Role of Ultrasound in the Diagnosis of Lymph Node Status in Axillary Lymph Node Metastases in Breast Cancer undergoing Neoadjuvant Chemotherapy

Background:Axillary ultrasound (AUS) is one of the important bases for evaluating the axillary status of breast cancer patients. And it would be helpful for the reassessment of axillary lymph node status in these patients after neoadjuvant chemotherapy(NAC) and guide the selection of their axillary surgical options.The purpose of this study was to evaluate the diagnostic performance of ultrasound,and to find out the factors related to the outcome of ultrasound. Methods:In this retrospective analysis, 172 patients (one bilateral breast cancer) with breast cancer and clinical positive axillary nodes, were enrolled. After NAC, all patients received mastectomy and axillary lymph node dissection (ALND). AUS was used before and after NAC to assess the axilla status. Results:Of the 173 axillae, 137 (79.19%) had pathological metastasis after NAC. The accuracy, sensitivity, specificity, positive predictive value and negative predictive value of axillary ultrasound in this cohort were 68.21%, 69.34%, 63.89%, 87.96% and 35.38% respectively. Univariate analysis showed that primary axillary lymph node(ALN) short axis, progesterone receptors, hormone receptors, the tumor status after NAC, tumor reduction rate, ALN short axis after NAC, physical examination of axilla after NAC and pN impacted the results of AUS(P = 0.000 ~ 0.040). Multivariate analysis of the above indicators showed that ALN short axis after NAC and pN associated with AUS results independently. Conclusion:AUS can accurately assess axilla status after NAC in most breast cancer patients. If the short axis of ALN≥10mm and AUS negative, SLNB could be chosen. However, AUS cannot detect residual lymph node disease after NAC in a short axis of the ALN <10mm.


Background
The status of axillary lymph node (ALN) is one of the most importent prognostic factors in patients with breast cancer.Sentinel lymph node biopsy (SLNB) instead of axillary lymph node dissection (ALND) is widely used to evaluate ALN status in clinically node-negative patients with breast cancer, with less morbidity of surgical complications compared with ALND [1].And its security had been tested by many trails in early-stage breast cancer [2,3].However, whether it is applicable to the clinical evaluation of patients with negative ALNs after neoadjuvant chemotherapy(NAC) from locally advanced breast cancer with positive initial axillary lymph nodes remains controversial [4].In ACOSOG Z1071 trial, in patients with breast cancer with clinical N1 stage receiving NAC,the false-negative rate(FNR) of SLN biopsy (SLNB) was 12.6% [5].And the secondary endpoint of this trial was to evaluate the ALN status of patients after NAC by ultrasound.SLNB was performed in patients with negative axillary lymph nodes by ultrasound, and FNR was 9.8% [6].Ultrasound is one of the important bases for evaluating the axillary status of breast cancer patients [7],it is an important means for evaluating the efficacy of chemotherapy [8].And it would be helpful for the reassessment of axillary lymph node status in these patients after NAC and guide the selection of their axillary surgical options.
The purpose of this study was to evaluate the diagnostic performance of ultrasound,and to find out the factors related to the outcome of ultrasound.

Patients
From August 2011 to December 2015,A retrospective analysis was performed, 172 patients with breast cancer who underwent ultrasound for assessment of tumour response during NAC were included.One of the 172 patiens had bilateral breast cancer,so we got 173 results of axillae.All 172 patients received a 21-day cycle NAC based on anthracycline and paclitaxel, followed by mastectomy and ALND.
Inclusion criteria:(1) primary unilateral or bilateral breast cancer in women, histologically confirmed as breast cancer;(2) patients with positive axillary lymph nodes diagnosed by ultrasonography in our hospital before neoadjuvant chemotherapy and with clinical stages of cT0~4,cN1~3 and M0;(3) complete NAC and ALND, and review breast and axillary ultrasound before surgery.

Ultrasound technique and interpretation
Ultrasound examinations had been performed in all patients by A group of radiologists with rich experience in breast imaging. Lymph nodes were classified as normal if the radiologist was unable to visualize any lymph nodes on AUS or indicated that the lymph nodes were normal in morphologic appearance. Lymph nodes with abnormal morphology on AUS were classified as suspicious.ALN metastasis was suspected if the LN had any of the following morphologic characteristics: eccentric or concentric cortical thickening >3 mm, absent fatty hilum, a transverse axis-tolongitudinal axis ratio more than two or increased blood flow in the thickened cortex on Doppler image.Referring to the above conditions, the grading system was divided into five categories (1) Normal without abnormal findings;(2) Benign abnormal findings;(3) Indeterminate or uncertain;(4) Suspicious of malignancy;(5) Highly suspicious of malignancy.

Histopathological evaluation
The specimens of tumour and ALNs were evaluated according to the following histopathological features: tumour size, histological type of carcinoma, ALN metastasis,immunohistochemistry(IHC) and so on.We defined ≥1% positive cells as the positive boundary value,according to the ER/PR immunohistochemical detection guidelines issued by the American society of clinical oncology (ASCO)/ American college of pathologists (CAP) in 2010 [6].The intensity of HER2 staining was scored as 0, 1+, 2+ or 3+. Tumours with a 3+ score were classified as HER2 positive, and tumours with a 0 or 1+ score were classified as negative. In tumours with a 2+ score, gene amplification by using fluorescence in situ hybridization(FISH) was used to determine HER2 status. And the HER2 gene copy number ≥6.0 or HER-2/CEP17 ≥2.0 was defined as HER2 positive.

Statistical analysis
We used SPSS 20.0 software for statistical analysis.Single-factor analysis was performed using chi-square test(χ2 test),and multiple-factor analysis was performed using logistical regression of the variables.The diagnostic performance of ultrasound for the evaluation of ALN after NAC was evaluated with receiver operating characteristic (ROC) curve analysis. The diagnostic accuracy was estimated by calculating the area under the ROC curve (Az value).P<0.05 was considered statistically significant.
The number of NAC cycles less than or equal to 2 was 31 cases (17.9%) ,meanwhile, 142 cases (82.1%) received more than 2 NAC cycles. During the postoperative evaluation,the number of pathological complete response (PCR) was 8 cases (4.6%), the number of partial response (Pr) was 89 cases (51.4%), the number of progressive disease (PD) was 7 cases (4.0%), and the number of stable disease (SD) was 67 cases (38.7%) .PCR and PR were regarded as effective for chemotherapy, and the effective rate of NAC was 56.0%. All of the patients' characteristics are listed in Table 1.  Further multiple-factor analysis was performed for the above results, independent factors associated with the AUS after NAC FNR is: after NAC ALN short diameter and pN in installment, the NAC ALN after short diameter on the judgement of the state of AUS after NAC significance is bigger (OR = 7.021), the FNR AUS after NAC ALN short diameter 10 mm OR the group of less than 10% (5.4%), short diameter < 10 mm in the two groups were greater than 10%.All the above data is listed in Table4 and Table5.  [4,[9][10][11][12].ACOSOG Z1071 test and a number of studies have shown [5,[13][14][15][16] that AUS combined with SLNB after NAC can reduce the FNR of axillary diagnosis and replace ALND to achieve the purpose of accurate assessment of axillary status.It has also been reported that [17] ultrasound and magnetic resonance imaging(MRI) have the same or more importent value in accurately measuring the size of residual lesions of breast cancer after NAC.
In the Z1071 test [5,18], the evaluation criteria of AUS for ALN were: a maximum cortical thickness > 3 mm, absence of fatty hilum significantly.The evaluation criteria in our institution were based on the shape is round, cortical thickness, the structure of the lymph node hilum of anomaly or disappear, blood flow is unusually rich, whether lymph node capsule indicators, combined with the above indicators, we simplified the grading system, the class 4 ~ 5 was used as the criterion for suspicious lymph nodes. This system effectively improved the ultrasonic diagnosis value and helped clinicians to quickly locate ALN.This study found that AUS was used to assess the axillary status, its accuracy was 68.21%, higher than that in the was 46.0%, with statistically significant difference (P = 0.000).The results were the same as that of ACOSOG Z1071 test [5].After NAC, 37 cases were clinically touched by enlarged lymph nodes, and 34 of which were pathologically confirmed to lymph node metastasis. AUS was diagnosed as negative in 3 cases of 34, and FNR was 8.8%, which was far lower than that (37.9%,P = 0.001) of those who were negative by clinical palpation.
In this study, FNR of AUS decreased successively, 73.7%, 32.9% and 5.4%, respectively, in the three subgroups with short diameter <5mm, d<10mm, and 10mm after NAC, P = 0.000.AUS, therefore, is not easy to find residual lymph node metastases with short diameter < 5 mm.The accuracy of these patients with lymph nodes between 5mm and 10mm needs to be improved by improving the definition of ultrasound examination equipment and the technical level of ultrasound examination physicians, which is consistent with Feu et al. 's opinion [19] that AUS is more accurate in evaluating lymph nodes with diameters greater than 10mm than those with diameters less than 10mm.In this study, 37 cases with short diameter of lymph nodes greater than or equal to 10mm after NAC were confirmed as lymph node metastasis, 2 of which showed no abnormalities in AUS and FNR was less than 10% (5.4%). Multi-factor analysis showed that short diameter of lymph nodes after NAC was an independent factor affecting AUS diagnosis.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.
Availability of data and material The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Figure 1
The ROC showed that the area under the curve(AUC) was 0.703(P=0.000 95%CI[0.605 0.800