多参数超声在判断浸润性乳腺癌 TIL（肿瘤浸润淋巴细胞）存在中的作用：来自 B 模式、SWE 和 SMI 模态的启示

To evaluate the relationship between tumor-infiltrating lymphocyte (TIL) levels and multiparametric ultrasonography (US) findings combining B-mode US, shear wave elastography (SWE), and superb microvascular imaging (SMI) in patients with invasive breast cancer, and to explore the potential of sonographic imaging modalities in predicting the tumor immune microenvironment.

This retrospective study included 148 patients diagnosed with invasive breast carcinoma between September 2021 and December 2024. Patient age, medical history, and immunohistopathological characteristics (grade, hormone positivity, Ki-67 ratio, subtype) of the lesions were recorded. TIL levels were assessed on hematoxylin-eosin (H&E) stained slides by pathologists following the International TILs Working Group guidelines, and lesions were categorized by different TIL levels (presence/absence, 10%, 20%, 30%). US evaluations were performed using a Toshiba Aplio A system (Canon, Tokyo, Japan) with a 12-16 MHz breast probe. Imaging assessments included B-mode ultrasound (morphology, echogenic halo sign), SWE (E-mean, E-ratio, stiff rim sign), and SMI (Adler classification, SMI vascular index). Associations between TIL levels and imaging parameters were analyzed using Chi-square tests for categorical and Student's t-tests for continuous variables (SWE and SMI).

TIL was detected in 121 of 148 lesions (81.8%). TIL value was >10% in 33 lesions, >20% in 12, and >30% in 8 lesions. On B-mode US, round/oval tumor shape (p = .003 at level of TIL > 20%, p = .001 at level of TIL > 30%) and non-parallel orientation (p = .023) were more prevalent in TIL positive lesions. On SWE, tumors with TIL levels 10% were significantly associated with higher E-mean values (130 24.7 vs. 107.9 36, p = .001) and the presence of a stiff rim sign (p < .001). Penetrating vascular structures were more commonly observed on SMI in lesions with TIL 10% (p = .023), along with a higher mean vascular index (p = .036). No significant difference was found in other US-SWE and SMI findings (all p > .4).

Our findings suggest that US features, particularly vascularity on SMI and stiffness on SWE, may reflect TIL presence in breast cancer. However, methodological variations and differing TIL levels across studies may influence inconsistent associations, especially with SWE. Further comprehensive studies are needed to clarify this relationship.