Which factors influence MRI-pathology concordance of tumour size measurements in breast cancer?

ABSTRACT

OBJECTIVES: To assess MRI-pathology concordance and factors influencing tumour size measurement in breast cancer. MATERIALS AND METHODS: MRI tumour size (greatest diameter in anatomical planes (MRI-In-Plane) and greatest diameter along main tumour axis (MRI-MPR)) of 115 consecutive breast lesions (59 invasive lobular carcinoma, 46 invasive ductal carcinoma, and 10 ductal carcinoma in situ) was retrospectively compared to size measured at histopathology (pT size (Path-TNM) and greatest tumour diameter as relevant for excision (Path-Diameter; reference standard)). Histopathological tumour types, preoperative palpability, surgical management, additional high-risk lesions, and BI-RADS lesion type (mass versus non-mass enhancements) were assessed as possible influencing factors. RESULTS: Systematic errors were most pronounced between MRI-MPR and Path-TNM (7.1 mm, limits of agreement (LoA) [-21.7; 35.9]), and were lowest between MRI-In-Plane and Path-Diameter (0.2 mm, LoA [-19.7; 20.1]). Concordance rate of MRI-In-Plane with Path-Diameter was 86% (97/113), overestimation 9% (10/113) and underestimation 5% (6/113); BI-RADS mass lesions were overestimated in 7% (6/81) versus 41% (13/32) for non-mass enhancements. On multivariate analysis only BI-RADS lesion type significantly influenced MRI-pathology concordance (p < 0.001). 2/59 (3%) ILC did not enhance. CONCLUSION: Concordance rate varies according to the execution of MRI and histopathological measurements. Beyond this only non-mass enhancement significantly predicted discordance. KEY POINTS • Execution and scope of MRI and histopathological size measurements influence concordance rate. • Non-mass like enhancement predicts discordance. • Additional high-risk lesions in proximity of tumour do not cause measurement discordance. • Low percentage of ILC do not enhance at all.