Advertisement

Contribution of endocrine therapy in oestrogen receptor-positive pT1a-b breast cancer: Results of a retrospective study

Open AccessPublished:October 01, 2022DOI:https://doi.org/10.1016/j.ejca.2022.08.027

      Highlights

      • No endocrine therapy (ET) was significantly associated with decreased disease-free survival in all patients (hazard ratio = 1.275).
      • No ET was significantly associated with decreased disease-free survival in grade 2–3 tumours (hazard ratio = 1.502).
      • ET omission was not associated with decreased survival in 3047 grade 1 tumours.
      • ET omission may be considered in some patients with subcentimetric oestrogen receptor-positive cancer.

      Abstract

      Introduction

      Few data have been reported regarding endocrine therapy (ET) in patients with small pT1a-b ER-postive breast cancer (BC). Thus, we conducted a study to detect possible survival improvements due to ET in such patients.

      Methods

      Our retrospective observational study included 5545 patients with pT1a-b ER-positive BC treated in 15 French centres, excluding patients with HER2-positive status, neoadjuvant chemotherapy, ER-negative status, unknown pN status or in situ BC. We estimated disease-free survival (DFS), recurrence-free survival (RFS) and overall survival (OS) via univariate analysis and multivariate Cox regression.

      Results

      Most patients (80.3%: 4453) received ET and–when compared to those without ET–experienced increases of 2.5% and 3.3% in DFS and 1.9% and 4.3% in RFS after 5 and 7 years of follow-up, respectively, with little difference in OS. In Cox regression analysis, no ET was significantly associated with decreased DFS (hazard ratio, HR = 1.275, p = 0.047, 95% CI[1.003–1.620]) but not OS or RFS in all patients, while in 2363 patients with pT1a-b ER-positive grade 2-3 BC, no ET was significantly associated with decreased DFS (HR = 1.502, p = 0.049, 95% CI[1.001–2.252]), but not OS (HR = 1.361, p = 0.272). ET omission was not significantly associated with decreased survival in 3047 patients with pT1a-b ER-positive grade 1 BC.

      Conclusion

      Our results indicate that while ET provided a beneficial impact on survival to patients with pT1a-bN0 ER-positive BC—and especially in those with grade 2–3 tumours—no such impact was observed in grade 1 tumours. Consequently, ET should be discussed with these patients, particularly in those with pT1a grade 1 tumours.

      Keywords

      1. Introduction

      Survival improvements have been well documented concerning the use of endocrine therapy (ET) in patients with oestrogen receptor (ER) positive breast cancer (BC), regardless of tumour size, patient age or type of ET [
      • Pan H.
      • Gray R.
      • Braybrooke J.
      • et al.
      20-Year risks of breast-cancer recurrence after stopping endocrine therapy at 5 years.
      ,
      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.
      ,
      Early Breast Cancer Trialists’ Collaborative Group (EBCTCG)
      Relevance of breast cancer hormone receptors and other factors to the efficacy of adjuvant tamoxifen: patient-level meta-analysis of randomised trials.
      ,
      Early Breast Cancer Trialists’ Collaborative Group (EBCTCG)
      Aromatase inhibitors versus tamoxifen in early breast cancer: patient-level meta-analysis of the randomised trials.
      ,
      • Burstein H.J.
      • Temin S.
      • Anderson H.
      • et al.
      Adjuvant endocrine therapy for women with hormone receptor–positive breast cancer: American society of clinical oncology clinical practice guideline focused update.
      ,
      • Sestak I.
      • Dowsett M.
      • Zabaglo L.
      • et al.
      Factors predicting late recurrence for estrogen receptor–positive breast cancer.
      ]. Currently, ET has an overall favourable toxicity profile, but may still induce significant side-effects that can negatively impact a patient's quality of life. These include rare but potentially life-threatening events such as endometrial carcinoma, thromboembolic (tamoxifen) and cardiovascular events, osteoporosis and increased fracture risk (aromatase inhibitors). Accordingly, ET for small tumours is indicated in various guidelines but may be discussed with the patient, particularly concerning pT1a tumours. That being said, few data have been reported regarding patients with small pT1a-b tumours, and we noted in a previous study that tumour size may not be the main prognostic factor in T1 BC [
      • Houvenaeghel G.
      • Goncalves A.
      • Classe J.M.
      • et al.
      Characteristics and clinical outcome of T1 breast cancer: a multicenter retrospective cohort study.
      ]. ET for ER-positive tumours may be avoid with patients presenting pT1a-b tumours without axillary involvement, but is rarely avoid for pN1 or even pN1mi tumours. Similarly, while the need for trastuzumab-based adjuvant chemotherapy may be disputed in pT1a-b ER-positive/HER2-positive tumours [
      • Bradley R.
      • Braybrooke J.
      • Gray R.
      • et al.
      Trastuzumab for early-stage, HER2-positive breast cancer: a meta-analysis of 13 864 women in seven randomised trials.
      ,
      • de Nonneville A.
      • Gonçalves A.
      • Zemmour C.
      • et al.
      Benefit of adjuvant chemotherapy with or without trastuzumab in pT1ab node-negative human epidermal growth factor receptor 2-positive breast carcinomas: results of a national multi-institutional study.
      ], ET is rarely omitted in small node-negative ER-positive/HER2-positive BC. With this in mind, we analysed the impact of ET in all patients with endocrine-sensitive pT1a-b BC, as well as additional subsets of patients with pN0, HER2-negative, or grade 2–3 disease, excluding those with unknown HER2 status in these subsets. We report herein the results of our multicentre retrospective study with the goal of detecting possible improvements in survival from ET for patients with pT1a-b ER-positive tumours.

      2. Methods and materials

      2.1 Patients

      We retrospectively analysed a cohort of 23,134 patients with BC and with pT1a-b ER-positive tumours across 15 centres. Our exclusion criteria included: in situ BC, neoadjuvant chemotherapy, ER-negative status (ER ≥ 10% was considered as positive), unknown pN status or HER2-positive status. Our analysis included age, tumour size (pT), Scarff Bloom Richardson (SBR) grade, tumour histology, HER2 status, lymphovascular invasion (LVI), lymph node status (pN), axillary and breast surgery, adjuvant chemotherapy, treatment period (1999–2003, 2004–2006, 2007–2009 and ≥2010) and treatment centre. All patients, except those lost to follow-up, were followed for a minimum of 6 years. Informed consent was waived since all data were de-identified and collected retrospectively from each centre.

      2.2 Statistical analysis

      Factors associated with ET were determined via binary logistic regression analysis. We performed survival analyses via univariate and multivariate Cox regression analyses of disease-free survival (DFS; defined as the length of time from the date of surgery to an event including death, local recurrence, axillary recurrence, metastasis, or contralateral BC), recurrence-free survival (RFS; date of surgery until an event including local recurrence, axillary recurrence, or metastasis) and overall survival (OS; date of surgery until death from any cause) for one or more of the following groups: (1) all patients, (2) those with pN0 or pN0(i+) status, (3) those with HER2-negative status and (4) patients with grade 2–3 tumours. We assessed the associations of BC outcomes with ET via Kaplan–Meier analyses and used Cox regression analysis to assess the association of event rate ratios with ET.
      In order to balance differences in prognostic variables associated with ET, we generated 1:1 ET and no ET matched cohorts. Coefficients of a logistic regression adjusted for age, tumour size, pN status, histology, LVI, axillary surgery, breast surgery, AC, regional node irradiation, periods and tumour grade were used to compute a propensity score. Patients with ET were then matched on this score to no ET patients using nearest-neighbour matching without replacement [
      • Rosenbaum P.R.
      • Rubin D.B.
      Constructing a control group using multivariate matched sampling methods that incorporate the propensity score.
      ,
      • Rosenbaum P.R.
      Modern algorithms for matching in observational studies.
      ]. Further details on propensity score matching approaches have been provided (Supplementary Fig. 1).
      Patient characteristics according to ET in the matched population have been reported (Supplementary Table 1). The impact of ET on DFS, MFS and OS was assessed in this matched population by log-rank tests stratified on the pairs [
      • Rosenbaum P.R.
      • Rubin D.B.
      The central role of the propensity score in observational studies for causal effects.
      ].

      3. Results

      3.1 Patient characteristics

      Of the 23,134 patients with BC treated in participating centres, 5545 patients were included in the study. Of these, 3999 patients were HER2-negative while 1546 patients had unknown HER2 status (Fig. 1).
      Fig. 1
      Fig. 1Flowchart of patient inclusion and exclusion, Abbreviations: NAC; neoadjuvant chemotherapy, DCIS, ductal carcinoma in situ.
      The clinical and pathological characteristics according to ET were recorded for 1064 pT1a tumours and 4481 pT1b tumours (Table 1). A majority of patients (80.3%: 4453) received ET, with a significantly higher rate of ET administered during the first time period compared to the last three periods combined (82.2%: 1432/1742 versus 79.4%: 3021/3,803, respectively; p < 0.02) for both pT1a (74.27%: 224/300 versus 72.9%: 557/764, respectively; nonsignificant) and pT1b (83.8%: 1208/1442 versus 81.1%: 2464/3039, respectively; p < 0.05) tumours. The administration of ET also differed significantly between treatment centres (p < 0.0001, Table 1), ranging from 41% to 100%, according to univariate analysis. Sentinel lymph node biopsies determined the axillary lymph node status of 5015 patients (90.4%) with complementary axillary lymph node dissection in 1039 patients (20.7%). The pathological axillary lymph node status was pN0 or pN0(i+) in 4782 patients (86.2%), pN1mi in 350 patients (6.3%) and pN1macro in 413 patients (7.4%).
      Table 1Patient characteristics according to endocrine therapy.
      Endocrine therapyYesNoχ2
      Nb%Nb%P
      Age≤401523.4403.70.004
      40.1–5080118.020618.9
      50.1–74.9324272.875068.9
      ≥752575.8938.5
      pTpT1a78117.528325.9<0.0001
      pT1b367282.580974.1
      Grade1227051.077771.2<0.0001
      2188642.425523.4
      32044.6181.6
      Unknown932.0423.9
      HistologyDuctal340876.589782.1<0.0001
      Lobular53712.112011.0
      Mixt641.420.2
      Others44410.0736.7
      pN statuspN0362381.4102293.6<0.0001
      pN (i+)1222.7151.4
      pN1mi3297.4211.9
      pN1 macro3798.5343.1
      HER2 statusHER2 <0325173.074868.50.002
      Unknown120227.034431.5
      LVINo342576.988080.60.004
      Yes3086.9484.4
      Unknown72016.216415.0
      Axillary surgerySLNB314570.783176.2<0.0001
      SLNB + ALND95221.4878.0
      ALND3548.017315.9
      Breast surgeryConservative400690.089882.2<0.0001
      Mastectomy3918.8878.0
      Unknown561.31079.8
      ChemotherapyYes61013.7393.6<0.0001
      No384386.3105396.4
      RTNo21053.75462.10.097
      MastectomyYes18146.33337.9
      RNINo297781.091896.9<0.0001
      Yes69719.013813.1
      Time periods1999–2003143233.231028.4<0.0001
      2004–2006139031.239035.7
      2007–200999822.420118.4
      ≥201063314.219117.5
      Centre<0.0001
      Abbreviations: LVI, lymphovascular invasion; SLNB, sentinel lymph node biopsy, ALND, axillary lymph node dissection; RT, radiation therapy; RNI, regional node irradiation.
      Patients with ET had proportionally fewer recurrences than those without (4.8% versus 7.4%, respectively; p = 0.001), though when analysed according to type, only local recurrence showed a significant difference (1.5% versus 4.8%, respectively; p < 0.0001, Supplementary Table 2).
      Patients who received adjuvant chemotherapy had clinical and biological characteristics that highlight a higher risk of recurrence (Supplementary Tables 3–4 for univariate and multivariate analysis) and patients without ET received lesser adjuvant chemotherapy (OR = 0.315, p < 0.0001). In matched population, there was no significant difference of adjuvant chemotherapy rate. Patients with axillary lymph node dissection had clinical and biological characteristics that highlight a higher risk of recurrence (Supplementary Tables 5–6 for univariate and multivariate analysis).

      3.2 Survival analysis in all patients

      Binary logistic regression analysis identified factors significantly associated with ET, which included age, grade, tumour histology, pN status, negative or unknown HER2 status and time period (Table 2).
      Table 2Factors associated with endocrine therapy in binary logistic regression.
      pOdds ratioCI 95%
      InferiorSuperior
      Age≤401
      40.1–500.1941.3080.8721.962
      50.1–74.90.0151.6091.0972.361
      ≥750.8231.0530.6691.658
      Tumour Size>5 versus ≤ 5 mm<0.00011.4681.2371.742
      Grade11
      2<0.00012.7262.3113.215
      3<0.00014.3292.6037.199
      No grade0.0014.9451.92212.72
      HistologyDuctal1
      Lobular0.8571.0210.8141.280
      Mixt0.0048.2391.93435.10
      Others<0.00012.3451.7813.088
      pN statuspN01
      pN (i+)0.0152.0191.1463.556
      pN1mi<0.00014.9773.1217.936
      pN1 macro<0.00012.8891.9884.199
      HER2Unknown versus < 0<0.00010.6130.5130.733
      LVINo1
      Yes0.7121.0670.7571.505
      Unknown0.0991.1890.9681.460
      SurgeryConservative1
      Mastectomy0.5940.9310.7151.211
      Periods1999–20031
      2004–20060.5140.9390.7791.133
      2007–20090.0970.8270.6601.035
      ≥2010<0.00010.5440.4280.693
      Abbreviations: LVI, lymphovascular invasion; SLNB, sentinel lymph node biopsy, ALND, axillary lymph node dissection; RT, radiation therapy; RNI, regional node irradiation.
      The median follow-up for all patients was 58.62 months (mean: 61.19, 95% confidence interval (CI) [60.1–62.3]). There were 194 events within the first 5 years of follow-up, with 2647 patients at risk of an event (47.7%) at 5 years. According to Kaplan–Meier analysis, the 5-year DFS rates in patients who either had or had not received ET were 95.7% (SD: 0.4) versus 93.2% (SD: 0.9), while 7-year DFS was 91.2% (SD: 0.6) and 87.9% (SD: 1.5), respectively (Log Rank: 0.052, Fig. 2).
      Fig. 2
      Fig. 2Disease-free survival according to endocrine therapy (univariate analysis).
      The RFS rates for those with or without ET at 5 and 7 years were 97.7% (SD: 0.3) versus 95.8% (SD: 0.7) and 95.5% (SD: 0.5) versus 91.2% (SD: 1.3), respectively (Log Rank: 0.016, Fig. 3).
      Fig. 3
      Fig. 3Recurrence-free survival according to endocrine therapy (univariate analysis).
      Overall survival rates in patients who had or had not received ET at 5 and 7 years were 97.7% (SD: 0.3) versus 97.2% (SD: 0.6) and 96.0% (SD: 0.4) versus 96.5% (SD: 0.7), respectively (Log Rank: 0.409, Supplementary Fig. 2).
      According to univariate analysis, factors significantly associated with decreased DFS included no ET, grade, LVI, HER2 status unknown, mastectomy and age; while OS and RFS included grade, LVI, and age and no ET, grade, LVI, mastectomy, age and pT1b, respectively (Supplementary Table 7).
      In Cox regression analysis stratified according to these factors and pN status, no ET was significantly associated with a decrease in DFS (hazard ratio, HR: 1.275, p = 0.047, 95% CI [1.003–1.620]; Table 3, Fig. 4) but not OS (HR: 1.179, p = 0.349, 95% CI [0.835–1.663]; Table 4) or RFS (HR: 1.306, p = 0.093, 95% CI [0.957–1.782]; Table 5) in all patients. Other significant factors for reduced DFS included LVI, HER2 status unknown, age ≥75 years old, and pN1 macrometastases; while significant factors for OS and RFS included grade 3, age ≥75 years old, and pN1 macrometastases; and LVI, age of 50.1–74.9 years and pT1b tumours, respectively.
      Table 3Multivariate Cox regression analysis of disease-free survival in all patients and subgroup of patients with grade 2–3 tumours.
      All patientsGrade 2-3
      DFSpHazard ratioCI 95%pHazard ratioCI 95%
      InferiorSuperiorInferiorSuperior
      EndocrineYes11
      TherapyNo0.0471.2751.0031.6200.0491.5021.0012.252
      Grade11
      20.7431.0370.8341.2911
      30.2111.3270.8522.0680.2111.3310.8502.085
      LVINo11
      Yes0.0021.6681.2072.3050.0451.5511.0092.382
      Age≤4011
      40.1–500.2780.7860.5081.2140.2460.7180.4111.256
      50.1–74.90.2980.8120.5481.2020.0800.6410.3891.055
      ≥75<0.00012.5181.5754.0250.0521.8900.9953.589
      pN statuspN011
      pN (i+)0.1701.4830.8442.6060.0232.1501.1114.161
      pN1mi0.8170.9530.6321.4360.8251.0700.5901.940
      pN1 macro0.0521.3460.9971.8180.0941.4240.9422.152
      HER2Unknown versus < 0<0.00010.6440.5140.805<0.00010.5170.3590.745
      SurgeryConservative11
      Mastectomy0.9471.0100.7461.3680.9250.9800.6401.499
      Abbreviations: LVI, lymphovascular invasion.
      Fig. 4
      Fig. 4Disease-free survival according to endocrine therapy for all patients in multivariate analysis.
      Table 4Multivariate Cox regression analysis of overall survival in all patients and subgroup of patients with grade 2–3 tumours.
      All patientsGrade 2-3
      OSpHazard ratioCI 95%pHazard ratioCI 95%
      InferiorSuperiorInferiorSuperior
      EndocrineYes11
      TherapyNo0.3491.1790.8351.6630.2721.3610.7852.360
      Grade11
      20.8370.9660.6971.3391
      30.0112.0091.1743.4400.0082.1221.2173.699
      LVINo11
      Yes0.2411.3590.8132.2720.3751.3630.6882.704
      Age≤4011
      40.1–500.6491.2240.5132.9210.8771.0890.3663.239
      50.1–74.90.0622.1200.9644.6620.1532.0160.7705.277
      ≥75<0.000110.424.49524.14<0.00017.9892.74623.25
      pN statuspN011
      pN (i+)0.9851.0100.3702.7570.2741.7810.6335.014
      pN1mi0.8311.0680.5851.9500.3880.5950.1831.936
      pN1 macro0.0161.6861.1042.5760.0591.7180.9793.015
      Abbreviations: LVI, lymphovascular invasion.
      Table 5Multivariate Cox regression analysis of recurrence-free survival in all patients and subgroup of patients with grade 2–3 tumours.
      All patientsGrade 2-3
      RFSpHazard ratioCI 95%PHazard ratioCI 95%
      InferiorSuperiorInferiorSuperior
      EndocrineYes11
      TherapyNo0.0931.3060.9571.7820.2511.3300.8172.164
      Grade11
      20.0761.3020.9731.7421
      30.0881.6170.9312.8100.3881.2760.7342.220
      LVINo11
      Yes0.0061.7601.1762.6350.1121.5410.9052.626
      Age≤4011
      40.1–500.3040.7690.4651.2690.8220.9290.4891.764
      50.1–74.90.0120.5470.3430.8730.0570.5570.3051.018
      ≥750.5050.7870.3891.5920.4830.6890.2441.950
      pN statuspN011
      pN (i+)0.1271.6690.8653.2180.0902.0040.8964.481
      pN1mi0.9791.0070.5861.7310.6961.1540.5642.360
      pN1 macro0.2841.2450.8341.8590.6001.1510.6801.950
      SurgeryConservative11
      Mastectomy0.6721.0880.7371.6060.6741.1180.6661.876
      pT sizepT1a11
      pT1b0.0020.6240.4600.8470.0050.5270.3380.822
      Abbreviations: LVI, lymphovascular invasion.
      There were no significant associations between no ET and decreased DFS (HR: 1.200, p = 0.267, 95% CI [0.870–1.655]), RFS (HR: 1.010, p = 0.964, 95% CI [0.650–1.569]) or OS (HR: 1.386, p = 0.157, 95% CI [0.882–2.176]) in 3999 patients with HER2-negative status (patients with unknown HER2 status excluded; Table 6).
      Table 6Multivariate Cox regression analysis of OS, DFS and RFS in patients with HER2-negative tumours.
      OSDFSRFS
      HER2 negative patientspHazard ratioCI 95%pHazard ratioCI 95%pHazard RatioCI 95%
      InfSupInfSupInfSup
      EndocrineYes111
      TherapyNo0.1571.3860.8822.1760.2671.2000.8701.6550.9641.0100.6501.569
      Grade1111
      20.9281.0190.6791.5290.4751.1050.8401.4550.2321.2520.8661.810
      30.0042.7821.3945.5530.0301.8831.0653.3290.0312.2141.0774.554
      LVINo111
      Yes0.3181.3940.7262.6780.0081.7561.1602.6580.0201.8441.1033.083
      Age≤40111
      40.1–500.2443.3570.43725.770.9520.9790.4991.9220.5121.3140.5812.971
      50.1–74.90.0955.3940.74838.920.9260.9710.5191.8150.9260.9640.4442.093
      ≥750.00223.863.207177.50.0032.9051.4495.8270.4441.4660.5503.905
      pN statuspN0111
      pN (i+)0.4491.4870.5334.1480.0112.1681.1943.9390.0202.3471.1424.824
      pN1mi0.8190.8990.3602.2410.3630.7510.4051.3920.5950.8100.3721.763
      pN1 macro0.1071.6210.9012.9160.8411.0460.6751.6210.9280.9740.5551.712
      SurgeryConservative11
      Mastectomy0.5350.8750.5731.3360.6651.1180.6761.848
      pT sizepT1a1
      pT1b0.2600.7990.5401.181
      Abbreviations: LVI, lymphovascular invasion.
      Based on the ET propensity score, 1318 patients with ET were matched to 761 patients without ET. The survival analysis of these cohorts revealed no prognostic impact on OS (Supplementary Fig. 3A), DFS (Supplementary Fig. 3B) or RFS (Supplementary Fig. 3C).

      3.3 Survival analysis in specific subsets of patients according to tumour grade, size and nodal status

      No ET was significantly associated with decreased DFS (HR: 1.502, p = 0.049, 95% CI [1.001–2.252]; Table 3, Fig. 5), but not OS (HR: 1.361, p = 0.272, 95% CI [0.785–2.360]; Table 4) or RFS (HR: 1.330, p = 0.251, 95% CI [0.817–2.164]; Table 5) in multivariate analysis of 2363 patients with pT1a-b ER-positive grade 2–3 tumours.
      Fig. 5
      Fig. 5Disease-free survival according to endocrine therapy for patients with grade 2–3 tumours in multivariate analysis.
      Whereas in 3047 patients with pT1a-b ER-positive grade 1 tumours, there were no significant associations between ET omission and decreased DFS, OS or RFS. These results were maintained upon stratifying the propensity score-matched cohort according to grade with decreased DFS (HR: 1.981, p = 0.034, 95% CI [1.044–3.758]; Supplementary Fig. 3F), but not OS (HR: 2.201, p = 0.102, 95% CI [0.856–5.660]; Supplementary Fig. 3E) or RFS (HR: 1.648, p = 0.180, 95% CI [0.793–3.423]; Supplementary Fig. 3G) in the 495 patients with pT1a-b ER-positive grade 2–3 tumours, whereas in 1584 patients with pT1a-b ER-positive grade 1 tumours, there were no significant associations between ET omission and decreased OS, DFS, or RFS (Supplementary Fig. 3H, I, and J).
      In 4782 pN0 or pN0(i+) patients (Table 7), there was no significant association between ET omission and DFS (HR: 1.188, p = 0.348, 95% CI [0.829–1.703]) or OS (HR: 1.210, p = 0.318, 95% CI [0.833–1.758]) or RFS (HR: 1.363, p = 0.082, 95% CI [0.961–1.933]). This was also the case in 413 patients with pN1macro where only 34 patients had not received ET.
      Table 7Overall survival, disease-free survival and recurrence-free survival multivariate Cox analysis in patients with pN0 or pN0(i+) axillary lymph nodes.
      pN0 & pN0(i+)OSDFSRFS
      pHazard ratioCI 95%pHazard ratioCI 95%pHazard ratioCI 95%
      InfSupInfSupInferiorSuperior
      EndocrineYes111
      TherapyNo0.3181.2100.8331.7580.3481.1880.8291.7030.0821.3630.9611.933
      Grade1111
      20.7251.0670.7421.5350.7211.0620.7631.4780.0661.3600.9791.888
      30.0512.0890.9974.3800.0252.1561.1044.2130.0671.9400.9563.938
      LVIYes versus No0.8711.0620.5122.2020.5841.1910.6372.2270.0521.6630.9962.777
      Age≤40111
      40.1–500.5491.5910.3487.2810.8010.8770.3172.4280.5360.8170.4321.547
      50.1–74.90.0573.9330.95816.150.2671.6750.6744.1650.1090.6170.3421.114
      ≥75<0.000117.864.21275.72<0.00016.7082.58717.390.4020.6980.3011.618
      pN statuspN (i+) versus pN00.9131.0580.3862.8950.8591.0850.4412.6700.1131.7100.8803.320
      Her2Unknown versus < 00.0010.5360.3760.7650.0430.7060.5040.989
      SurgeryMastectomy versus BCS0.7660.9230.5421.5690.8111.0590.6601.701
      pT1b vs pT1a0.0130.6490.4610.913
      Abbreviations: LVI, lymphovascular invasion; BCS, breast-conserving surgery.
      Comparison of pT1a versus pT1b tumours did not reveal any significant difference for either DFS (HR: 0.846, p = 0.180, 95% CI [0.663–1.080]) or OS (HR: 1.006, p = 0.976, 95% CI [0.695–1.455]).

      4. Discussion

      Among patients with ER-positive pT1a-b BC, those who had received ET experienced a 2.5% and 3.3% increase in DFS and a 1.9% and 4.3% increase in RFS after 5 and 7 years of follow-up, respectively, with little to no difference in OS, compared to patients that had not received ET. This benefit in DFS was confirmed via Cox regression analysis stratified according to ET administration and other significant factors, such as LVI, unknown HER2 status, age ≥75 years old and pN1 macrometastases. Furthermore, no ET was significantly associated with reduced DFS in both all patients and a subset of patients with grade 2–3 tumours. However, in a subset of patients with grade 1 tumours, there was no prognostic impact observed on DFS, OS or RFS. These findings were confirmed in both subsets of patients in propensity scoring stratified according to grade.
      While most guidelines recommend adjuvant ET in almost all patients with invasive ER-positive BC [
      • Burstein H.J.
      • Curigliano G.
      • Thürlimann B.
      • et al.
      Customizing local and systemic therapies for women with early breast cancer: the St. Gallen International Consensus Guidelines for treatment of early breast cancer 2021.
      ], the prescribing of ET in subcentimeter tumours lacks a strong consensus as reflected by our results. A significantly higher rate of ET was administered to patients with pT1b tumours compared to pT1a tumours (81.9% versus 73.4%, respectively), and to those with pN1 disease (94.0% for pN1mi and 91.8% for pN1macro) compared to pN0 disease (78.0% for pN0 and 89.0% for pN0(i+)), even though we found no difference in DFS and OS between pT1a and pT1b tumours in multivariate analysis. This lack of consensus was further illustrated by a significant difference in ET administration according to treatment centre. Interestingly, ET was also administered less frequently after the year 2010 compared to the initial period of 1999–2003 (OR: 0.544, 76.8% versus 82.2%, respectively, p = 0.002), which may be due to recent efforts aimed at deescalating therapeutic intensity in patients with early BC since ET may significantly impact the quality of life for a relatively small–though still significant–benefit, as indicated by our results in both our current and previous studies [
      • Houvenaeghel G.
      • Lambaudie E.
      • Cohen M.
      • et al.
      Therapeutic escalation – de-escalation: data from 15.508 early breast cancer treated with upfront surgery and sentinel lymph node biopsy (SLNB).
      ].
      Our primary goal was to identify potential improvements due to ET in patients with ER-positive pT1a-b BC in order to expand current knowledge and aide clinicians in the decision-making for ET administration in these tumour types. We found that while there was significant improvement in both DFS and RFS and recurrences were less common in patients with ET, the absolute benefits were small, and OS was unaffected. Still, the results of our study were consistent with those of others. In the EBCTCG study of patients with early-stage ER-positive BC and who received ET for 5 years, the risk of distant recurrence in T1N0 BC at 10 and 15 years was 3% and 7%, respectively, for 5527 pT1a-bN0 tumours and 4% and 9%, respectively, for 13,875 pT1cN0 tumours (HR: 0.67, p < 0.001). The risk of any breast cancer event was 6% and 13% for pT1a-bN0 tumours versus 7% and 14% for pT1cN0 tumours at 10 and 15 years, respectively (HR: 0.85, p = 0.01).(2) Our DFS rates of 95.7% and 91.8% at 5 and 7 years, respectively, closely resembled these data. It is also clear that in both studies, patients with pT1a-b ER-positive grade 1 disease do not experience a significant benefit from ET, and as such, ET could potentially be omitted in these patients and should be discussed in accordance with the patient's co-morbidities and/or tolerance of ET (if such treatment had been started).
      Another relevant setting in which omitting ET could be considered appropriate concerns patients with pN0 status. In our study, the multivariate analysis of patients with pN0-pN0(i+) disease did not reveal any significant association between ET and OS or DFS. In fact, in terms of pN status, only pN1macro disease had any significant association with survival. A similar result was found in the EBCTCG study, where patients with T1N1, T1N2, T2N1 and T2N2 BC, all saw a minimum risk of recurrence of 20%. However, recurrences after five years of follow-up are not rare, and the authors themselves concluded that after 5-years of treatment ‘even low-grade T1N0 disease carries an appreciable risk of distant recurrence and contralateral breast cancer, a risk that is sufficient for at least the consideration of extended ET’[
      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.
      ].
      Concerning intraoperative radiotherapy itself increased local recurrence has been observed in the absence of hormone therapy [
      • Houvenaeghel G.
      • de Nonneville A.
      • Cohen M.
      • et al.
      Isolated ipsilateral local recurrence of breast cancer: predictive factors and prognostic impact.
      ,
      • Tallet A.
      • Racadot S.
      • Boher J.M.
      • et al.
      The actual benefit of intraoperative radiation therapy using 50 kV x-rays in early breast cancer: a retrospective study of 676 patients.
      ]. In the TARGIT-A trial, the omission of whole-breast radiotherapy compared to intraoperative radiotherapy corresponded with an absolute increase in the local recurrence rate of approximately 7% at 5 years and 8% at 10 years, without any detectable increase in breast cancer mortality [
      • Vaidya J.S.
      • Wenz F.
      • Bulsara M.
      • et al.
      Risk-adapted targeted intraoperative radiotherapy versus whole-breast radiotherapy for breast cancer: 5-year results for local control and overall survival from the TARGIT-A randomised trial.
      ]. Consequently, ET should be recommended for patients who have received only partial-breast irradiation, such as intraoperative radiotherapy, and the omission of ET should be likewise discouraged in patients benefiting from strategies of therapeutic de-escalation which have been validated in the context of ET.
      Our study has several limitations. Concerning ER status, tumours were considered as ER-positive only when 10% or more of cells stained positive. However, some patients have between 1% and 10% of cells staining positive [
      • Harvey J.M.
      • Clark G.M.
      • Osborne C.K.
      • Allred D.C.
      Estrogen receptor status by immunohistochemistry is superior to the ligand-binding assay for predicting response to adjuvant endocrine therapy in breast cancer.
      ,
      • Khoshnoud M.R.
      • Löfdahl B.
      • Fohlin H.
      • et al.
      Immunohistochemistry compared to cytosol assays for determination of estrogen receptor and prediction of the long-term effect of adjuvant tamoxifen.
      ], which may be considered as ER-positive according to US guidelines and thus allow for adjuvant endocrine treatment, but were not included in our study. Yet accumulating evidence demonstrates that their prognosis is quite similar to patients with ER-negative BCand with limited benefit from adjuvant ET [
      • Fujii T.
      • Kogawa T.
      • Dong W.
      • et al.
      Revisiting the definition of estrogen receptor positivity in HER2-negative primary breast cancer.
      ,
      • Villegas S.L.
      • Nekljudova V.
      • Pfarr N.
      • et al.
      Therapy response and prognosis of patients with early breast cancer with low positivity for hormone receptors – an analysis of 2765 patients from neoadjuvant clinical trials.
      ,
      • Dieci M.V.
      • Griguolo G.
      • Bottosso M.
      • et al.
      Impact of estrogen receptor levels on outcome in non-metastatic triple negative breast cancer patients treated with neoadjuvant/adjuvant chemotherapy.
      ,
      • Paakkola N.M.
      • Karakatsanis A.
      • Mauri D.
      • Foukakis T.
      • Valachis A.
      The prognostic and predictive impact of low estrogen receptor expression in early breast cancer: a systematic review and meta-analysis.
      ]. In addition, due to the length of our inclusion period, 1546 patients lacked analysis of HER2 status. The follow-up in our study was also relatively short and while only a small benefit was seen, the absolute difference between patients with or without ET could actually be higher after 10 or 15 years of follow-up. Furthermore, our study lacked information concerning compliance with ET, as only a limited amount of data were detailed enough to include adherence. In addition, due to the retrospective design of the study, we lacked stratification according to the menopausal status and the type of ET. The survival rates of patients who had not fully completed the therapy would therefore be somewhat lower than those in our study. On the other hand, one potential benefit of a shorter follow-up is that only a few of the patients received a BC diagnosis before the year 2000, which somewhat limits differences in treatments due to technological advances and changes to existing guidelines. Our study also benefitted from a high percentage (90.4%) of patients that underwent Sentinel lymph node biopsies and thus had a more extensive analysis of axillary lymph nodes than in older studies where axillary lymph node status was instead determined by axillary dissection. With over 5000 patients included in the study across 15 different centres, our results also carry greater statistical power than those of smaller studies, though–as noted previously–the administration of ET differed significantly between centres, and this could have affected the results.

      5. Conclusions

      In our study, adjuvant ET was significantly associated with a survival benefit in all patients with pT1a-bN0 ER-positive BC and particularly in those with grade 2–3 tumours, even with a relatively short follow-up for these patients who can recur even after 20–25 years and little data concerning compliance. However, the lack of consensus on ET administration resulted in largely different rates of ET for pT1a-b ER-positive BC across the various centres, and additional studies are needed to determine potential effects on survival in these patients. In addition, there was no beneficial impact of ET on survival in patients with pT1a-b ER-positive grade 1 BC, and as such, ET should be discussed–particularly for pT1a tumours–according to these patient's co-morbidities and/or tolerance (if ET had already been started).

      Author statement

      Gilles Houvenaeghel: Conceptualisation, Data curation; Formal analysis; Investigation; Methodology; Project administration; Supervision; Validation; Visualisation; Roles/Writing - original draft; Writing - review & editing.
      Alexandre De Nonneville: Conceptualisation, Data curation; Formal analysis; Investigation; Methodology; Project administration; Supervision; Validation; Visualisation; Roles/Writing - original draft; Writing - review & editing.
      Anthony Gonçalves: Conceptualisation, Data curation; Formal analysis; Investigation; Methodology; Project administration; Supervision; Validation; Visualization; Roles/Writing - original draft; Writing - review & editing.
      Monique Cohen, Jean-Marc Classe, Fabien Reyal, Chafika Mazouni, Christelle Faure, Alejandra Martinez, Marie-Pierre Chauvet, Emile Daraï, Charles Coutant, Pierre-Emmanuel Colombo, Pierre Gimbergues, Anne-Sophie Azuar, Roman Rouzier, Christine Tunon De Lara, Patrice Crochet, Sandrine Rua: Data curation.
      Funding acquisition: Not applicable
      Software: SPSS

      Funding

      This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

      Conflict of interest statement

      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      Acknowledgements

      The authors would like to thank JRS Translations for their assistance in writing the article. Funding for writing assistance was provided by the Institut Paoli-Calmettes.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Pan H.
        • Gray R.
        • Braybrooke J.
        • et al.
        20-Year risks of breast-cancer recurrence after stopping endocrine therapy at 5 years.
        N Engl J Med. 2017; 377: 1836-1846https://doi.org/10.1056/NEJMoa1701830
        • 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: 1687-1717https://doi.org/10.1016/S0140-6736(05)66544-0
        • Early Breast Cancer Trialists’ Collaborative Group (EBCTCG)
        Relevance of breast cancer hormone receptors and other factors to the efficacy of adjuvant tamoxifen: patient-level meta-analysis of randomised trials.
        Lancet. 2011; 378: 771-784https://doi.org/10.1016/S0140-6736(11)60993-8
        • Early Breast Cancer Trialists’ Collaborative Group (EBCTCG)
        Aromatase inhibitors versus tamoxifen in early breast cancer: patient-level meta-analysis of the randomised trials.
        Lancet. 2015; 386: 1341-1352https://doi.org/10.1016/S0140-6736(15)61074-1
        • Burstein H.J.
        • Temin S.
        • Anderson H.
        • et al.
        Adjuvant endocrine therapy for women with hormone receptor–positive breast cancer: American society of clinical oncology clinical practice guideline focused update.
        JCO. 2014; 32: 2255-2269https://doi.org/10.1200/JCO.2013.54.2258
        • Sestak I.
        • Dowsett M.
        • Zabaglo L.
        • et al.
        Factors predicting late recurrence for estrogen receptor–positive breast cancer.
        J Natl Cancer Inst. 2013; 105: 1504-1511https://doi.org/10.1093/jnci/djt244
        • Houvenaeghel G.
        • Goncalves A.
        • Classe J.M.
        • et al.
        Characteristics and clinical outcome of T1 breast cancer: a multicenter retrospective cohort study.
        Ann Oncol. 2014; 25: 623-628https://doi.org/10.1093/annonc/mdt532
        • Bradley R.
        • Braybrooke J.
        • Gray R.
        • et al.
        Trastuzumab for early-stage, HER2-positive breast cancer: a meta-analysis of 13 864 women in seven randomised trials.
        Lancet Oncol. 2021; 22: 1139-1150https://doi.org/10.1016/S1470-2045(21)00288-6
        • de Nonneville A.
        • Gonçalves A.
        • Zemmour C.
        • et al.
        Benefit of adjuvant chemotherapy with or without trastuzumab in pT1ab node-negative human epidermal growth factor receptor 2-positive breast carcinomas: results of a national multi-institutional study.
        Breast Cancer Res Treat. 2017; 162: 307-316https://doi.org/10.1007/s10549-017-4136-5
        • Rosenbaum P.R.
        • Rubin D.B.
        Constructing a control group using multivariate matched sampling methods that incorporate the propensity score.
        Am Statistician. 1985; 39: 33-38https://doi.org/10.1080/00031305.1985.10479383
        • Rosenbaum P.R.
        Modern algorithms for matching in observational studies.
        Annu Rev Stat App. 2020; 7: 143-176https://doi.org/10.1146/annurev-statistics-031219-041058
        • Rosenbaum P.R.
        • Rubin D.B.
        The central role of the propensity score in observational studies for causal effects.
        Biometrika. 1983; 70: 41-55https://doi.org/10.2307/2335942
        • Burstein H.J.
        • Curigliano G.
        • Thürlimann B.
        • et al.
        Customizing local and systemic therapies for women with early breast cancer: the St. Gallen International Consensus Guidelines for treatment of early breast cancer 2021.
        Ann Oncol. 2021; 32: 1216-1235https://doi.org/10.1016/j.annonc.2021.06.023
        • Houvenaeghel G.
        • Lambaudie E.
        • Cohen M.
        • et al.
        Therapeutic escalation – de-escalation: data from 15.508 early breast cancer treated with upfront surgery and sentinel lymph node biopsy (SLNB).
        Breast. 2017; 34: 24-33https://doi.org/10.1016/j.breast.2017.04.008
        • Houvenaeghel G.
        • de Nonneville A.
        • Cohen M.
        • et al.
        Isolated ipsilateral local recurrence of breast cancer: predictive factors and prognostic impact.
        Breast Cancer Res Treat. 2019; 173: 111-122https://doi.org/10.1007/s10549-018-4944-2
        • Tallet A.
        • Racadot S.
        • Boher J.M.
        • et al.
        The actual benefit of intraoperative radiation therapy using 50 kV x-rays in early breast cancer: a retrospective study of 676 patients.
        Breast J. 2020; 26: 2145-2150https://doi.org/10.1111/tbj.13827
        • Vaidya J.S.
        • Wenz F.
        • Bulsara M.
        • et al.
        Risk-adapted targeted intraoperative radiotherapy versus whole-breast radiotherapy for breast cancer: 5-year results for local control and overall survival from the TARGIT-A randomised trial.
        Lancet. 2014; 383: 603-613https://doi.org/10.1016/S0140-6736(13)61950-9
        • Harvey J.M.
        • Clark G.M.
        • Osborne C.K.
        • Allred D.C.
        Estrogen receptor status by immunohistochemistry is superior to the ligand-binding assay for predicting response to adjuvant endocrine therapy in breast cancer.
        JCO. 1999; 17 (1474-1474)https://doi.org/10.1200/JCO.1999.17.5.1474
        • Khoshnoud M.R.
        • Löfdahl B.
        • Fohlin H.
        • et al.
        Immunohistochemistry compared to cytosol assays for determination of estrogen receptor and prediction of the long-term effect of adjuvant tamoxifen.
        Breast Cancer Res Treat. 2011; 126: 421-430https://doi.org/10.1007/s10549-010-1202-7
        • Fujii T.
        • Kogawa T.
        • Dong W.
        • et al.
        Revisiting the definition of estrogen receptor positivity in HER2-negative primary breast cancer.
        Ann Oncol. 2017; 28: 2420-2428https://doi.org/10.1093/annonc/mdx397
        • Villegas S.L.
        • Nekljudova V.
        • Pfarr N.
        • et al.
        Therapy response and prognosis of patients with early breast cancer with low positivity for hormone receptors – an analysis of 2765 patients from neoadjuvant clinical trials.
        Eur J Cancer. 2021; 148: 159-170https://doi.org/10.1016/j.ejca.2021.02.020
        • Dieci M.V.
        • Griguolo G.
        • Bottosso M.
        • et al.
        Impact of estrogen receptor levels on outcome in non-metastatic triple negative breast cancer patients treated with neoadjuvant/adjuvant chemotherapy.
        npj Breast Cancer. 2021; 7: 1-7https://doi.org/10.1038/s41523-021-00308-7
        • Paakkola N.M.
        • Karakatsanis A.
        • Mauri D.
        • Foukakis T.
        • Valachis A.
        The prognostic and predictive impact of low estrogen receptor expression in early breast cancer: a systematic review and meta-analysis.
        ESMO Open. 2021; 6https://doi.org/10.1016/j.esmoop.2021.100289