Sanjay Ramanlal Patel MBBS, MRCS, Presenter: Nothing to Disclose

Prasanna Perera MBBS, MRCP, Abstract Co-Author: Nothing to Disclose

Paul Malcolm MRCP, FRCR, Abstract Co-Author: Nothing to Disclose

Anne Girling, Abstract Co-Author: Nothing to Disclose

Melanie Shaw MBBS, FRCR, Abstract Co-Author: Nothing to Disclose

The decision to proceed to wide local excision or mastectomy depends on several factors including the size of the primary tumour. With the advent of new treatment modalities the need for accurate preoperative assessment has increased. The purpose of this study is to measure the accuracy of MR measurement of the primary tumour by comparison with post-operative macroscopic and microscopic histology  

Consecutive patients over a six month period were studied retrospectively. All 21 patients had a biopsy proven diagnosis of breast carcinoma and had undergone MR examination prior to surgery. Patients who underwent neo-adjuvant chemotherapy were excluded. The maximum measurement of the primary lesion identified at MR and confirmed at histology was recorded. This size was compared with the final histopathological macroscopic and microscopic measurements.

Histology included ductal and lobular carcinoma, with both invasive and carcinoma in situ pattern. Some tumours were of mixed histology. MR measurement was within 5mm of the microscopic tumour size measurement in 14 patients ( 66.6%). MR overestimated tumour size In 6 patients (28.6%) and underestimated size in 1 case (4.8%). The mean difference between MRI and microscopic histology for the overestimated lesions was 2.12cm±2.4cm standard deviation (SD). The macroscopic evaluation by histopathologist was compared to microscopic size measurement. There was correspondence within 5mm in 12 patients (57.1%). In 8 cases (38.1%) size was overestimated but under estimated in one case (4.8%) corresponding to the case underestimated on MR. The mean difference in the measurements for MRI and macroscopic examination was 2.27 ± 4.31cm. In the group in which MR overestimated lesion size compared with microscopy, the difference in MRI measurements with macroscopic histological examination with mean of 1.9 ± 1.98.      

This study has shown good correspondence between MR and microscopic lesion measurement but a tendency to overestimate lesion size at MR in a minority of cases. This is similar to the degree of overestimation at histological macroscopic measurement.

Pre-operative MR is currently used in a minority of cases for particular indications. Correspondence to microscopic lesion size is generally good but MR overestimates the size of some lesions.

Patel, S, Perera, P, Malcolm, P, Girling, A, Shaw, M, The Value of MRI in Preoperative Assessment of Size of Primary Breast Cancer in Comparison with Histology.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8007841.html