Quantification of islet loss during immune rejection using iron-labeled islet cells by 3T MRI in the rat model

S. Borot¹, L. Crowe², F. Ris¹, P. Morel¹, S. Nielles-Vallespin³, P. Speier³, S. Masson-Charvier¹, C. Toso¹, D. Bosco¹, J.P. Vallée², T. Berney^1
1 Geneva University Hospitals, Islet Isolation and Transplantation Center, GENEVA 14, Switzerland; ² Radiology, Geneva University Hospital, Geneva, Switzerland; ³ Siemens AG Medical Solutions, Erlangen, Germany

Background: Monitoring the fate of transplanted islets remains a major challenge in clinical islet transplantation (IT). We developed a MRI imaging protocol (3D difference ultra-short echo-time = dUTE), resulting in positive contrast images of superparamagnetic iron-oxide (SPIO) nanoparticles-labeled islets transplanted in rat. Our aim was to compare the evolution of the MRI signal in 3 types of IT in the rat model.

Methods: Syngeneic and allogeneic SPIO-labeled islets (ferucarbotran, 280µg/ml iron) were injected intraportally into streptozotocin-induced diabetic Lewis rats. Xenogeneic human islets were transplanted into normoglycaemic rats and graft functionality was evaluated by serum human C-peptide levels. Images were performed on a 3T clinical scanner, from day 0 up to day 106. An intensity threshold was applied within the liver region, giving automatically the number of dUTE-enhanced pixels, allowing quantification . Histological studies included insulin, CD4 and CD8 staining and iron detection.

Results: Decay rates for the 3 types of transplantation were different (Figure 1). The syngeneic graft signal showed a 20% decrease during the first 2 weeks and remained stable up thereafter. For allogeneic transplantation, islet rejection (G>20 mmol/l) occurred at day 7.7±0.5 and 41%±12 of the initial signal was lost. In the xenogeneic model, 43%±8 of the initial signal was lost by day 3, when significant basal and stimulated human C-peptide levels were not detected any longer. Islet rejection was confirmed by islet CD4+ and CD8+ cell infiltration and loss of insulin staining.

Conclusion: After intra-portal IT and during immune rejection, the loss of SPIO-labeled islets can be monitored with clinical-grade 3T MRI imaging using a semi-automatic quantification method. The decay of the signal is correlated with the graft function and histological findings.