Preserving and evaluating hearts with ex vivo machine perfusion: an avenue to improve early graft performance and expand the donor pool

Abstract

Cardiac transplantation remains the first choice for the surgical treatment of end stage heart failure. An inadequate supply of donor grafts that meet existing criteria has limited the application of this therapy to suitable candidates and increased interest in extended criteria donors. Although cold storage (CS) is a time-tested method for the preservation of hearts during the ex vivo transport interval, its disadvantages are highlighted in hearts from the extended criteria donor. In contrast, transport of high-risk hearts using hypothermic machine perfusion (MP) provides continuous support of aerobic metabolism and ongoing washout of metabolic byproducts. Perhaps more importantly, monitoring the organ's response to this intervention provides insight into the viability of a heart initially deemed as extended criteria. Obviously, ex vivo MP introduces challenges, such as ensuring homogeneous tissue perfusion and avoiding myocardial edema. Though numerous groups have experimented with this technology, the best perfusate and perfusion parameters needed to achieve optimal results remain unclear. In the present review, we outline the benefits of ex vivo MP with particular attention to how the challenges can be addressed in order to achieve the most consistent results in a large animal model of the ideal heart donor. We provide evidence that MP can be used to resuscitate and evaluate hearts from animal and human extended criteria donors, including the non-heart beating donor, which we feel is the most compelling argument for why this technology is likely to impact the donor pool.

Fig. 1

Representative traces of signal intensity vs time (SIVT) curves during machine perfusion (MP). SIVT were generated using gadolinium-DPTA MRI standard first-pass imaging methodology for each of four areas of interest (posterior, anterior, lateral, and septal walls) in each slice for a total of 12 SIVT curves per heart. Variable penetration of gadolinium-DPTA in hearts preserved with high potassium perfusate (panel A) indicates inhomogeneous perfusion. Hearts perfused with a low potassium perfusate (panel B) showed homogenous perfusion in hearts. Reproduced with permission, Circ J [10].

Fig. 2

Weight gain during ex vivo MP (measured as a % of baseline weight) for hearts perfused at a constant, non-variable perfusion pressure and hearts perfused with biologically variable perfusion pressure. By mimicking physiologic perfusion pressures, biologically variable perfusion significantly reduces weight gain during ex vivo preservation.

Fig. 3

Three biopsies were obtain for all hearts preserved with either ex vivo machine perfusion (MP) or cold storage (CS); the first biopsy was obtained in situ prior to preservation, the second after ex vivo preservation and the third after 1 h of reperfusion with whole blood on a Langendorff apparatus. Hearts preserved with MP showed improved ATP levels after preservation with a return to baseline ATP levels after 1 h of whole blood reperfusion (panel A). Similarly, hearts preserved with MP showed significantly lower levels of endothelin-1 (marker of endothelial injury, panel B), malondialdehyde (MDA, marker of oxidative damage, panel C) and caspase-3 (marker of apoptosis, panel D) compared with hearts preserved with CS.

Fig. 4

To simulate a NHBD in a large animal model, canine hearts were exposed to 60 min of global warm ischemia prior to cardiectomy then preserved with 6 h of continuous ex vivo perfusion (WI + MP group). Hearts exposed to warm ischemia followed by cold storage (WI + CS) and hearts preserved with 6 h of machine perfusion (MP) or cold storage (CS) without any prior warm ischemia served as control groups. Despite 60 min of warm ischemia, mean functional recovery of the WI + MP group, based on the developed pressure (panel A), rate of pressure generation (+dP/dt, panel B) and rate of relaxation (−dP/dt, panel C) was not significantly different from hearts preserved without prior warm ischemia (MP group). However functional recovery in the WI + MP group was highly variable indicating that not all hearts in this homogeneous canine population are able to recovery from such severe ischemic injury. Hearts in the WI + CS group showed no functional recovery.

Fig. 5

Representative real time LV intramyocardial pH tracing during ex vivo preservation. Hearts that showed good functional recovery following restoration of whole blood flow had demonstrated improving anterior and posterior LV myocardial pH during ex vivo preservation (light tracings, panel A), in contrast hearts that showed poor functional recovery did not demonstrate ex vivo pH improvement and were acidotic at the time of blood reperfusion (dark tracings, panel A). Tissue pH may thus be an important variable to identify, during the ex vivo preservation period, hearts that are at risk for primary nonfunction. Hearts preserved with cold storage (panel B) showed a gradual decline in anterior and posterior tissue pH during the preservation period indicative of anaerobic metabolism.

Fig. 6

Correlation of fractional anisotropy with markers of functional recovery following restoration of whole blood flow. Fractional anisotropy, assessed ex vivo using diffusion tensor MRI, provides a direct measure of cellular viability. The strong correlations with developed pressure (panel A), rate of pressure generation (+dP/dt, panel B) and rate of relaxation (−dP/dt, panel C) suggest that this technology could be used to assess graft viability during the ex vivo period. Reproduced with permission, J Heart Lung Transplant [50].