Normothermic Machine Perfusion of Discarded Liver Grafts—What Is Viable?


To the Editor:

We read the paper by op den Dries et al. [1] analyzing Ex Vivo Normothermic Machine Perfusion and Viability Testing of Discarded Human Donor Livers with great interest. This paper is noteworthy for its careful analysis and honest reporting.

We congratulate the authors for (a) the outstanding technical approach to preserve human livers during normothermic ex vivo liver perfusion; (b) highlighting the importance of using marginal grafts to expand the donor pool; and (c) protection of the graft during ex vivo liver perfusion.

Notwithstanding these strengths, we believe that this study has minor weaknesses, which should be addressed.

  1. The normothermic ex vivo perfusion showed a strong rise of ALT and lactate during the first hour of perfusion. This could be interpreted as ischemia due to low perfusion flow of the graft in the initial phase of perfusion. AST as a marker of Zone III ischemia might be a better predictor of liver ischemia [2]. Kelly et al. [3] showed that a portal venous flow below 100 cm3/min/100 g liver tissue and a hepatic artery flow less 17 cm3/min/100 g liver tissue was associated with severe reperfusion injury in liver transplant recipients. The portal venous and hepatic artery flow in the current study only reached 35 and 15 cm3/min/100 g, respectively. The low perfusion flow together with the low hematocrit might contribute to the ALT release.
  2. The perfusion system used by op den Dries et al. did not contain mediators of reperfusion injury, such as leukocytes and platelets. The true magnitude of graft injury might have been masked and will only become apparent after reperfusion in a setting that contains all components of the injury cascade.
  3. The high LDH concentration in the bile together with the severe mucosa injury in the histology might indicate significant ischemic type bile duct injury. It remains unclear if the high LDH levels in bile are a result of washout after ischemic injury, or represent ongoing damage during perfusion. The extent of the biliary injury might only become apparent after reperfusion with whole blood. Providing a phospholipid to bile acid ratio would have been of interest.
  4. Interestingly, the authors performed PAS staining after diastase digestion, but no PAS staining for glycogen itself. Warm ischemia results in a depletion of glycogen stores and restoring the hepatic energy content could be an attractive target for normothermic perfused preservation.

The human ex vivo liver perfusion model of op den Dries et al. is a promising approach to improve marginal grafts and to better assess livers prior to transplantation.

  • J. M. Knaak1, V. N. Spetzler1, N. Selzner2 and M. Selzner1,*

  • 1Department of Surgery, Multi Organ Transplant Program, Toronto General Hospital, Toronto, Canada

  • 2Department of Medicine, Multi Organ Transplant Program, Toronto General Hospital, Toronto, Canada

  • * Corresponding author: Markus Selzner,


The authors of this letter have no conflicts of interest to disclose as described by the American Journal of Transplantation.