Benign-Malign Differentiation of Axillary Lymph Nodes: The Role Of Superb Microvascular Imaging

MATERIAL AND METHOD Patient The hospital's ethics committee approved this single-center prospective study (E-93471371-514.10). 139 female patients, ages 18 to 74, who had suspicious lymph nodes in the axilla during a breast ultrasound examination at our clinic between April 2021 and April 2022 were included in the study. 20 patients did not come back for follow-ups, 4 patients' ultrasonographic images were not sufficient diagnostic quality, and 13 patients were excluded from the study because the biopsy result was not diagnostic and the patient did not undergo re-biopsy. As a result, 108 lymph nodes from 102 patients were included in the study, and 37 patients were excluded.

Patients who were male, under the age of 18, underwent axillary lymph node dissection, had active infection in the axillary region, were diagnosed with histopathologically confirmed axillary lymph node metastasis, had received known chemotherapy or radiotherapy to the axillary region, or did not sign the informed consent form were excluded from the study. Patients who had their pathology results unavailable or who later decided not to participate in the study were also excluded.

Ultrasonographic Evaluation Suspicious axillary lymph nodes in the patients included in the study were evaluated with Toshiba Aplio a (Toshiba Medical Systems Corporation, Tokyo, Japan) ultrasound (US) device and 14-18 MHz high frequency linear probe. Gray scale US, PDUS and SMI images of the lymph nodes were stored. PRF and gain settings were adjusted to prevent from artifacts and to obtain the highest amount of signal while images were taken. In both Doppler methods, these parameters remained unchanged. The routine examination time for each patient extended 1-2 minutes by the SMI evaluation. At the end of the evaluation, a histological assessment using a 21G fine needle aspiration (FNA) biopsy or core biopsy (16-18G) from the relevant lymph nodes, or follow-up with US was performed, depending on the clinical suspicion. Any lymph nodes that regressed or disappeared throughout the US follow-up period of 1-3 months were considered benign. The lymph nodes that remained stable or progressed were evaluated through histopathological sampling. In case of non-diagnostic biopsy, patients were recalled and FNA/core biopsy was repeated. Cytology and pathology specimens were evaluated by a pathologist with at least 10 years of experience. Before the histopathological diagnosis, all patients were prospectively evaluated with ultrasonography by a radiologist blinded to the patient's clinic and history, and images were taken. After the pathology results, PDUS and SMI images were re-evaluated and the findings were compared with each other and pathology results. Finally, two radiologists with 22 years and 4 years of experience evaluated the images in terms of separately determined features without knowing the patient's clinical information and pathology result. The results were compared with each other in order to assess interobserver agreement.

Both PDUS and SMI assessed vascular patterns based on the number of internal vessels, the distribution (avascular/central/peripheral/mixed) and appearance (normal, dysplasic, avascular) of internal vessels. On SMI, both monochrome (mSMI) and color (cSMI) images were assessed. If there was diffuse cortical thickening of the lymph node, the entire cortex was included the images. Likewise if there was an asymmetric cortical thickening, images focused on that part.

Cortical flow signals were taken into account while internal vascularity was quantified. The lymph nodes with no vascular signals in the cortex were classified as avascular, the nodes with peripheral vascular signals as peripheral, the nodes with vascular signals extending from the hilus to the cortex as central, and the nodes with vascular signals in both the periphery and the central as mixed. To evaluate the morphology of the vessels, the signals with smooth, symmetrical hilar extension were considered as normal, the eccentric signals were identified as diplasic, and the lymph nodes without a signal were regarded as avascular.

Statistical Analysis In the study, mean and standard deviation were used for continuous variables, while frequency and percentage values were used when specifying categorical variables. The Kolmogorov-Smirnov and Shapiro Wilk test was used to evaluate the homogeneity of the patient distribution. Student t-test and Mann Whitney U test were used for continuous variables. Chi-square test or Fisher's exact chi-square test was used to evaluate categorical features In order to evaluate the diagnostic performance of each investigated feature of Doppler methods in the differentiation of benign and malignant lymph nodes, the area under the curve (AUC) was calculated separately by receiver operating characteristic (ROC) analysis, depending on whether they were used individually or together. Logistic regression analysis was used to determine the significance value when the features were used together. In addition, sensitivity, specificity, positive predictive value(PPV), negative predictive value (NPV) and accuracy of each method were evaluated.

The Kohen Kappa test was used for categorical features in both SMI and PDUS to assess interobserver agreement. Intragroup correlation coefficient test was used to compare vascular numbers. Chi-square test or Fisher's exact chi-square test was used to evaluate vascular properties between benign and malignant categories after consensus.

All statistical analyses was performed with SPSS version 25.0 (IBM Corp. Armonk, NY: USA. Released 2017). P value \<0.05 was considered statistically significant.