Retrospective study.  All patients who had histopathological analysis of a surgical specimen of ILC in NHS Grampian in 2016/17 were reviewed (screening and symptomatic patients).  Pre-operative conventional imaging of mammogram and ultrasound followed by MRI was standard protocol.  Additional findings on MRI were investigated via second look ultrasound +/- biopsy and some patients required MRI guided biopsy.  Twenty eight patients were excluded as they did not have imaging as per protocol and / or imaging and reports were not available to view. 

All MRI scans were performed with the Siemens Avanto 1.5 Tesla scanner.  Imaging and reports were reviewed via the Soliton Reporting+^TM system (version 1.2.11.10).  Pathological reports were reviewed via the online electronic patient record (NHS Grampian SCI Store version 8.5). 

Mammographic density was assessed by a single radiologist who is an experienced registrar.  This was graded quantitatively according to the Breast Imaging-Reporting and Data System (BI-RADS), 4th edition, which documents area-based percent MD in quartiles [9].  Categories 1 and 2 were considered low density and categories 3 and 4 as high density.

In order to compare the tumour on conventional imaging with MRI we recorded the largest diameter of the largest lesion and documented whether the disease was unifocal or more than one lesion.  If the index lesion was larger by 5 mm or more on MRI compared with conventional imaging this was considered to represent an 'increase in size of the index lesion'.  The other categories were, 'No additional disease', 'Additional ipsilateral lesion(s)' and 'Additional contralateral lesion(s)'. 

Statistical analysis was performed.