BACKGROUND

Invasive lobular carcinoma (ILC) is the second most common subtype of breast cancer, comprising 10-15% of cases. Due to its diffuse growth pattern, conventional imaging techniques have decreased sensitivity for ILC. While breast magnetic resonance imaging (MRI) is often recommended for ILC, its accuracy following neoadjuvant therapy is unknown. We evaluated the accuracy of post-treatment MRI and examined the impact on surgical outcomes.

PATIENTS AND METHODS

We retrospectively analyzed 129 patients with ILC who underwent neoadjuvant chemotherapy (NAC) or endocrine therapy (NET) and had post-treatment MRI. We considered a 0.5 cm difference between longest tumor diameter on MRI and pathologic tumor size to be discrepant. Tumor imaging phenotype was categorized as mass, non-mass enhancement (NME), or mass + NME. We evaluated concordance between imaging and pathology by tumor phenotype and associations with positive margin rates using Stata 18.0.

RESULTS

Post-treatment MRI underestimated tumor size in 52.5% of cases, was concordant in 25.3%, and overestimated in 22.2%. The presence of NME was associated with a higher rate of tumor size underestimation (62.5% versus 39.5% for mass only, p = 0.023); mean underestimation was 3.4 cm in those with NME. Underestimation was associated with higher positive margin rates following breast-conserving surgery (p = 0.018). Finally, among patients with complete imaging response on MRI, 93.3% had residual invasive disease on pathology.

CONCLUSIONS

Following neoadjuvant therapy, post-treatment MRI frequently underestimates tumor size in ILC, particularly in tumors with NME. Surgeons should consider these imaging limitations when planning resection, which could improve surgical outcomes.