Axillary lymph node status is important for staging,prognosis and determining the optimal treatment forbreast cancer. Axillary lymph node metastasis occurs in30-40% of all patients with breast cancers (1, 2). Aphysical examination alone is neither a sensitive nor re-liable way to ascertain lymph node status because the

metastatic lymph nodes are often not palpable and reac-tive lymph nodes may be mistaken for metastasis (3, 4).

There are several imaging techniques that have beenused to evaluate lymph nodes, including mammogra-phy, computed tomography (CT), MR imaging, lym-phangiography and scintigraphy, and all these methodshave unsatisfactory sensitivity and specificity. For ex-ample, the sensitivity of mammography ranges from 18-41% and the specificity ranges from 80-90% (4-6). Thesensitivity of CT is 50% and its specificity is about 75%(7). Among the imaging modalities, ultrasound has beenthe most frequently investigated because of its noninva-sive nature and the ease of its use for evaluating the axil-

J Korean Radiol Soc 2005;52:45-49

─ 45 ─

The Value of Ultrasonographic Detection forMetastatic Axillary Lymph Nodes in Breast Cancer1

Jung Hee Shin, M.D., Asiry Hwang, M.D., Hye-Young Choi, M.D., Seung Yon Baek, M.D.

Purpose: We evaluated the significance and accuracy of sonographic detection ofmetastatic axillary lymph nodes (LNs) in breast cancer.Materials and Methods: We retrospectively reviewed the sonographic findings andpostoperative results of axillary LNs in 47 patients with breast cancer. The sonograph-ic criteria for metastatic LNs were defined as the loss of the echogenic hilum and anyuneven cortical thickness of over 3 mm. We analyzed the correlation between the pre-operative sonographic findings and the postoperative results of the LNs.Results: Out of 47 patients, 22 patients showed 43 sonographic metastatic LNs.Among these 22 patients, 18 patients had 183 histopathologically proven metastaticLNs. The pathological examination of the remaining 25 patients revealed metastaticLNs in 6 patients. The overall sensitivity, specificity and accuracy of ultrasonographyfor detecting metastatic axillary LNs in breast cancer were 75%, 82.6% and 78.7%, re-spectively. As the number of metastatic LNs detected on sonography increased, thenumber of histologically proven metastatic LNs increased.Conclusion: Ultrasonographic evaluation of axillary LNs in breast cancer can providerelatively accurate information about the presence or absence of metastasis.Therefore, it is useful to decide the initial staging and treatment planning of patientswith breast cancer.

Index words : Breast neoplasmsUltrasonographyLymph Nodes

1Department of Radiology, College of Medicine, Ewha WomansUniversityReceived March 2, 2004 ; Accepted Decemebr 1, 2004Address reprint requests to : Hye-Young Choi, M.D., Department ofRadiology, Ewha Womans University, Mokdong Hospital, 911-1 Mok-dong, Yangcheon-gu, Seoul 158-710, Korea.Tel. 82-2-2650-5173 Fax. 82-2-2650-5302E-mail:choihy@hanmail.net

la. It is used to detect alterations in the size, shape andcontours of lymph nodes, as well as detecting changes inthe morphology and texture of the nodal cortex: it canthereby identify the presence of the underlying metasta-sis (6).

The purpose of this study was to evaluate the signifi-cance and accuracy for the sonographic detection ofmetastatic axillary lymph nodes in breast cancer.

Materials and Methods

During a 3-year period, forty-seven patients withbreast cancer underwent formal axillary lymph nodedissection together with or immediately after lumpecto-my or mastectomy, and all the patients were histologi-cally proved as having an invasive breast cancer. Inthese 47 patients (46 females and one male), the preop-erative ultrasonographic (US) images and postoperativehistologic findings of the axillary lymph nodes were ana-lyzed. We excluded those patients with unavailable pre-operative US findings. The patients’ ages ranged from 25to 75 years (mean age: 46 years).

All the scans were performed by an experienced radi-ologist with real-time ultrasound using a 10 MHz lineararray probe XP 10 (Acuson, Moutain view, CA), or anHDI 5000 (Advanced Technology Laboratories, Bothell,WA). A careful search for adenopathy was performedalong the axilla in the longitudinal and transverse planesusing the axillary vessels as landmarks. On preoperativeUS, we evaluated the size of the breast cancer, the pres-ence of echogenic hilum, the uneven degree of corticalthickness and the number of axillary lymph nodes.Sono-graphic criteria of metastatic lymph nodes weredefined as the loss of the echogenic fatty hilum and anyuneven cortical thickness over 3 mm. We then analyzedthe correlation between the preoperative US findingsthat were suggestive of metastatic lymph nodes and thepostoperative results. For the control study (48 lymphnodes in 31 negative cases), the ratio of the shortest axisto the longest axis (S/L), the cortical thickness, and theobliteration of the echogenic hilum of the lymph nodeswere compared between the metastatic and benignlymph nodes. For statistical analysis, Spearman’s corre-lation coefficients were obtained. Correlation was con-sidered significant at the 0.01 level.

Results

Out of 47 patients, 22 patients showed 43 lymph

nodes that were suggestive of metastasis on US exami-nation. Axillary dissection revealed metastasis in 24 of47 patients (51.1%). The correlation between the ultra-sonography and histopathology is presented in Table 1.Among the 22 sonographically positive patients, 18 pa-tients (18/22, 81.8%) had histopathologically provenbreast cancers with 183 metastatic lymph nodes (Fig. 1).The remaining 25 patients revealed no evidence ofmetastatic lymph nodes on US (Fig. 2), but 6 of these pa-tients (6/25, 24%) were histologically confirmed as hav-ing 11 metastatic lymph nodes. Each of these 6 patientshad three or less metastatic lymph nodes on histologicexamination. One of four false positive patients wasconfirmed histologically as having tuberculous lym-phadenitis. Three false positive patients showed reac-tive hyperplasia. The overall sensitivity of ultrasonogra-phy for the detection of metastatic axillary lymph nodesin breast cancer was 75%, the specificity was 82.6%, theaccuracy was 78.7%, the positive predictive value was81.8% and the negative predictive value was 76%. Thegreater the number of nodes that were suggestive ofmetastasis on US, the greater was the number of histo-

Jung Hee Shin, et al : The Value of Ultrasonographic Detection for Metastatic Axillary Lymph Nodes in Breast Cancer

─ 46 ─

Table 1. Correlation between Ultrasonography and Histology

Histologic Findings

No. of patients Positive Negative Total

US Positive 18 04 22Negative 06 19 25

Total 24 23 47

No.: Number, US: Ultrasonography

Table 2. Correlation between Metastatic Lymph Nodes on Ultra-sonography and Histologically Positive Lymph Nodes

No. of metastaticNo. of patients Mean No. of histologically

LNs on US positive LNs

0 25 0.441 10 4

2-3 10 5.7≥4 02 23

No.: Number, LNs: lymph nodes, US: Ultrasonography

Table 3. Correlation between Tumor Size and Number ofMetastatic Lymph Nodes

Tumor size (cm)No. /A total of patients Mean No.with positive LNs (%) of positive LNs

≤1 1 / 5 (20) 0.21-2 5 / 14 (35.7) 1.572-5 17 / 27 (63) 3.78>5 1/1 (100) 19

No.: Number, LNs: lymph nodes

logically proven metastatic lymph nodes that were iden-tified (p<0.01)(Table 2). Moreover, there was a statisti-cally significant correlation between the size of thebreast mass and the number of histologically provenmetastatic lymph nodes (p<0.01) (Table 3). When com-paring the control and metastatic lymph nodes, the ratioof the shortest axis to the longest axis (S/L), and the corti-cal thickness were 0.38 and 1.46 mm, respectively, inthe benign nodes and 0.74 and 6.11 mm, respectively, inthe metastatic nodes. Twenty-seven (81.8%) of thirty-

three lymph nodes that were suggestive of metastasis onUS showed the obliteration of the fatty hilum. All threeof the criteria we selected for the presence of metastasisshowed statistically significant results (p<0.01).

Discussion

The axillary lymph node status is, in the absence ofdistant metastasis, the single most important factor forpredicting breast cancer patient survival. Axillary

J Korean Radiol Soc 2005;52:45-49

─ 47 ─

A BFig. 2. A 43-year-old woman with breast cancer.A. An indistinct hypoechoic lesion is noted in the right subareolar area that measured about 2.2 cm in width.B. An axillary lymph node shows the even cortical thickness and the visualization of the echogenic hilum, representing no evi-dence of metastasis on US, but this node revealed the presence of metastasis on pathology.

A

BFig. 1. A 56-year-old woman with breast cancer.A. An inhomogeneous hypoechoic mass is noted in the upper outer portion of theright breast. This mass measured about 4.3×4.1 cm in size.B. An ipsilateral axillary lymph node of the same patient shows uneven corticalthickness of more than 3 mm (arrow) with preservation of echogenic hilum, whichwas compatible with the malignant criteria used in our study. This node hadproven pathologically metastasis of the infiltrating ductal carcinoma.

lymph node dissection is routinely performed as a partof the surgical treatment for breast cancer, and it ismainly done for staging and planning the adjuvant ther-apy according to the number of metastatic lymph nodesthat are found (1, 2). However, in most patients, axillarydissection has proved to have been unnecessary afterthe histological node examination. Moreover, axillarynode dissection has significant complications, such aslymphedema, pain in the arm, nerve injury and a pro-longed hospital stay. The newer surgical approaches,such as sentinel node dissection or selective node dissec-tion, have been proposed to reduce the number of un-necessary axillary dissections and these techniques arecurrently under investigation. An accurate noninvasivepreoperative diagnostic method to assess the axillarylymph node status would help reduce the need for axil-lary operations.

US has been proved to have a higher sensitivity (45-84%) and specificity (72-97%) for the detection ofmetastatic axillary lymph nodes in breast cancer than aclinical examination or the other imaging techniques(3-6, 8). Similar to the results of earlier studies, the sen-sitivity, specificity and accuracy of ultrasonographywere 75%, 82.6% and 78.7%, respectively, in our cur-rent study. According to previous studies, false negativecases had three or less microscopically invaded lymphnodes on histologic examination (9, 10). In our study,the six false negative patients had less than threemetastatic lymph nodes. It appears that the commonfactors associated with the discrepancy between the USfindings and the pathological results are the failure to vi-sualize lymph nodes during US examination of the axil-la, the small number of metastatic lymph nodes and thepresence of micrometastasis. Ultrasonography alonecannot always differentiate reactive or inflammatoryconditions from the malignant causes due to the overlapof their sonographic features. The four false positive cas-es in our study were considered to be benign reactivehyperplasia, and one of these cases was confirmed byhistology as tuberculous lymphadenitis. As for the caus-es of the other false positive nodes, pathological sinusoidhyperplasia and irregular, thickened capsules represent-ed the uneven sonographic cortical thickening. Also, therich vascular structures within the lymph node cancause the compression or absence of hilum. US-guidedfine-needle aspiration (FNA) may increase the specifici-ty or reliability in such cases. In a study by Kirshna-murthy et al, the overall sensitivity, specificity, diagnos-tic accuracy, positive predictive value and negative pre-

dictive value of US-guided FNA were 86.4%, 100%,79%, 100% and 67%, respectively (10).

There was a positive correlation between the numberof nodes that were suggestive of metastasis on US andthe number of histologically proven metastatic lymphnodes that were identified (p<0.01).

According to Lam and associates and also Yang andcolleagues, a normal axillary lymph node was defined ashaving an ovoid hypoechoic C-shaped rim of lymphoidtissue (usually 1 to 2 mm thick) around a centralechogenic fatty hilum, and lymph nodes that were sug-gestive of metastasis were rounded and hypoechoicwith or without the associated eccentric cortical hyper-trophy and obliteration of the fatty hilum (5, 6). Size wasnot considered an important factor. In an in vitro USstudy for axillary node analysis by Feu et al., absence ofthe hilum was found to be the most specific sonographicfeature for the diagnosis of metastasis (12). The in-creased long-to-short axis ratio was the finding thatcaused the most false-negative interpretations, indicat-ing that lymph nodes appearing elongated or ovoid canbe metastatic. Yet in other study by Tateishi et al, a cir-cular shape (i.e., a ratio between 0.5 and 1.0 for theshortest axis to the longest axis) was the best single fea-ture for distinguishing metastatic from nonmetastaticlymph nodes (13). In our study, we defined metastaticlymph nodes as those nodes having obliteration of thefatty hilum and if they had over 3mm of uneven corticalthickness.

The prevalence of axillary nodal involvement inbreast cancer is 40-70%, and the prevalence of axillarynodal involvement has been shown to be related to tu-mor size. The risk for having a positive node is known tobe 30% if the primary tumor is larger than 1 cm, and it’s15% if the primary tumor is less than 1 cm. In patientswith T3 cancer, up to 60% will have axillary nodalmetastasis at presentation (14). In a current study, therewas a statistically significant correlation between thesize of the breast mass and the number of provenmetastatic lymph nodes (p<0.01).

The accuracy of ultrasonographic axillary lymph nodedetection will probably improve with the support ofmore advanced sonographic scanners and better-de-fined nodal differential criteria. Additional studies willbe necessary to determine the value of ultrasonographicevaluation for the detection of metastatic axillary lymphnodes due to the limited number of cases in this study.

In conclusion, if the axillary lymph nodes in patientswith breast cancer show an increased ratio of the short-

Jung Hee Shin, et al : The Value of Ultrasonographic Detection for Metastatic Axillary Lymph Nodes in Breast Cancer

─ 48 ─

est axis to the longest axis, the obliteration of fatty hilumand over 3 mm of uneven cortical thickness on US, thenthey should be considered as having metastasis. As thenumber of metastatic LNs detected on US was higher orthe size of breast mass was larger, the number of histo-logically proven metastatic nodes increased. The rela-tively high accuracy and the positive and negative pre-dictive values of US for detecting metastatic axillarylymph nodes indicate that it is a useful modality for theinitial staging of breast cancer, and this US modality canbe immensely valuable for planning the appropriatemanagement of breast cancer patients.

References

1. Fisher B, Bauer M, Wickerham L, Redmond C, Fisher E. Relationof number of positive axillary nodes to the prognosis of patientswith primary breast cancer. Cancer 1983;52:1551-1557

2. Ruffin WK, Stacey-Clear A, Younger J, Hoover HC Jr. Rationalefor routine axillary dissection in carcinoma of the breast. J Am CollSurg 1995;180:245-251

3. DeFreitas R Jr, Costa MV, Schneider SV, Nicolau MA, Marussi E.Accuracy of ultrasound and clinical examination in the diagnosisof axillary lymph node metastasis in breast cancer. Eur J SurgOncol 1991;17:240-244

4. Pamilo M, Soiva M, Lavast EM. Real time ultrasound, axillarymammography and clinical examination in the detection of axil-lary lymph node metastases in breast cancer patients. J UltrasoundMed 1989;8:115-120

5. Lam WW, Yang WT, Chan YL, Stewart IE, Metrewelic, King W.

Detection of axillary lymph node metastases in breast carcinomaby technetium-99m sestamibi breast scintigraphy, ultrasound andconventional mammography. Eur J Nucl Med 1996;23:498-503

6. Yang WT, Ahuja A, Tang A, Suen M, King W, Metrewelic. Highresolution sonographic detection of axillary lymph node metas-tases in breast cancer. J Ultrasound Med 1996;15:241-246

7. March DE, Wechsler RJ, Kurtz AB, Rosenberg AL, Needleman L.CT pathologic correlation of axillary lymph nodes in breast carci-noma. J Comput Assist Tomogr 1991;15:440-444

8. Bruneton JN, Caramella E, Hery M, Aubanel D, Manzino JJ,Picard JL. Axillary lymph node metastasis in breast cancer : preop-erative detection with US. Radiology 1986;158:325-326

9. Verbanck J, Vandewiele I, De Winter H, Tytgat J, Van Aelst F,Tanghe W. Value of axillary ultrasonography and sonographicallyguided puncture of axillary nodes: a prospective study in 144 con-secutive patients. J Clin Ultrasound 1997;25:53-56

10. Krishnamurthy S, Sneige N, Bedi DG, Edieken BS, Fornage BD,Kuerer HH, et al. Role of ultrasound-guided fine-needle aspirationof indeterminate and suspicious axillary lymph nodes in the initialstaging of breast carcinoma. Cancer 2002;95:982-988

11. Yang WT, Ahuja A, Tang A, Suen M, King W, Metreweli C.Ultrasonographic demonstration of normal axillary lymph nodes -a learning curve. J Ultrasound Med 1995;14:823-827

12. Feu J, Tresserra F, Fabregas R, Navarro B, Grases PJ, Suris JC, etal. Metastatic breast carcinoma in axillary lymph nodes: in vitroUS detection. Radiology 1997;205: 831-835

13. Tateishi T, Machi J, Feleppa EJ, Oishi R, Furumoto N, McCarthyLJ, et al. In vitro B-mode ultrasonographic criteria for diagnosingaxillary lymph node metastasis of breast cancer. J Ultrasound Med1999;18:349-356

14. Carter CL, Allen C, Henson DE. Relation of tumor size, lymphnode status, and survival in 24,740 breast cancer cases. Cancer1989;63:181-187

J Korean Radiol Soc 2005;52:45-49

─ 49 ─

대한영상의학회지 2005;52:45-49

유방암의액와림프절전이에대한초음파검사의가치1

1이화여자대학교 의과대학 이대목동병원 방사선과

신정희·황아실이·최혜영·백승연

목적: 유방암 환자에서 액와 림프절의 전이 유무를 알아보기 위한 초음파 검사의 진단적 정확성과 가치를 알아 보고자

하였다.

대상과 방법: 수술 전 초음파 검사에서 액와 림프절 전이 유무와 수술 후 전이 여부를 알고 있는 47명의 유방암 환자를

대상으로 하였다. 전이된 림프절의 초음파적 진단 기준은 에코성의 문의 소실과 3 mm 이상의 비균일한 피질 두께를 가

질 경우로 정의하였다. 수술 전 림프절의 초음파 소견과 수술 후 결과 사이의 관계를 분석했다.

결과: 47명 중 22명의 환자에서 초음파 검사로 43개 림프절의 전이가 진단되었다. 22명의 환자 중 18명에서 병리조직

학적으로 183개의 액와 림프절 전이가 확진되었다. 초음파 검사에서 나머지 25명은 전이된 림프절이 발견되지 않았으

나 이 중 6명은 병리조직학적으로 7개의 림프절 전이가 확진되었다. 유방암에서 전이된 액와 림프절의 발견에 대한 초

음파의 진단적 예민도, 특이도, 그리고 정확도는 각각 75%, 82.6%, 78.7% 였다. 초음파 검사에서 전이된 림프절의 수

가 많이 보일수록, 병리적으로도 전이된 림프절의 수가 증가하였다.

결론: 유방암에서 액와 림프절의 초음파 검사는 전이의 유무에 대한 비교적 정확한 정보를 제공한다. 그러므로 유방암

환자의 액와 림프절 검사는 초기 병기와 치료 계획을 결정하는 데 도움을 주는 필요한 검사라고 생각된다.