Purpose: The purpose of the study was to estimate the receptor-ligand binding of an arginine-glycine-aspartic acid (RGD) peptide in somatic tumours. To this aim, we employed dynamic positron emission tomography (PET) data obtained from breast cancer patients with metastases, studied with the α(v)β(3/5) integrin receptor radioligand [(18)F]fluciclatide.
Methods: First, compartmental modelling and spectral analysis with arterial input function were performed at the region of interest (ROI) level in healthy lung and liver, and in lung and liver metastases; compartmental modelling was also carried out at the pixel level. The selection of the most appropriate indexes for tumour/healthy tissue differentiation and for estimation of specific binding was then assessed.
Results: The two-tissue reversible model emerged as the best according to the Akaike Information Criterion. Spectral analysis confirmed the reversibility of tracer kinetics. Values of kinetic parameters, estimated as mean from parametric maps, correlated well with those computed from ROI analysis. The volume of distribution V(T) was on average higher in lung metastases than in the healthy lung, but lower in liver metastases than in the healthy liver. In agreement with the expected higher α(v)β(3/5) expression in pathology, k(3) and k(3)/k(4) were both remarkably higher in metastases, which makes them more suitable than V(T) for tumour/healthy tissue differentiation. The ratio k(3)/k(4), in particular, appeared a reasonable measure of specific binding.
Conclusion: Besides establishing the best quantitative approaches for the analysis of [(18)F]fluciclatide data, this study indicated that the k(3)/k(4) ratio is a reasonable measure of specific binding, suggesting that this index can be used to estimate α(v)β(3/5) receptor expression in oncology, although further studies are necessary to validate this hypothesis.