Magnetic resonance imaging assessment of functional differences between kidneys in vivo and during ex vivo normothermic machine perfusion
Tim Hamelink1,4,6, Veerle Lantinga1,6, Baran Ogurlu1, Carol Pamplona1, Sigrid Bennedsgaard4, Haiyun Qi3, Johannes Castelein2, Marco Eijken3, Chris Jaynes1,7, Ulrike Dydak6, Bente Jespersen3, Henri Leuvenink1, Esben Hansen5, Christopher Laustsen5, Stephen Ringgaard5, Anna Krarup Keller4, Ronald Borra2, Cyril Moers1.
1Surgery, University Medical Center Groningen, Groningen, Netherlands; 2Radiology, University Medical Center Groningen, Groningen, Netherlands; 3Renal Medicine, Aarhus University Hospital, Aarhus, Aarhus, Denmark; 4Urology, Aarhus University Hospital, Aarhus, Denmark; 5Clinical Medicine, Aarhus University Hospital, Aarhus, Denmark; 6School of Health Sciences, Purdue University, West Lafayette, United States; 734, Lives, West Lafayette, United States
Background: Normothermic machine perfusion (NMP) is a promising method for pre-transplant donor kidney quality assessment. Although its potential is increasingly being recognized, it remains unclear to what extend NMP parameters convey information about graft viability. To compare renal biology during NMP with in vivo renal function we combined in vivo non-invasive functional magnetic resonance imaging (MRI) with renal normothermic perfusion in a porcine model. This project aims to determine regional renal tissue oxygenation, blood flow in the microcirculation, and diffusion patterns.
Methods: Pigs (n=30) weighing 80 kg were anesthetized and brought into a clinical-grade MRI scanner (Siemens Skyra 3T). In vivo MRI scans were performed to provide information about regional tissue oxygenation using T2* mapping, blood flow in the microcirculation using arterial spin labeling (ASL), and water diffusion patterns using diffusion-weighted imaging (DWI). Subsequently, a bilateral nephrectomy was performed to retrieve kidney pairs, which were randomized to sustain either no warm ischemia (WI) or 75 min WI. After 4-5 hours of cold machine preservation, both kidneys were simultaneously connected to an MRI-compatible NMP circuit and perfused for 6 hours. Hourly, T2* maps, ASL, and DWI images were acquired from both kidneys. Regions of interest were drawn in the cortex and medulla to calculate the mean signal intensity.
Results: In vivo mean T2* corticomedullar (CM) ratio (1.7±0.2) differed significantly from the mean ex vivo CM ratio of the no WI group (0.6±0.1, P<0.0001) and the 75 min WI group (0.6±0.2, P<0.0001). Calculated ASL CM ratios were significantly lower in the 75 min WI group (5.4±0.9) compared to the in vivo measurements (8.7±0.6, P<0.0001) and the no WI group (8.6±1.6, P < 0.0001). In vivo cortical DWI values (2.1±0.15 x 10-3 mm2/s) were significantly higher compared to the no WI group (1.46±0.25, P<0.0001) and the 75 min WI group (1.76±0.09, P=0.002).
Conclusion: These results provide the first evidence for remarkable differences in regional tissue oxygenation, blood flow in the microcirculation, and diffusion patterns between a normal physiological in vivo environment and during ex vivo normothermic machine perfusion. These findings highlight that renal function during ex vivo perfusion differs from what we are used to in our in vivo reference frame. Therefore, organ viability assessment during NMP should likely consider parameters other than those functional markers that are common in vivo.
From May to July 2023, the results from the pig study will be validated in discarded human donor kidneys (n=25) at Purdue University in collaboration with 34 Lives and the University Medical Center Groningen. These results will be available in September 2023.
