Résumé : Purpose: Cell therapy using bone marrow mesenchymal stem cells (BMSCs) shows promise in the treatment of myocardial infarction (MI) but accurate cell delivery within MI areas remains critical. In the present study, we tested the feasibility of in vivo pinhole SPECT imaging for monitoring the sites of intramyocardial implanted BMSCs in relation to targeted MI areas in rats. Methods: BMSCs were labelled with 111In-oxine and injected within the fibrotic areas of 3-month-old MI in ten rats. Two days later, dual 111In/ 99mTc-sestamibi pinhole SPECT was recorded for localisation of 111In-BMSCs on a 15-segment left ventricular (LV) division. Additional 99mTc-sestamibi pinhole SPECT had been performed 1 month earlier and on the day before transplantation. In vitro counting on histological sections was used to validate the pinhole SPECT determination of 111In-BMSC activity within LV segments. Results: The underperfused MI area (segments with <70% uptake) was stable between the 99mTc- sestamibi SPECT study recorded at 1 month (4.6±1.9 segments) and at 1 day (4.7±2.3 segments) before transplantation. 111In-BMSCs were detected by dual-energy SPECT in 56 segments: 33 (59%) were underperfused MI segments but 23 (41%) were not (14 adjacent and nine remote segments). Finally, 111In-labelled BMSCs were not detected in 14 out of the 47 (30%) underperfused MI segments. Conclusion: When BMSCs are injected within MI areas in rats, sites of early cell retention do not always match the targeted MI areas. The dual-energy pinhole SPECT technique may be used for monitoring the sites of early retention of implanted BMSCs and the data obtained may have critical importance when analysing the effects of cardiac cell therapy. © Springer-Verlag 2006.