Key PointsQuestion听
Can high nodal burden (鈮�4 axillary metastases/鈮2) be predicted without completion axillary lymph node dissection (CALND) in patients with luminal ERBB2-negative tumors and, separately, in those with invasive lobular carcinoma, and 1 or 2 sentinel lymph node (SLN) macrometastases?
Findings听
In this diagnostic/prognostic study including the CALND arm (n鈥�=鈥�1010) of the randomized SENOMAC trial, 2 prediction models for the identification of patients with 鈮2 were developed, including number of SLN macrometastases, additional SLN micrometastases, SLN ratio, extracapsular extension, and tumor size (only luminal model).
Meaning听
The prediction models may be used to identify patients at risk of high nodal burden eligible for intensified systemic treatment strategies without performing CALND.
Importance听
In patients with clinically node-negative (cN0) breast cancer and 1 or 2 sentinel lymph node (SLN) macrometastases, omitting completion axillary lymph node dissection (CALND) is standard. High nodal burden (鈮�4 axillary nodal metastases) is an indication for intensified treatment in luminal breast cancer; hence, abstaining from CALND may result in undertreatment.
Objective听
To develop a prediction model for high nodal burden in luminal ERBB2-negative breast cancer (all histologic types and lobular breast cancer separately) without CALND.
Design, Setting, and Participants听
The prospective Sentinel Node Biopsy in Breast Cancer: Omission of Axillary Clearance After Macrometastases (SENOMAC) trial randomized patients 1:1 to CALND or its omission from January 2015 to December 2021 among adult patients with cN0 T1-T3 breast cancer and 1 or 2 SLN macrometastases across 5 European countries. The cohort was randomly split into training (80%) and test (20%) sets, with equal proportions of high nodal burden. Prediction models were developed by multivariable logistic regression in the complete luminal ERBB2-negative cohort and a lobular breast cancer subgroup. Nomograms were constructed. The present diagnostic/prognostic study presents the results of a prespecified secondary analysis of the SENOMAC trial. Herein, only patients with luminal ERBB2-negative tumors assigned to CALND were selected. Data analysis for this article took place from June 2023 to April 2024.
Exposure听
Predictors of high nodal burden.
Main Outcomes and Measures听
High nodal burden was defined as 鈮�4 axillary nodal metastases. The luminal prediction model was evaluated regarding discrimination and calibration.
Results听
Of 1010 patients (median [range] age, 61 [34-90] years; 1006 [99.6%] female and 4 [0.4%] male), 138 (13.7%) had a high nodal burden and 212 (21.0%) had lobular breast cancer. The model in the training set (n鈥�=鈥�804) included number of SLN macrometastases, presence of SLN micrometastases, SLN ratio, presence of SLN extracapsular extension, and tumor size (not included in lobular subgroup). Upon validation in the test set (n鈥�=鈥�201), the area under the receiver operating characteristic curve (AUC) was 0.74 (95% CI, 0.62-0.85) and the calibration was satisfactory. At a sensitivity threshold of 鈮�80%, all but 5 low-risk patients were correctly classified corresponding to a negative predictive value of 94%. The prediction model for the lobular subgroup reached an AUC of 0.74 (95% CI, 0.66-0.83).
Conclusions and Relevance听
The predictive models and nomograms may facilitate systemic treatment decisions without exposing patients to the risk of arm morbidity due to CALND. External validation is needed.
Trial Registration听
ClinicalTrials.gov Identifier:
Sentinel lymph node (SLN) biopsy is the standard procedure for axillary staging in patients with clinically node-negative (cN0) breast cancer. While completion axillary lymph node dissection (CALND) was previously routine care in patients with positive SLNs, it is today safely omitted in patients with 1 or 2 SLN metastases.1-3 However, some systemic treatment recommendations in estrogen receptor (ER)鈥損ositive Erb-B2 receptor tyrosine kinase 2 (ERBB2)鈥搉egative breast cancer rely on the identification of high nodal burden (ie, 鈮�4 metastatic axillary lymph nodes [ALNs]), which can only be determined by CALND.4 The recent Abemaciclib Plus Endocrine Therapy for Hormone Receptor鈥揚ositive, HER2-Negative, Node-Positive, High-Risk Early Breast Cancer (MONARCHE) trial5,6 showed that patients with high-risk early ER-positive breast cancer, defined as 4 or more metastatic ALNs or 1 to 3 metastatic ALNs together with either tumor size of 5 cm or greater or histologic grade 3, had improved invasive disease-free survival after 2 years of adjuvant CDK 4/6 inhibitor treatment in addition to endocrine therapy compared to the control group. In the Microarray in Node-Negative and 1 to 3 Positive Lymph Node Disease May Avoid Chemotherapy (MINDACT) trial,7 axillary nodal burden was indicative of systemic treatment. In addition, screening for distant metastases and locoregional radiotherapy target extension are considered in breast cancer with high nodal burden.8 High nodal burden can thus alter adjuvant treatment decisions. Therefore, CALND omission may lead to undertreatment.
Negative outcomes of CALND, such as arm lymphedema (affecting up to 25% of patients), numbness, and restricted shoulder mobility, are well documented.3 Arm morbidity is considerably worse after CALND than after SLN biopsy only.9,10 Therefore, alternative strategies to correctly identify high nodal burden without exposing patients to the risk of long-term arm morbidity are warranted.
Known indicators of high nodal burden are larger tumor size and a higher number of suspicious nodes at axillary imaging.11 Risk factors for additional non-SLN metastases in patients with metastatic SLNs (ie, not necessarily a high nodal burden) also include a high ratio of metastatic to excised SLNs (SLN ratio), extracapsular extension, and lymphovascular invasion in the primary tumor.12
Invasive lobular carcinoma (ILC) is most often of luminal A subtype and accounts for 10% to 15% of all breast cancer cases.13-16 In comparison to invasive breast cancer of no special type (NST), ILC is associated with a higher nodal burden and a higher rate of non-SLN metastases.17-19 In ILC, axillary ultrasonography, fine-needle aspiration cytology, and core needle biopsy have a lower sensitivity for detection of nodal metastases than in NST.20-22 In addition, SLN biopsy has a higher false negative rate in ILC than in NST.23 Unfortunately, no published axillary de-escalation trials have presented subgroup analyses addressing ILC.1-3 Additionally, ILC is associated with late recurrences, which may not be fully captured by follow-up times ranging from 5.0 to 9.3 years.1-3
Prediction models aimed at detecting high nodal burden offer a tool to identify individuals who require intensified adjuvant treatments while avoiding CALND.24-28 However, existing models are based on retrospective data encompassing any histopathological and molecular subtypes. None of the models have focused on luminal breast cancer, where nodal burden influences adjuvant strategies and no distinction between NST and ILC has been made. The primary aim of the present study was to develop prediction models for high nodal burden in patients with 1 or 2 SLN macrometastases and luminal ERBB2-negative breast cancer and ILC, respectively, which could reduce the risk of undertreatment without exposing patients to CALND.
The Sentinel Node Biopsy in Breast Cancer: Omission of Axillary Clearance After Macrometastases (SENOMAC) trial is a randomized (1:1), prospective, international, noninferiority trial. The study protocol was approved by the ethics committee at Karolinska Institutet, Stockholm, Sweden, and the trial was conducted according to good clinical practice. Enrollment was from January 2015 to December 2021. Randomized assignment was to CALND or its omission. Adjuvant treatment was according to national guidelines. Informed written consent was obtained from all individual participants prior to study inclusion.
The SENOMAC trial design and end points have been previously described, and initial results have been presented.29,30 Adult patients diagnosed with primary invasive cN0 T1-T3 breast cancer and 1 or 2 SLN macrometastases were eligible. Additional SLN micrometastases were allowed. Preoperative axillary ultrasonography was mandatory. While patients having an SLN biopsy prior to preoperative systemic therapy were eligible, 55 such patients were excluded from the present analysis since their CALND results, obtained after systemic therapy, may not reflect the initial nodal status. Overall, 1010 patients with luminal ERBB2-negative breast cancer randomized to CALND constituted the present study cohort, 212 of whom (21.0%) had ILC (eFigure in Supplement 1). Data analysis for this article took place from June 2023 to April 2024. The Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis () reporting guideline was followed.
The TNM classification system was used. Thus, pathological nodal status (pN) was pN1 in case of 1 to 3 nodal metastases, pN2 for 4 to 9, and pN3 for 10 or more metastases.31 High nodal burden was defined as pN2 or pN3. Histopathological tumor type was categorized into 3 groups: NST, ILC, or other/mixed.32
Sample size calculations were performed for the primary and secondary outcomes of the SENOMAC trial, but not for the present analysis.29 Patient and tumor demographic characteristics were summarized using standard descriptive statistical measures. Differences between the subgroups were evaluated using 蠂2 tests for categorical variables, linear trend tests for characteristics measured on an ordinal scale, and t tests for continuous variables.
Two prediction models for pN2-3 vs pN1 were developed: one for the entire luminal ERBB2-negative cohort and another for the ILC subgroup. For the development of the prediction model for the luminal ERBB2-negative cohort, patients were randomly split into a training set (80%) and a test set (20%) for validation. The split was stratified on outcome (鈮N2) to get the same prevalence of high nodal burden in both datasets. To minimize the risk of overfitting, the prediction model was developed in the training set and evaluated in the test set. The ILC subgroup was not split due to its smaller size.
Routinely accessible variables after SLN biopsy were selected based on published literature and considered as candidates for the prediction models (eAppendix 1 in Supplement 1). Univariable logistic regression analysis was used to estimate associations between candidate predictors and high nodal burden in the luminal ERBB2-negative cohort and the ILC subgroup, respectively.
The model development steps are described in detail in eAppendix 1 in Supplement 1. Briefly, backward logistic regression analysis was combined with bootstrap to select a subset of predictor variables for the final model. Since the fraction of missing data was low, models were developed using data from patients with complete data.
The diagnostic performance of the prediction model for the luminal ERBB2-negative cohort was evaluated in terms of discrimination (area under the receiver operating characteristic curve [AUC]) and calibration (Hosmer-Lemeshow plots, calibration slope, and intercept). The model for the luminal ERBB2-negative cohort was validated in the test set. Since no test set was set aside for the ILC subgroup, this model was not internally validated and should be regarded as exploratory. The equations for the final prediction models are presented in eAppendix 2 in Supplement 1. Nomograms were created. Sensitivity, specificity, positive predictive value, negative predictive value, true negative, true positive, false negative, and false positive rates for the models were calculated for clinically relevant predetermined cutoffs.
The prognostic value of the model for the luminal ERBB2-negative cohort was evaluated with recurrence-free survival (RFS) as the end point, using Kaplan-Meier estimates together with 1-df log-rank test for trend over the risk groups, and Cox regression. RFS was defined according to the Standardized Definitions for Efficacy End Points (STEEP) criteria.33 The proportional hazards assumption in the Cox models was tested using Schoenfeld residuals.
The evidence against each null hypothesis was quantified by a P value and the word significant used for 2-sided P鈥�<鈥�.05. The reported P values have not been adjusted for multiple comparisons. R version 4.2.2 (R Foundation) was used for all statistical analyses.
Of 1010 patients with luminal ERBB2-negative breast cancer assigned to CALND in the SENOMAC trial (luminal cohort; median [range] age, 61 [34-90] years; 1006 [99.6%] female and 4 [0.4%] male), 138 (13.7%) had a high nodal burden (Table). In 846 patients with 1 SLN macrometastasis, 83 (9.8%) had 鈮N2, while in 164 with 2 SLN macrometastases, 55 had 鈮N2 (33.5%) (P鈥�&濒迟;鈥�.001). The median (range) tumor size was significantly smaller in patients with a low nodal burden (19.0 mm [1.1-136.0] vs 25.0 mm [4.0-155.0]; P鈥�&濒迟;鈥�.001). Additional SLN micrometastases were less common in the low nodal burden group (84 of 872 [9.6%] vs 22 of 138 [15.9%]; P鈥�=鈥�.07), and SLN extracapsular extension significantly less prevalent (296 of 868 [34.1%] vs 66 of 137 [48.2%]; P鈥�=鈥�.002). No difference was seen between low vs high nodal burden for lymphovascular invasion in the primary tumor (242 of 870 [27.8%] vs 38 of 137 [27.7%]; P鈥�&驳迟;鈥�.99).
ILC constituted 21.0% of the entire luminal ERBB2-negative cohort. High nodal burden was more common in ILC than other histopathological tumor types (42 of 212 [19.8%] vs 96 of 798 [12.0%]; P鈥�=鈥�.005) (eTable in Supplement 1). The median (range) tumor size was significantly larger in patients with ILC (32.0 mm [6.0-155.0] vs 18.0 mm [1.1-97.0]; P鈥�&濒迟;鈥�.001).
Luminal ERBB2-Negative Cohort
Two vs 1 SLN macrometastases, higher SLN ratio, additional SLN micrometastases, SLN extracapsular extension, and larger tumor size were independently associated with high nodal burden and selected for the multivariable prediction model (eAppendix 2 in Supplement 1), which reached an AUC of 0.77 (95% CI, 0.72-0.82) in the training set and 0.74 (95% CI, 0.62-0.85) in the test set (Figure 1A and B). The calibration plot showed that the mean predicted probabilities of high nodal burden did not deviate systematically from observed fractions in deciles based on the predicted probabilities of high nodal burden in the test set (minor deviations from the dashed identity line, which indicates perfect calibration). This was also supported by the summary measures calibration intercept 0.05 and calibration slope 0.83, which are close to the optimal values 0 and 1.00, respectively, for a perfectly calibrated model (Figure 1C). At a sensitivity threshold of 80% or greater, 111 of the 201 patients in the test set (55.2%) were identified as high risk (Figure 1B). While this proportion seems high compared to the observed prevalence for 鈮N2 of 13.4%, the negative predictive value was 94.4% (85/90), correctly classifying all but 5 patients as low risk at this threshold. The model is visualized in a nomogram (Figure 2A).
In the ILC subgroup, 2 SLN macrometastases vs 1 had an odds ratio (OR) of 4.96 (95% CI, 2.17-11.47) in the final multivariable model for high vs low nodal burden (eAppendix 2 in Supplement 1). All independent factors identified in the luminal ERBB2-negative cohort, apart from tumor size, were associated with high nodal burden in ILC also and were included in the lobular prediction model with an AUC of 0.74 (95% CI, 0.66-0.83) in the development set (Figure 1D). The model is visualized in a nomogram (Figure 2B).
To investigate the prognostic potential of the prediction model, the total nomogram score was related to RFS in the luminal ERBB2-negative cohort. Cox regression analysis revealed that a 10-unit increase in the score corresponded to a 7% increase in the incidence of RFS events (hazard ratio [HR], 1.07; 95% CI, 1.02-1.13; P鈥�=鈥�.01) (Figure 3A). When categorizing the score into quartiles, the HR for the high-risk group (quartile 4) vs the low-risk group (quartile 1) was 2.11 (95% CI, 1.02-4.35; P鈥�=鈥�.04) and the HR for the medium-risk group (quartiles 2-3) vs quartile 1 was 1.59 (95% CI, 0.81-3.14; P鈥�=鈥�.18). The corresponding Kaplan-Meier estimates of cumulative incidence of RFS events are shown in Figure 3B.
In this diagnostic/prognostic study, the predictive models and corresponding nomograms for high nodal burden (鈮N2) in patients with 1 or 2 SLN macrometastases and luminal ERBB2-negative breast cancer and ILC, respectively, may help select patients for intensified adjuvant therapies while avoiding CALND. At a threshold corresponding to a sensitivity of at least 80%, all but 5 low-risk patients were correctly classified, corresponding to a negative predictive value of 94%. In the luminal ERBB2-negative cohort, the predicted probability of high nodal burden (鈮N2) was also prognostic of RFS, underscoring the clinical relevance of the prediction model as an alternative to CALND.
A minority of patients with 1 or 2 macrometastases on SLN biopsy have a high nodal burden: in the American College of Surgeons Oncology Group Z0011 (ACOSOG Z0011) trial,1 47 of 343 (13.7%) had 4 or more metastatic ALNs; in the European Organisation for Research and Treatment of Cancer 10981-22023 (AMAROS) trial,3 52 of 672 (8%) had 4 or more positive additional nodes (besides SLN metastasis); and in the Optimal Treatment of the Axilla鈥擲urgery or Radiotherapy After Positive Sentinel Lymph Node Biopsy in Early-Stage Breast Cancer (OTOASOR) trial,34 54 of 244 (22%) had 鈮N2a. Despite the presence of 鈮N2 disease in these trials, the omission of CALND did not increase long-term locoregional recurrence rates. However, there are currently no data from randomized clinical trials comparing CALND vs its omission in a population of patients with verified 鈮N2 disease. The proportion of 鈮N2 status ranges from 3.5% to 16% in cN0 breast cancer.35,36 Thus, the therapeutic effect of CALND is debatable. Here we present a tool for procuring similar diagnostic information considering multiple risk factors of 鈮N2.
Prediction Models鈥擟omparison to Previous Studies
In the multicenter tool for risk prediction of 4 or more ALN metastases in SLN-positive breast cancer developed by Meretoja et al,28 based on retrospective data from 675 patients, the AUC reached 0.77 (95% CI, 0.74-0.81) in the external validation using a retrospective cohort (n鈥�=鈥�760). The included variables were site prevalence of 鈮N2, number of metastatic and benign SLNs, tumor size, and extracapsular extension. Murata et al25 presented a model for prediction of high nodal burden based on 804 patients who were cN0 SLN positive. The AUC was 0.89 (95% CI, 0.84-0.93) in the test set, with good discrimination; however, as this was a retrospective single-institution study, external validation is needed. In another prediction model by Kim et al,26 lymphovascular invasion, extranodal extension, T stage, and a high SLN ratio were independently associated with high nodal burden in a retrospective analysis of 1437 patients. The AUC was 0.82 in the test set (no CI provided; no external validation). In a nomogram by Yang et al24 incorporating the number of SLN metastases, extracapsular extension, pT stage, and tumor location in the breast in 1480 patients, the AUC was 0.80 (95% CI, 0.76-0.85) in the external validation set. In a small study by Corsi et al27 (n鈥�=鈥�271), SLN ratio, extranodal extension, and multifocal disease were predictors of more than 2 metastatic ALNs, with an AUC of 0.74 (no CI provided; no external validation).
The prevalence of high nodal burden in the above-mentioned studies ranged from 5.7% to 20.7% (compared to 13.7% in this study). While the models based on retrospective cohorts by Meretoja et al and Yang et al were externally validated, none have been evaluated in terms of recurrence or survival analyses.
ILC is a distinct biological entity which warrants specific focus. The presented prediction model for ILC is, to our knowledge, the first model for high nodal burden in ILC and 1 or 2 SLN macrometastases. Studies are rarely analyzed by histopathological tumor type,37 and treatment recommendations usually address NST and ILC as 1 entity.4 Only in recent years, ILC-specific clinical trials have been initiated.16
Tumor size was selected in the luminal ERBB2-negative cohort but not in the ILC subgroup. Lee et al36 and Rivers et al38 found tumors greater than 20 mm to be associated with high nodal burden. However, in our study larger tumor size was independently associated with high nodal burden in the luminal ERBB2-negative but not the ILC subgroup. Considering the suggested threshold at greater than 20 mm, the significantly larger tumor size in ILC compared to other histopathological tumor types (median, 32 mm vs 18 mm) might explain why tumor size was not a strong independent predictive factor in the ILC subgroup.
Imaging is highly indicative of high nodal burden and can rule out stage 鈮N2 breast cancer with a negative predictive value of 96%.11,39 The sensitivity of axillary ultrasonography for predicting 鈮�3 metastatic ALN is, however, lower in ILC than in NST.20 Despite ILC being associated with more ALN metastases,40,41 nodal burden is more often underestimated preoperatively.42 Other imaging modalities, such as magnetic resonance imaging or 18F-fluorodeoxyglucose positron emission tomography/computed tomography, are of limited clinical value.42-44 These limitations underline the need for prediction models developed exclusively in ILC.
Altogether, ILC poses diagnostic staging challenges, and the risk of occult ALN involvement remains challenging. The use of a decision support tool might thus be more important in ILC than in NST.
Strengths and Limitations
A major strength of this study is that it is based on meticulously curated prospectively collected data from a large randomized clinical trial. Moreover, we address both global performance measures of discrimination and calibration. In addition, applied sensitivity cutoffs are derived from clinical reasoning rather than being data driven. We exclusively incorporated routinely accessible variables, available at the intended time point for using the nomogram. To further enable clinical use, a nomogram is presented for each target population. While prediction models were not developed to predict RFS, we show clinical validity by differentiation between prognostic strata based on nomogram scores. Longer follow-up will be necessary to monitor separation of RFS curves given the potential of luminal breast cancer to recur late.
The foremost limitation of this study is the lack of validation in an external cohort and the fact that no internal validation was conducted for the ILC subgroup due to limited sample size. Although the inclusion criteria in SENOMAC are considered broad, generalization beyond this population should be used with caution. Another limitation is the absence of some variables of putative relevance for modeling, such as mode of detection (eg, screening or clinically detected), race, ethnicity, and body mass index. Progesterone receptor status and proliferation rate (Ki67) were not considered because of poor standardization and varying availability across centers. The 21-gene Oncotype DX Breast Recurrence Score may inform decisions on adjuvant chemotherapy in postmenopausal luminal ERBB2-negative breast cancer with 1 to 3 metastatic ALNs without CALND. However, the clinical value of Oncotype DX for guidance of adjuvant abemaciclib is not yet determined, and a nomogram can be applied irrespective of the availability of genomic tests.45 In the SENOMAC trial, multigene signatures were not registered. While acknowledging this as a limitation, we emphasize that the presented models are clinically feasible also in economically diverse conditions and not dependent on the availability of tissue for analysis. No imaging-derived variables were considered for the models, enabling application in routine clinical settings where axillary imaging is not part of diagnostic workup for patients with cN0 T1-T2 breast cancer eligible for upfront surgery.11,46
The presented model predicts high nodal burden in patients with luminal ERBB2-negative T1-T3 breast cancer and 1 or 2 SLN macrometastases with good performance measures. Use of the nomogram may facilitate systemic treatment decisions without exposing patients to CALND, thus minimizing the risk of arm morbidity. External validation is necessary.
Accepted for Publication: July 9, 2024.
Published Online: September 25, 2024. doi:10.1001/jamasurg.2024.3944
Open Access: This is an open access article distributed under the terms of the CC-BY License. 漏 2024 Skarping I et al. JAMA Surgery.
Corresponding Author: Ida Skarping, MD, PhD, Division of Oncology, Department of Clinical Sciences, Lund University, 223 81 Lund, Sweden (ida.skarping@med.lu.se).
Author Contributions: Drs Szulkin and de Boniface had full access to all data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Ryd茅n and Boniface are co鈥搒enior authors responsible for study oversight.
Concept and design: Skarping, Bendahl, Szulkin, Andersson, Filtenborg Tvedskov, Frisell, Lundstedt, Vrou Offersen, Olofsson Bagge, Sund, Ryd茅n, de Boniface.
Acquisition, analysis, or interpretation of data: Skarping, Szulkin, Alkner, Andersson, Bergkvist, Christiansen, Frisell, Gentilini, Kontos, K眉hn, Lundstedt, Vrou Offersen, Olofsson Bagge, Reimer, Sund, Ryd茅n, de Boniface.
Drafting of the manuscript: Skarping, Bendahl, Szulkin, Frisell, Lundstedt, Ryd茅n, de Boniface.
Critical review of the manuscript for important intellectual content: Skarping, Szulkin, Alkner, Andersson, Bergkvist, Christiansen, Filtenborg Tvedskov, Frisell, Gentilini, Kontos, K眉hn, Lundstedt, Vrou Offersen, Olofsson Bagge, Reimer, Sund, Ryd茅n, de Boniface.
Statistical analysis: Bendahl, Szulkin.
Obtained funding: Skarping, Ryd茅n, de Boniface.
Administrative, technical, or material support: Skarping, Frisell, Lundstedt, Olofsson Bagge, Ryd茅n, de Boniface.
Supervision: Skarping, Bergkvist, Frisell, Gentilini, Lundstedt, Vrou Offersen, Olofsson Bagge, Reimer, Ryd茅n, de Boniface.
Conflict of Interest Disclosures: Dr Gentilini reported compensation for consultation/advisory board service for Merck Sharp & Dohme, AstraZeneca, Eli Lilly, BD Biosciences, and Bayer. Dr K眉hn reported personal fees from Merck Sharp & Dohme, Eli Lilly, Endomag, Sirius Medical, Merit Medical, and AstraZeneca outside the submitted work. Dr Olofsson Bagge reported an institutional research grant from Endomagnetics, a speaker honorarium from Pierre-Fabre, and shareholdings with Sahlgrenska Translational Melanoma Group Ventures outside the submitted work. No other disclosures were reported.
Funding/Support: This trial is supported by grants from the Swedish Research Council, the Swedish Cancer Society, the Swedish Society of Medicine, the Swedish Breast Cancer Association, and the Nordic Cancer Union. In addition, we acknowledge funding from the Governmental Funding of Clinical Research in the National Health Services (Avtal om l盲karutbildning och forskning young researcher Dr Skarping) and Erling Persson Foundation (Drs Bendahl and Ryd茅n).
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Data Sharing Statement: See Supplement 2.
Additional Contributions: We thank all the participants in the trial.
1.Giuliano
听AE锘�, Ballman
听KV锘�, McCall
听L锘�,
听et al. 听Effect of axillary dissection vs no axillary dissection on 10-year overall survival among women with invasive breast cancer and sentinel node metastasis: the ACOSOG Z0011 (Alliance) randomized clinical trial.听锘� 听闯础惭础. 2017;318(10):918-926. doi:
2.Galimberti
听V锘�, Cole
听BF锘�, Zurrida
听S锘�,
听et al; International Breast Cancer Study Group Trial 23-01 investigators. 听Axillary dissection versus no axillary dissection in patients with sentinel-node micrometastases (IBCSG 23-01): a phase 3 randomised controlled trial.听锘� 听Lancet Oncol. 2013;14(4):297-305. doi:
3.Donker
听M锘�, van Tienhoven
听G锘�, Straver
听ME锘�,
听et al. 听Radiotherapy or surgery of the axilla after a positive sentinel node in breast cancer (EORTC 10981-22023 AMAROS): a randomised, multicentre, open-label, phase 3 non-inferiority trial.听锘� 听Lancet Oncol. 2014;15(12):1303-1310. doi:
4.Burstein
听HJ锘�, Curigliano
听G锘�, Loibl
听S锘�,
听et al; Members of the St. Gallen International Consensus Panel on the Primary Therapy of Early Breast Cancer 2019. 听Estimating the benefits of therapy for early-stage breast cancer: the St. Gallen International Consensus Guidelines for the primary therapy of early breast cancer 2019.听锘� 听Ann Oncol. 2019;30(10):1541-1557. doi:
5.Johnston
听SRD锘�, Harbeck
听N锘�, Hegg
听R锘�,
听et al; MONARCHE Committee Members and Investigators. 听Abemaciclib combined with endocrine therapy for the adjuvant treatment of HR+, HER2-, node-positive, high-risk, early breast cancer (MONARCHE).听锘� 听J Clin Oncol. 2020;38(34):3987-3998. doi:
6.Rastogi
听P锘�, O鈥橲haughnessy
听J锘�, Martin
听M锘�,
听et al. 听Adjuvant abemaciclib plus endocrine therapy for hormone receptor-positive, human epidermal growth factor receptor 2-negative, high-risk early breast cancer: results from a preplanned MONARCHE overall survival interim analysis, including 5-year efficacy outcomes.听锘� 听J Clin Oncol. 2024;42(9):987-993. doi:
7.Cardoso
听F锘�, van鈥檛 Veer
听LJ锘�, Bogaerts
听J锘�,
听et al; MINDACT Investigators. 听70-Gene signature as an aid to treatment decisions in early-stage breast cancer.听锘� 听N Engl J Med. 2016;375(8):717-729. doi:
8.Cardoso
听F锘�, Kyriakides
听S锘�, Ohno
听S锘�,
听et al; ESMO Guidelines Committee. Electronic address:
clinicalguidelines@esmo.org. 听Early breast cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up.听锘�
听Ann Oncol. 2019;30(8):1194-1220. doi:
9.Fleissig
听A锘�, Fallowfield
听LJ锘�, Langridge
听CI锘�,
听et al. 听Post-operative arm morbidity and quality of life. Results of the ALMANAC randomised trial comparing sentinel node biopsy with standard axillary treatment in the management of patients with early breast cancer.听锘� 听叠谤别补蝉迟 Cancer Res Treat. 2006;95(3):279-293. doi:
10.Appelgren
听M锘�, Sackey
听H锘�, Wengstr枚m
听Y锘�,
听et al; SENOMAC Trialists鈥� Group. 听Patient-reported outcomes one year after positive sentinel lymph node biopsy with or without axillary lymph node dissection in the randomized SENOMAC trial.听锘� 听叠谤别补蝉迟. 2022;63:16-23. doi:
11.Chang
听JM锘�, Leung
听JWT锘�, Moy
听L锘�, Ha
听SM锘�, Moon
听WK锘�. 听Axillary nodal evaluation in breast cancer: state of the art.听锘� 听搁补诲颈辞濒辞驳测. 2020;295(3):500-515. doi:
12.van la Parra
听RF锘�, Peer
听PG锘�, Ernst
听MF锘�, Bosscha
听K锘�. 听Meta-analysis of predictive factors for non-sentinel lymph node metastases in breast cancer patients with a positive SLN.听锘� 听Eur J Surg Oncol. 2011;37(4):290-299. doi:
13.Pramod
听N锘�, Nigam
听A锘�, Basree
听M锘�,
听et al. 听Comprehensive review of molecular mechanisms and clinical features of invasive lobular cancer.听锘� 听翱苍肠辞濒辞驳颈蝉迟. 2021;26(6):e943-e953. doi:
14.Van Baelen
听K锘�, Geukens
听T锘�, Maetens
听M锘�,
听et al. 听Current and future diagnostic and treatment strategies for patients with invasive lobular breast cancer.听锘� 听Ann Oncol. 2022;33(8):769-785. doi:
15.McCart Reed
听AE锘�, Kutasovic
听JR锘�, Lakhani
听SR锘�, Simpson
听PT锘�. 听Invasive lobular carcinoma of the breast: morphology, biomarkers and 'omics.听锘� 听叠谤别补蝉迟 Cancer Res. 2015;17(1):12. doi:
16.McCart Reed
听AE锘�, Kalinowski
听L锘�, Simpson
听PT锘�, Lakhani
听SR锘�. 听Invasive lobular carcinoma of the breast: the increasing importance of this special subtype.听锘� 听叠谤别补蝉迟 Cancer Res. 2021;23(1):6. doi:
17.Narbe
听U锘�, Bendahl
听PO锘�, Fern枚
听M锘�, Ingvar
听C锘�, Dihge
听L锘�, Ryd茅n
听L锘�. 听St Gallen 2019 guidelines understage the axilla in lobular breast cancer: a population-based study.听锘� 听Br J Surg. 2021;108(12):1465-1473. doi:
18.Adachi
听Y锘�, Sawaki
听M锘�, Hattori
听M锘�,
听et al. 听Comparison of sentinel lymph node biopsy between invasive lobular carcinoma and invasive ductal carcinoma.听锘� 听叠谤别补蝉迟 Cancer. 2018;25(5):560-565. doi:
19.Caudle
听AS锘�, Kuerer
听HM锘�, Le-Petross
听HT锘�,
听et al. 听Predicting the extent of nodal disease in early-stage breast cancer.听锘� 听Ann Surg Oncol. 2014;21(11):3440-3447. doi:
20.Morrow
听E锘�, Lannigan
听A锘�, Doughty
听J锘�,
听et al. 听Population-based study of the sensitivity of axillary ultrasound imaging in the preoperative staging of node-positive invasive lobular carcinoma of the breast.听锘� 听Br J Surg. 2018;105(8):987-995. doi:
21.Topps
听A锘�, Clay
听V锘�, Absar
听M锘�,
听et al. 听The sensitivity of pre-operative axillary staging in breast cancer: comparison of invasive lobular and ductal carcinoma.听锘� 听Eur J Surg Oncol. 2014;40(7):813-817. doi:
22.Hackney
听L锘�, Williams
听S锘�, Bajwa
听S锘�, Morley-Davies
听AJ锘�, Kirby
听RM锘�, Britton
听I锘�. 听Influence of tumor histology on preoperative staging accuracy of breast metastases to the axilla.听锘� 听叠谤别补蝉迟 J. 2013;19(1):49-55. doi:
23.Guo
听R锘�, Brabham
听CE锘�, Fahrner-Scott
听K锘�,
听et al. 听Accuracy of sentinel lymph node biopsy in invasive lobular carcinoma of the breast.听锘� 听J Clin Oncol. 2020;38(15)(suppl):e12604-e12604. doi:
24.Yang
听Z锘�, Lan
听X锘�, Huang
听Z锘�,
听et al. 听Development and external validation of a nomogram to predict four or more positive nodes in breast cancer patients with one to three positive sentinel lymph nodes.听锘� 听叠谤别补蝉迟. 2020;53:143-151. doi:
25.Murata
听T锘�, Watase
听C锘�, Shiino
听S锘�,
听et al. 听Development and validation of a pre- and intra-operative scoring system that distinguishes between non-advanced and advanced axillary lymph node metastasis in breast cancer with positive sentinel lymph nodes: a retrospective study.听锘� 听World J Surg Oncol. 2022;20(1):314. doi:
26.Kim
听I锘�, Ryu
听JM锘�, Kim
听JM锘�,
听et al. 听Development of a nomogram to predict N2 or N3 stage in T1-2 invasive breast cancer patients with no palpable lymphadenopathy.听锘� 听J Breast Cancer. 2017;20(3):270-278. doi:
27.Corsi
听F锘�, Sorrentino
听L锘�, Albasini
听S锘�,
听et al. 听Prediction of nodal staging in breast cancer patients with 1-2 sentinel nodes in the Z0011 era.听锘� 听Medicine (Baltimore). 2020;99(35):e21721. doi:
28.Meretoja
听TJ锘�, Audisio
听RA锘�, Heikkil盲
听PS锘�,
听et al. 听International multicenter tool to predict the risk of four or more tumor-positive axillary lymph nodes in breast cancer patients with sentinel node macrometastases.听锘� 听叠谤别补蝉迟 Cancer Res Treat. 2013;138(3):817-827. doi:
29.de Boniface
听J锘�, Frisell
听J锘�, Andersson
听Y锘�,
听et al; SENOMAC Trialists鈥� Group. 听Survival and axillary recurrence following sentinel node-positive breast cancer without completion axillary lymph node dissection: the randomized controlled SENOMAC trial.听锘� 听BMC Cancer. 2017;17(1):379. doi:
30.de Boniface
听J锘�, Filtenborg Tvedskov
听T锘�, Ryd茅n
听L锘�,
听et al; SENOMAC Trialists鈥� Group; SENOMAC Trialists鈥� Group. 听Omitting axillary dissection in breast cancer with sentinel-node metastases.听锘� 听N Engl J Med. 2024;390(13):1163-1175. doi:
31.Sobin
听L锘�, Gospodarowicz
听M锘�, Wittekind
听C锘�. 听TNM Classification of Malignant Tumours. Wiley-Blackwell; 2009.
32.Canas-Marques
听R锘�, Schnitt
听SJ锘�. 听E-cadherin immunohistochemistry in breast pathology: uses and pitfalls.听锘� 听贬颈蝉迟辞辫补迟丑辞濒辞驳测. 2016;68(1):57-69. doi:
33.Tolaney
听SM锘�, Garrett-Mayer
听E锘�, White
听J锘�,
听et al. 听Updated Standardized Definitions for Efficacy End Points (STEEP) in adjuvant breast cancer clinical trials: STEEP version 2.0.听锘� 听J Clin Oncol. 2021;39(24):2720-2731. doi:
34.S谩volt
听脕锘�, P茅ley
听G锘�, Polg谩r
听C锘�,
听et al. 听Eight-year follow up result of the OTOASOR trial: the Optimal Treatment of the Axilla鈥擲urgery or Radiotherapy After Positive Sentinel Lymph Node Biopsy in Early-Stage Breast Cancer: a randomized, single centre, phase III, non-inferiority trial.听锘� 听Eur J Surg Oncol. 2017;43(4):672-679. doi:
35.Rocco
听N锘�, Ghilli
听M锘�, Curcio
听A锘�,
听et al. 听Is routine axillary lymph node dissection needed to tailor systemic treatments for breast cancer patients in the era of molecular oncology? a position paper of the Italian National Association of Breast Surgeons (ANISC).听锘� 听Eur J Surg Oncol. 2024;50(2):107954. doi:
36.Lee
听MK锘�, Montagna
听G锘�, Pilewskie
听ML锘�, Sevilimedu
听V锘�, Morrow
听M锘�. 听Axillary staging is not justified in postmenopausal clinically node-negative women based on nodal disease burden.听锘� 听Ann Surg Oncol. 2023;30(1):92-97. doi:
37.Mouabbi
听JA锘�, Hassan
听A锘�, Lim
听B锘�, Hortobagyi
听GN锘�, Tripathy
听D锘�, Layman
听RM锘�. 听Invasive lobular carcinoma: an understudied emergent subtype of breast cancer.听锘� 听叠谤别补蝉迟 Cancer Res Treat. 2022;193(2):253-264. doi:
38.Rivers
听AK锘�, Griffith
听KA锘�, Hunt
听KK锘�,
听et al. 听Clinicopathologic features associated with having four or more metastatic axillary nodes in breast cancer patients with a positive sentinel lymph node.听锘� 听Ann Surg Oncol. 2006;13(1):36-44. doi:
39.Neal
听CH锘�, Daly
听CP锘�, Nees
听AV锘�, Helvie
听MA锘�. 听Can preoperative axillary US help exclude N2 and N3 metastatic breast cancer?听锘� 听搁补诲颈辞濒辞驳测. 2010;257(2):335-341. doi:
40.Chen
听BF锘�, Tsai
听YF锘�, Lien
听PJ锘�,
听et al. 听Clinical characteristics and treatment outcomes of invasive ductal and lobular carcinoma: analyses of 54,832 Taiwan cancer registry index cases.听锘� 听叠谤别补蝉迟 Cancer Res Treat. 2023;201(3):547-560. doi:
41.Oesterreich
听S锘�, Nasrazadani
听A锘�, Zou
听J锘�,
听et al. 听Clinicopathological features and outcomes comparing patients with invasive ductal and lobular breast cancer.听锘� 听J Natl Cancer Inst. 2022;114(11):1511-1522. doi:
42.Kawaguchi
听S锘�, Kinowaki
听K锘�, Tamura
听N锘�,
听et al. 听High-accuracy prediction of axillary lymph node metastasis in invasive lobular carcinoma using focal cortical thickening on magnetic resonance imaging.听锘� 听叠谤别补蝉迟 Cancer. 2023;30(4):637-646. doi:
43.Gilardi
听L锘�, Air貌 Farulla
听LS锘�, Curigliano
听G锘�, Corso
听G锘�, Leonardi
听MC锘�, Ceci
听F锘�. 听FDG and Non-FDG radiopharmaceuticals for PET imaging in invasive lobular breast carcinoma.听锘� 听叠颈辞尘别诲颈肠颈苍别蝉. 2023;11(5):1350. doi:
44.Schumacher
听K锘�, Inciardi
听M锘�, O鈥橬eil
听M锘�,
听et al. 听Is axillary imaging for invasive lobular carcinoma accurate in determining clinical node staging?听锘� 听叠谤别补蝉迟 Cancer Res Treat. 2021;185(3):567-572. doi:
45.Kalinsky
听K锘�, Barlow
听WE锘�, Gralow
听JR锘�,
听et al. 听21-Gene assay to inform chemotherapy benefit in node-positive breast cancer.听锘� 听N Engl J Med. 2021;385(25):2336-2347. doi:
46.Saksena
听M锘�, Jimenez
听R锘�, Coopey
听S锘�, Harris
听K锘�. 听Axillary ultrasound evaluation in breast cancer patients: a multidisciplinary viewpoint and middle ground.听锘� 听J Breast Imaging. 2021;3(6):672-675. doi: