Abstract
BACKGROUND: Ultrasound (US) is routinely used in breast cancer assessment, but the prognostic significance of paracancerous US features remains unclear. The 21-gene recurrence risk score (RS) is widely used to evaluate recurrence risk in patients with luminal A invasive breast cancer. This study aimed to clarify the association between US characteristics of the paracancerous area (PA) and RS in patients with early-stage luminal A invasive breast cancer. METHODS: A retrospective analysis was employed to examine the clinical, pathological, and 21-gene RS data of 194 patients with early-stage luminal A invasive breast cancer. All patients underwent surgical treatment with histopathologically confirmed outcomes. US characteristics of the PA were analyzed, including conventional US (margin, echogenic halo, and peritumoral vascularity), elastography (UE) (enlarged stiff area and elastic strain ratio), and contrast-enhanced US (CEUS) [boundary, penetrating vessels, enhancement intensity, enhancement scope, radial peripheral perfusion, peak intensity (PI), the time to peak (TTP), mean transit time (MTT), sharpness, area under the curve (AUC), area under the wash-in curve, and area under the washout curve]. Patients were categorized based on the RS score into a high-recurrence risk group (RS >25) and a low-recurrence risk group (RS ≤25). A logistic regression analysis model was used to assess the correlation between paracancerous US features and different recurrence risk groups based on RS. RESULTS: A comparison between the RS ≤25 and RS >25 groups revealed significant differences. The incidence of echogenic halo was 78.8% (41/52) in the RS >25 group and 55.6% (79/142) in the RS ≤25 group (P=0.003). The presence of an enlarged stiff area was observed in 76.9% (40/52) of patients in the RS >25 group, which was significantly higher than the 56.3% (80/142) in the RS ≤25 group (P=0.009). CEUS showed that the rates of enhancement scope and radial peripheral perfusion were 82.7% (43/52) (P=0.010) and 65.4% (34/52) (P=0.047) in the RS ≤25 group, respectively, which were significantly higher than those in the RS >25 group. Meanwhile, the RS >25 group exhibited higher PI values (P=0.038) and lower time-to-peak values (P=0.030) compared to the RS ≤25 group. Logistic regression analysis identified echogenic halo, margin, enlarged stiff area, enhancement scope, and PI values as independent diagnostic factors for RS >25. In the combined logistic regression analysis, the predicted AUC for RS >25 was 0.815 (P<0.001), with a sensitivity of 0.750 and a specificity of 0.803. CONCLUSIONS: The combined assessment of the PA using conventional US, UE, and CEUS holds significant value for preoperative evaluation and postoperative prognostic prediction.