- Email: [email protected]
Sensitivity, Specificity and Accuracy of Ultrasound in Diagnosis of Breast Cancer Metastasis to the Axillary Lymph Nodes in Chinese Patients
Ultrasound in Med. & Biol., Vol. 41, No. 7, pp. 1835–1841, 2015 Copyright Ó 2015 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/$ - see front matter
http://dx.doi.org/10.1016/j.ultrasmedbio.2015.03.024
d
Original Contribution SENSITIVITY, SPECIFICITY AND ACCURACY OF ULTRASOUND IN DIAGNOSIS OF BREAST CANCER METASTASIS TO THE AXILLARY LYMPH NODES IN CHINESE PATIENTS YAN-NA ZHANG,* CHANG-JUN WANG,* YING XU,* QING-LI ZHU,y YI-DONG ZHOU,* JING ZHANG,y FENG MAO,* YU-XIN JIANG,y and QIANG SUN* * Department of Breast Surgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, Beijing, China; and y Department of Ultrasound, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, Beijing, China (Received 23 May 2014; revised 5 March 2015; in final form 27 March 2015)
Abstract—The use of ultrasound in the diagnosis of axillary lymph node metastases from breast cancer in a Chinese population was investigated. Data for 1,049 with breast cancer were retrospectively collected. All patients had undergone pre-operative axillary ultrasound and then axillary lymph node dissection. The sensitivity, specificity and accuracy of axillary ultrasound in this cohort were 69.4%, 81.8% and 77.0%, respectively. The overall false-negative rate of ultrasound images was 30.6% (123/402). False-negative ultrasound rates for pathologic N1, N2 and N3 patients were 46.2%, 21.8% and 9.3%, respectively. In patients with stage T1 disease and fewer than three metastatic lymph nodes, the false-negative ultrasound rate was 52.2% (47/90). Moreover, breast cancer patients with a false-negative axillary ultrasound were more likely to have a large tumor (p , 0.001) and high tumor grade (p 5 0.009). However, there were no statistically significant differences between accuracy of axillary ultrasound and age of patients or experiences of ultrasound practitioners. In conclusion, the sensitivity, specificity and accuracy of ultrasound in the diagnosis of breast cancer metastasis to the axillary lymph nodes in Chinese patients were assessed. These data could help us to carefully use axillary ultrasound to diagnose and predict breast cancer axillary lymph node metastasis. (E-mail: [email protected] or [email protected]) Ó 2015 World Federation for Ultrasound in Medicine & Biology. Key Words: Breast cancer, Ultrasound diagnosis, Lymph node metastasis, Sensitivity, Specificity, Accuracy.
(such as patient age, hormone receptor status, tumor stage, tumor size and presence of lymph node and distant metastases) (Goldhirsch et al. 2013). Among these, axillary lymph node status is one of the most important prognostic factors in breast cancer. Thus, pre-operative axillary ultrasound is crucial to staging and management of breast cancer in many institutions (Gentilini and Veronesi 2012; Vaidya et al. 1996). For example, accurate evaluation of axillary lymph node status could avoid an unnecessary sentinel lymph node biopsy (SLNB) or even the maximally invasive radical resection (Moorman et al. 2014). However, data on false-negative axillary ultrasound in Chinese women are sparse. In addition, a recent large-scale randomized, controlled, multicentric clinical trial (SOUND [Sentinel Node versus Observation after Axillary Ultrasound] trial) of patients with negative axillary ultrasound compared the clinical outcome of SLNB with that of observation only (Gentilini and Veronesi 2012). In this trial,
INTRODUCTION Breast cancer is one of the most significant health problems in the world and is the second leading cause of cancer-related death among women in the United States (American Cancer Society 2014). During the past three decades, both advancements in screening for early stages of breast cancer and improvement in treatment options have significantly reduced breast cancer mortality and improved quality of life, but many patients still develop metastatic disease and consequently die (Miao et al. 2014). A number of factors can affect the outcome of breast cancer, for example, clinicopathologic features
Address correspondence to: Qiang Sun, Department of Breast Surgery, Peking Union Medical College Hospital, 1 Shuaifuyuan, Wangfujing, Beijing 100730, China. E-mail: [email protected] or [email protected] Conflicts of interest: The authors declare that there is no conflict of interest in this study. 1835
1836
Ultrasound in Medicine and Biology
Table 1. Clinicopathologic characteristics of breast cancer patients Variable
Number (%) of cases
Age ,65 y $65 y Tumor size (cm) #2 .2 Axillary lymph node metastasis N1 N2 N3 ER/PR expression Positive Negative HER-2 expression Low High Triple negative Yes No p53 expression Negative Positive Ki-67 expression Negative Positive Histologic grade I II III
945 (90.1) 104 (9.9) 638 (60.8) 411 (39.2) 195 (18.6) 110 (10.5) 97 (9.2) 707 (67.4) 342 (32.6) 613 (58.4) 436 (41.6) 156 (14.9) 893 (85.1) 696 (66.3) 353 (33.7) 399 (38.0) 650 (62.0) 213 (20.3) 467 (44.5) 369 (35.2)
ER/PR 5 estrogen receptor/progesterone receptor.
physicians used the inclusion criteria of negative axillary ultrasound and clinically negative axillary lymph nodes
Volume 41, Number 7, 2015
to rule out an evident or suspicious lymph node metastasis of breast cancer and, thus, perform conservative surgery instead of maximally invasive radical resection. However, a question was immediately raised as to whether ultrasound is useful in excluding axillary lymph node metastases from breast cancer. The reliability of the ongoing SOUND trial also needs to be considered with respect to the false-negative rate of axillary ultrasound. Indeed, ultrasound imaging is widely used to evaluate the breast and axillae pre-operatively and can guide physicians in performance of core needle biopsy. Several previous studies (Motomura et al. 2001; Vaidya et al. 1996; Verbanck et al. 1997; Yang et al. 1996) on the accuracy of ultrasound in breast cancer staging reported that high-resolution ultrasound had broad ranges of sensitivity, specificity and overall accuracy of 50%–92%, 90%–100% and 76%–92%, respectively. Their finding indicates that ultrasound has moderate sensitivity, but high specificity in the detection of axillary metastases. However, other studies have reported on the impact of false-negative results in preoperative axillary ultrasound (Choi et al. 2012; Johnson et al. 2011; Neal et al. 2010; Park et al. 2013) on prognosis and choice of surgery. Thus, in this study, we assessed the diagnostic value of ultrasound imaging in the detection of breast cancer metastases to the axillary lymph nodes. Our aim was to investigate whether there is an identifiable subset of breast cancers associated with a higher risk of false-negative axillary ultrasound results.
Fig. 1. Ultrasound images of normal and metastatic axillary lymph nodes. (a) Normal lymph node. (b) Blood flow in normal lymph node. (c) Metastatic lymph node (long-to-short axis ratio ,2, compression of the fatty hilum, cortical thickening and asymmetry). (d) Blood flow in metastatic lymph node (rich blood flow signal).
US diagnoses of breast cancer metastasis to axillary lymph nodes d Y.-N. ZHANG et al.
1837
chemotherapy. All enrolled patients underwent preoperative ultrasound of the axilla on both sides. The ultrasound was performed by three advanced practitioners with more than 4 y of breast imaging experience and three senior practitioners with less than 4 y of breast imaging experience. Clinicopathologic features of these patients were collected from their medical records and included patient age; TNM (tumor/node/metastasis) classification; tumor size, axillary lymph node status, histologic grade and expression of estrogen receptor (ER), progesterone receptor (PR), Her-2, p53 and Ki-67. Pre-operative ultrasound images and post-operative pathologic results were also documented for each patient. This study was approved by the Peking Union Medical College Hospital, and each participant signed an informed consent form to participate in this study. Ultrasound images Ultrasound (ACUSON S2000, Siemens Medical Solutions, Mountain View, CA, USA, with an 18L6 probe; and LOGIQ 7, GE Ultra Solutions, Ontario, CA, USA, with a 10L probe) was used to scan both axillae in patients. Both B-mode (the fundamental mode was used routinely, the harmonic mode was used if necessary) and color Doppler flow imaging (CDFI) were used to characterize lymph node metastases. Any abnormal lymph nodes were noted as suspicious. The criteria used to diagnose abnormal lymph nodes under ultrasound included: long axis-to-short axis ratio ,2, unclear margin or irregular shape, hypo-echoic mass, compression or disappearance of the fatty hilum, cortical thickening or asymmetry, merged lymph nodes and rich blood flow signal. Any lymph nodes that met three or more of the aforementioned criteria were considered to have a tumor metastasis. Fig. 2. Ultrasound images of false negative and positive axillary lymph nodes. (a) False-negative lymph node (a little cortical thickening, considered to be normal lymph node by ultrasound but metastatic on pathology). (b) False-positive lymph node (long-to-short axis ratio ,2, compression of the fatty hilum, cortical thickening, considered to have a tumor metastasis by ultrasound but normal on pathology). (c) Blood flow in falsepositive lymph node.
METHODS Patients In this study, we retrospectively enrolled 1,049 consecutive primary breast cancer patients from Peking Union Medical College Hospital between January 2010 and December 2011. All of these patients had undergone axillary lymph node dissection; however, exclusion criteria were metastasis of other primary cancers to the mammary glands and treatment with neo-adjuvant
Statistical analysis Sensitivity was calculated as the proportion of patients with pathologic metastatic lymph nodes who had an abnormal axillary lymph node in ultrasound images. Specificity was calculated as the proportion of patients with no pathologic metastatic lymph node who had a normal lymph node in ultrasound images. Accuracy was calculated as the proportion of patients whose axillary lymph node status was correctly predicted by ultrasound. The overall sensitivity, specificity and accuracy of ultrasound were analyzed using multinomial logistic regression and binary logistic regression tests. The difference was considered statistically significant when the p value was #0.05. All statistical analyses were performed using SPSS software (Version 19.0, SPSS, Chicago, IL, USA) by a statistician who was blinded to the pathologic records.
1838
Ultrasound in Medicine and Biology
Volume 41, Number 7, 2015
Table 2. Comparison of ultrasound results with lymph node involvement (diagnosed pathologically) in breast cancer patients False-negative True-negative False-positive True-positive p value* Accuracy (%) Sensitivity (%) Specificity (%) Total (n 5 1,049) Age ,65 y (n 5 945) $65 y (n 5 104) Tumor size #2 cm (n 5 638) .2 cm (n 5 411) Grade Low (n 5 213) High (n 5 836) ER/PR expression Positive (n 5 707) Negative (n 5 342) HER-2 expression Low (n 5 613) High (n 5 436) p53 expression Negative (n 5 696) Positive (n 5 353) Ki-67 index #15% .15% Practitioner Experienced (n 5 649) Inexperienced (n 5 400)
123
529
118
279
77.0
69.4
81.8
112 11
472 57
105 13
256 23
0.445
77.0 76.9
69.6 67.6
81.8 81.4
70 53
391 138
60 58
117 162
,0.001
77.3 77.0
70.7 69.1
81.3 81.9
53 16
138 139
58 33
162 25
,0.001
76.9 79.6
83.3 62.6
73.5 86.7
107 100
390 378
85 59
254 170
0.009
73.0 77.0
75.3 61.0
70.4 80.8
23 83
151 313
59 65
109 152
0.268
77.0 77.5
70.4 63.0
82.1 86.5
85 38
72 46
364 165
175 104
0.705
76.0 75.9
82.6 64.7
71.9 82.8
44 79
40 78
237 292
78 201
0.07
78.7 77.4
76.0 67.3
80.3 83.5
83 40
336 193
62 56
168 111
0.251
76.2 78.9
73.2 63.9
78.2 85.6
ER/PR 5 estrogen receptor/progesterone receptor. * p Value refers to the comparison of accuracy in each study group (Age, Tumor size, Tumor grade, etc.).
Table 3. Comparison of clinical and pathologic influencing factors between patients with false-predictive US results and those with true-predictive axillary ultrasound results False-negative US (N 5 123)
Age ,65 y $65 y Tumor size #2 cm .2 cm Grade I II or III ER/PR expression Negative Positive HER-2 expression Low High p53 expression Negative Positive Ki-67 index #15% .15% Practitioner Experienced Inexperienced
True-negative US (N 5 529)
N
%
N
%
112 11
19.2 16.2
472 57
81.2 83.8
70 53
15.2 27.7
391 138
84.8 72.3
6 108
7.5 21.7
74 390
92.5 78.3
22 101
13.8 21.1
138 377
86.2 78.9
83 40
22.0 15.6
295 216
78.0 84.4
85 38
19.5 18.5
352 167
80.5 81.5
44 79
17.0 21.2
215 293
83.0 78.8
83 40
19.8 17.2
336 193
80.2 82.8
p value*
Odds ratio (95% CI)
0.308
1.515 (0.68–3.36)
0.000
0.389 (0.24–0.60)
0.009
0.329 (0.14–0.75)
0.464
0.807 (0.45–1.43)
0.165
1.415 (0.86–2.30)
0.311
1.279 (0.79–2.05)
0.234
0.754 (0.47–1.20)
0.860
1.039 (0.68–1.58)
CI 5 confidence interval; ER/PR 5 estrogen receptor/progesterone receptor; US 5 ultrasound. * p Value refers to the comparison of false-negative and true-negative results in each study group (Age, Tumor size, Tumor grade, etc.).
US diagnoses of breast cancer metastasis to axillary lymph nodes d Y.-N. ZHANG et al.
1839
Table 4. Association of ultrasound results with pathologic axillary lymph node stages False-negative US (n 5 123)
True-positive US (n 5 279)
Pathologic axillary lymph node stage
N
%
N
%
p value*
Odds ratio (95% CI)
N1 (n 5 195) N2 (n 5 110) N3 (n 5 97)
90 24 9
46.2 21.8 9.3
105 86 88
53.8 78.2 90.7
0.002 0.019
0.196 (0.07–0.54) 0.346 (0.14–0.83)
CI 5 confidence interval; US 5 ultrasound. * p Value refers to the comparison of false-negative and true-positive results in N1, N2 and N3.
RESULTS Characteristics of patients In this study, we analyzed ultrasound assessment of axillary lymph node metastases in 1,049 breast cancer patients (Table 1). The mean age of this cohort was 50.3 y (range: 21–88 y). Four hundred and two patients (38.3%) had axillary lymph node metastases, whereas 638 (60.8%) patients had stage T1 disease and 156 (14.9%) had triple-negative tumors (Table 1). The reason we chose 65 y as the cutoff age is because patients older than 65 y are considered to be senile breast cancer patients. A tumor size of 2 cm was used as a cutoff point because this size is considered stage T1 and is one of the inclusion criteria for the SOUND trial. Ultrasound image analysis of the axilla lymph nodes Three hundred ninety-seven (37.8%) of the 1,049 patients were found to have tumor axillary lymph node metastases by ultrasound imaging analysis. In Figure 1 are representative images of the normal and metastatic lymph nodes of these patients. However, pathology data on surgically resected axillary lymph nodes indicated that 402 cases (38.3%) were positive for breast cancer metastasis to the axillary lymph nodes. We then compared the ultrasound images with the data on surgically resected lymph nodes and found that 123 patients were proven to have false-negative (pathologic metastatic lymph nodes with normal ultrasound images) results. In contrast, 118 cases were found to be false positive as no metastatic axillary lymph node was discovered after surgery. In Figure 2 are representative images of falsenegative and false-positive lymph nodes from these
patients. The sensitivity, specificity and overall accuracy of axillary ultrasound were 69.4%, 81.8% and 77.0%, respectively. The false-negative rate (negative ultrasound images but positive pathology) was 30.6% (123/402), and the false-positive rate (positive ultrasound images but negative pathology) was 18.2% (118/647) in these patients. We then compared ultrasound rates for falsepositive or -negative axillary lymph nodes with other clinicopathologic data of the patients. Ultrasound accuracy differed with tumor size, tumor grade and ER/PR expression (Table 2). Patients with false-negative axillary ultrasound results were more likely to have larger (p , 0.001) and higher-grade (p 5 0.009) tumors (Table 3). However, expression of p53 and Ki-67 was not significantly associated with false-negative axillary ultrasound results (Table 3). Moreover, the accuracy of axillary ultrasound was not statistically associated with the patient’s age or the practitioner’s experience (Table 3). Table 4 compares the false-negative rates among various pathologic stages; false-negative ultrasound rates for stage N1, N2 and N3 patients were 46.2%, 21.8% and 9.3%, respectively. The differences between N1 and N3 (p 5 0.002) and between N2 and N3 (p 5 0.019) are statistically significant (correlation coefficient 5 0.246). This indicates that advanced N stage is associated with a lower false-negative rate. In addition, we compared falsenegative rates between stage T1 patients with one or two metastatic lymph nodes and those with three or more metastatic lymph nodes according to the approved conclusion of the American College of Surgeons Oncology Group Z0011 trial (Giuliano et al. 2011) (Table 5). There was no difference in the false-negative rate between the two groups of patients (p 5 0.117). In
Table 5. Association of ultrasound results with numbers of pathologic axillary lymph nodes (T1 tumor) False-negative US (n 5 70)
True-positive US (n 5 117)
Number of pathologic lymph nodes
N
%
N
%
p value*
Odds ratio (95% CI)
1 or 2 (n 5 90) $3 (n 5 97)
47 23
52.2 23.7
43 74
47.8 76.3
0.117
0.535 (0.24–1.16)
CI 5 confidence interval; US 5 ultrasound. * p Value refers to the comparison of false-negative and true-positive results in the study group.
1840
Ultrasound in Medicine and Biology
patients with stage T1 disease and one or two metastatic lymph nodes, the false-negative ultrasound rate was 52.2% (47/90). DISCUSSION Clinical assessment of breast cancer stage is one of the most important pre-surgery procedures in defining the surgical resection of tumor and metastasized lymph nodes (i.e., radical versus conservative surgery). In this regard, pre-operative axillary ultrasound is crucial to staging and selection of management options for breast cancer patients. Accurate axillary evaluation may avoid an unnecessary SLNB and maximally invasive radical resection. In this retrospective study, we evaluated the diagnostic value of ultrasound in detection of axillary lymph node metastases of breast cancer and the clinicopathologic factors associated with the false-negative ultrasound results in Chinese patients. Our data revealed that ultrasound had 69.4% sensitivity, 81.8% specificity and 77.0% overall accuracy in the diagnosis of breast cancer axillary lymph metastases in this Chinese cohort. These data are consistent with previously reported results in Western countries (Motomura et al. 2001; Vaidya et al. 1996; Verbanck et al. 1997; Yang et al. 1996). The falsenegative rate of axillary ultrasound was 30.6%, which was within the ranges of previous studies in Western countries (Cools-Lartigue and Meterissian 2012; Vaidya et al. 1996; van Rijk et al. 2006; Yang et al. 1996). We also found that the false-negative rate of axillary ultrasound was associated with tumor size, histologic grade and ER/PR expression. We concluded that axillary ultrasound has diagnostic value in detecting positive axillary lymph nodes in Chinese women. Thus, improvement of ultrasound sensitivity, specificity and accuracy or use of other techniques is warranted. There is an urgent demand for non-invasive modalities to differentiate stage N1 from advanced N stage patients. As an example, the Z0011 trial proved that patients with fewer than three positive SLNs could be spared from axillary lymph node dissection. Additionally, the ongoing SOUND trial, which represents a step forward in the conservative approach to breast cancer, adopted negative ultrasound of the axilla as an inclusion criterion and assigned patients randomly to either an observation group or an SLN 6 axillary dissection group. Thus, in the future, ultrasound may serve as an alternative tool for detection of axillary metastases in clinical practice. However, the contemporary false-negative rate of N2/3 disease has scarcely been reported. For example, Park et al. (2013) reported that the actual false-negative rate of ultrasound for patients with N2 and N3 disease was 66.7% (6/9), whereas Neal et al. (2010) suggested that pre-operative axillary ultrasound excluded 96% of N2
Volume 41, Number 7, 2015
and N3 invasive ductal metastases. Our study revealed that up to 15.9% (33/207) of patients with N2 and N3 metastases had false-negative ultrasound results. Thus, use of axillary ultrasound alone is considered inadequate to detect N2 and N3 axillary metastases given the unacceptably high false-negative rate (approximately 16%). Furthermore, expression of ER/PR, Her2 and Ki-67 protein and other clinicopathologic parameters are crucial prognostic indicators in breast cancer patients. However, correlation between these parameters and false-negative ultrasound results remains controversial. A previous retrospective study by Park et al. (2013) indicated that ER and PR positivity were strongly correlated with the high false-negative rate. In contrast, Johnson et al. (2011) reported that there were no statistically significant differences in false-negative ultrasound results with respect to ER/PR status. To our knowledge, no study has been conducted on the relationship of HER-2, p53 and Ki-67 with the false-negative ultrasound rate. On the basis of this study, which enrolled more than 1,000 patients, we believe that hormonal receptor status has an impact on the accuracy of ultrasound. This may be explained by the fact that ER/PR-positive breast cancer usually involves large tumors without clinically significant axillary metastases; thus, the advanced T stage of the primary breast lesion may contribute to a high false-negative rate. Therefore, we should pay more attention to those ER/PR-positive breast cancer patients with high histologic grade and false-negative axillary ultrasound results. Additionally, our study also indicated that expression of p53 and Ki67 was not significantly associated with the false-negative ultrasound results. The present study does have some limitations. For example, this retrospective study may compromise the internal and external validation control numbers compared with a prospective study. In the data analysis, we did not use the exact number of positive axillary lymph nodes in analyzing the association with false-negative ultrasound results. Thus, any future study should focus on a combination of ultrasound images and ultrasound-guided fine needle aspiration to increase the sensitivity, specificity and accuracy of axillary ultrasound. Given our results and those of others, we advocate the careful use of axillary ultrasound to rule out axillary lymph node metastases, especially with larger and/or highergrade tumors. More studies on the accuracy of breast cancer axillary lymph node ultrasound are warranted.
REFERENCES American Cancer Society. Cancer facts & figures 2014. Atlanta, GA: Author; 2014. Choi JS, Kim MJ, Moon HJ, Kim EK, Yoon JH. False negative results of pre-operative axillary ultrasound in patients with invasive breast
US diagnoses of breast cancer metastasis to axillary lymph nodes d Y.-N. ZHANG et al. cancer: Correlations with clinicopathologic findings. Ultrasound Med Biol 2012;38:1881–1886. Cools-Lartigue J, Meterissian S. Accuracy of axillary ultrasound in the diagnosis of nodal metastasis in invasive breast cancer: A review. World J Surg 2012;36:46–54. Gentilini O, Veronesi U. Abandoning sentinel lymph node biopsy in early breast cancer? A new trial in progress at the European Institute of Oncology of Milan (SOUND: Sentinel node vs Observation after axillary UltraSouND). Breast 2012;21:678–681. Giuliano AE, Hunt KK, Ballman KV, Beitsch PD, Whitworth PW, Blumencranz PW, Leitch AM, Saha S, McCall LM, Morrow M. Axillary dissection vs no axillary dissection in women with invasive breast cancer and sentinel node metastasis: A randomized clinical trial. JAMA 2011;305:569–575. Goldhirsch A, Winer EP, Coates AS, Gelber RD, Piccart-Gebhart M, Thurlimann B, Senn HJ. Personalizing the treatment of women with early breast cancer: Highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2013. Ann Oncol 2013;24:2206–2223. Johnson S, Brown S, Porter G, Steel J, Paisley K, Watkins R, Holgate C. Staging primary breast cancer. Are there tumour pathologic features that correlate with a false-negative axillary ultrasound? Clin Radiol 2011;66:497–499. Miao H, Hartman M, Bhoo-Pathy N, Lee SC, Taib NA, Tan EY, Chan P, Moons KG, Wong HS, Goh J, Rahim SM, Yip CH, Verkooijen HM. Predicting survival of de novo metastatic breast cancer in Asian women: Systematic review and validation study. PLoS One 2014; 9:e93755.
1841
Moorman AM, Bourez RL, Heijmans HJ, Kouwenhoven EA. Axillary ultrasonography in breast cancer patients helps in identifying patients preoperatively with limited disease of the axilla. Ann Surg Oncol 2014;21:2904–2910. Motomura K, Inaji H, Komoike Y, Kasugai T, Nagumo S, Hasegawa Y, Noguchi S, Koyama H. Gamma probe and ultrasonographicallyguided fine-needle aspiration biopsy of sentinel lymph nodes in breast cancer patients. Eur J Surg Oncol 2001;27:141–145. Neal CH, Daly CP, Nees AV, Helvie MA. Can pre-operative axillary US help exclude N2 and N3 metastatic breast cancer? Radiology 2010; 257:335–341. Park SH, Kim EK, Park BW, Kim SI, Moon HJ, Kim MJ. False negative results in axillary lymph nodes by ultrasonography and ultrasonography-guided fine-needle aspiration in patients with invasive ductal carcinoma. Ultraschall Med 2013;34:559–567. Vaidya JS, Vyas JJ, Thakur MH, Khandelwal KC, Mittra I. Role of ultrasonography to detect axillary node involvement in operable breast cancer. Eur J Surg Oncol 1996;22:140–143. van Rijk MC, Deurloo EE, Nieweg OE, Gilhuijs KG, Peterse JL, Rutgers EJ, Kroger R, Kroon BB. Ultrasonography and fineneedle aspiration cytology can spare breast cancer patients unnecessary sentinel lymph node biopsy. Ann Surg Oncol 2006;13:31–35. Verbanck J, Vandewiele I, De Winter H, Tytgat J, Van Aelst F, Tanghe W. Value of axillary ultrasonography and sonographically guided puncture of axillary nodes: A prospective study in 144 consecutive patients. J Clin Ultrasound 1997;25:53–56. Yang WT, Ahuja A, Tang A, Suen M, King W, Metreweli C. High resolution sonographic detection of axillary lymph node metastases in breast cancer. J Ultrasound Med 1996;15:241–246.
http://dx.doi.org/10.1016/j.ultrasmedbio.2015.03.024
d
Original Contribution SENSITIVITY, SPECIFICITY AND ACCURACY OF ULTRASOUND IN DIAGNOSIS OF BREAST CANCER METASTASIS TO THE AXILLARY LYMPH NODES IN CHINESE PATIENTS YAN-NA ZHANG,* CHANG-JUN WANG,* YING XU,* QING-LI ZHU,y YI-DONG ZHOU,* JING ZHANG,y FENG MAO,* YU-XIN JIANG,y and QIANG SUN* * Department of Breast Surgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, Beijing, China; and y Department of Ultrasound, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, Beijing, China (Received 23 May 2014; revised 5 March 2015; in final form 27 March 2015)
Abstract—The use of ultrasound in the diagnosis of axillary lymph node metastases from breast cancer in a Chinese population was investigated. Data for 1,049 with breast cancer were retrospectively collected. All patients had undergone pre-operative axillary ultrasound and then axillary lymph node dissection. The sensitivity, specificity and accuracy of axillary ultrasound in this cohort were 69.4%, 81.8% and 77.0%, respectively. The overall false-negative rate of ultrasound images was 30.6% (123/402). False-negative ultrasound rates for pathologic N1, N2 and N3 patients were 46.2%, 21.8% and 9.3%, respectively. In patients with stage T1 disease and fewer than three metastatic lymph nodes, the false-negative ultrasound rate was 52.2% (47/90). Moreover, breast cancer patients with a false-negative axillary ultrasound were more likely to have a large tumor (p , 0.001) and high tumor grade (p 5 0.009). However, there were no statistically significant differences between accuracy of axillary ultrasound and age of patients or experiences of ultrasound practitioners. In conclusion, the sensitivity, specificity and accuracy of ultrasound in the diagnosis of breast cancer metastasis to the axillary lymph nodes in Chinese patients were assessed. These data could help us to carefully use axillary ultrasound to diagnose and predict breast cancer axillary lymph node metastasis. (E-mail: [email protected] or [email protected]) Ó 2015 World Federation for Ultrasound in Medicine & Biology. Key Words: Breast cancer, Ultrasound diagnosis, Lymph node metastasis, Sensitivity, Specificity, Accuracy.
(such as patient age, hormone receptor status, tumor stage, tumor size and presence of lymph node and distant metastases) (Goldhirsch et al. 2013). Among these, axillary lymph node status is one of the most important prognostic factors in breast cancer. Thus, pre-operative axillary ultrasound is crucial to staging and management of breast cancer in many institutions (Gentilini and Veronesi 2012; Vaidya et al. 1996). For example, accurate evaluation of axillary lymph node status could avoid an unnecessary sentinel lymph node biopsy (SLNB) or even the maximally invasive radical resection (Moorman et al. 2014). However, data on false-negative axillary ultrasound in Chinese women are sparse. In addition, a recent large-scale randomized, controlled, multicentric clinical trial (SOUND [Sentinel Node versus Observation after Axillary Ultrasound] trial) of patients with negative axillary ultrasound compared the clinical outcome of SLNB with that of observation only (Gentilini and Veronesi 2012). In this trial,
INTRODUCTION Breast cancer is one of the most significant health problems in the world and is the second leading cause of cancer-related death among women in the United States (American Cancer Society 2014). During the past three decades, both advancements in screening for early stages of breast cancer and improvement in treatment options have significantly reduced breast cancer mortality and improved quality of life, but many patients still develop metastatic disease and consequently die (Miao et al. 2014). A number of factors can affect the outcome of breast cancer, for example, clinicopathologic features
Address correspondence to: Qiang Sun, Department of Breast Surgery, Peking Union Medical College Hospital, 1 Shuaifuyuan, Wangfujing, Beijing 100730, China. E-mail: [email protected] or [email protected] Conflicts of interest: The authors declare that there is no conflict of interest in this study. 1835
1836
Ultrasound in Medicine and Biology
Table 1. Clinicopathologic characteristics of breast cancer patients Variable
Number (%) of cases
Age ,65 y $65 y Tumor size (cm) #2 .2 Axillary lymph node metastasis N1 N2 N3 ER/PR expression Positive Negative HER-2 expression Low High Triple negative Yes No p53 expression Negative Positive Ki-67 expression Negative Positive Histologic grade I II III
945 (90.1) 104 (9.9) 638 (60.8) 411 (39.2) 195 (18.6) 110 (10.5) 97 (9.2) 707 (67.4) 342 (32.6) 613 (58.4) 436 (41.6) 156 (14.9) 893 (85.1) 696 (66.3) 353 (33.7) 399 (38.0) 650 (62.0) 213 (20.3) 467 (44.5) 369 (35.2)
ER/PR 5 estrogen receptor/progesterone receptor.
physicians used the inclusion criteria of negative axillary ultrasound and clinically negative axillary lymph nodes
Volume 41, Number 7, 2015
to rule out an evident or suspicious lymph node metastasis of breast cancer and, thus, perform conservative surgery instead of maximally invasive radical resection. However, a question was immediately raised as to whether ultrasound is useful in excluding axillary lymph node metastases from breast cancer. The reliability of the ongoing SOUND trial also needs to be considered with respect to the false-negative rate of axillary ultrasound. Indeed, ultrasound imaging is widely used to evaluate the breast and axillae pre-operatively and can guide physicians in performance of core needle biopsy. Several previous studies (Motomura et al. 2001; Vaidya et al. 1996; Verbanck et al. 1997; Yang et al. 1996) on the accuracy of ultrasound in breast cancer staging reported that high-resolution ultrasound had broad ranges of sensitivity, specificity and overall accuracy of 50%–92%, 90%–100% and 76%–92%, respectively. Their finding indicates that ultrasound has moderate sensitivity, but high specificity in the detection of axillary metastases. However, other studies have reported on the impact of false-negative results in preoperative axillary ultrasound (Choi et al. 2012; Johnson et al. 2011; Neal et al. 2010; Park et al. 2013) on prognosis and choice of surgery. Thus, in this study, we assessed the diagnostic value of ultrasound imaging in the detection of breast cancer metastases to the axillary lymph nodes. Our aim was to investigate whether there is an identifiable subset of breast cancers associated with a higher risk of false-negative axillary ultrasound results.
Fig. 1. Ultrasound images of normal and metastatic axillary lymph nodes. (a) Normal lymph node. (b) Blood flow in normal lymph node. (c) Metastatic lymph node (long-to-short axis ratio ,2, compression of the fatty hilum, cortical thickening and asymmetry). (d) Blood flow in metastatic lymph node (rich blood flow signal).
US diagnoses of breast cancer metastasis to axillary lymph nodes d Y.-N. ZHANG et al.
1837
chemotherapy. All enrolled patients underwent preoperative ultrasound of the axilla on both sides. The ultrasound was performed by three advanced practitioners with more than 4 y of breast imaging experience and three senior practitioners with less than 4 y of breast imaging experience. Clinicopathologic features of these patients were collected from their medical records and included patient age; TNM (tumor/node/metastasis) classification; tumor size, axillary lymph node status, histologic grade and expression of estrogen receptor (ER), progesterone receptor (PR), Her-2, p53 and Ki-67. Pre-operative ultrasound images and post-operative pathologic results were also documented for each patient. This study was approved by the Peking Union Medical College Hospital, and each participant signed an informed consent form to participate in this study. Ultrasound images Ultrasound (ACUSON S2000, Siemens Medical Solutions, Mountain View, CA, USA, with an 18L6 probe; and LOGIQ 7, GE Ultra Solutions, Ontario, CA, USA, with a 10L probe) was used to scan both axillae in patients. Both B-mode (the fundamental mode was used routinely, the harmonic mode was used if necessary) and color Doppler flow imaging (CDFI) were used to characterize lymph node metastases. Any abnormal lymph nodes were noted as suspicious. The criteria used to diagnose abnormal lymph nodes under ultrasound included: long axis-to-short axis ratio ,2, unclear margin or irregular shape, hypo-echoic mass, compression or disappearance of the fatty hilum, cortical thickening or asymmetry, merged lymph nodes and rich blood flow signal. Any lymph nodes that met three or more of the aforementioned criteria were considered to have a tumor metastasis. Fig. 2. Ultrasound images of false negative and positive axillary lymph nodes. (a) False-negative lymph node (a little cortical thickening, considered to be normal lymph node by ultrasound but metastatic on pathology). (b) False-positive lymph node (long-to-short axis ratio ,2, compression of the fatty hilum, cortical thickening, considered to have a tumor metastasis by ultrasound but normal on pathology). (c) Blood flow in falsepositive lymph node.
METHODS Patients In this study, we retrospectively enrolled 1,049 consecutive primary breast cancer patients from Peking Union Medical College Hospital between January 2010 and December 2011. All of these patients had undergone axillary lymph node dissection; however, exclusion criteria were metastasis of other primary cancers to the mammary glands and treatment with neo-adjuvant
Statistical analysis Sensitivity was calculated as the proportion of patients with pathologic metastatic lymph nodes who had an abnormal axillary lymph node in ultrasound images. Specificity was calculated as the proportion of patients with no pathologic metastatic lymph node who had a normal lymph node in ultrasound images. Accuracy was calculated as the proportion of patients whose axillary lymph node status was correctly predicted by ultrasound. The overall sensitivity, specificity and accuracy of ultrasound were analyzed using multinomial logistic regression and binary logistic regression tests. The difference was considered statistically significant when the p value was #0.05. All statistical analyses were performed using SPSS software (Version 19.0, SPSS, Chicago, IL, USA) by a statistician who was blinded to the pathologic records.
1838
Ultrasound in Medicine and Biology
Volume 41, Number 7, 2015
Table 2. Comparison of ultrasound results with lymph node involvement (diagnosed pathologically) in breast cancer patients False-negative True-negative False-positive True-positive p value* Accuracy (%) Sensitivity (%) Specificity (%) Total (n 5 1,049) Age ,65 y (n 5 945) $65 y (n 5 104) Tumor size #2 cm (n 5 638) .2 cm (n 5 411) Grade Low (n 5 213) High (n 5 836) ER/PR expression Positive (n 5 707) Negative (n 5 342) HER-2 expression Low (n 5 613) High (n 5 436) p53 expression Negative (n 5 696) Positive (n 5 353) Ki-67 index #15% .15% Practitioner Experienced (n 5 649) Inexperienced (n 5 400)
123
529
118
279
77.0
69.4
81.8
112 11
472 57
105 13
256 23
0.445
77.0 76.9
69.6 67.6
81.8 81.4
70 53
391 138
60 58
117 162
,0.001
77.3 77.0
70.7 69.1
81.3 81.9
53 16
138 139
58 33
162 25
,0.001
76.9 79.6
83.3 62.6
73.5 86.7
107 100
390 378
85 59
254 170
0.009
73.0 77.0
75.3 61.0
70.4 80.8
23 83
151 313
59 65
109 152
0.268
77.0 77.5
70.4 63.0
82.1 86.5
85 38
72 46
364 165
175 104
0.705
76.0 75.9
82.6 64.7
71.9 82.8
44 79
40 78
237 292
78 201
0.07
78.7 77.4
76.0 67.3
80.3 83.5
83 40
336 193
62 56
168 111
0.251
76.2 78.9
73.2 63.9
78.2 85.6
ER/PR 5 estrogen receptor/progesterone receptor. * p Value refers to the comparison of accuracy in each study group (Age, Tumor size, Tumor grade, etc.).
Table 3. Comparison of clinical and pathologic influencing factors between patients with false-predictive US results and those with true-predictive axillary ultrasound results False-negative US (N 5 123)
Age ,65 y $65 y Tumor size #2 cm .2 cm Grade I II or III ER/PR expression Negative Positive HER-2 expression Low High p53 expression Negative Positive Ki-67 index #15% .15% Practitioner Experienced Inexperienced
True-negative US (N 5 529)
N
%
N
%
112 11
19.2 16.2
472 57
81.2 83.8
70 53
15.2 27.7
391 138
84.8 72.3
6 108
7.5 21.7
74 390
92.5 78.3
22 101
13.8 21.1
138 377
86.2 78.9
83 40
22.0 15.6
295 216
78.0 84.4
85 38
19.5 18.5
352 167
80.5 81.5
44 79
17.0 21.2
215 293
83.0 78.8
83 40
19.8 17.2
336 193
80.2 82.8
p value*
Odds ratio (95% CI)
0.308
1.515 (0.68–3.36)
0.000
0.389 (0.24–0.60)
0.009
0.329 (0.14–0.75)
0.464
0.807 (0.45–1.43)
0.165
1.415 (0.86–2.30)
0.311
1.279 (0.79–2.05)
0.234
0.754 (0.47–1.20)
0.860
1.039 (0.68–1.58)
CI 5 confidence interval; ER/PR 5 estrogen receptor/progesterone receptor; US 5 ultrasound. * p Value refers to the comparison of false-negative and true-negative results in each study group (Age, Tumor size, Tumor grade, etc.).
US diagnoses of breast cancer metastasis to axillary lymph nodes d Y.-N. ZHANG et al.
1839
Table 4. Association of ultrasound results with pathologic axillary lymph node stages False-negative US (n 5 123)
True-positive US (n 5 279)
Pathologic axillary lymph node stage
N
%
N
%
p value*
Odds ratio (95% CI)
N1 (n 5 195) N2 (n 5 110) N3 (n 5 97)
90 24 9
46.2 21.8 9.3
105 86 88
53.8 78.2 90.7
0.002 0.019
0.196 (0.07–0.54) 0.346 (0.14–0.83)
CI 5 confidence interval; US 5 ultrasound. * p Value refers to the comparison of false-negative and true-positive results in N1, N2 and N3.
RESULTS Characteristics of patients In this study, we analyzed ultrasound assessment of axillary lymph node metastases in 1,049 breast cancer patients (Table 1). The mean age of this cohort was 50.3 y (range: 21–88 y). Four hundred and two patients (38.3%) had axillary lymph node metastases, whereas 638 (60.8%) patients had stage T1 disease and 156 (14.9%) had triple-negative tumors (Table 1). The reason we chose 65 y as the cutoff age is because patients older than 65 y are considered to be senile breast cancer patients. A tumor size of 2 cm was used as a cutoff point because this size is considered stage T1 and is one of the inclusion criteria for the SOUND trial. Ultrasound image analysis of the axilla lymph nodes Three hundred ninety-seven (37.8%) of the 1,049 patients were found to have tumor axillary lymph node metastases by ultrasound imaging analysis. In Figure 1 are representative images of the normal and metastatic lymph nodes of these patients. However, pathology data on surgically resected axillary lymph nodes indicated that 402 cases (38.3%) were positive for breast cancer metastasis to the axillary lymph nodes. We then compared the ultrasound images with the data on surgically resected lymph nodes and found that 123 patients were proven to have false-negative (pathologic metastatic lymph nodes with normal ultrasound images) results. In contrast, 118 cases were found to be false positive as no metastatic axillary lymph node was discovered after surgery. In Figure 2 are representative images of falsenegative and false-positive lymph nodes from these
patients. The sensitivity, specificity and overall accuracy of axillary ultrasound were 69.4%, 81.8% and 77.0%, respectively. The false-negative rate (negative ultrasound images but positive pathology) was 30.6% (123/402), and the false-positive rate (positive ultrasound images but negative pathology) was 18.2% (118/647) in these patients. We then compared ultrasound rates for falsepositive or -negative axillary lymph nodes with other clinicopathologic data of the patients. Ultrasound accuracy differed with tumor size, tumor grade and ER/PR expression (Table 2). Patients with false-negative axillary ultrasound results were more likely to have larger (p , 0.001) and higher-grade (p 5 0.009) tumors (Table 3). However, expression of p53 and Ki-67 was not significantly associated with false-negative axillary ultrasound results (Table 3). Moreover, the accuracy of axillary ultrasound was not statistically associated with the patient’s age or the practitioner’s experience (Table 3). Table 4 compares the false-negative rates among various pathologic stages; false-negative ultrasound rates for stage N1, N2 and N3 patients were 46.2%, 21.8% and 9.3%, respectively. The differences between N1 and N3 (p 5 0.002) and between N2 and N3 (p 5 0.019) are statistically significant (correlation coefficient 5 0.246). This indicates that advanced N stage is associated with a lower false-negative rate. In addition, we compared falsenegative rates between stage T1 patients with one or two metastatic lymph nodes and those with three or more metastatic lymph nodes according to the approved conclusion of the American College of Surgeons Oncology Group Z0011 trial (Giuliano et al. 2011) (Table 5). There was no difference in the false-negative rate between the two groups of patients (p 5 0.117). In
Table 5. Association of ultrasound results with numbers of pathologic axillary lymph nodes (T1 tumor) False-negative US (n 5 70)
True-positive US (n 5 117)
Number of pathologic lymph nodes
N
%
N
%
p value*
Odds ratio (95% CI)
1 or 2 (n 5 90) $3 (n 5 97)
47 23
52.2 23.7
43 74
47.8 76.3
0.117
0.535 (0.24–1.16)
CI 5 confidence interval; US 5 ultrasound. * p Value refers to the comparison of false-negative and true-positive results in the study group.
1840
Ultrasound in Medicine and Biology
patients with stage T1 disease and one or two metastatic lymph nodes, the false-negative ultrasound rate was 52.2% (47/90). DISCUSSION Clinical assessment of breast cancer stage is one of the most important pre-surgery procedures in defining the surgical resection of tumor and metastasized lymph nodes (i.e., radical versus conservative surgery). In this regard, pre-operative axillary ultrasound is crucial to staging and selection of management options for breast cancer patients. Accurate axillary evaluation may avoid an unnecessary SLNB and maximally invasive radical resection. In this retrospective study, we evaluated the diagnostic value of ultrasound in detection of axillary lymph node metastases of breast cancer and the clinicopathologic factors associated with the false-negative ultrasound results in Chinese patients. Our data revealed that ultrasound had 69.4% sensitivity, 81.8% specificity and 77.0% overall accuracy in the diagnosis of breast cancer axillary lymph metastases in this Chinese cohort. These data are consistent with previously reported results in Western countries (Motomura et al. 2001; Vaidya et al. 1996; Verbanck et al. 1997; Yang et al. 1996). The falsenegative rate of axillary ultrasound was 30.6%, which was within the ranges of previous studies in Western countries (Cools-Lartigue and Meterissian 2012; Vaidya et al. 1996; van Rijk et al. 2006; Yang et al. 1996). We also found that the false-negative rate of axillary ultrasound was associated with tumor size, histologic grade and ER/PR expression. We concluded that axillary ultrasound has diagnostic value in detecting positive axillary lymph nodes in Chinese women. Thus, improvement of ultrasound sensitivity, specificity and accuracy or use of other techniques is warranted. There is an urgent demand for non-invasive modalities to differentiate stage N1 from advanced N stage patients. As an example, the Z0011 trial proved that patients with fewer than three positive SLNs could be spared from axillary lymph node dissection. Additionally, the ongoing SOUND trial, which represents a step forward in the conservative approach to breast cancer, adopted negative ultrasound of the axilla as an inclusion criterion and assigned patients randomly to either an observation group or an SLN 6 axillary dissection group. Thus, in the future, ultrasound may serve as an alternative tool for detection of axillary metastases in clinical practice. However, the contemporary false-negative rate of N2/3 disease has scarcely been reported. For example, Park et al. (2013) reported that the actual false-negative rate of ultrasound for patients with N2 and N3 disease was 66.7% (6/9), whereas Neal et al. (2010) suggested that pre-operative axillary ultrasound excluded 96% of N2
Volume 41, Number 7, 2015
and N3 invasive ductal metastases. Our study revealed that up to 15.9% (33/207) of patients with N2 and N3 metastases had false-negative ultrasound results. Thus, use of axillary ultrasound alone is considered inadequate to detect N2 and N3 axillary metastases given the unacceptably high false-negative rate (approximately 16%). Furthermore, expression of ER/PR, Her2 and Ki-67 protein and other clinicopathologic parameters are crucial prognostic indicators in breast cancer patients. However, correlation between these parameters and false-negative ultrasound results remains controversial. A previous retrospective study by Park et al. (2013) indicated that ER and PR positivity were strongly correlated with the high false-negative rate. In contrast, Johnson et al. (2011) reported that there were no statistically significant differences in false-negative ultrasound results with respect to ER/PR status. To our knowledge, no study has been conducted on the relationship of HER-2, p53 and Ki-67 with the false-negative ultrasound rate. On the basis of this study, which enrolled more than 1,000 patients, we believe that hormonal receptor status has an impact on the accuracy of ultrasound. This may be explained by the fact that ER/PR-positive breast cancer usually involves large tumors without clinically significant axillary metastases; thus, the advanced T stage of the primary breast lesion may contribute to a high false-negative rate. Therefore, we should pay more attention to those ER/PR-positive breast cancer patients with high histologic grade and false-negative axillary ultrasound results. Additionally, our study also indicated that expression of p53 and Ki67 was not significantly associated with the false-negative ultrasound results. The present study does have some limitations. For example, this retrospective study may compromise the internal and external validation control numbers compared with a prospective study. In the data analysis, we did not use the exact number of positive axillary lymph nodes in analyzing the association with false-negative ultrasound results. Thus, any future study should focus on a combination of ultrasound images and ultrasound-guided fine needle aspiration to increase the sensitivity, specificity and accuracy of axillary ultrasound. Given our results and those of others, we advocate the careful use of axillary ultrasound to rule out axillary lymph node metastases, especially with larger and/or highergrade tumors. More studies on the accuracy of breast cancer axillary lymph node ultrasound are warranted.
REFERENCES American Cancer Society. Cancer facts & figures 2014. Atlanta, GA: Author; 2014. Choi JS, Kim MJ, Moon HJ, Kim EK, Yoon JH. False negative results of pre-operative axillary ultrasound in patients with invasive breast
US diagnoses of breast cancer metastasis to axillary lymph nodes d Y.-N. ZHANG et al. cancer: Correlations with clinicopathologic findings. Ultrasound Med Biol 2012;38:1881–1886. Cools-Lartigue J, Meterissian S. Accuracy of axillary ultrasound in the diagnosis of nodal metastasis in invasive breast cancer: A review. World J Surg 2012;36:46–54. Gentilini O, Veronesi U. Abandoning sentinel lymph node biopsy in early breast cancer? A new trial in progress at the European Institute of Oncology of Milan (SOUND: Sentinel node vs Observation after axillary UltraSouND). Breast 2012;21:678–681. Giuliano AE, Hunt KK, Ballman KV, Beitsch PD, Whitworth PW, Blumencranz PW, Leitch AM, Saha S, McCall LM, Morrow M. Axillary dissection vs no axillary dissection in women with invasive breast cancer and sentinel node metastasis: A randomized clinical trial. JAMA 2011;305:569–575. Goldhirsch A, Winer EP, Coates AS, Gelber RD, Piccart-Gebhart M, Thurlimann B, Senn HJ. Personalizing the treatment of women with early breast cancer: Highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2013. Ann Oncol 2013;24:2206–2223. Johnson S, Brown S, Porter G, Steel J, Paisley K, Watkins R, Holgate C. Staging primary breast cancer. Are there tumour pathologic features that correlate with a false-negative axillary ultrasound? Clin Radiol 2011;66:497–499. Miao H, Hartman M, Bhoo-Pathy N, Lee SC, Taib NA, Tan EY, Chan P, Moons KG, Wong HS, Goh J, Rahim SM, Yip CH, Verkooijen HM. Predicting survival of de novo metastatic breast cancer in Asian women: Systematic review and validation study. PLoS One 2014; 9:e93755.
1841
Moorman AM, Bourez RL, Heijmans HJ, Kouwenhoven EA. Axillary ultrasonography in breast cancer patients helps in identifying patients preoperatively with limited disease of the axilla. Ann Surg Oncol 2014;21:2904–2910. Motomura K, Inaji H, Komoike Y, Kasugai T, Nagumo S, Hasegawa Y, Noguchi S, Koyama H. Gamma probe and ultrasonographicallyguided fine-needle aspiration biopsy of sentinel lymph nodes in breast cancer patients. Eur J Surg Oncol 2001;27:141–145. Neal CH, Daly CP, Nees AV, Helvie MA. Can pre-operative axillary US help exclude N2 and N3 metastatic breast cancer? Radiology 2010; 257:335–341. Park SH, Kim EK, Park BW, Kim SI, Moon HJ, Kim MJ. False negative results in axillary lymph nodes by ultrasonography and ultrasonography-guided fine-needle aspiration in patients with invasive ductal carcinoma. Ultraschall Med 2013;34:559–567. Vaidya JS, Vyas JJ, Thakur MH, Khandelwal KC, Mittra I. Role of ultrasonography to detect axillary node involvement in operable breast cancer. Eur J Surg Oncol 1996;22:140–143. van Rijk MC, Deurloo EE, Nieweg OE, Gilhuijs KG, Peterse JL, Rutgers EJ, Kroger R, Kroon BB. Ultrasonography and fineneedle aspiration cytology can spare breast cancer patients unnecessary sentinel lymph node biopsy. Ann Surg Oncol 2006;13:31–35. Verbanck J, Vandewiele I, De Winter H, Tytgat J, Van Aelst F, Tanghe W. Value of axillary ultrasonography and sonographically guided puncture of axillary nodes: A prospective study in 144 consecutive patients. J Clin Ultrasound 1997;25:53–56. Yang WT, Ahuja A, Tang A, Suen M, King W, Metreweli C. High resolution sonographic detection of axillary lymph node metastases in breast cancer. J Ultrasound Med 1996;15:241–246.