Application of molecular imaging to address current clinical challenges in breast cancer

• Main Author: Glendenning, Jennifer Louise
• Published: King's College London (University of London) 2017
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.733375

Breast cancer has multiple sub-types, designated in clinical practice by the presence or absence of prognostic and predictive biomarkers: oestrogen receptor, progesterone receptor and the human epidermal growth receptor 2 (HER2). Molecular imaging offers the opportunity firstly to non-invasively and dynamically interrogate in vivo tumour sites to provide prognostic and predictive evaluation at clinically meaningful on-treatment time points. Secondly, it may address diagnostic challenges in the metastatic setting by enabling assessment of inter- and intra-lesion heterogeneity without requirement for tissue acquisition from multiple sites. This thesis describes pre-clinical evaluation of a novel HER2 targeted DARPin radiotracer as an in vivo diagnostic across multiple tumour sites in murine models. Metastatic breast xenograft models were developed and validated using bioluminescence imaging and definitive histology for in vivo evaluation of a novel HER2 targeted DARPin radiotracer. In subsequent preclinical testing the DARPin radio-tracer failed to differentiate HER2 status of pre-clinical tumour xenografts models and this data raises significant questions regarding suitability of the DARPin radiotracer for clinical evaluation as a HER2 diagnostic. Additionally this work reports the set-up of a Phase 2 imaging feasibility study designed in two parts to evaluate post-cycle 1 PET response using the FLT- and FDG-PET imaging tracers to address clinical questions concerning tracer selection, scan acquisition and interpretation for validation of Positron Emission Tomography (PET) response as a predictive biomarker of neoadjuvant response in the triple negative breast cancer (TNBC) phenotype. Part A (participant recruitment completed) delivers the first phenotype specific repeatability constraints for the most commonly reported standardised uptake parameters (SUV); maximum (SUVmax), mean (SUVmean), peak (SUVpeak) and lean body mass corrected peak (SULpeak), assessed at conventional (90 minutes) and exploratory (120 and 180 minute) acquisition time points. The TNBC SUV intrinsic variability was 12-24% in both tracers and is dependent on scan acquisition time and SUV parameter. The FDG tracer has progressed to the second phase, Part B, to provide the first TNBC phenotype specific response data at a post-cycle 1 time point. The data suggests SUV change can predict later residual cancer burden and that >40% threshold change will be required to differentiate RCB 0-1 vs 2-3 response, a change that exceeds current EORTC/PERCIST recommendations for solid tumour chemotherapy response prediction. The study will inform future use of early FDG-PET as an exploratory biomarker in window of opportunity and novel therapy neo-adjuvant trials in TNBC.