Multiparametric MRI and clinical factors for predicting early recurrence of breast cancer after modified radical mastectomy

Abstract

Objective

To determine whether preoperative multiparametric MRI features and clinicopathological factors can predict early recurrence in breast cancer patients following modified radical mastectomy (MRM).

Materials and methods

This retrospective study included 132 patients who underwent MRM at two hospitals from January 2012 to December 2022. Patients were classified into early recurrence (disease-free survival ≤3 years, n = 86) or late recurrence (>3 years, n = 46). We collected preoperative MRI features (tumor size, morphology, enhancement patterns, apparent diffusion coefficient (ADC) values, early enhancement rate, time-signal intensity curves) and clinical/pathological factors (age, menstruation status, neoadjuvant chemotherapy, T/N/TNM stage, tumor grade, axillary nodal involvement, vascular/nerve/nipple/skin invasion, ER, PR, HER-2, Ki-67, molecular subtype, and adjuvant therapies). Univariate and multivariate analyses (binary logistic and Cox regression) identified independent predictors of early recurrence and their impact on disease-free survival. Predictive accuracy was assessed by ROC curves.

Results

Significant factors associated with early recurrence (P < 0.05) included tumor size, ADC value, early enhancement rate, axillary lymph node metastasis, T and TNM stage, vascular invasion, ER, PR, HER-2, molecular subtype, and endocrine therapy. Multivariate analysis identified early enhancement rate, lower ADC value, and larger tumor diameter as independent predictors (AUC = 0.899). These MRI factors, together with lack of endocrine therapy, also independently influenced disease-free survival.

Conclusion

Preoperative multiparametric MRI features, especially early enhancement rate and tumor size, combined with endocrine therapy status, can predict early recurrence, enhancing prognosis for patients undergoing MRM.