high-risk groups with 4% difference in DMFS (low-risk-95.4% and high-risk-91.3%, P = 0.1039) which was statistically nonsignificant.
|Figure 2: Kapla-Meier survival analysis for biomarker based prognostic tools: Kaplan-Meier survival plots of distant metastasis free survival for low- and high-risk groups categorized by Ki67 (a); luminal subtyping (b); IHC4 (c); PREDICT (d)|
Click here to view
Following this, we examined the performance of the luminal subtyping as a prognostic tool. Luminal subtyping is based on the expression of ER, PR, Ki67, and HER2/neu biomarkers (see methods). “Luminal-A” tumors represent 60% of luminal tumors and are known to have a low risk of recurrence. Total cohort studied was classified into luminal A and B subtypes equally. The Kaplan–Meier survival analysis shown in [Figure 2]b indicated that Luminal A and B subtyping could significantly separate patients based on the risk of recurrence (P = 0.0321). Patients of “Luminal-A” subgroup had a DMFS of 96.5%, and “Luminal-B” subgroup had DMFS of 91.4%.
Next, we tested the performance of the IHC4 tool in predicting prognosis. IHC4 uses the markers-ER, PR, HER2, and Ki67 in predicting prognosis using an algorithm that weights expression of each of these markers individually. [Figure 2]c shows that IHC4 can segregate patients into “low-,” “intermediate-,” and “high-” risk groups (P = 0.0248). Overall DMFS for intermediate + high-risk groups was 91.3% and in low-risk group was 96.88% (P = 0.0207) [Supplementary Figure 1 [Additional file 3]]b. The DMFS values shown for Ki67, luminal A/B subtyping, and IHC4 in the Kaplan-Meier survival curves represent data at 5 years.
Finally, we evaluated the performance of the online tool PREDICT. PREDICT is yet another prognostic tool that predicts overall survival (OS) for 10 years by incorporating the ER, HER2, and Ki67 levels along with tumor size, grade, and node status. We used 90% OS as the cutoff, patients with >90% OS were called low-risk and those with <90% OS were called high-risk for recurrence. PREDICT was unable to stratify patients into low- and high-risk categories in a statistically significant manner (P = 0.6289). The difference in OS at 5 years between the two groups was 1.85% (94.8% low-risk versus 93% high-risk) [Figure 2]d.
Risk stratification in chemotherapy-naïve cohort
Risk of recurrence is the best assessed in patients who have not received CT (called as CT-naïve in this study), as this group does not have the confounding effect of the benefit of CT treatment in preventing recurrence. In the CT-naive cohort of this study (n = 157) “Ki67 low” and “Ki67 high” groups were not separated with statistical significance (95.8% vs. 88.5%, P = 0.0832) [Figure 3]a. Examination of the “Luminal-A” and “Luminal-B” subtypes in the CT-naïve subgroup revealed statistically significant, well-separated Kaplan–Meier survival curves for DMFS (97.5% in “Luminal A” versus 88.4% in “Luminal B,” P = 0.0271) [Figure 3]b. IHC4 was able to stratify patients of this cohort into distinct low-, intermediate, high-risk groups which were well separated with statistically significant “P” value (P = 0.0021) [Figure 3]c. Low-risk patients had a DMFS of 98.8% (n = 89), high-risk group patients had DMFS of 92.3% (n = 13), and intermediate-risk group patients had DMFS of 83.6% (n = 56) [Figure 3]c. PREDICT could not significantly stratify the low- (97.3%) and high-risk (94.5%) groups in the CT-naive cohort (P = 0.1205) [Figure 3]d.
|Figure 3: Kaplan-Meier survival analysis for biomarker based prognostic tools in chemotherapy-naive cohort: Kaplan-Meier survival plots of distant metastasis free survival for low-and high-risk groups categorized by Ki67 (a); luminal subtyping (b); IHC4 (c); PREDICT (d)|
Click here to view
Thus, in the total and CT-naïve cohorts, out of the four biomarker tools tested, IHC4 and luminal A and B subtyping could segregate patients into low-risk and high-risk groups.
Next, we compared the hazard ratios for these tools using univariate and multivariate analysis. In the univariate Cox proportional hazards analysis [Table 2]a IHC4 tool was a good predictor of prognosis with a hazard ratio of 2.847 (95% CI, 1.160–5.751, P = 0.0202) for high + intermediate-versus low-risk groups. Luminal subtyping had a hazard ratio of 2.530 (95% CI, 1.078–5.342, P = 0.0321). The hazard ratio for intermediate-versus low-risk groups of IHC4 was a value of 2.541 (95% CI, 1.026–5.46, P = 0.0445, suggesting that risk of recurrence in the intermediate-risk group is as high as that in the high-risk group [Table 2]a. The high- and intermediate-risk groups together, constitute a significant 52% (n = 208) proportion of the total cohort (n = 401). The recurrence rates appear to be overestimated by this stratification as the actual risk recurrence rate in endocrine therapy treated HR-positive disease is 15%–20%., Thus, we believe that many of the patients in the intermediate-risk zone are being over-treated. In the multivariate Cox proportional hazard analysis, both luminal subtypes (HR = 2.2439, 95% CI, 0.8281–6.0800, P = 0.1139) and IHC4 tool (HR = 2.2571, 95% CI, 0.7878–6.4666, P = 0.1315) had comparable hazard ratios [Table 2]b.
Comparison of the performance of risk stratification tools
To determine the best prognostic tool among those evaluated here, we compared Ki67, luminal subtyping and IHC4 tool for their prognostic performance by their ability to re-stratify risk groups. We did not include PREDICT in this comparison as it predicts OS rather than metastasis-free survival. “Luminal-B” subtype contains tumor cells expressing Ki67 less than and greater than 14%, and thus, we attempted to see if this subtype can be re-stratified further by Ki67 as low- and high-risk groups for recurrence. However, the analysis showed that patients with “Luminal-B” tumors could not be re-stratified into low- or high-risk groups by Ki67 and that both low- and high-risk groups have similar DMFS of 91.5% [Figure 4]a. “Luminal-A” tumors are by definition Ki67-low, and therefore could not be analyzed in this manner. There was no concordance in the “Luminal-A” versus IHC4 low-risk groups and “Luminal-B” versus IHC4 intermediate- and high-risk groups by kappa correlation (see supplementary material). Kappa correlation has been used here to measure the extent of agreement or concordance between the risk groups stratified by luminal subtyping and IHC4. No concordance between these two tools observed by low kappa coefficient (0.134) implies that “Luminal-A/B” is a mixture of low-, intermediate- and high-risk groups that can be re-stratified by IHC4.
|Figure 4: Kaplan–Meier survival analysis for re-stratification of Luminal A and B subtypes by Ki67 and IHC4: Kaplan–Meier survival plots of distant metastasis free survival for low- and high-risk for Ki67 in luminal-B cohort (a); low, intermediate and high-risk groups obtained by IHC4 re-stratification of cohort comprising chemotherapy-naïve patients belonging to either Luminal A (b) or luminal-B (c) subtypes|
Click here to view
However, IHC4 tool was unable to re-stratify either the CT-naive or total cohort of “Luminal-A” patient subgroup with statistical significance [Figure 4]b and [Supplementary Figure 2[Additional file 5]]a, respectively. Interestingly, IHC4 tool re-stratified “Luminal-B” subtype patients from both CT naïve and total cohorts. Patients of CT-naïve “Luminal-B” subtype were stratified with statistically significant P value (DMFS: 100% in low-risk versus 77.7% in intermediate-risk and 92.3% in high-risk, P = 0.0203) [Figure 4]c, but the same was not statistically significant in total cohort (DMFS: 95.8 in low-risk versus 94.3% in high-risk, P = 0.1897) [Supplementary Figure 2 [Additional file 4]]b. This re-stratification results reveal that IHC4 tool can identify high-risk subgroups more accurately than intrinsic molecular subtyping.
In the total cohort of “Luminal-A” tumor subtype, 3.46% of patients (n = 7) were re-stratified by IHC4 tool as high-risk group [Figure 5]. However, about 24.12% (n = 48) of “Luminal-B” subtype patients were stratified as low-risk group [Figure 5] by IHC4 tool, suggesting that this small subset of Luminal B patients need not be given CT, which is otherwise a general choice of treatment for patients of “Luminal-B” subtype.
|Figure 5: Stacked bar diagram for re-stratification of “Luminal-A” and “Luminal-B” subtypes by IHC4|
Click here to view
| Discussion|| |
Clinical management of breast cancer is complex as the disease represents a heterogeneous group of tumors with different morphologic and biological features and varied response to treatments. Therapeutic decisions are based on the information provided by various prognostic and predictive factors. Patients are stratified into risk groups based on a combination of clinicopathological variables such as tumor size, node status, histological grade, and molecular expression of hormone receptors, proliferation markers, signaling proteins and hormone and growth factor receptor proteins, etc. These risk categories are useful for guiding optimum treatment planning. In this study, we compared easily available, low-cost IHC based and traditional prognostic risk of recurrence stratification tools for their performance. Node status is known to be a powerful prognostic tool. Our results show that patients with lymph node involvement have lower DMFS compared to patients without lymph node involvement. Earlier studies have revealed the correlation between lymph node involvement and mortality., There is a direct relationship between the number of involved axillary nodes involved and the risk for distant recurrence and survival period., Systemic therapy improves the survival of node-positive patients. However, studies have shown that 25%–30% of node-positive patients can have metastasis-free survival without adjuvant CT and hormone therapy would suffice. Studies have also indicated that some node-negative women also get benefitted by adjuvant therapy revealing that node status is not an absolute predictor of prognosis.
We have previously shown that Ki67 is not a significant predictor of prognosis in early-stage hormone receptor breast cancer Indian cohort. This study reiterated the finding. Ki67 can stratify patients into two risk groups, but they are separated by a 4% difference in metastasis-free survival which is perhaps not useful in treatment planning. Ki67 biomarker is not an ideal predictor due to the lack of compelling evidence for its role in prognosis, lack of robust analytical validity regarding its staining protocols, and lack of convincing data for prediction of CT benefit.
Luminal subtyping which includes expression levels of Ki67 marker also was a better prognosticator in separating the low- and high-risk groups than Ki67 marker alone. “Luminal-B” tumors could not be re-stratified by Ki67 marker, establishing the superiority of luminal subtyping over Ki67 marker. Patients with “Luminal-A” subtype respond to endocrine therapy and CT is necessary for “Luminal-B” subtype patients., The differences in their response to different treatments can be attributed to molecular differences that exist in both the types. At the molecular level, “Luminal-B” type has more expression of proliferation and cell cycle-related proteins. In comparison, “Luminal-A” subtype has a lower number of mutations and lower number of chromosomal changes. IHC4 was able to re-stratify patients from “Luminal-B” subtypes (CT-naïve and treated cohort) further into low-, intermediate- and high-risk groups, thus adding additional prognostic information to luminal subtyping. We think that the nonsignificance of re-stratification of CT treated cohort could be due to the confounding effect of the benefit of CT to the patients in the treated group.
IHC4 with significant independent prognostic value and its ability to further re-stratify the luminal subtypes certainly adds value as a risk stratification tool. Despite its merits, IHC4 has multiple limitations. (1) IHC4 uses a mathematical equation to arrive at a risk score that uses the expression levels of four biomarkers, and this equation is available for open source. Lack of standardized methods for quantitation of each of these four biomarkers by IHC which arise due to differences in test's sensitivity and dynamic ranges across laboratories can lead to ambiguous results. (2) IHC4 tool uses proliferation markers to assess the risk of recurrence. Multiple recent reports mention that tumor recurrence is a process regulated by multiple pathways but not only proliferation alone. A stratification tool which uses biomarkers from signaling pathways involved in cancer progression and metastasis thus reflective of good versus bad tumor biology beyond proliferation would be more beneficial in optimal therapy planning. Available multi-gene tests do solve this issue, however, are not impactful as they are prohibitively expensive. 3) IHC4 is primarily validated in postmenopausal women. Young age is the major risk factor among breast cancer patients in India with the majority falling between 40 and 49 years of age, with a significant proportion below 30 years of age., Thus, IHC4 tool although useful may not be the most appropriate test for patients of Indian and other Asian countries where the occurrence of breast cancer is at a much earlier age compared to the Western population. 4) Another issue associated with IHC4 is the presence of wide intermediate-risk zone. IHC4 has a wider intermediate-risk zone compared to low- and high-risk zones. Treatment options for the patients in this zone are not well defined and are dependent on a dialog between patient and clinician. As a result, many patients of from this zone are prescribed CT which could result in overtreatment.
| Conclusions|| |
That in the absence of an affordable, accurate prognostic tool, IHC4 can be used as an alternate prognostic tool for patients since IHC4 could stratify the patients into three risk groups with statistical significance. However, the imitations of IHC4 listed above, reinforce the need for an accurate, cost-effective risk of recurrence stratification tool based on biomarkers reflective of tumor biology beyond proliferation and which can be used in pre- and post-menopausal early breast cancer HR-positive patients.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Rakha EA, Reis-Filho JS, Baehner F, Dabbs DJ, Decker T, Eusebi V, et al.
Breast cancer prognostic classification in the molecular era: The role of histological grade. Breast Cancer Res 2010;12:207.
Saez RA, McGuire WL, Clark GM. Prognostic factors in breast cancer. Semin Surg Oncol 1989;5:102-10.
Sotiriou C, Wirapati P, Loi S, Harris A, Fox S, Smeds J, et al.
Gene expression profiling in breast cancer: Understanding the molecular basis of histologic grade to improve prognosis. J Natl Cancer Inst 2006;98:262-72.
Crowe JP Jr., Patrick RJ, Rybicki LA, Grundfest-Broniatowski S, Kim JA, Lee KB, et al.
Race is a fundamental prognostic indicator for 2325 Northeastern Ohio women with infiltrating breast cancer. Breast J 2005;11:124-8.
Cianfrocca M, Goldstein LJ. Prognostic and predictive factors in early-stage breast cancer. Oncologist 2004;9:606-16.
Siewertsz van Reesema LL, Lee MP, Zheleva V, Winston JS, O'Connor CF, Perry RR, et al.
RAS pathway biomarkers for breast cancer prognosis. Clin Lab Int 2016;40:18-23.
Cuzick J, Dowsett M, Pineda S, Wale C, Salter J, Quinn E, et al.
Prognostic value of a combined estrogen receptor, progesterone receptor, Ki-67, and human epidermal growth factor receptor 2 immunohistochemical score and comparison with the genomic health recurrence score in early breast cancer. J Clin Oncol 2011;29:4273-8.
Filipits M, Rudas M, Jakesz R, Dubsky P, Fitzal F, Singer CF, et al.
A new molecular predictor of distant recurrence in ER-positive, HER2-negative breast cancer adds independent information to conventional clinical risk factors. Clin Cancer Res 2011;17:6012-20.
Dowsett M, Nielsen TO, A'Hern R, Bartlett J, Coombes RC, Cuzick J, et al.
Assessment of Ki67 in breast cancer: Recommendations from the international Ki67 in breast cancer working group. J Natl Cancer Inst 2011;103:1656-64.
Tang P, Gary M, Tse M. Immunohistochemical surrogates for molecular classification of breast carcinoma: A 2015 update. Arch Pathol Lab Med 2016;140:806-14.
Beadle BM, Woodward WA, Buchholz TA. The impact of age on outcome in early-stage breast cancer. Semin Radiat Oncol 2011;21:26-34.
Kroman N, Jensen MB, Wohlfahrt J, Mouridsen HT, Andersen PK, Melbye M, et al.
Factors influencing the effect of age on prognosis in breast cancer: Population based study. BMJ 2000;320:474-8.
Viale G, Regan MM, Mastropasqua MG, Maffini F, Maiorano E, Colleoni M, et al.
Predictive value of tumor Ki-67 expression in two randomized trials of adjuvant chemoendocrine therapy for node-negative breast cancer. J Natl Cancer Inst 2008;100:207-12.
Nishimura R, Osako T, Okumura Y, Hayashi M, Toyozumi Y, Arima N, et al.
Ki-67 as a prognostic marker according to breast cancer subtype and a predictor of recurrence time in primary breast cancer. Exp Ther Med 2010;1:747-54.
O'Brien KM, Cole SR, Tse CK, Perou CM, Carey LA, Foulkes WD, et al.
Intrinsic breast tumor subtypes, race, and long-term survival in the carolina breast cancer study. Clin Cancer Res 2010;16:6100-10.
Bartlett JM, Christiansen J, Gustavson M, Rimm DL, Piper T, van de Velde CJ, et al.
Validation of the IHC4 breast cancer prognostic algorithm using multiple approaches on the multinational TEAM clinical trial. Arch Pathol Lab Med 2016;140:66-74.
Wishart GC, Azzato EM, Greenberg DC, Rashbass J, Kearins O, Lawrence G, et al.
PREDICT: A new UK prognostic model that predicts survival following surgery for invasive breast cancer. Breast Cancer Res 2010;12:R1.
Early Breast Cancer Trialists' Collaborative Group (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;365:1687-717.
Early Breast Cancer Trialists' Collaborative Group (EBCTCG), Peto R, Davies C, Godwin J, Gray R, Pan HC, et al.
Comparisons between different polychemotherapy regimens for early breast cancer: Meta-analyses of long-term outcome among 100,000 women in 123 randomised trials. Lancet 2012;379:432-44.
Hugh J, Hanson J, Cheang MC, Nielsen TO, Perou CM, Dumontet C, et al.
Breast cancer subtypes and response to docetaxel in node-positive breast cancer: Use of an immunohistochemical definition in the BCIRG 001 trial. J Clin Oncol 2009;27:1168-76.
Arriagada R, Le MG, Dunant A, Tubiana M, Contesso G. Twenty-five years of follow-up in patients with operable breast carcinoma: Correlation between clinicopathologic factors and the risk of death in each 5-year period. Cancer 2006;106:743-50.
Weiss RB, Woolf SH, Demakos E, Holland JF, Berry DA, Falkson G, et al.
Natural history of more than 20 years of node-positive primary breast carcinoma treated with cyclophosphamide, methotrexate, and fluorouracil-based adjuvant chemotherapy: A study by the cancer and leukemia group B. J Clin Oncol 2003;21:1825-35.
Joensuu H, Pylkkänen L, Toikkanen S. Long-term survival in node-positive breast cancer treated by locoregional therapy alone. Br J Cancer 1998;78:795-9.
Ramkumar C, Prakash C, Madhav L, Kumar A, Basavaraj C, Prathima R, et al.
Asssessment of Ki67 as a prognostic marker in hormone receptor positive breast cancer: A retrospective study on an Indian cohort. J Mol Biomark Diagn 2017;8:2155-9929.
Andre F, Arnedos M, Goubar A, Ghouadni A, Delaloge S. Ki67 – No evidence for its use in node-positive breast cancer. Nat Rev Clin Oncol 2015;12:296-301.
Creighton CJ. The molecular profile of luminal B breast cancer. Biologics 2012;6:289-97.
Györffy B, Schäfer R. Meta-analysis of gene expression profiles related to relapse-free survival in 1,079 breast cancer patients. Breast Cancer Res Treat 2009;118:433-41.
Cancer Genome Atlas Network. Comprehensive molecular portraits of human breast tumours. Nature 2012;490:61-70.
Malvia S, Bagadi SA, Dubey US, Saxena S. Epidemiology of breast cancer in Indian women. Asia Pac J Clin Oncol 2017;13:289-95.
Khokhar A. Breast cancer in India: Where do we stand and where do we go? Asian Pac J Cancer Prev 2012;13:4861-6.
Yeo B, Zabaglo L, Hills M, Dodson A, Smith I, Dowsett M, et al.
Clinical utility of the IHC4+C score in oestrogen receptor-positive early breast cancer: A prospective decision impact study. Br J Cancer 2015;113:390-5.
Elsamany S, Elmorsy S, Alzahrani A, Rasmy A, Abozeed WN, Mohammed AA, et al.
Predictive value of IHC4 score for pathological response to neoadjuvant chemotherapy in hormone receptor-positive breast cancer. Asian Pac J Cancer Prev 2015;16:7975-9.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
[Table 1], [Table 2]