Breast cancer (BC) is the most frequent carcinoma di-agnosed and the leading cause of cancer death in fe-
males worldwide.1 BC treatment has improved due to thediscovery of specific prognostic and predictive biomarkersthat enable more individualized therapies to differentmolecular subgroups. In this scenario, estrogen receptor(ER), progesterone receptor (PR), and human epidermalgrowth factor receptor 2 (HER2) have played a significant
role of selecting patients to hormonal and HER2 target-ing therapy. Unfortunately adjuvant systemic therapyfails to prevent recurrence in some patients with high-riskhormone receptor positive, and many lower risk patientsmay have been subjected to side effects without benefitfrom such interventions. Thus, better discrimination be-tween patients at high risk for recurrence from patients atlow risk is needed.
BC recurrence is an important marker related topoor outcome.2 Several articles showed different variablesassociated with BC recurrence and prognosis, suggestingthat we could more accurately identify patients that willbenefit from adjuvant treatment, like chemotherapy.More recently, BC molecular classification had been as-sociated with an important prognostic tool.2–18 Classi-fication can be based on gene expression profiling(Oncotype DX; Genomic Health, Redwood City, CA orMammaPrint; Agendia, Amsterdam, the Netherlands) orimmunohistochemical (IHQ) analysis.9,12,13,17,19–21 Al-though most of these studies reported survival differences,only a few have associated local (LR) and distant (DR)relapses with different molecular subtypes.11,14,16
In this retrospective cohort, we proposed that LRand DR can be reliably predicted by BC subtype classi-fication with 4 IHQ markers.
MATERIALS AND METHODSAll patient submitted to surgical treatment to BC
between 2000 and 2005 were searched in our computerdatabank at the university (Hospital Santa Casa deMisericordia de Porto Alegre). The cohort included 410patients. Including criteria were woman sex, older than 18years, invasive BC (lobular or ductal), and potentiallycurative breast resections with clear surgical margins.Those who did not had their follow-up scheduled at ourinstitution were excluded.
IHQ technique staining for ER, PR, HER2, and Ki-67 were performed as previously described,22 but ER andPR were considered positive when Z1% of the nucleiwere stained in 10 high-power field, and HER2 wasconsidered positive only when IHQ score 3+. Patientswere classified into the following categories: luminal A(ER or PR positive, HER2 negative, Ki-67<14%);luminal B (ER or PR positive, HER2 negative,Ki-67Z14%); luminal-HER2 (ER or PR positive, HER2positive); HER2 enriched (ER and PR negative, HER2
Received for publication February 7, 2013; accepted March 9, 2013.From the Departments of *Surgical Oncology; wMastology, Hospital
Nossa Senhora da Conceicao; zDepartment of Pathology, HospitalSanta Casa de Misericordia; and yDepartment of Gynecological,Hospital de Clınicas de Porto Alegre, Porto Alegre, RS, Brazil.
The authors declare no conflict of interest.Reprints: Ricardo P. Cruz, MSc, Av. Willy Eugenio Fleck 1500, casa 29,
positive); and triple negative (ER and PR negative, HER2negative).
For local treatment of BC, patients underwenteither breast-conserving surgery (BCS) or mastectomy,followed or not by adjuvant radiotherapy. Adjuvanttherapy was anthracycline based. LR was defined as dis-ease recurrence in the ipsilateral breast, chest wall, axil-lary or supraclavicular ipsilateral nodes. DR was definedas disease recurrence in internal mammary nodes, and anyother metastatic disease. Patients were censored at thetime of last follow-up, on the date of DR, or at the time ofdeath. Descriptive analysis was performed to variables,with mean and SD to continuous variables, andpercentage to categorical ones. The w2 was used in thebivariable analysis. For bivariable survival analysis, LR-free survival and DR-free survival were estimated usingKaplan-Meier curves, and survival differences were as-sessed using log-rank tests. Multivariable analysis wasperformed with Cox proportional hazard models to allcovariates with P<0.20 acquired in the bivariable anal-ysis. All statistical analyses were carried out using Stat-istical Package for the Social Sciences version 17.0 (SPSS,Chicago, IL) having a P<0.05 as significant. This study
was approved by the Research Ethics Committee of theUniversidade Federal de Ciencias Medicas de PortoAlegre.
RESULTSFrom January 2000 to December 2005, a total of
410 BC were identified. From these, 48 were excludedbecause they did not meet the study criteria: 9 hadmetastatic disease at the initial diagnosis, 6 had under-gone BC therapy previously, 6 were submitted to BCS butdid not receive adjuvant radiotherapy, 4 were male, and23 had their follow-up in other institution. The finalcohort included 362 cases, with a mean follow-up of4.8±2.37 years. Four died before 5 years of follow-upfor another reason not related to BC.
BC relapses were noticed in 124 (34.3%) patients.LR only were present in 26 (7.2%) patients, and 53(14.6%) patients had DR only. Patients that had LR andDR totalized 45 (12.4%). Relapses were compared withclinical and pathologic features of this cohort listedin Table 1.
No statistical differences (P=0.187) were evidencedin bivariate analysis of locoregional relapses among
TABLE 1. Distribution of Clinical and Pathologic Characteristics of the Cohort According to Relapses
No. Patients (%)
Characteristics (n) Relapses Group (n=124) Nonrelapses Group (n=238) v2 (P)
Age (y) (362)40 and younger 17 (53.1) 15 (46.9)Above 40 107 (32.4) 223 (67.6) 0.018
patients treated with BCS and RT (23 of 129 patients,15.1%), mastectomy (17 of 73 patients, 23.3%), or mas-tectomy and RT (31 of 137 patients, 22.6%).
Lymph node dissection was performed in 295(81.5%) patients. The mean number of lymph nodesdissected was 16. Separate analysis of regional relapses(axillar relapses) was performed (data not shown). Nev-ertheless, no statistical differences were found and theywere analyzed together with LR; perhaps a larger samplemay evidence statistical relapses differences.
Eighty-two of 345 (22.7%) patients received neo-adjuvant chemotherapy. Adjuvant chemotherapy, radio-therapy, and hormone therapy were performed in 208(57.5%, of 345), 289 (79.8%, of 345), and 236 (65.2%, of347) patients, respectively. Hormonal treatment was ad-ministered to 91.7% of patients with ER or PR positive.Seven (1.9%) patients received trastuzumab. Eighteen(5%) patients did not receive any systemic treatment.
BC molecular subtypes achieved by IHQ analysisshowed 119 (32.9%) luminal A, 73 (20.2%) luminal B, 67(18.5%) luminal-HER2, 41 (11.3%) HER2 enriched, and 62(17.1%) triple negative. The clinical and pathologic featureswere compared between the subtype cohorts (Table 2).
Bivariate survival analysis revealed statistically sig-nificant differences in LR (Fig. 1A) and DR (Fig. 1B)among the molecular BC subtypes. For both LR and DR,patients with luminal A tumors had the most favorableprognosis, with LR and DR rates in 5 years of only 6.7%and 15.1%, respectively. Conversely, HER2-enriched andtriple-negative groups exhibited the highest rates of LR(26.8% and 38.7%, respectively) and DR (34.1% and40.3%, respectively).
Multivariable Cox analysis revealed statisticallysignificant differences in LR for tumors >5 cm, N2 or N3lymph node stages, and triple-negative phenotype (TNP)molecular subtype (Table 3). Luminal B subtype showedstatistical tendency (P=0.053) to LR. Variables pre-dictors of DR were tumors >5 cm, N2 or N3 lymph nodestages, and local radiotherapy. Hormonal therapy(tamoxifen) was associated with reduced DR rates(P=0.017), but did not with LR (P=0.979).
BCS Subgroups AnalysesBivariate survival analysis of patients treated with
BCS revealed statistically significant differences in wholerelapses (Fig. 2A) and LR (Fig. 2B) among the molecular
TABLE 2. Distribution of Clinical and Pathologic Characteristics Among Breast Cancer Molecular Subtypes
No. Patients (%)
Characteristics (n) Luminal A Luminal B Luminal-HER2 HER2 Enriched Triple Negative v2 (P)
FIGURE 1. A, Bivariate analysis of local recurrence-free survival among patients reveals statistically significant differences amongbreast cancer intrinsic subtypes (log-rank test, P < 0.0001). B, Bivariate analysis of distant recurrence-free survival among patientsreveals statistically significant differences among breast cancer intrinsic subtypes (log-rank test, P < 0.0001). Blue line, luminal A;red, luminal human epidermal growth factor receptor 2 (luminal-HER2); green, luminal B; violet, HER2 enriched; yellow, triplenegative (basal and nonbasal).
Cruz et al Appl Immunohistochem Mol Morphol � Volume 00, Number 00, ’’ 2013
4 | www.appliedimmunohist.com r 2013 Lippincott Williams & Wilkins
BC subtypes. Recurrence-free survival (RFS) is shownon Table 4. For both LR and DR, patients with luminalA tumors had the most favorable prognosis, with LR andDR rates of only 4.5% and 9.1%, respectively. Con-versely, HER2-enriched and triple-negative groups ex-hibited the highest rates of LR (36.4% and 42.9%,respectively). Triple-negative and luminal B groupsshowed the highest rates of DR (21.4%, and 21.9%,respectively).
Multivariable Cox analysis revealed statisticallysignificant differences in LR for N3 lymph node stage,luminal B, and HER2-enriched molecular subtypes(Table 5). Variables predictors of DR were age 40 yearsand younger, N2 and N3 lymph node stages, and luminalB molecular subtype.
Mastectomy Subgroups AnalysesBivariate survival analysis of patients treated with
mastectomy revealed statistically significant differences inwhole relapses (Fig. 3A) and LR (Fig. 3B) among themolecular BC subtypes. RFS are shown in Table 4. Forboth LR and DR, patients with luminal A tumors had themost favorable prognosis, with LR and DR rates of
only 9.4% and 22.6% in 5 years, respectively. Conversely,luminal B and triple-negative groups exhibited the highestrates of LR (24.4% and 37.5%, respectively), followed byHER2-enriched group (23.3%). HER2-enriched and tri-ple-negative groups exhibited the highest rates of DR(40% and 45.8%, respectively).
Multivariable Cox analysis revealed statistically sig-nificant differences in LR and DR for N2 and N3 lymphnode stages (Table 6), respectively. N3 lymph node stage hada tendency to LR (P=0.058). Molecular subtypes were notpredictors of recurrence after mastectomy. Hormonal ther-apy (tamoxifen) was associated with reduced DR rate.
DISCUSSIONThis study demonstrates the importance of BC
subtype classification on risk assessment to LR and DR.We demonstrated that biological subtyping with 4 IHQbiomarkers (ER, PR, HER2, and Ki67) had statisticallysignificant value that was independent of standard clin-icopathologic parameters (including age at diagnosis, tu-mor size, grade, lymphovascular invasion, and axillarylymph node status) to identify high-risk LR women,
TABLE 3. Multivariable Cox Analysis of Local and Distant Relapse
Local Relapse (n=345) Distant Relapses (n=345)
Variables HR 95% CI P HR 95% CI P
Age (y)Above 40 1.0 1.040 and younger 0.9 0.4-2.0 0.793 0.8 0.4-1.6 0.595
FIGURE 2. A, Bivariate analysis of recurrence-free survival in patients treated with breast-conserving therapy reveals significantdifferences among breast cancer intrinsic subtypes (log-rank test, P = 0.03). B, Bivariate analysis of local recurrence-free survival inpatients treated with breast-conserving therapy reveals significant differences among breast cancer intrinsic subtypes (log-ranktest, P = 0.013). Blue line, luminal A; red, luminal human epidermal growth factor receptor 2 (luminal-HER2); green, luminal B;violet, HER2 enriched; yellow, triple negative (basal and nonbasal).
Cruz et al Appl Immunohistochem Mol Morphol � Volume 00, Number 00, ’’ 2013
6 | www.appliedimmunohist.com r 2013 Lippincott Williams & Wilkins
mainly on BCS. Some studies had evidenced an asso-ciation between BC molecular subtype and recur-rences.3,6–8,11,12,14,16 However, other works had not foundrelationship of different subtypes to LR risk.23,24
Cheang et al12 did not make separate analysis ofmolecular subtypes and BC local therapies. They found inmultivariable analysis that luminal B and luminal-HER2were independently associated with increased risk ofcancer relapse and death when comparable with luminalA cancers. A case-control study evidenced that basal-likeTNP (ER, PR and HER2 negative, cytokeratin 5/6 pos-itive) and HER2 enriched (ER and PR negative, HER2positive) were associated with more aggressive pre-sentations of BC.7 A retrospective cohort compared TNPwith other BC subtypes.9 TNP was associated with pooreroverall survival and RFS.
The importance of BC subtype classification inpatients treated with BCS to predict relapse risk and
prognosis have been demonstrated.11,14,16 A cohort of 793patients treated with BCS found that HER2-enriched andTNP tumors were associated with an increased LR onmultivariable analysis.11 Multivariable analysis alsoshows increased DR to luminal B and TNP basal-likesubtypes. Millar and colleagues compared recurrencerates among 498 early-stage BCs treated with BCS. Theyfound a bigger 5-year locoregional recurrence to TNPtumors when compared with luminal A tumors on mul-tivariable Cox analysis.14
A cohort of 482 patients treated with BCS did notfind difference in LR rates between TNP and otherssubtypes of BCs, although a higher DR and poor prog-nosis were observed in TNP cancers versus other sub-types.23 Another study also did not find LR ratesdifference to TNP BC in 1601 patients, but they proposedthat TNP have a more aggressive transient clinical coursethan other forms of BC.24 Other cohort exhibited lower
FIGURE 3. A, Bivariate analysis of recurrence-free survival in patients treated with mastectomy therapy reveals significant dif-ferences among breast cancer intrinsic subtypes (log-rank test, P = 0.005). B, Bivariate analysis of local recurrence-free survival inpatients treated with mastectomy therapy reveals significant differences among breast cancer intrinsic subtypes (log-rank test,P = 0.003). Blue line, luminal A; red, luminal human epidermal growth factor receptor 2 (luminal-HER2); green, luminal B; violet,HER2 enriched; yellow, triple negative (basal and nonbasal).
Cruz et al Appl Immunohistochem Mol Morphol � Volume 00, Number 00, ’’ 2013
8 | www.appliedimmunohist.com r 2013 Lippincott Williams & Wilkins
rates of LR and regional relapses and better prognosis toluminal A BC subtype.16 At the same sample they foundthat HER2 enriched was associated with a higher LR andregional relapses after BCS on multivariable analysis.
Other articles also showed the importance of sub-type classification on the relapse patients treated withmastectomy.10,16 Kyndi et al10 found that TNP andHER2-enriched subtypes were independent predictors tolocoregional relapse. In other study, luminal B, luminal-HER2, HER2 enriched, and TNP basal-like were asso-ciated with higher risk of LR and regional relapses aftermastectomy on multivariable analysis.16
In the current study, luminal A tumors exhibit thesmallest LR and DR rates, 6.7% and 15.1%, respectively.TNP and HER2 enriched exhibit the highest LR rates,38.7% and 26.8%, respectively. On multivariable anal-ysis, TNP was associated with higher LR, but none wasassociated with DR. Luminal B subgroup had a tendencyto LR (P=0.053). When the sample was divided by BCSand mastectomy, BC subtype classification was an in-dependent variable to relapse only in the BCS group.Multivariable analysis showed that in the BCS group,luminal B and HER2-enriched subtypes were related toincreased LR rates, but only luminal B achieved statistical
significance to DR. These results indicate an importantrole of Ki-67 in BC subtype classification. We adopted the14% cutoff, detaching patients with hormone receptorpositive and HER2 negative that presents worse RFS.Luminal B was the second largest BC subtype (20.2%) ofthis sample. Other studies also pointed the importance ofKi-67 on relapse prediction and worse prog-nosis.12,16,21,25–29 Cheang et al12 defined the Ki-67 cut-point of 13.25% to differentiate luminal B from luminal Atumors. They associated luminal B with smaller survivaland relapse-free survival when compared with luminal Atumors. Cutoff points for Ki-67 varied among otherstudies, described as 10% to 20%.16,29–32
Subtype classification related with regional relapsewas also described.16 In our sample, axillary dissectionwas performed in 295 patients (81.5%). Same as pre-viously described, this could explain the low axillar re-lapse rate (only 2.2%) evidenced in our sample.33–35 Nostatistical difference was found between subtype classi-fication and axillary relapses.
Same as Voduc et al16 and Nguyen et al,11 a majorlimitation of this cohort is the underuse of systemic ofadjuvant trastuzumab, only in 7 (1.9%) patients. Use oftrastuzumab would probably reduce relapses’ risk.
In conformity with previous works,11,12,14,16 wedemonstrated that a small panel of IHQ markers canidentify patients at increased relapse risk. However, stilltoday, clinical decision making is based basically on a fewprognostic factors (tumor size, lymph node status, histo-logic grade, lymphovascular invasion, hormone receptorstatus, HER2 status, and age).27,36,37
Randomized controlled trials should not only eval-uate new surgical, interventional, and systemic therapiesand their combinations within a multidisciplinary setting,but also assess the clinical value of biological markers toidentify responders to specific therapies.
Although further follow-up is needed, these resultsmay be useful in counseling patients about their antici-pated outcome after BC diagnosis.
CONCLUSIONSBC subtype classification is an important outcome
predictor tool. Luminal A tumors had best prognosis andthe lowest rate of LR and DR. TNP and HER2-enrichedtumors had higher relapses rates and worse RFS.
Use of Ki-67 to luminal B subtype classification wasa valuable prognostic marker, showing increased relapserisk. BC classification using 4 IHQ markers can identifypatients at an increased risk of LR and DR after BCS,contributing to decide about the extension of surgicalresection.
Despite the existence of gene expression profiling(Oncotype DX, MammaPrint), in the current study weanalyzed 4 IHQ markers as an alternative to identifyhigher recurrence risk patients. Randomized trials be-tween gene expression profiling and IHQ markers areneeded to indicate the most cost-benefit pattern of BCmolecular classification.
TABLE 6. Multivariable Cox Analysis of Local and DistantRelapse After Mastectomy
Local Relapse
(n=193)
Distant Relapses
(n=193)
Variables HR 95% CI P HR 95% CI P
Age (y)Above 40 1.0 1.040 and younger 0.2 0.06-1.2 0.092 0.4 0.1-1.1 0.081
REFERENCES1. Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA
Cancer J Clin. 2011;61:69–90.2. Haffty BG, Hauser A, Choi DH, et al. Molecular markers for
prognosis after isolated postmastectomy chest wall recurrence.Cancer. 2004;100:252–263.
3. Sørlie T, Perou CM, Tibshirani R, et al. Gene expression patterns ofbreast carcinomas distinguish tumor subclasses with clinicalimplications. Proc Natl Acad Sci USA. 2001;98:10869–10874.
4. van’t Veer LJ, Dai H, van de Vijver MJ, et al. Gene expressionprofiling predicts clinical outcome of breast cancer. Nature.2002;415:530–536.
5. Sotiriou C, Neo SY, McShane LM, et al. Breast cancer classificationand prognosis based on gene expression profiles from a population-based study. Proc Natl Acad Sci USA. 2003;100:10393–10398.
6. Brenton JD, Carey LA, Ahmed AA, et al. Molecular classificationand molecular forecasting of breast cancer: ready for clinicalapplication? J Clin Oncol. 2005;23:7350–7360.
7. Carey LA, Perou CM, Livasy CA, et al. Race, breast cancersubtypes, and survival in the Carolina Breast Cancer Study. JAMA.2006;295:2492–2502.
8. Fan C, Oh DS, Wessels L, et al. Concordance among gene-expression-based predictors for breast cancer. N Engl J Med.2006;355:560–569.
9. Rakha EA, El-Sayed ME, Green AR, et al. Prognostic markers intriple-negative breast cancer. Cancer. 2007;109:25–32.
10. Kyndi M, Sørensen FB, Knudsen H, et al. Estrogen receptor,progesterone receptor, HER-2, and response to postmastectomyradiotherapy in high-risk breast cancer: the Danish Breast CancerCooperative Group. J Clin Oncol. 2008;26:1419–1426.
11. Nguyen PL, Taghian AG, Katz MS, et al. Breast cancer subtypeapproximated by estrogen receptor, progesterone receptor, andHER-2 is associated with local and distant recurrence after breastconserving therapy. J Clin Oncol. 2008;26:2373–2378.
12. Cheang MC, Chia SK, Voduc D, et al. Ki67 index, HER2 status,and prognosis of patients with luminal B breast cancer. J NatlCancer Inst. 2009;101:736–750.
13. Cianfrocca M, Gradishar W. New molecular classifications of breastcancer. CA Cancer J Clin. 2009;59:303–313.
14. Millar EK, Graham PH, O’Toole SA, et al. Prediction of localrecurrence, distant metastases, and death after breast-conservingtherapy in early stage invasive breast cancer using a five-biomarker panel. J Clin Oncol. 2009;27:4701–4708.
15. Khan SA, Eladoumikdachi F. Optimal surgical treatment of breastcancer: implications for local control and survival. J Surg Oncol.2010;101:677–686.
16. Voduc KD, Cheang MC, Tyldesley S, et al. Breast cancer subtypesand the risk of local and regional relapse. J Clin Oncol.2010;28:1684–1691.
17. Bhargava R, Beriwal S, Striebel JM, et al. Breast cancer molecularclass ERBB2: preponderance of tumors with apocrine differ-entiation and expression of basal phenotype markers CK5, CK5/6,and EGFR. Appl Immunohistochem Mol Morphol. 2010;18:113–118.
18. Niemiec JA, Adamczyk A, Ma"ecki K, et al. Relationships betweenimmunophenotype, Ki-67 index, microvascular density, Ep-CAM/P-cadherin, and MMP-2 expression in early-stage invasive ductalbreast cancer. Appl Immunohistochem Mol Morphol. 2012;20:550–560.
19. Nielsen TO, Hsu FD, Jensen K, et al. Immunohistochemical andclinical characterization of the basal-like subtype of invasive breastcarcinoma. Clin Cancer Res. 2004;10:5367–5374.
20. Cheang MC, Voduc D, Bajdik C, et al. Basal like breast cancerdefined by five biomarkers has superior prognostic value than triple-negative phenotype. Clin Cancer Res. 2008;14:1368–1376.
21. Weigel MT, Dowsett M. Current and emerging biomarkers in breastcancer: prognosis and prediction. Endocr Relat Cancer. 2010;17:R245–R262.
22. Schorr MC, Pedrini JL, Savaris RF, et al. Are the pure in situ breastductal carcinomas and those associated with invasive carcinomathe same? Appl Immunohistochem Mol Morphol. 2010;18:51–54.
23. Haffty BG, Yang Q, Reiss M, et al. Locoregional relapse and distantmetastasis in conservatively managed triple negative early-stagebreast cancer. J Clin Oncol. 2006;24:5652–5657.
24. Dent R, Trudeau M, Pritchard KI, et al. Triple negative breastcancer: clinical features and patterns of recurrence. Clin Cancer Res.2007;13:4429–4434.
25. Elkhuizen PH, Voogd AC, van den Broek LC, et al. Risk factors forlocal recurrence after breast conserving therapy for invasivecarcinomas: a case-control study of histological factors andalterations in oncogene expression. Int J Radiat Oncol Biol Phys.1999;45:73–83.
26. Colleoni M, Rotmensz N, Peruzzotti G, et al. Minimal and smallsize invasive breast cancer with no axillary lymph node involvement:the need for tailored adjuvant therapies. Ann Oncol. 2004;15:1633–1639.
27. de Azambuja E, Cardoso F, de Castro G Jr. et al. Ki-67 asprognostic marker in early breast cancer: a meta-analysis ofpublished studies involving 12,155 patients. Br J Cancer.2007;96:1504–1513.
28. Stuart-Harris R, Caldas C, Pinder SE, et al. Proliferation markersand survival in early breast cancer: a systematic review and meta-analysis of 85 studies in 32,825 patients. Breast. 2008;17:323–334.
29. Clahsen PC, van de Velde CJ, Duval C, et al. The utility of mitoticindex, oestrogen receptor and Ki-67 measurements in the creation ofnovel prognostic indices for node-negative breast cancer. Eur J SurgOncol. 1999;25:356–363.
30. Keshgegian AA, Cnaan A. Proliferation markers in breastcarcinoma. Mitotic figure count, S-phase fraction, proliferating cellnuclear antigen, Ki-67 and MIB-1. Am J Clin Pathol. 1995;104:42–49.
31. Bevilacqua P, Verderio P, Barbareschi M, et al. Lack of prognosticsignificance of the monoclonal antibody Ki-S1, a novel marker ofproliferative activity, in node-negative breast carcinoma. BreastCancer Res Treat. 1996;37:123–133.
32. Joensuu H, Isola J, Lundin M, et al. Amplification of erbB2 anderbB2 expression are superior to estrogen receptor status as riskfactors for distant recurrence in pT1N0M0 breast cancer: anationwide population-based study. Clin Cancer Res. 2003;9:923–930.
33. Rutgers EJ. Guidelines to assure quality in breast cancer surgery.Eur J Surg Oncol. 2005;31:568–576.
34. Sanghani M, Balk EM, Cady B. Impact of axillary lymph nodedissection on breast cancer outcome in clinically node negativepatients: a systematic review and meta-analysis. Cancer. 2009;115:1613–1620.
35. Yildirim E. Locoregional recurrence in breast carcinoma patients.Eur J Surg Oncol. 2009;35:258–263.
36. Colozza M, Azambuja E, Cardoso F, et al. Proliferative markers asprognostic and predictive tools in early breast cancer: where arewe now? Ann Oncol. 2005;16:1723–1739.
37. Hayes DF. Prognostic and predictive factors revisited. Breast.2005;14:493–499.
Cruz et al Appl Immunohistochem Mol Morphol � Volume 00, Number 00, ’’ 2013
10 | www.appliedimmunohist.com r 2013 Lippincott Williams & Wilkins
