An update from the University of Dundee on the outcomes from the Ruth Prosser 2023 award:
I can confirm that following successful ethical approval we have been able to conduct the study as planned. We have performed the preliminary analysis and have had an abstract accepted for presentation at the British Society of Breast Radiologist (BSBR) annual meeting. We intend to submit two further abstracts to upcoming meetings; Symposium Mammographicum and the Scottish Breast Radiology Forum (SBRF) to ensure wide dissemination of results.
Abstract:
Accuracy of abbreviated MRI (abMRI) for predicting pathological complete response (pCR) following neoadjuvant chemotherapy (NACT): a retrospective image analysis study
Background:
With developments in oncological treatment, more women are receiving NACT. Imaging response assessment is essential for surgical decision-making. Full protocol MRI (fpMRI) is considered gold-standard for predicting pCR, but is time-consuming limiting its availability and patient tolerability. AbMRI utilises fewer MRI sequences and is quicker to perform and report. Emerging evidence shows abMRI has similar accuracy to fpMRI for screening. This pilot study investigates whether abMRI could offer an alternative for NACT response assessment.
Methods:
Following ethical approval, MRIs from sixty consecutive NACT patients from 01/01/2017 were identified. Pseudonymised end-of-treatment abMRIs were synthesised and interpreted by a radiologist blinded to surgical pathology. Only pseudonymised pre-treatment fpMRI were available for comparison. A researcher extracted pathology and fpMRI report data. FpMRI and abMRI findings were compared with surgical pathology (at breast level). Accuracy of predicting pCR: ypT0 (no residual disease) and ypTis (residual in situ disease) was calculated.
Results:
59 patients with 63 cancers (four bilateral) were included. pCR occurred in 9(ypT0) and 16(ypTis), breasts. For ypT0: Sensitivity:88.9% vs 100%, specificity:96.3% vs 77.8%, and accuracy:95.2% vs 81.0%, for fpMRI vs abMRI. For ypTis: sensitivity:60.0% vs 81.3%, specificity:97.9% vs 83.0%, and accuracy:88.9% vs 82.54%, for fpMRI vs abMRI.
Summary:
AbMRI has lower accuracy for pCR than fpMRI. FpMRI tended to over-estimate residual disease. AbMRI tended to under-estimate residual disease, possibly due to absence of later sequences to capture subtle type 1 enhancement. Furthermore, unlike in a clinical setting, the blinded reader did not have access to conventional imaging which may influence reporting.