Monitoring kidney viability before transplantation by means of 31P CSI and oxygenated hypothermic perfusion

Introduction Monitoring viability of kidney before transplantation has become critical with the increasing need of marginal organs. 31P MRS is able to assess energy metabolites such as nucleotide triphosphates (NTP), inorganic phosphate (Pi) and cell phosphomonoester (PME). During cold storage, ATP decreases rapidly and is partially metabolized in PME. Therefore PME/Pi has been proposed as useful indicator to assess kidney viability [1]. In this paper, we show that ATP can be directly detected in the presence of O2, which provides direct evidence of cell integrity. This new approach can be used for kidney monitoring during cold storage before transplantation. In this paper, we have tested the effect of cold static storage (CSS) versus oxygenated perfusion on kidney metabolic status. Material and Methods HPP perfusion MRI compatible oxygenated hypothermic pulsatile perfusion (O2+HPP) system was used to maintain the kidney in cold storage (4° C) using 25 mmole phosphorus medium (KPS-1). Oxygen tension of pO2 ≥ 100kPa was maintained during HPP. Two storage conditions were tested: In the first experience, kidney was perfused (O2+HPP) immediately after resection for 8 hours, than was maintained in cold static storage (CSS) for 10 hours, and reperfused (O2+HPP) for 10 hours. In the second experiment, the kidney was maintained under CSS for 18 hours after resection, following by 10 hours of O2+HPP. MR acquisition MRI/MRS was performed at 3T (Trio, Siemens), with home made 31P interface and surface coil. 1H imaging and shimming were performed with the body coil. T2 sequence (TSE, TR 5000ms, TE 108 ms, 3 mm slices) was used as localizer. 31P MRS consisted of 3D CSI, 16x16x8, resolution 1.56x1.56x2.5 cm, TR 1000 ms, weighted k-space, acquisition time 44 min. Data Analysis Processing was performed with SAGE software (General Electric Medical Systems) and consisted of zero filling, 10 Hz exponential apodization, and FFT. Acquisition delay (2.5 ms) was corrected with a first order phase correction and sinc deconvolution. Individual spectra were corrected for frequency shift and averaged. The assessment of the metabolites concentrations was performed using the Pi peak as internal reference (25 mMolar).