It is difficult to obtain biopsies of islets transplanted into the liver. Islets are spread throughout the liver parenchyma, and liver biopsies per se are not free of risk. However, in the muscle, a biopsy would be simple to obtain. Preemptively dividing the transplanted islet mass, designating a specific islet mass to a separate site intended for potential biopsies, would allow subsequent studies to identify early markers for rejection.
The feasibility of imaging transplanted islets through PET techniques depends to a large extent on the site of implantation. This can be understood by examining the composition of a PET tracer signal emitted within a given region in vivo. In the main, a signal containing a composite of specific binding is produced (i), or the PET probe bound to the receptor of interest is proportional to receptor density. Displaying nonspecific binding (ii), the PET probe may also bind to other structures, revealing vascular contribution (iii). Vascular contribution is vital, as the tissue uptake of tracer is limited by local perfusion, potentiating an underestimation of uptake in grafts with low revascularization. All of these factors influence our ability to study islet grafts in different tissues.
With these issues in mind, we can see that longitudinal noninvasive visualization of hepatic grafts presents a considerable challenge, with substantial graft dilution (i). In addition, it is difficult to design PET tracers with a low hepatic background signal. Most tracers are metabolized in the liver (ii). The perfusion (iii), however, is sufficient to transport the tracer to engrafted islets in hepatic sinusoids shortly after transplantation.
This paradigm is reversed when considering intramuscular islet grafts. Grafts are generally pure and concentrated (i), with the above-mentioned PET tracers effecting a low to negligible uptake in muscle tissue (ii). Revascularization is therefore the limiting factor for visualization of the graft. It has recently been shown that intramuscular islet grafts attain a vascularization comparable to that in pancreatic islets after approximately 2 weeks [
32]. Considering these illustrations, quantification of intramuscular islet grafts containing an adequate volume is potentially achievable after engraftment and revascularization has been completed.
Progress with visualizing has been made both in preclinical and clinical studies of intrahepatic and intramuscularly transplanted islets using PET. The hepatic distribution and survival of islets during the peritransplant phase, following intraportal islet transplantation, has been studied with ex vivo labeling of islets by [
18F]FDG prior to infusion in murine [
45] and porcine [
46] models as well as in the clinic [
7]. However, longitudinal studies are not yet possible using the ex vivo labeling methodology.
Uptake in an intramuscular islet graft of a vesicular monoamine transporter 2-(VMAT2-) ligand, [
11C]DTBZ, correlated well to the observable decrease in blood sugar of a preclinical STZ mouse model in a study by Witkowski et al. [
47]. However, interpretation of the results and translation to the clinical situation is problematic when considering that the islets were implanted using a bioscaffold.
Pattou et al. [
48] showed proof-of-principle in the clinic by visualizing islets transplanted to the brachioradialis muscle in a type 1 diabetic patient by intravenous administration of a [
111In]-labeled exendin-4, a GLP-1R ligand, using SPECT, an imaging modality related to PET.
The VMAT2 ligand [18F]FE-DTBZ-d4 (see ) and the catecholamine precursor [18F]L-DOPA (unpublished data) are currently being investigated as biomarkers for transplanted beta cells in two ongoing preclinical studies of intramuscularly transplanted islets in mice.