Introduction

Lung transplantation is a curative treatment for many with end-stage lung disease, but a shortage of transplantable donor lungs means that only slightly more than 30% of waiting list patients receive transplant, and many die before organs become available. To address this critical issue, numerous efforts aimed at expanding the organ pool by utilizing lungs from donors who are traditionally disqualified—e.g., age>45, smoking history, brain dead, etc.—are currently underway¹. Ex-Vivo Lung Perfusion (EVLP) can be used to preserve both ideal and extended criteria donor lungs by delivering a flowing perfusate through the organ, thereby providing a mechanism for therapeutic intervention, oxygenation, nutrient delivery, and assessment that has the potential to improve transplant outcomes². EVLP differs from the current gold standard method of donor lung maintenance, which involves flushing and statically storing the excised lung in a bag of preservation solution, this bag is then placed in two bags of saline in a 4-8^oC icebox³. Evaluating metabolic biomarkers in donor lungs undergoing preservation may provide an early signal of lung viability. In this study in a rat model, magnetic resonance spectroscopy was used to compare preservation using un-ventilated normothermic EVLP and the gold standard (GS) of cold static storage over a period of extended (5-hour) ischemia by monitoring lungs’ energy status.

Method

Sprague Dawley (n=8) male rats were mechanically ventilated with oxygen for 10 minutes to remove gaseous nitrogen from the lungs (VT=3.0 ml, f=38 min^-1, FiO₂=1.0, PEEP=3 cmH2O, 10% sigh, 33:67 I:E). Lungs were perfused through the pulmonary artery at a 10 ml/min flow rate with 500 ml of an extracellular type low potassium perfusate: modified Krebs-Henseleit buffer (119 mM NaCl, 25 mM NaHCO₃, 1.3 mM CaCl₂, 1.2 mM MgSO₄, 4.7 mM KCl, 10 mM glucose, 2 mM lactate, 0.2 mM pyruvate, and 3% (w/v) fatty-acid-free bovine serum albumin). The trachea was sutured upon expiration, and lungs were excised and placed in a 20 mm NMR tube before being transferred to the MRI machine (AVANCE III 9.4T, Bruker Inc.). This setup is shown in Figure 1. The perfusate was passed through an oxygenating column under a constant flow of oxygen and warmed via passage through heated water-jacketed tubing to maintain a temperature of ~37^oC in the NMR tube. pH was periodically adjusted to ~7.4 by adding 1.2 N HCl or NaOH to the perfusate reservoir. For the gold standard cold storage model (GS), six lungs were removed from the magnet after 1 hour of perfusion, flushed and statically stored in 20 ml of 4^oC Perfadex+^R (XVIVO Perfusion) for 3 hours, then gradually rewarmed, perfused at 37^oC, and returned to the magnet for 1 hour. The remaining lungs (n=2) remained in the spectrometer throughout the entire 5-hour study (EVLP model). ³¹P spectra were obtained using a 25 mm dual-tuned (¹H/³¹P) coil (Bruker Inc.) with the following acquisition parameters: TR=1 s, FA=60^o, SW=8 kHz, NP=2048, NA=1024, with a total acquisition time of ~17 minutes. Data were processed using custom routines in MATLAB2018 and RStudio 1.2.13.

Results and Discussion

Figure 2 displays representative ³¹P spectra for both EVLP and GS models. The ex-vivo lung’s energy status and mitochondrial function declined over time, as can be seen spectrographically from the decline in β-ATP as well as the rise in P_i and phosphomonoester signals. Figure 3 displays a normalized time series of the ratio of β-ATP/P_i, quantifying this shift from high to low energy phosphates. In the admittedly small number of studies reported here, the EVLP model exhibited a slower rate of decline in β-ATP/P_i than the GS model, suggesting that EVLP was better at supporting lung ATP. However, substantial variability was observed in the GS results, complicating their interpretation. Specifically, two of the GS experiments exceeded the 5-hour timeframe by well over 30 mins, which may account for the low β-ATP/P_i ratios observed near their endpoints. Further highlighting the potential impact of extended ischemic time, a study by Shaghaghi, et al. found that the rate of decline in ex-vivo lung β-ATP peak intensity accelerated once β-ATP fell below 50% of its baseline value⁴. The GS model also involved more experimental steps, such as removing the lungs from the magnet, restarting pumps, etc., which may have contributed to lung insult or injury. While EVLP was better at preserving β-ATP/P_i, this metric’s connection to clinical lung transplant viability remains unclear—once donor lungs have been screened, approved, and transported to transplant centers, surgeons review the organ’s history and inspect the lungs, but do not assess ³¹P biomarkers.

Conclusion

Analysis of ³¹P spectra and β-ATP/P_i time series suggests that normothermic EVLP may improve upon cold static storage in preserving donor lungs’ ATP energy status prior to transplantation. However, this finding is complicated both by our study’s small sample size as well substantial variability in β-ATP/P_i values due to some GS experiments exceeding 5 hours in length. To further compare EVLP with cold static storage and assess β-ATP/P_i’s value as a metric for clinical donor lung evaluation, future testing will focus on re-ventilating the collapsed lung after 5-hour preservation, transplanting lungs at various β-ATP/P_i values, and attenuating the observed β-ATP/P_i variability due to experiment length.

Acknowledgements

No acknowledgement found.