To the editor:

In vivo visualization of transplanted pancreatic islets might advance islet transplantation as treatment for diabetes, helping to confirm technical success and diagnose rejection. Evgenov et al.1 described a method of labeling human islets with a superparamagnetic contrast agent and the subsequent detection of the islets using magnetic resonance imaging after transplantation into the liver or the renal subcapsular space in mice. The islets or islet clusters were apparent as hypointense spots on T2-weighted magnetic resonance images for up to 180 days.

Although we appreciate the technical excellence and possible application of this method in clinical practice, we disagree with the interpretation of our previous report offering very similar data using the clinically approved superparamagnetic magnetic resonance contrast agent ferucarbotran in a rat model. We clearly showed that transplantation of islets in vitro–labeled with iron successfully treated experimental diabetes and that the presence of iron-labeled islets transplanted into the liver could be successfully monitored on serial magnetic resonance scans for as long as 22 weeks after transplantation2.

In addition, we attempted to estimate the average iron content in islets following a 2-day culture in the presence of ferucarbotran. We object to the characterization of our result as “seem[ing] unrealistic” as a value for iron uptake per cell, as we had not reported uptake by single cells. The amount of iron per cell was calculated from the iron measured in islet samples. Soon after placement in the culture medium, iron within an islet is localized not only to intracellular but also to extracellular spaces, which we stated clearly2.

We were also surprised that Evgenov et al.1 referred to our paper only regarding the quantification of islet iron content and the impact of ferucarbotran on in vitro islet function. We were the first to report on the use of magnetic resonance imaging for in vivo islet visualization, and this fact was not included in their article. Although ours was only a pilot study, our results clearly showed that in vivo imaging of transplanted islets is feasible. As ferucarbotran labeling did not substantially impair in vitro and in vivo islet function, we believe that our technique of islet visualization using this agent is closer to application in human islet transplantation.