Sage Journals: Discover world-class research

Abstract

BACKGROUND:

The p53 mutation in breast cancer confers a worse prognosis and is usually associated with p53 overexpression (p53+) on immunohistochemistry. Previous studies have shown that p53+ tumors could be associated with low axillary tumor burden (ATB).

OBJECTIVE:

We aimed to evaluate the association between p53+ and ATB in a large series of breast cancers as an aid to personalizing axillary surgical treatment.

METHODS:

We retrieved 1762 infiltrating breast carcinomas from our database that were treated with upfront surgery in Hospital del Mar from 2004 to 2018. We compared p53+ and p53-negative (p53-) tumors in terms of the percentage of cases with high ATB and overall survival. This comparison was made overall and for each immunophenotype.

RESULTS:

Overall, 18.7% of breast tumors were p53+. High ATB was less common in p53+ tumors than in p53- tumors in the luminal B-Her2-negative immunophenotype (6.2% versus 16.9%, respectively, P = 0.025), but not in the other immunophenotypes or overall. Overall survival was worse in patients with p53+ breast cancer (P = 0.002).

CONCLUSION:

p53+ breast cancers were associated with worse overall survival. However, low ATB was more common in these tumors than in p53- tumors in the luminal B-Her2-negative subtype. Information on p53 expression could be of use to predict ATB in some breast cancer tumors.

Keywords

p53 axillary tumor burden breast cancer immunophenotypes

1. Background

Breast cancer outcomes are related to well-known biological factors, such as tumor size, the presence of axillary lymph node invasion, metastasis, histological subtype, the number of foci [1], estrogen (ER) and progesterone receptor (PR) status, human epidermal receptor 2 (Her 2) status, and the Ki67 proliferation index [2].

The TP53 mutation is detected in 23% of breast tumors [3,4] and is usually associated with p53 overexpression (p53+) on immunohistochemistry. It is also related to breast cancer outcomes, as it confers a worse prognosis, including more frequent loco-regional recurrence [5–7].

The histological and molecular characteristics of the tumor are helpful for individualizing local breast cancer treatment [2,8]. Since the publication of the Z0011 trial in 2011 [9], there has been ongoing research to predict axillary tumor burden (ATB) with the aim of personalizing axillary surgery. Identified predictors of ATB include ultrasound features [10], a previous positive axillary biopsy [11], and the HER2+ subtype [12]. These predictors have been used in several predictive nomograms with variable accuracy [13].

p53+ has also been evaluated as a feature that could predict ATB, but the findings are discrepant. While some authors have found no relationship between p53+ and ATB [14,15], our previous studies revealed a significant association between these two biomarkers [16,17].

Given that the contradictory results found in previous studies could be due to the different features of the cohorts studied, the aim of this study was to evaluate p53 expression as a predictor of ATB in a large series of breast tumors for breast cancer as a whole and for each immunophenotype separately.

2. Materials and methods

2.1. Study type and ethics

We performed a retrospective cohort study of data prospectively entered into our Institutional Tumor Registry. The trial was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and was approved by the Clinical Research Ethics Committee of the IMIM, Spanish acronym for Instituto Municipal de Investigación Médica (Municipal Institute of Medical Research) id 2015/6282/I.

The study population comprised all women with a diagnosis of invasive breast carcinoma and upfront surgery treated in Hospital del Mar from January 1, 2004, to September 30, 2018.

Exclusion criteria comprised treatment in another center, in situ carcinoma, neoadjuvant treatments before surgery, and patients with insufficient information in the pathology report.

2.2. Pathology processing

Our standardized institutional protocol was followed to process breast surgery specimens. Briefly, all specimens were stained and sectioned at 3-mm intervals and then fixed in 10% neutral buffered formalin for a period ranging from 24–48 h. If possible, the entire specimen underwent paraffin processing. When not possible, the specimen was sampled using systematic mapping. Hematoxylin and eosin (H&E) slides were evaluated. Lesion extent, pathological type and histological grade were systematically reported. Sentinel lymph nodes were examined by serial sectioning: they were cut into 2-mm slices and, for each slice, six sections of 4-μm were obtained, leaving a 20-μm gap between them. Three alternate sections were stained with routine H&E and, if negative, the remaining sections were studied using cytokeratin. For axillary lymph node dissection (ALND), 3-mm slices were obtained for each node and, for each slice, a single 4-μm section was stained with routine H&E.

Immunohistochemical analysis was systematically performed including ER, PR, p53, Her2Neu, and Ki67. ER, PR and p53 were reported using a semi-quantitative scale, describing the percentage of positive cells and staining intensity. Ki67 was reported describing the percentage of positive cells. Her-2-Neu expression by immunohistochemistry was reported as negative (0,1+), positive (3+) or borderline (2+); in situ hybridization (ISH) was conducted in borderline cases to rule out Her-2-Neu gene amplification.

2.3. Association of p53 with other features and survival

We evaluated the association between p53 expression and tumor size, stage, number of foci, ATB, immunophenotype, and overall survival.

We divided ATB in two categories: low ATB, defined as ≤2 positive axillary nodes, and high ATB, defined as >2 positive axillary nodes. We chose this cut-off following the concept of low and high ATB used in the Z0011 trial [9,18] and the recommendations of international guidelines [19].

We defined five immunophenotypes following the recommendations of the 2013 St Gallen conference [2]: Luminal A-like (ER⁺, PR ≥20%, Her2-, ki67 <14%), luminal B-Her2- (ER⁺ and/or PR⁺ but PR <20%, and/or ki67 ≥ 14% and Her2-), luminal B-Her2+ (ER and/or PR⁺ and HER2+), Her2+-like (ER⁻, PR⁻ and HER2+) and triple-negative-like (ER⁻, PR⁻, HER2-). Luminal B tumors were analyzed separately according to Her2 expression as their prognosis and management usually differ [20].

2.4. Variables evaluated as predictors of ATB

The variables analyzed as candidate predictors of ATB were p53 expression, histological subtype, histological grade, lympho-vascular invasion, Ki67, number of foci, tumor size, ER, PR, Her2 and immunophenotype.

2.5. Statistical analysis

We performed the statistical analysis using PAWS version 18.0. Continuous variables were compared using the Kruskall--Wallis or Spearmans Rho tests. We analyzed categorical variables with thechi-square and Fisher exact tests in the univariate analysis. Variables that were significant in the univariate analysis or significant in previously published research were entered into a multivariateanalysis, performed using multiple linear regression. Overall survival was analyzed using Kaplan-Meyer tests. Differences were considered significant if the p value was ≤0.05.

3. Results

3.1. Cohort description

During the study period, 2766 patients underwent surgery for primary breast cancer in Hospital del Mar. We excluded 699 patients treated with upfront systemic treatment and 342 patients because of missing pathological information on axillary nodes. The remaining 1725 patients underwent 1762 interventions (37 bilateral cases). The mean age was 61 years (range 23–92) and the mean tumor size was 18.91 mm (SD 14.55).

The median number of affected axillary nodes was 0.93 (range 0–30). Cases with a high ATB were less frequent in the luminal A subtype (19 of 362, 5.2%) compared to the Her2+ subtype (12 0f 79, 15.2%, P = 0.007). Other features of the study population are summarized in Table 1.

Table 1
Main pathology features of the cohort

Features Subtypes N(%)

Histological subtype

Non-special type 1443 (81.9%)

Lobular 174 (9.9%)

Other 145 (8.2%)

Immunophenotypes

Luminal A 362 (20.5%)

Luminal B-HER2- 569 (32.3%)

Luminal B-HER2+ 137 (7.8%)

HER2+ 79 (4.5%)

Triple negative 177 (10.0%)

Luminal, not specified 438 (24.9%)

Histological grade

Grade I 600 (38.7%)

Grade II 563 (36.3%)

Grade III 388 (25.0%)

Lympho-vascular invasion

No 1218 (80%)

Yes 304 (20%)

Number of foci

Unifocal 1507 (85.5%)

Multifocal/Multicentric 254 (14.5%)

Features	Subtypes	N(%)
Histological subtype
	Non-special type	1443 (81.9%)
	Lobular	174 (9.9%)
	Other	145 (8.2%)
Immunophenotypes
	Luminal A	362 (20.5%)
	Luminal B-HER2-	569 (32.3%)
	Luminal B-HER2+	137 (7.8%)
	HER2+	79 (4.5%)
	Triple negative	177 (10.0%)
	Luminal, not specified	438 (24.9%)
Histological grade
	Grade I	600 (38.7%)
	Grade II	563 (36.3%)
	Grade III	388 (25.0%)
Lympho-vascular invasion
	No	1218 (80%)
	Yes	304 (20%)
Number of foci
	Unifocal	1507 (85.5%)
	Multifocal/Multicentric	254 (14.5%)

3.2. Association of p53 with other features and survival

Overall, 329 of 1762 tumors (18.7%) were p53+. P53+ was more frequent in larger tumors (P = 0.001) and those in more advanced stages (P = 0.003). No differences were found between unifocal and multifocal/multicentric tumors (P = 0.669). These data are shown in Table 2.

Table 2
Main pathology features in relation to p53 expression

Pathology features P53- P53+ p

pT

pT1 1032 (72%) 196 (59.6%) 0,001

pT2 345 (24.1%) 115 (34.8%)

pT3 56 (3.9%) 18 (5.5%)

Stage

I 852 (59.9%) 160 (49.8%) 0,003

II 478 (33.6%) 140 (43.6%)

III 92 (6.5%) 21 (6.5%)

Number of foci

Unifocal 85.4% 86.3% 0,669

Multifocal/Multicentric 14.6% 13.7%

Pathology features	P53-	P53+	p
pT
pT1	1032 (72%)	196 (59.6%)	0,001
pT2	345 (24.1%)	115 (34.8%)
pT3	56 (3.9%)	18 (5.5%)
Stage
I	852 (59.9%)	160 (49.8%)	0,003
II	478 (33.6%)	140 (43.6%)
III	92 (6.5%)	21 (6.5%)
Number of foci
Unifocal	85.4%	86.3%	0,669
Multifocal/Multicentric	14.6%	13.7%

Data on p53 expression and ATB are summarized in Table 3. High ATB was more frequent in p53- than in p53+ tumors (12.7% versus 4.5%, P = 0.025) in the luminal B Her2- subtype, but there were no significant differences in the remaining immunophenotypes or for the overall population.

Table 3

p53 expression in cases of low and high ATB, overall and for each immunophenotype

Immunophenotypes	p53	Low ATB	High ATB	p
Luminal B HER2-	p53-	419 (87.3%)	61 (12.7%)	0,025
	p53+	85 (95.5%)	4 (4.5%)
Luminal B HER2+	p53-	84 (87.5%)	12 (12.5%)	0,647
	p53+	37 (90.2%)	4 (9.8%)
Luminal A	p53-	324 (94.5%)	19 (5.5%)	0,292
	p53+	19 (100%)	0 (0%)
HER2+	p53-	26 (83.9%)	5 (16.1%)	0,852
	p53+	41 (85.4%)	7 (14.6%)
Triple negative	p53-	76 (85.4%)	13 (14.6%)	0,377
	p53+	79 (89.8%)	9 (10.2%)
Overall	p53-	1283 (89.5%)	150 (10.5%)	0,162
	p53+	303 (92.1%)	26 (7.9%)

In the luminal A subtype, there were 19 p53+ tumors (15.6%), 89 in the luminal B-Her2- subtype (15.6%), 41 in the luminal B Her2+ subtype (29.9%), 48 in the Her2+ subtype (60.8%), and 88 in the triple-negative subtype (49.7%).

We registered 198 deaths during the study period (11.98%). Overall survival was significantly higher in p53- than in p53+ patients (P = 0.0018), as shown in Fig. 1.

Fig. 1.

Differences in overall survival between patients with p53- and patients with p53+ breast tumors.

3.3. Features evaluated as predictors of ATB

In the univariate analysis, variables associated with high ATB were lobular histological subtype, tumor size, lympho-vascular invasion, Ki67 > 14%, high histological grade, and multifocality/multicentricity, but not p53 expression (data shown in Table 4).

Table 4
Pathology features and their relation with axillary tumor burden

Features Low ATB^∗ High ATB^∗ p

p53

P53- 1283 (89.5%) 150 (10.5%) 0.162

P53+ 303 (92.1%) 26 (7.9%)

Histological subtype

Carcinoma NST^∗∗ 1283 (90.2%) 139 (9.8%) 0.003

Lobular 145 (83.3%) 29 (16.7%)

Tubular 20 (95.2%) 1 (4.8%)

Other 138 (95.2%) 7 (4.8%)

pT

T1 1169 (95.2%) 59 (4.8%) 0.001

T2 370 (80.4%) 90 (19.6%)

T3 47 (63.5%) 27 (36.5%)

Lympho-vascular infiltration

Absent 1158 (95.1%) 60 (4.9%) 0.001

Present 208 (68.4%) 96 (31.6%)

Unknown 220 (91.7%) 20 (8.3%)

Grade

Grade I 560 (93.3%) 40 (6.7%) 0.006

Grade II 502 (89.2%) 61 (10.8%)

Grade III 337 (86.9%) 51 (13.1%)

Number of foci

Unifocal 1387 (91.2%) 133 (8.8%) 0.001

Multifocal/Multicentric 211 (83.1%) 43 (16.9%)

Ki proliferation index

Ki ≤ 14% 453 (94.2%) 28 (5.8%) 0.002

Ki > 14% 406 (88.5%) 53 (11.5%)

Estrogen receptor

Positive 237 (86.8%) 36 (13.2%) 0.055

Negative 1349 (90.6%) 140 (9.4%)

Progesterone Receptor

Positive 1133 (90.5%) 57 (11.2%) 0.283

Negative 452 (88.8%) 119 (9.5%)

HER2

HER2- 1398 (90.4%) 148 (9.6%) 0.120

HER2+ 188 (87.0%) 28 (13.0%)

Features	Low ATB^∗	High ATB^∗	p
p53
P53-	1283 (89.5%)	150 (10.5%)	0.162
P53+	303 (92.1%)	26 (7.9%)
Histological subtype
Carcinoma NST^∗∗	1283 (90.2%)	139 (9.8%)	0.003
Lobular	145 (83.3%)	29 (16.7%)
Tubular	20 (95.2%)	1 (4.8%)
Other	138 (95.2%)	7 (4.8%)
pT
T1	1169 (95.2%)	59 (4.8%)	0.001
T2	370 (80.4%)	90 (19.6%)
T3	47 (63.5%)	27 (36.5%)
Lympho-vascular infiltration
Absent	1158 (95.1%)	60 (4.9%)	0.001
Present	208 (68.4%)	96 (31.6%)
Unknown	220 (91.7%)	20 (8.3%)
Grade
Grade I	560 (93.3%)	40 (6.7%)	0.006
Grade II	502 (89.2%)	61 (10.8%)
Grade III	337 (86.9%)	51 (13.1%)
Number of foci
Unifocal	1387 (91.2%)	133 (8.8%)	0.001
Multifocal/Multicentric	211 (83.1%)	43 (16.9%)
Ki proliferation index
Ki ≤ 14%	453 (94.2%)	28 (5.8%)	0.002
Ki > 14%	406 (88.5%)	53 (11.5%)
Estrogen receptor
Positive	237 (86.8%)	36 (13.2%)	0.055
Negative	1349 (90.6%)	140 (9.4%)
Progesterone Receptor
Positive	1133 (90.5%)	57 (11.2%)	0.283
Negative	452 (88.8%)	119 (9.5%)
HER2
HER2-	1398 (90.4%)	148 (9.6%)	0.120
HER2+	188 (87.0%)	28 (13.0%)

^∗ATB, Axillary tumor burden. ^∗∗ Non-specified type.

In the multivariate analysis, the factors that remained significantly associated with high ATB were lymphovascular invasion (P = 0.001), Ki67 > 14% (P = 0.04), and p53- (P = 0.029).

4. Discussion

In this study, ATB was lower in p53+ than in p53- tumors in the luminal B Her2- subtype. Importantly, in the multivariate analysis using the whole cohort, p53- expression was a predictor of high ATB.

Personalizing surgical treatments, particularly in axillary surgery, is becoming of the utmost importance to minimize the number of unnecessary axillary dissections and their unwanted secondary effects. Therefore, a better understanding of the different factors that predict ATB could be of use to better select the optimal technique for the staging and treatment of the axilla. Currently, ongoing trials are being conducted on tailored axillary surgery [21] and on the avoidance of axillary staging in good prognosis tumors [22]. The results of this study indicate that both approaches would be much more precise if information on molecular factors, such as p53 expression, were to be included in those trials.

A surprising finding was that p53+ was not associated with low ATB in the univariate analysis including all the subtypes, and then showed an association in the multivariate analysis. An explanation is that this finding reflects a previously described statistical phenomenon called the Simpson paradox [23], in which a specific correlation between two outcomes can take an inverse direction when more parameters are combined.

When we analyzed the immunophenotypes separately, the fact that p53+ was predictive of low ATB only in the luminal B-Her2- subtype strongly suggests that nomograms predicting ATB should be built for each immunophenotype. This is also supported by previous findings [24].

The rates of p53+ tumors overall and for each immunophenotype reported in this study are concordant with those described in previously published articles [25]. This homogeneity between different research groups is reassuring as it suggests that evaluation of p53 expression with immunohistochemistry could be comparable between different centers.

The fact that p53+ was associated with low ATB in luminal tumors on the one hand, but with worse overall survival on the other, could be deemed contradictory. We hypothesize that p53+ tumors have a higher propensity for hematological but not for lymphatic dissemination, a feature already described for basal-like breast cancer [26].

A strength of this study is that, to our knowledge, it reports the largest case series with information on p53 expression. Another strength is the long follow-up. However, it also has some limitations. The first is that the number of patients in some subgroup analyses was insufficient. The second is that patient selection was biased to good prognosis immunophenotypes and stages, as we included only patients with upfront surgery. The reason for excluding patients receiving neoadjuvant treatment was that chemotherapy could alter both ATB and p53 expression, as observed by other authors [27]. Finally, another limitation is that we did not include any data on recurrences due to the limited information we retrieved on this subject.

5. Conclusion

In this study, p53+ was associated with low ATB in the luminal B-Her2- immunophenotype. In the era of personalized cancer treatments, this information could be used when individualizing axillary surgery. Further research is needed to determine whether the inclusion of information on p53 expression in ATB predictive nomograms is useful.

Footnotes

Acknowledgments

We thank Gail Craigie for her help in editing this manuscript.

We thank Sergi Mojal for his advice and support regarding the statistical analysis.

The authors would like to thank Josep M. Corominas, former pathologist in the Hospital del Mar, Barcelona, for his advice on the conduct of this study.

The study was funded by the Breast Diseases Unit of Hospital del Mar, Barcelona, Spain.

Conflict of interest

The authors declare that they have no conflicts of interest.

Data availability

The data supporting the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Author contributions

Conception: Pau Nicolau, Maria Vernet-Tomás.

Data interpretation or analysis: All authors.

Preparation of the manuscript: Pau Nicolau, Maria Vernet-Tomás.

Revision for important intellectual content: All authors.

Supervision: Maria Vernet-Tomás.

References

Amin

, Greene

, Edge

, Compton

, Gershenwald

, Brookland

, The Eighth Edition AJCC Cancer Staging Manual: Continuing to build a bridge from a population-based to a more “personalized” approach to cancer staging, CA. Cancer J. Clin. , 67: 93, 2017. doi:10.3322/caac.21388.

Burstein

, Curigliano

, Thürlimann

, Weber

, Poortmans

, Regan

, Panelists of the St Gallen Consensus Conference, 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. Off. J. Eur. Soc. Med. Oncol. , 32: 1216–1235, 2021. doi:10.1016/j.annonc.2021.06.023.

Giacomelli

, Yang

, Lintner

, Mcfarland

, Duby

, Kim

, Mutational processes shape the landscape of TP53 mutations in human cancer, Nat. Genet. , 50: 1381–1387, 2018. doi:10.1038/s41588-018-0204-y.

Koifman

, Shetzer

, Eizenberger

, Solomon

, Rotkopf

, Molchadsky

, A mutant p53-dependent embryonic stem cell gene signature is associated with augmented tumorigenesis of stem cells, Cancer Res , 78: 5833–5847, 2018. doi:10.1158/0008-5472.CAN-18-0805.

Sadighi

, Zokaasadi

, Kasaeian

, Maghsudi

, Jahanzad

, Kamranzadeh Fumani

, The effect of immunohistochemically detected p53 accumulation in prognosis of breast cancer; A retrospective survey of outcome, PLoS One , 12: e0182444, 2017. doi:10.1371/journal.pone.0182444.

, Zhang

, Qiu

, Pang

, Huang

, Zeng

, Comparisons of p53, KI67 and BRCA1 expressions in patients with different molecular subtypes of breast cancer and their relationships with pathology and prognosis, J. BUON. , 24: 2361–2368, 2019. http://www.ncbi.nlm.nih.gov/pubmed/31983107 (accessed February 12, 2023).

Bertheau

, Lehmann-Che

, Varna

, Dumay

, Poirot

, Porcher

, p53 in breast cancer subtypes and new insights into response to chemotherapy, Breast , 22(Suppl 2): S27–S29, 2013. doi:10.1016/j.breast.2013.07.005.

Fragomeni

, Sciallis

, Jeruss

, Molecular subtypes local-regional control of breast cancer, Surg. Oncol. Clin. N. Am. , 27: 95–120, 2018. doi:10.1016/j.soc.2017.08.005.

Giuliano

, Hunt

, Ballman

, Beitsch

, Whitworth

, Blumencranz

, Axillary dissection versus no axillary dissection in women with invasive breast cancer and sentinel node metastasis: a randomized clinical trial, JAMA , 305: 569–575, 2011. doi:10.1001/jama.2011.90.

10.

Chang

, Leung

JWT

, Moy

, Ha

, Moon

, Axillary nodal evaluation in breast cancer: state of the art, Radiology , 295: 500–515, 2020. doi:10.1148/radiol.2020192534.

11.

Huang

, Zheng

, Lin

, Accuracy and utility of preoperative ultrasound-guided axillary lymph node biopsy for invasive breast cancer: A systematic review and meta-analysis, Comput. Intell. Neurosci. , 2022: 3307627, 2022. doi:10.1155/2022/3307627.

12.

Zheng

, Cai

, Song

, Wang

, Han

, Wu

, Positive non-sentinel axillary lymph nodes in breast cancer with 1–2 sentinel lymph node metastases, Medicine (Baltimore) , 97: e13015, 2018. doi:10.1097/MD.0000000000013015.

13.

Madekivi

, Karlsson

, Boström

, Salminen

, Are breast cancer nomograms still valid to predict the need for axillary dissection? Oncology , 99: 397–401, 2021. doi:10.1159/000514616.

14.

Kondov

, Kondov

, Spirovski

, Milenkovikj

, Colanceski

, Petrusevska

, Prognostic Factors on the Positivity for Metastases of the Axillary Lymph Nodes from Primary Breast Cancer, Pril. (Makedonska Akad. Na Nauk. i Umet. Oddelenie Za Med. Nauk.) , 38: 81–90, 2017. doi:10.1515/prilozi-2017-0011.

15.

Shokouh

, Ezatollah

, Barand

, Interrelationships between Ki67, HER2/neu, p53, ER, and PR status and their associations with tumor grade and lymph node involvement in breast carcinoma subtypes: Retrospective-observational analytical study, Medicine (Baltimore) , 94: e1359, 2015. doi:10.1097/MD.0000000000001359.

16.

Vernet-Tomás

, Baños

, Sabadell

, Corominas

J-M

, Mestre-Fusco

, Suárez-Piñera

, p53 expression in breast cancer predicts tumors with low probability of non-sentinel nodes infiltration, J. Obstet. Gynaecol. Res. , 41: 1115–1121, 2015. doi:10.1111/jog.12670.

17.

Nicolau

, Gamero

, Rodríguez-Arana

, Plancarte

, Alcántara

, Carreras

, Imaging and pathology features to predict axillary tumor load in breast cancer, J. Obstet. Gynaecol. Res. , 44: 331–336, 2018. doi:10.1111/jog.13490.

18.

Giuliano

, Ballman

, Mccall

, Beitsch

, Brennan

, Kelemen

, Effect of axillary dissection versus no axillary dissection on 10-year overall survival among women with invasive breast cancer and sentinel node metastasis, JAMA , 318: 918, 2017. doi:10.1001/jama.2017.11470.

19.

NCCN guidelines, version 2.2023, https://www.nccn.org/guidelines/guidelines. Accessed October the 1st 2023.

20.

Z-H

, Hu

P-H

, Tu

J-H

, Yu

N-S

, Luminal B breast cancer: patterns of recurrence and clinical outcome, Oncotarget , 7: 65024–65033, 2016. doi:10.18632/oncotarget.11344.

21.

Weber

, Matrai

, Hayoz

, Tausch

, Henke

, Zwahlen

, Tailored axillary surgery in patients with clinically node-positive breast cancer: Pre-planned feasibility substudy of TAXIS (OPBC-03, SAKK 23/16, IBCSG 57-18, ABCSG-53, GBG 101), Breast , 60: 98–110, 2021. doi:10.1016/j.breast.2021.09.004.

22.

Ahmed

, Beyond the speed of SOUND, Breast Cancer , 27: 793–795, 2020. doi:10.1007/s12282-020-01127-7.

23.

Teuscher

, The quantification of Simpson’s paradox and other contributions to contingency table theory, PLoS One , 17: e0262502, 2022. doi:10.1371/journal.pone.0262502.

24.

Gao

, Luo

, He

, Yang

, Nomogram models for stratified prediction of axillary lymph node metastasis in breast cancer patients (cN0), Front. Endocrinol. (Lausanne) , 13: 967062, 2022. doi:10.3389/fendo.2022.967062.

25.

Duffy

, Synnott

, Crown

, Mutant p53 in breast cancer: potential as a therapeutic target and biomarker, Breast Cancer Res. Treat. , 170: 213–219, 2018. doi:10.1007/s10549-018-4753-7.

26.

Shriver

, Hueman

, Ellsworth

, Molecular signatures of lymph node status by intrinsic subtype: gene expression analysis of primary breast tumors from patients with and without metastatic lymph nodes, J. Exp. Clin. Cancer Res , 33: 116, 2014. doi:10.1186/s13046-014-0116-3.

27.

Peng

J-H