Kinetics of protein-based in vivo imaging tracers for positron emission tomography

Abstract

Within the framework of the “Sel-tag imaging project”, a novel method was used to rapidly label protein tracers and the in vivo targeting abilities of these tracers were studied in animal models of cancer using a preclinical positron emission tomography (PET) camera. To first evaluate and optimize preclinically the use of PET tracers can facilitate their translation to and implementation in human patient studies. The ultimate goal of the different projects within the Sel-tag imaging project was to find imaging biomarkers that could potentially be used for individualizing cancer treatment and thereby improve the therapeutic results.

This thesis focuses on methods employed to describe the distribution of these protein-based tracers in human xenografts. Many of the techniques used had been developed for other imaging circumstances. Therefore verification for these imaging applications was an important aspect of these papers.

Paper I examined the distribution in a tumour of a medium-sized AnnexinA5-based tracer that targeted phosphatidylserine externalised during cell death in tumours in two cases; first, with no pre-treatment (baseline) and, second, after pre-treatment with a chemotherapeutic agent. Small differences between tracer uptakes in the two cases required a macro parameter analysis method for quantifications. Evaluations of the influence of the enhanced permeability and retention effect by using a size-matched control were introduced. The AnnexinA5 results were compared to those of the metabolic tracer [18F]FDG and complemented with circulating serum markers to increase sensitivity.

Paper II extended the analysis in paper I to incorporate more verifications that were also more thorough. The choice of input (blood or reference tissue) and the statistical significance of intergroup comparisons when using conventional uptake measurements and the more involved macro parameter analyses like in paper I were compared. We also proposed that distribution volume ratio was a more appropriate quantification parameter concept for these protein-based tracers with relatively large non-specific uptake.

Paper III assessed the smaller Affibody™ tracer ZHER2:342 as an imaging biomarker for human epidermal growth factor 2 (HER2), whose overexpressions are associated with a poor prognosis for breast cancer patients. In order to demonstrate specific binding to HER2, pre-treatment of the tumour with unlabelled protein and uptake in xenografts with low HER2 expression was evaluated. Ex vivo immunohistochemistry of expression levels supported the imaging results.

Paper IV examined a radiopharmaceutical that targeted the epidermal growth factor receptor (EGFR), whose overexposure in tumours is associated with a negative prognosis. Again an Affibody™ molecule, (ZEGFR:2377), was used and, as in in paper I, a size-matched control was also used to estimate the non-specific uptake. Uptakes, quantified by conventional uptake methods, varied in tumours with different EGFR expression levels. Ex vivo analyses of expression levels were also performed.

Paper V addressed the non-uniform (heterogeneous) uptake of different tracers in a tumour tissue. An algorithm was written that aimed at incorporating all relevant aspects that will influence non-uniformity. Histograms were generated that visualized how the frequency and spread of deviations contributed to the heterogeneity. These aspects could not always be attended in a direct manner, but instead had to be handled in an indirect way. The effect of varying imaging parameters was examined as part of the validation procedure. The method developed is a robust, user-friendly tool for comparing heterogeneity in similar volume preclinical tumor tissues.