Abstract
Background
The relationship between imaging features and nonsentinel lymph node (NSLN) metastasis is not clear.
Objectives
To determine whether imaging features could predict NSLN metastasis in sentinel lymph node (SLN)-positive breast cancer patients and to provide new clues for avoiding unnecessary axillary lymph node dissection.
Method
171 patients with clinically negative axillary lymph nodes and a pathologically positive SLN were recruited between January 2007 and January 2014. According to the Breast Imaging Reporting and Data System (BI-RADS), the effects of clinicopathological factors, especially imaging features, on NSLN metastases were assessed by univariate and multivariate statistical analyses.
Results
The average number of dissected SLNs was 2.11 (range, 1–6); 56 of the 171 (32.75%) patients exhibited NSLN metastases. In univariate analysis, tumor size, number of positive SLNs, ratio of positive SLNs, mammographic mass margins, ultrasonographic mass margins, and ultrasonographic vascularity were significantly correlated with NSLN involvement. Furthermore, through multivariate analysis, tumor size, number of positive SLNs, mammographic mass margins, and ultrasonographic vascularity were still independent predictors of NSLN involvement. Additionally, in SLN-positive patients, number of positive SLNs and ultrasonographic vascularity could also predict the tumor burden in NSLN.
Conclusions
In addition to tumor size and the number of positive SLNs, mammographic mass margins and ultrasonographic vascularity were also independent predictors of NSLN metastases in SLN-positive patients of breast cancer. The number of positive SLNs and ultrasonographic vascularity could also predict the tumor burden in NSLN.
Keywords: Breast neoplasms, Sentinel lymph node biopsy, Mammography, Ultrasonography, Predictive value of tests
Introduction
Sentinel lymph node biopsy (SLNB) is the standard treatment for breast cancer patients with clinically negative axillary lymph nodes. Although axillary lymph node dissection (ALND) has higher morbidities than SLNB, it is still a standard procedure for patients with a positive SLN, except for patients that fit the ACOSOG Z0011 criteria [1]. However, 40–60% of patients have been found to lack nonsentinel lymph node (NSLN) involvement after ALND, indicating that these patients received unnecessary ALND [2]. Thus, it is quite important to identify patients with SLN involvement only but without NSLN metastases.
Over the past few years, multiple studies have been conducted to identify the clinicopathological factors that could predict NSLN metastases, including the method of detection (hematoxylin and eosin, HE), SLN metastases >2 mm in size, extracapsular extension in the SLN, >1 positive SLN, ratio of positive sentinel nodes >50%, tumor size >2 cm, and lymphovascular invasion in the primary tumor [2, 3]. Furthermore, some mathematical models have been developed to predict the NSLN status [4, 5, 6]. However, the predicted probability of these models was not very high. Therefore, new factors are needed to help improve the predicted probability.
Mammography and ultrasonography images could reflect alterations in breast anatomy and pathology. Previous studies showed that mammographic tumor features were associated with axillary lymph node metastases in breast cancer patients [7, 8, 9, 10], and studies also revealed mammographic spiculation as a positive prognostic marker due to the correlation with low-grade tumors and improved survival [11, 12, 13]. Besides, vessels in tumors which could be detected by color Doppler ultrasound were also correlated with lymph node metastases. Based on the above theory, we supposed that mammographic and ultrasonographic tumor features could also predict NSLN metastases.
In this study, we investigated the imaging features that could predict NSLN metastases in breast cancer patients with clinically negative axillary lymph nodes. Through this study, we expected to provide more information for the prediction of NSLN metastases.
Materials and Methods
Patient History
This was a retrospective review study including 537 breast cancer patients with clinically negative axillary lymph nodes who received SLNB from January 2007 through January 2014. Patients treated with neoadjuvant chemotherapy were excluded. No enlarged lymph nodes were palpable, and absence of any suspicious findings on ultrasound was considered axillary negative [14]. Of these patients with clinically negative axillary lymph nodes, 171 patients with a pathologically positive SLN received ALND. The medical records of these 171 patients were retrieved from our registry, and their clinicopathological characteristics were analyzed. All images for this study were reviewed by 2 breast radiologists in our hospital. The mammographic and ultrasonographic features of the tumors in the current study were evaluated according to BI-RADS criteria [15].
Sentinel Lymph Node Biopsy
SLN mapping was performed using lymphoscintigraphy with methylene blue dye. On the day of operation, technetium-99 sulfur colloid (Beijing Shihong Pharmaceutical Development Center, Beijing, China) was injected intradermally above the tumor, peritumorally, at the areola of the breast, or at the surgical site of the previous biopsy. Scans of the involved breast and axilla were acquired 2 h after tracer injection. Methylene blue dye (Jumpcan, Taixing, China) was injected 15 min before surgery. During surgery, the SLN was localized by using a γ-probe (Neoprobe Corporation, Dublin, OH, USA). The SLN was defined as a blue lymph node and/or a lymph node with an ex vivo radioactive count ≥10% of the ex vivo radioactive count of the hottest lymph node. In addition, any clinically suspicious, palpable lymph nodes were also removed.
Pathological Examinations
All SLNs were subjected to standard frozen section evaluation with HE-stained section. The first SLN defined as the bluest or hottest lymph node was bisected longitudinally and frozen; other SLNs were frozen intactly. Frozen sections were taken with a microtome setting of 4 μm. The remaining nodal tissue was fixed in 10% formalin and embedded in paraffin. After this fixation, serial sections of the SLN were obtained for definitive analysis. The HER2 overexpression status was determined according to the American Society of Clinical Oncology guidelines [16]. According to different combinations of ER, PR, HER2 status, and Ki67, patients were categorized into 4 subgroups as follows: luminal A, luminal B, HER2 overexpression, and triple negative [17].
Statistical Analyses
Statistical analyses were performed using SPSS 19.0 (IBM Corporation, New York, NY, USA). A χ2 test, Fisher's exact test, and analysis of variance were used to compare patients and tumor characteristics. Multivariable analyses were performed on variables with p <0.050 from the univariable analyses by logistic regression. Independent sample t tests were used to compare the factors which influenced the number of NSLN metastases. A p value of 0.050 or smaller was considered statistically significant.
Results
Baseline Characteristics of the Patients
A total of 171 patients with invasive breast cancer had a positive SLN, and all received ALND. The average number of total metastatic lymph nodes was 1.95 (range, 1–11). The median age of these patients was 49 years (range, 24–75 years). The average number of SLNs dissected was 2.11 (range, 1–6). Of these 171 patients, 56 patients (32.75%) had at least 1 positive NSLN. In patients with 1, 2, or more SLN metastases, the average numbers of NSLN metastases were 0.59 ± 1.258 and 1.18 ± 1.749 (p = 0.027). The proportion of patients who had the luminal A subtype was 42.69%, and the proportion of patients with luminal B, HER2 overexpression, and triple-negative subtypes was 37.43, 6.43, and 6.43%, respectively. The subtypes of another 7.02% patients were unclear because of HER2++ on immunohistochemistry without fluorescence in situ hybridization test. Baseline characteristics of these 171 breast cancer patients are shown in Table 1.
Table 1.
Characteristics of the breast cancer patients with positive SLNs and univariate analysis of NSLN metastases
| Variable | Total | Negative | Positive | P |
|---|---|---|---|---|
| (n = 171) | NSLN (n = 115) | NSLN (n = 56) | ||
| Age | 0.639 | |||
| ≤50 years | 99 | 68 | 31 | |
| >50 years | 72 | 47 | 25 | |
| Tumor size | 0.022 | |||
| T1 | 112 | 81 | 30 | |
| T2 | 55 | 31 | 25 | |
| NA | 4 | 3 | 1 | |
| Histological grade | 0.134 | |||
| I | 37 | 29 | 8 | |
| II | 108 | 71 | 37 | |
| III | 24 | 13 | 11 | |
| NA | 2 | 2 | 0 | |
| SLNs | 0.935 | |||
| 1 | 53 | 37 | 16 | |
| 2 | 70 | 46 | 24 | |
| 3 | 31 | 20 | 11 | |
| ≥4 | 17 | 12 | 5 | |
| Positive SLNs | 0.000 | |||
| 1 | 137 | 101 | 36 | |
| ≥2 | 34 | 14 | 20 | |
| Ratio of positive SLN | 0.034 | |||
| ≤50% | 87 | 65 | 22 | |
| >50% | 84 | 50 | 34 | |
| Molecular subtype | 0.425 | |||
| Luminal A | 73 | 44 | 29 | |
| Luminal B | 64 | 45 | 19 | |
| HER2 overexpression | 11 | 9 | 2 | |
| Triple negative | 11 | 8 | 3 | |
| NA | 12 | 9 | 3 |
NA, not available; SLN, sentinel lymph node; HER2, human epidermal growth factor receptor 2; NSLN, non-SLN.
Of these 171 patients, 8 patients received tumor resection biopsy in other hospitals, so imaging data were not available. Tables 2 and 3 list the distribution of mammographic and ultrasonographic appearances among 163 patients with available information, and they also show the correlations between these appearances and lymph node status.
Table 2.
Mammography findings of breast cancer patients with positive SLNs and univariate analysis of NSLN metastases
| Mammography findings | Total | Negative NSLN | Positive NSLN | Ρ |
|---|---|---|---|---|
| (n = 163) | (n = 108) | (n = 55) | ||
| Breast density | ||||
| 1 | 27 | 17 | 10 | 0.527 |
| 2 | 37 | 28 | 9 | |
| 3 | 64 | 42 | 22 | |
| 4 | 35 | 21 | 14 | |
| Result | ||||
| Normal (BI-RADS 1 or 2) | 12 | 11 | 1 | 0.061 |
| Abnormal | 151 | 97 | 54 | |
| Lesion type | ||||
| Mass | 89 | 61 | 28 | 0.192 |
| Mass with calcifications | 49 | 30 | 19 | |
| Calcification only | 11 | 6 | 5 | |
| Other imaging | 2 | 0 | 2 | |
| Mass shape | ||||
| Oval | 15 | 7 | 8 | 0.095 |
| Irregular | 123 | 84 | 39 | |
| Mass margins | ||||
| Circumscribed | 7 | 6 | 1 | |
| Microlobulated | 8 | 6 | 2 | |
| Obscure | 6 | 3 | 3 | |
| Spiculated | 45 | 24 | 21 | 0.030a |
| Indistinct | 72 | 52 | 20 | |
| Calcifications | ||||
| No | 104 | 73 | 31 | 0.125b |
| Typically benign | 1 | 1 | 0 | |
| Amorphous | 5 | 3 | 2 | |
| Coarse heterogeneous | 6 | 3 | 3 | |
| Fine pleomorphic | 43 | 25 | 18 | |
| Fine linear or branching | 4 | 3 | 1 | |
| Associated features | ||||
| Skin retraction | ||||
| Yes | 5 | 3 | 2 | 1.000 |
| No | 158 | 105 | 53 | |
| Nipple retraction | ||||
| Yes | 12 | 7 | 5 | 0.541 |
| No | 151 | 101 | 50 | |
| Skin thickening | ||||
| Yes | 13 | 7 | 6 | 0.366 |
| No | 150 | 101 | 49 | |
| Mass size, cm | 2.056±0.707 | 2.355±0.773 | 0.024 |
Circumscribed, microlobulated, obscure, and indistinct numbers were grouped together to provide adequate numbers for comparison with spiculated numbers.
No calcification and typically benign numbers were grouped together to provide adequate numbers for comparison with other types of calcifications.
Table 3.
Ultrasonographic findings of breast cancer patients with positive SLNs and univariate analysis of NSLN metastases
| Ultrasound findings | Total | Negative NSLN | Positive NSLN | Ρ |
|---|---|---|---|---|
| (n = 163) | (n = 108) | (n = 55) | ||
| Mass shape | ||||
| Round | 1 | 1 | 0 | 0.664a |
| Oval | 4 | 3 | 1 | |
| Irregular | 158 | 104 | 54 | |
| Mass margins | ||||
| Circumscribed | 10 | 8 | 2 | 0.029b |
| Indistinct | 56 | 40 | 16 | |
| Microlobulated | 20 | 17 | 3 | |
| Angular | 60 | 33 | 27 | |
| Spiculated | 17 | 10 | 7 | |
| Orientation | ||||
| Parallel | 104 | 68 | 36 | 0.754 |
| Not parallel | 59 | 40 | 19 | |
| Echo pattern | ||||
| Hypoechoic | 2 | 2 | 0 | |
| Hyperechoic | 1 | 1 | 0 | |
| Heterogeneous | 159 | 104 | 55 | 0.301c |
| Complex cystic and solid | 1 | 1 | 0 | |
| Posterior feature | ||||
| No features | 119 | 80 | 39 | |
| Shadowing | 42 | 26 | 16 | 0.489d |
| Enhancement | 1 | 1 | 0 | |
| Combined pattern | 1 | 1 | 0 | |
| Calcifications | ||||
| No | 114 | 78 | 36 | 0.373 |
| In mass | 49 | 30 | 19 | |
| Vascularity | ||||
| Absent | 36 | 29 | 7 | 0.040e |
| Internal | 91 | 55 | 36 | |
| Rim | 22 | 16 | 6 | |
| Internal + rim | 14 | 8 | 6 | |
| Mass size, cm | 1.934±0.661 | 2.429±0.883 | 0 |
Round and oval numbers were grouped together to provide adequate numbers for comparison with irregular numbers.
Proportions of microlobulated and circumscribed margins compared with proportions of indistinct, angular, or spiculated margins.
Hypoechoic, hyperechoic, and complex cystic and solid were grouped together to provide adequate numbers for comparison with heterogeneous.
Category enhancement, no features, and combined pattern were grouped together to allow adequate group size for comparison.
Internal, rim, and internal + rim were grouped together to provide adequate numbers for comparison.
Factors Predictive of NSLN Involvement
In the univariate analysis of NSLN metastases (Tables 1, 2, 3), larger tumors were more common in the NSLN-positive group than the NSLN-negative group (p = 0.022). The number of positive SLNs in the NSLN-positive group was higher than in the NSLN-negative group (p = 0.000). However, age at diagnosis (p = 0.639), histological grade (p = 0.134), SLN number examined (p = 0.935), and molecular subtypes (p = 0.425) did not show significant differences between these 2 groups. Patients with the luminal A subtype seemed to have a higher risk of NSLN metastases than other subtypes, but there was no statistically significant difference.
On mammography, circumscribed, microlobulated, obscure, and indistinct features were grouped together to provide adequate numbers for comparison with spiculated numbers. We found that masses with spiculated margins were more commonly observed in the NSLN-positive group compared with the NSLN-negative group (p = 0.030); breast density, mass shape, calcifications, and associated features did not reveal statistical differences between these 2 groups (Table 2).
Ultrasonographically, proportions of microlobulated and circumscribed margins were merged to compare them with proportions of indistinct, angular, or spiculated margins. Similarly, masses with indistinct, angular, or spiculated margins were more commonly found in the NSLN-positive group (p = 0.029). Besides this, Doppler color imaging revealed that masses with internal, rim, or internal plus rim vascularity were more commonly found in the NSLN-positive group, and absence of vascularity occurred more often in the NSLN-negative group (p = 0.040) (Table 3).
In multivariate analysis, tumor size (p = 0.027), number of positive SLNs (p = 0.003), mammographic mass margins (p = 0.038), and ultrasonographic vascularity (p = 0.036) were all significant predictive factors of NSLN metastasis, while the ratio of positive SLN and ultrasonographic mass margins showed no statistical differences (Table 4).
Table 4.
Multivariate analysis of factors affecting NSLN metastases
| Variable | OR | 95% CI | p |
|---|---|---|---|
| Tumor size | |||
| T1 | 1 | Reference | |
| T2 | 2.466 | 1.109–5.484 | 0.027 |
| Positive SLNs | |||
| 1 | 1 | Reference | |
| ≥2 | 4.737 | 1.712–13.107 | 0.003 |
| Mammographic mass margins | |||
| Circumscribed + microlobulated + obscure + indistinct | 1 | Reference | |
| Spiculated | 2.37 | 1.049–5.353 | 0.038 |
| Ultrasonographic vascularity | |||
| Absent | 1 | Reference | |
| Internal, rim, or internal + rim | 3.128 | 1.080–9.059 | 0.036 |
| Ratio of positive SLN | |||
| ≤50% | 1 | Reference | |
| >50% | 1.005 | 0.433–2.335 | 0.99 |
| Ultrasonographic mass margins | |||
| Circumscribed + microlobulated | 1 | Reference | |
| Indistinct + angular + speculated | 2.162 | 0.709–6.595 | 0.175 |
OR, odds ratio; CI, confidence interval.
At last, we analyzed which factors could predict the tumor burden in NSLN. Table 5 indicates that ultrasonographic vascularity (p = 0.032) can also predict the tumor burden in NSLN besides the number of positive SLNs (p = 0.027). Tumor size (p = 0.107) and mammographic mass margins (p = 0.390) also tend to predict the NSLN burden, but there was no statistical significance.
Table 5.
Analysis of factors affecting tumor burden in NSLN
| Variable | Cases | Positive NSLN (x ± s), n | p |
|---|---|---|---|
| Tumor size | |||
| T1 | 111 | 0.59±1.351 | |
| T2 | 56 | 0.96±1.464 | 0.107 |
| Positive SLNs | |||
| 1 | 137 | 0.59±1.258 | |
| ≥2 | 34 | 1.18±1.749 | 0.027 |
| Mammographic mass margins | |||
| Circumscribed + microlobulated + obscure + indistinct | 93 | 0.62±1.351 | |
| Spiculated | 45 | 0.82±1.072 | 0.39 |
| Ultrasonographic vascularity | |||
| Absent | 36 | 0.39±0.934 | |
| Internal, rim, or internal + rim | 127 | 0.83±1.505 | 0.032 |
Discussion
SLNB is widely used to assess axillary status in patients with early-stage invasive breast cancer. Generally, ALND is the standard treatment for most patients with a positive SLN. However, the therapeutic efficacy of this procedure is controversial because 40–60% of patients lack NSLN involvement after ALND [2]. In order to avoid unnecessary ALND, many authors have tried to predict in which patients ALND can be safely omitted, and various risk factors for probable NSLN metastases have been identified [3]. In addition, some nomograms have been developed to predict NSLN status [5, 18]. However, the predictive value of these models does not reach the intended high accuracy level. Therefore, new factors and methods are needed to improve the predicted probability.
In this study, we found additional nodal involvement in only 32.75% of patients, which is lower than the reported range of 40–60% [2]. The variation in NSLN involvement can partly be explained by differences in the size of study populations and rigorous use of preoperative axillary ultrasound. Another reason may be that patients with suspicious lymph nodes on ultrasound who received core needle biopsy were excluded in our study.
Similar to previous research, we also revealed that primary tumor size and the number of positive SLNs were predictors of NSLN metastases. Patients with a tumor size >2 cm had a significantly higher risk of NSLN metastases. Besides, the NSLN metastasis risk also increased with increasing number of positive SLNs. However, the ratio of positive SLNs and histological grade did not reach significant predictive values. A previous study found significant differences in different subtypes in terms of NSLN metastases [19]. Zhou et al. [20] reported that patients with luminal A and B subtypes had a higher risk of NSLN metastases than those with triple-negative breast cancer. However, in a study by Gülben et al. [21], luminal A subtype of breast cancer had the lowest risk of NSLN metastases, while patients with the luminal/HER2+ subtype had a higher risk of NSLN metastases than those with the luminal A subtype, and the HER2-overexpressing subtype was significantly associated with NSLN metastases. In our previous studies [19, 20, 21], molecular subtypes were divided based on ER, (and/or) PR, and HER2 status. Ki67 was also used to distinguish luminal A and B (HER2-negative) subtypes in our study which was different from the above research. Our results revealed patients with the luminal A subtype had a higher risk of NSLN metastases than those with other subtypes, but this difference was not statistically significant, which was in agreement with the study by Zhou et al. [20].
Tumor features on mammography were also related to axillary lymph node metastases in breast cancer patients. Regarding the nodal involvement, previous data demonstrated that pleomorphic calcifications [7, 10], overlying skin thickening, or dimpling on mammogram were more frequently present in the lymph node-positive group [7]. Recently, a study showed that higher probability of malignant calcifications was associated with more lymph node involvement compared with patients with spiculated masses, and a nonstellate mass had a lower risk of lymphatic metastases [8]. However, another study showed that tumors appearing as calcifications were more often ALN negative compared with tumors appearing as a distinct mass; tumors presenting as an ill-defined mass and spiculated appearance tended to be more often ALN positive than tumors which presented as a distinct mass [9]. In our present study, spiculated mammographic margins had a significantly higher risk of NSLN metastases, which was basically consistent with the above 2 studies. Besides, patients with malignant calcification also had a tendency to be more prone to lymph node metastases but without statistical significance. Why was there a relationship between spiculated mammographic margins and NSLN metastases? As previously reported, loss of adhesion factors on carcinoma cells was considered to play a role in the histological appearance of invasive carcinoma as loosely dispersed linear columns of cells and a typical discrete mass [22, 23]. This more diffuse infiltrative pattern may explain some of the typical imaging appearances of tumors, such as spiculation and distortion [22]. Thus, based on this theory, we considered that loss of adhesion factors may increase potential lymph node metastases in spiculated mammographic tumors. As to why mammographic spiculation may be a good prognostic marker, another study showed that the spiculation feature was associated with high ER and PR expression [24]. Therefore, we thought that due to the better treatment, especially endocrine therapy, spiculated mammographic tumors could still achieve a good prognosis, although lymphatic metastases are more likely to occur.
Regarding tumor blood supply, it is well established that microvessel formation in tumors is closely correlated with lymph node metastases. Several studies confirmed this theory by Doppler sonography [25, 26, 27]. Tumor vascularity revealed by power Doppler sonography was strongly correlated with the detection of lymph node involvement and lymphatic vascular invasion, with sensitivities of 93 and 90%, respectively [25], especially for small tumors in predicting lymph node metastasis [26, 27]. In our study, ultrasonographic vascularity was an independent predictive factor of NSLN involvement and NSLN tumor burden, which was in agreement with previous studies.
In conclusion, our study indicated that tumor size, number of positive SLNs, mammographic mass margins, and ultrasonographic vascularity were independent predictors of NSLN metastases in SLN-positive breast cancer. Furthermore, the number of positive SLNs and ultrasonographic vascularity could also predict the tumor burden in NSLN. These results support the development of new nomograms including mammographic mass margins and ultrasonographic vascularity to increase predictive accuracy.
The limitations of this study include the small sample size and the retrospective design of the study. However, the conclusions of the present study are useful when planning additional studies. Future prospective studies in a larger patient cohort are needed to validate our findings.
Statement of Ethics
The protocol was approved by the Hospital Human Ethical Committee (201521). Informed consent had been obtained from all patients before surgery and specimen sampling were conducted.
Disclosure Statement
The authors have no conflicts to disclose.
Funding Sources
The study was supported by the key project of research and development plan of Shandong Province (No. 2018GSF118125) and Yantai City (No. 2017YD007).
Acknowledgments
The authors thank all staff members of the Departments of Imaging and Pathology who were involved in diagnosing our patients.
References
- 1.Giuliano AE, Hunt KK, Ballman KV, Beitsch PD, Whitworth PW, Blumencranz PW, et al. Axillary dissection vs no axillary dissection in women with invasive breast cancer and sentinel node metastasis: a randomized clinical trial. JAMA. 2011 Feb;305((6)):569–75. doi: 10.1001/jama.2011.90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.van la Parra RF, Peer PG, Ernst MF, Bosscha K. Meta-analysis of predictive factors for non-sentinel lymph node metastases in breast cancer patients with a positive SLN. Eur J Surg Oncol. 2011 Apr;37((4)):290–9. doi: 10.1016/j.ejso.2011.01.006. [DOI] [PubMed] [Google Scholar]
- 3.Moghaddam Y, Falzon M, Fulford L, Williams NR, Keshtgar MR. Comparison of three mathematical models for predicting the risk of additional axillary nodal metastases after positive sentinel lymph node biopsy in early breast cancer. Br J Surg. 2010 Nov;97((11)):1646–52. doi: 10.1002/bjs.7181. [DOI] [PubMed] [Google Scholar]
- 4.Canavese G, Bruzzi P, Catturich A, Vecchio C, Tomei D, Del Mastro L, et al. A risk score model predictive of the presence of additional disease in the axilla in early-breast cancer patients with one or two metastatic sentinel lymph nodes. Eur J Surg Oncol. 2014 Jul;40((7)):835–42. doi: 10.1016/j.ejso.2014.03.005. [DOI] [PubMed] [Google Scholar]
- 5.Koca B, Kuru B, Ozen N, Yoruker S, Bek Y. A breast cancer nomogram for prediction of non-sentinel node metastasis - validation of fourteen existing models. Asian Pac J Cancer Prev. 2014;15((3)):1481–8. doi: 10.7314/apjcp.2014.15.3.1481. [DOI] [PubMed] [Google Scholar]
- 6.Kuo YL, Chen WC, Yao WJ, Cheng L, Hsu HP, Lai HW, et al. Validation of Memorial Sloan-Kettering Cancer Center nomogram for prediction of non-sentinel lymph node metastasis in sentinel lymph node positive breast cancer patients an international comparison. Int J Surg. 2013;11((7)):538–43. doi: 10.1016/j.ijsu.2013.05.005. [DOI] [PubMed] [Google Scholar]
- 7.Kollias J, Gill PG, Beamond B, Rossi H, Langlois S, Vernon-Roberts E. Clinical and radiological predictors of complete excision in breast-conserving surgery for primary breast cancer. Aust N Z J Surg. 1998 Oct;68((10)):702–6. doi: 10.1111/j.1445-2197.1998.tb04655.x. [DOI] [PubMed] [Google Scholar]
- 8.Li JN, Xu J, Wang J, Qing C, Zhao YM, Liu PF. Correlation between mammograghic findings and clinical/ pathologic features in women with small invasive breast carcinomas. Asian Pac J Cancer Prev. 2014;15((24)):10643–6. doi: 10.7314/apjcp.2014.15.24.10643. [DOI] [PubMed] [Google Scholar]
- 9.Sartor H, Borgquist S, Hartman L, Olsson Å, Jawdat F, Zackrisson S. Do mammographic tumor features in breast cancer relate to breast density and invasiveness, tumor size, and axillary lymph node involvement? Acta Radiol. 2015 May;56((5)):536–44. doi: 10.1177/0284185114532081. [DOI] [PubMed] [Google Scholar]
- 10.Sun SS, Zhang B, Zhao HM, Cao XC. Association between mammographic features and clinicopathological characteristics in invasive ductal carcinoma of breast cancer. Mol Clin Oncol. 2014 Jul;2((4)):623–9. doi: 10.3892/mco.2014.297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Alexander MC, Yankaskas BC, Biesemier KW. Association of stellate mammographic pattern with survival in small invasive breast tumors. AJR Am J Roentgenol. 2006 Jul;187((1)):29–37. doi: 10.2214/AJR.04.0582. [DOI] [PubMed] [Google Scholar]
- 12.Evans AJ, Pinder SE, James JJ, Ellis IO, Cornford E. Is mammographic spiculation an independent, good prognostic factor in screening-detected invasive breast cancer? AJR Am J Roentgenol. 2006 Nov;187((5)):1377–80. doi: 10.2214/AJR.05.0725. [DOI] [PubMed] [Google Scholar]
- 13.Tabar L, Tony Chen HH, Amy Yen MF, Tot T, Tung TH, Chen LS, et al. Mammographic tumor features can predict long-term outcomes reliably in women with 1-14-mm invasive breast carcinoma. Cancer. 2004 Oct;101((8)):1745–59. doi: 10.1002/cncr.20582. [DOI] [PubMed] [Google Scholar]
- 14.Reynders A, Brouckaert O, Smeets A, Laenen A, Yoshihara E, Persyn F, et al. Prediction of non-sentinel lymph node involvement in breast cancer patients with a positive sentinel lymph node. Breast. 2014 Aug;23((4)):453–9. doi: 10.1016/j.breast.2014.03.009. [DOI] [PubMed] [Google Scholar]
- 15.D'Orsi CJ, Sickles EA, Mendelson EB, Morris EA, et al. ACR BI-RADS® Atlas, Breast Imaging Reporting and Data System. Reston, VA, American College of Radiology, 2013 [Google Scholar]
- 16.Wolff AC, Hammond ME, Hicks DG, Dowsett M, McShane LM, Allison KH, et al. American Society of Clinical Oncology. College of American Pathologists Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American Society of Clinical Oncology/College of American Pathologists clinical practice guideline update. Arch Pathol Lab Med. 2014 Feb;138((2)):241–56. doi: 10.5858/arpa.2013-0953-SA. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Goldhirsch A, Wood WC, Coates AS, Gelber RD, Thürlimann B, Senn HJ, Panel members Strategies for subtypes—dealing with the diversity of breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2011. Ann Oncol. 2011 Aug;22((8)):1736–47. doi: 10.1093/annonc/mdr304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Al-Masri M, Darwazeh G, El-Ghanem M, Hamdan B, Sughayer M. Assessment of an existing and modified model for predicting non sentinel lymph node metastasis in breast cancer patients with positive sentinel node biopsy. Gulf J Oncolog. 2013 Jan;1((13)):15–22. [PubMed] [Google Scholar]
- 19.Reyal F, Belichard C, Rouzier R, de Gournay E, Senechal C, Bidard FC, et al. Non-sentinel lymph node metastasis prediction in breast cancer with metastatic sentinel lymph node: impact of molecular subtypes classification. PLoS One. 2012;7((10)):e47390. doi: 10.1371/journal.pone.0047390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zhou W, He Z, Xue J, Wang M, Zha X, Ling L, et al. Molecular subtype classification is a determinant of non-sentinel lymph node metastasis in breast cancer patients with positive sentinel lymph nodes. PLoS One. 2012;7((4)):e35881. doi: 10.1371/journal.pone.0035881. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Gülben K, Berberoğlu U, Aydoğan O, Kınaş V. Subtype is a predictive factor of nonsentinel lymph node involvement in sentinel node-positive breast cancer patients. J Breast Cancer. 2014 Dec;17((4)):370–5. doi: 10.4048/jbc.2014.17.4.370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Doyle S, Evans AJ, Rakha EA, Green AR, Ellis IO. Influence of E-cadherin expression on the mammographic appearance of invasive nonlobular breast carcinoma detected at screening. Radiology. 2009 Oct;253((1)):51–5. doi: 10.1148/radiol.2531090045. [DOI] [PubMed] [Google Scholar]
- 23.Gastl G, Spizzo G, Obrist P, Dünser M, Mikuz G. Ep-CAM overexpression in breast cancer as a predictor of survival. Lancet. 2000 Dec;356((9246)):1981–2. doi: 10.1016/S0140-6736(00)03312-2. [DOI] [PubMed] [Google Scholar]
- 24.Moriuchi H, Yamaguchi J, Hayashi H, Ohtani H, Shimokawa I, Abiru H, et al. Cancer Cell Interaction with Adipose Tissue: Correlation with the Finding of Spiculation at Mammography. Radiology. 2016 Apr;279((1)):56–64. doi: 10.1148/radiol.2015142191. [DOI] [PubMed] [Google Scholar]
- 25.Mehta TS, Raza S. Power Doppler sonography of breast cancer: does vascularity correlate with node status or lymphatic vascular invasion? AJR Am J Roentgenol. 1999 Aug;173((2)):303–7. doi: 10.2214/ajr.173.2.10430124. [DOI] [PubMed] [Google Scholar]
- 26.Lee WJ, Chu JS, Huang CS, Chang MF, Chang KJ, Chen KM. Breast cancer vascularity: color Doppler sonography and histopathology study. Breast Cancer Res Treat. 1996;37((3)):291–8. doi: 10.1007/BF01806510. [DOI] [PubMed] [Google Scholar]
- 27.Okuyama N, Murakuni H, Ogata H. The use of Doppler ultrasound in evaluation of breast cancer metastasis to axillary lymph nodes. Oncol Rep. 2004 Feb;11((2)):389–93. [PubMed] [Google Scholar]