Effectiveness of Quantitative Shear Wave Elastography for the Prediction of Axillary Lymph Node Metastasis

The value of microvascular Doppler ultrasound technique, qualitative or quantitative shear-wave elastography of breast lesions for predicting axillary nodal burden in patients with breast cancer

Background

The status of the axillary lymph node (ALN) in patients with breast cancer can critically inform clinical decision-making and prognosis. Preoperative evaluation of limited nodal burden (0-2 metastatic ALNs) and high nodal burden (≥3 metastatic ALNs) is vital for individual treatment in patients with breast cancer. Thus, this study aimed to evaluate the value of Angio-PLUS (AP; Aixplorer, SuperSonic Imagine) and the qualitative and quantitative shear-wave elastography (SWE) of breast lesions to predict limited or high axillary nodal burden and to develop a model for predicting limited or high axillary nodal burden.

Methods

From March 2020 to November 2022, a total of 232 consecutive patients with breast cancer comprising 232 breast lesions were enrolled retrospectively from Yueyang Central Hospital. The sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), accuracy, and area under the receiver operating characteristic curve (AUC) of AP, qualitative SWE, quantitative SWE, and the predictive model for evaluating limited or high axillary nodal burden were compared.

Results

There was no significant difference in AP patterns between the limited nodal burden group and high nodal burden group. The best cutoff values of Emin (the minimal value of the first Q-box), Emean (the mean value of the first Q-box), Emax (the maximum value of the first Q-box), Eratio (ratio of the first Q-Box and the second Q-Box) and standard deviation for predicting limited or high nodal burden were 80.85 KPa, 133.45 KPa, 153.40 KPa, 9.95, and 19.25 KPa, respectively. The Emax had the highest AUC, and its sensitivity, specificity, PPV, NPV, accuracy, and AUC were 71.64%, 56.36%, 40.00%, 83.04%, 60.78%, and 0.640 [95% confidence interval (CI): 0.575-0.702], respectively. The sensitivity, specificity, PPV, NPV, accuracy, and AUC of seven color patterns for qualitative SWE were 71.64%, 74.55%, 53.33%, 86.62%, 73.71%, and 0.731 (95% CI: 0.669-0.787), respectively, which was significantly higher than all the other quantitative SWE parameters. ALN evaluation in ultrasound and qualitative SWE were independent risk factors for predicting limited or high nodal burden according to a binary logistics regression analysis. The AUC of the predictive model based on independent risk factors was 0.820 (95% CI: 0.765-0.867), which was significantly higher than that of the other independent risk factors.

Conclusions

The seven color patterns in the qualitative SWE of breast lesions were valuable for predicting limited or high nodal burden for patients with breast cancer. Compared with quantitative SWE, qualitative SWE exhibited a better diagnostic performance. Breast lesions present no findings, vertical stripes, and spot patterns were important indicators for limited nodal burden. The predictive model developed in this study could be a simple, noninvasive, and convenient method for predicting limited or high nodal burden, which would be beneficial for clinical decision-making and individual treatment to improve prognosis.

An improved nomogram including elastography for the prediction of non-sentinel lymph node metastasis in breast cancer patients with 1 or 2 sentinel lymph node metastases

Background

The rate of breast-conserving surgery is very low in China, compared with that in developed countries; most breast cancer patients receive mastectomy. It is great important to explore the possibility of omitting axillary lymph node dissection (ALND) in early-stage breast cancer patients with 1 or 2 positive sentinel lymph nodes (SLNs) in China. The aim of this study was to develop a nomogram based on elastography for the prediction of the risk of non-SLN (NSLN) metastasis in early-stage breast cancer patients with 1 or 2 positive SLNs.

Methods

A total of 601 breast cancer patients were initially recruited. According to the inclusion and exclusion criteria, 118 early-stage breast cancer patients with 1 or 2 positive SLNs were finally enrolled and were assigned to the training cohort (n=82) and the validation cohort (n=36), respectively. In the training cohort, the independent predictors were screened by logistic regression analysis and then were used to conducted the nomogram for the prediction of NSLN metastasis in early-stage breast cancer patients with 1 or 2 positive SLNs. The calibration curves, concordance index (C-index), the area under the receiver operating characteristic (ROC) curve (AUC), and Decision curve analysis (DCA) were used to verified the performance of the nomogram.

Results

The multivariable analysis showed that the enrolled patients with positive HER2 expression (OR=6.179, P=0.013), Ki67≥14% (OR=8.976, P=0.015), larger lesion size (OR=1.038, P=0.045), and higher Emean (OR=2.237, P=0.006) were observed to be the independent factors of NSLN metastasis. Based on the above four independent predictors, a nomogram was conducted to predict the risk of the NSLN metastasis in early-stage breast cancer patients with 1 or 2 positive SLNs. The nomogram showed good discrimination in the prediction of NSLN metastasis, with bias-corrected C-index of 0.855 (95% CI, 0.754-0.956) and 0.853 (95% CI, 0.724-0.983) in the training and validation cohorts, respectively. Furthermore, the AUC was 0.877 (95%CI: 0.776- 0.978) and 0.861 (95%CI: 0.732-0.991), respectively, indicating a good performance of the nomogram. The calibration curve suggested a satisfactory agreement between the predictive and actual risk in both the training (χ2 = 11.484, P=0.176, HL test) and validation (χ2 = 6.247, p = 0.620, HL test) cohorts, and the obvious clinical nets were revealed by DCA.

Conclusions

We conducted a satisfactory nomogram model to evaluate the risk of NSLN metastasis in early-stage breast cancer patients with 1 or 2 SLN metastases. This model could be considered as an ancillary tool to help such patients to be selectively exempted from ALND.