The Pre-Operative Imaging Risk Factors for Lymphovascular Invasion in Breast Cancer as Determined by Mammography and Ultrasound Findings of Primary Breast Masses

Abstract

Objective: To determine pre-operative mammographic and ultrasound features predictive of lymphovascular invasion in patients with breast cancer presenting to a tertiary care center in Pakistan.
Methodology: This retrospective descriptive study was conducted at SIUT Breast Oncology OPD and Radiology department. Patients with histopathologically proven breast cancer whose pre- treatment mammograms, ultrasounds and histopathology reports are available, patients who have undergone upfront mastectomy or breast-conserving surgery and LVI status confirmed by immunohistochemistry were included. Imaging features assessed included mass margin, shape, echo pattern, posterior acoustic features, microcalcification pattern, axillary lymph node morphology on ultrasound, and mammographic density and distortion. LVI status was determined on histopathology and correlated with imaging findings using univariate and multivariate logistic regression.
Results: Of the 350 cases, LVI was present in an estimated 112 (32%) patients. Irregular tumor margins, non-parallel orientation, and abnormal axillary lymph node morphology on ultrasound were significantly more frequent in the LVI-positive group (78% vs. 41%, P;0.001; 71% vs. 38%, p<0.001; 65% vs. 29%, p<0.001, respectively). Spiculated margins and posterior acoustic shadowing also showed strong association with LVI (p=0.002 and p=0.014). On mammography, pleomorphic/fine linear microcalcifications and architectural distortion were significantly associated with LVI (p=0.008 and p=0.021). On multivariate analysis, irregular margin (OR 3.1, 95% CI 1.8–5.4, p<0.001) and abnormal axillary nodal morphology (p=0.001) emerged as independent predictors of LVI.
Conclusion: Specific mammographic and sonographic features, particularly tumor margin irregularity and abnormal axillary lymph node appearance, are significantly associated with lymphovascular invasion in breast cancer patients in the local population.
Keywords: Breast cancer; lymphovascular invasion; mammography; ultrasound; pre-operative prediction; Pakistan