Using a cyclic RGD peptide, P975 conjugated to a Gd-DOTA chelate, Klink et al. demonstrated enhanced MRI signal of the carotid artery in vivo using a murine model of arachidonic acid induced thrombosis. T1-weighted MRI was used to image the black blood of the healthy left carotid and an occluded right carotid artery. After injection of P975, increased in Gd signal intensity was observed at 30 min and 120 min, where no signal enhancement was noted for the untargeted Gd or in the healthy carotid artery (Klink et al., 2010). Two other radiolabelled RGD-based constructs have proceeded to clinical trials for the diagnosis of deep vein thrombosis (DVT) via SPECT imaging (Muto et al., 1995; Lister-James et al., 1996; Taillefer et al., 1999, 2000; Klem et al., 2000). Unfortunately, both 99mTc-P280 and99mTc-DMP444 have been discontinued because of their low sensitivity and specificity for the diagnosis of PE, contributed by prolonged radioactivity readout in both the chest region and the blood pool, as well as their low uptake by the thrombus (Carretta et al., 1999; Taillefer et al., 2000; Oliveira and Caravan, 2017). However, RGD-based contrast agents have two major limitations. 1) As RGD peptides do not provide specificity for activated GPIIb/IIIa , RGD-based contrast agents will bind to all circulating platelets. Thrombus imaging can be only achieved by the comparison of the static thrombus signal to the moving blood signal. 2) Several other integrins act as receptors for RGD containing ligands (Wang and Peter, 2017). Indeed, this has been used for molecular imaging approaches since RGD peptides have been used to target the vitronectin receptor (integrin αvβ3, CD51/CD61) (Golestani et al., 2015; Sun Yoo et al., 2015) and are widely used in cancer imaging (Alonso et al., 2009; Melemenidis et al., 2015; Withofs et al., 2015).
Using the antibody fragment abciximab (Reopro®), Alonso et al. demonstrated that preincubation of abciximab-microbubbles with human thrombi generated ex vivo resulted in enhanced ultrasound signals after i.v. injection. Visualisation of the human clot in the artery was achieved using ultrasound imaging, and as a negative control, upon ultrasonic bubble destruction, the clot lost its echogenicity/brightness (Alonso et al., 2007). However, Reopro has been developed as an anti-platelet drug and is known to bind to GPIIb/IIIa on all human platelets, independent of their activation state (Coller, 1999; Schwarz et al., 2002). The use of Reopro as an imaging agent faces the same limitations as RGD-based contrast agents with binding to all circulating platelets and therefore the excistence of a circulating blood background signal. Also abciximab blocks fibrinogen binding and therefore inhibits the function of circulating platelets, potentially resulting in an increased bleeding risk (Peter et al., 2000; Lele et al., 2001).
The development of a single-chain antibody (scFv) that specifically targets the activated form of the GPIIb/IIIa receptor (scFvanti-GPIIb/IIIa) has provided the unique opportunity to do molecular imaging of activated platelets. This approach has been successfully used to image diseases with central involvement of activated platelets such as thrombosis, inflammation and malignant diseases (Wang and Peter, 2017; Yap et al., 2017). This scFvanti-GPIIb/IIIa does not bind to the resting platelets (Figure 6) in the circulation and does not cause bleeding complications (Schwarz et al., 2006; Wang et al., 2012). The scFvanti-GPIIb/IIIa has been successfully used with a wide range of contrast agents, such as ultrasound microbubbles (Wang et al., 2012, 2014, 2016) and MPIOs (von Elverfeldt et al., 2014), and demonstrated highly favourable binding capabilities to microthrombi under flow conditions (Figure 7 and 8).