Functional Recovery After Human Heart Transplantation Is Related to the Metabolic Condition of the Hypothermic Donor Heart
Background Although strict selection criteria are being used for the acceptance of human donor hearts for transplantation, problems with respect to functional recovery on reperfusion sometimes still occur. Therefore, evaluation of the viability of a human donor heart before implantation during heart transplantation may be of great value.
Methods and Results In the present study, the energy metabolism of 25 excised human donor hearts arrested with St Thomas' Hospital No. 2 cardioplegic solution was evaluated noninvasively by use of 31P magnetic resonance spectroscopy (MRS) before implantation and was correlated with myocardial function measured with thermodilution in heart transplant patients. No significant correlation was observed between the cardiac index of heart transplant patients during the first hours after transplantation and the phosphocreatine/ATP (r=.13, P=.54), inorganic phosphate/ATP (r=.26, P=.21), phosphomonoesters/ATP (r=.02, P=.92), or phosphocreatine/inorganic phosphate (r=.16, P=.44) ratio or the intracellular pH (r=.06, P=.78) at the time of reperfusion. However, 1 week after transplantation, a significant correlation was observed between the cardiac index and the phosphocreatine/ATP (r=.49, P=.01), phosphomonoesters/ATP (r=.45, P=.02), and phosphocreatine/inorganic phosphate (r=.40, P=.05) ratios at the time of reperfusion. In contrast, the inorganic phosphate/ATP (r=.10, P=.63) ratio and pH (r=.31, P=.13) at the time of reperfusion showed a poor correlation with the cardiac index 1 week after transplantation.
Conclusions Functional recovery after human heart transplantation is related to the metabolic condition of the hypothermic donor heart.
Recovery of myocardial function is essential for the success of a human cardiac transplantation. Strict selection criteria and optimal management of the donor, therefore, remain essential for the acceptance of human donor hearts for transplantation. Echocardiography should show normal left ventricular wall motion, there should be no ECG signs of ischemia, and direct coronary palpation and inspection should show no signs of significant coronary atherosclerosis.
However, despite taking the aforementioned precautions during donor selection, problems with respect to functional recovery of the transplanted human donor heart sometimes still occur. Retrospectively, an explanation is not always present for these problems, but they may be related to both the clinical condition of the brain-dead donor and the cardioplegic arrest with subsequent preservation. A simple, reliable, and noninvasive test of myocardial viability during cold cardioplegic preservation would therefore be of great value. In addition, because the total cold ischemic time of a donor heart from cardioplegic arrest to reperfusion currently may not exceed 4 hours in general, a test of functional viability should not be a time-consuming method. 31P MRS permits noninvasive and rapid investigations of myocardial high-energy phosphate metabolism.1
It has been shown that myocardial energy indexes, ie, PCr/ATP ratios and PCr/Pi ratios, are related to myocardial functional performance under ischemic conditions in animal hearts in vivo.2 3 4 It also has been demonstrated that ischemic PME/ATP ratios in rat hearts in vitro are related to myocardial performance during reperfusion.5 Likewise, the metabolic condition of human donor hearts during preservation may be an important factor for functional recovery after reperfusion. Until now, little attention has been paid to the high-energy phosphate state of the hypothermic cardioplegic heart,6 7 especially the human heart,8 during preservation before implantation.
In the present study, the high-energy phosphate metabolism of human donor hearts was investigated before implantation by means of 31P MRS and correlated with functional performance during the first hours and 1 week after heart transplantation.
The energy metabolism of 25 excised human donor hearts was evaluated noninvasively by use of 31P MRS before implantation of the hearts. Characteristics of the donors are summarized in Table 1⇓. All donor hearts were explanted under standard conditions for whole heart transplantation and were arrested and preserved with St Thomas' Hospital No. 2 cardioplegic solution. The hearts were collected in sterile plastic bags and transported on crushed ice in an insulated container to the recipient hospital for implantation.
The 31P MRS examinations were performed on a 1.5-T Philips Gyroscan whole-body MR system. The heart, within the plastic bag, was placed on crushed ice (0°C) in a small plastic box and positioned on a 14-cm-diameter 31P surface coil (left ventricle down). To minimize total ischemic time, optimization of the procedure, ie, shimming and tuning, was performed on a phantom before the arrival of the donor heart at the hospital. The total time needed for the 31P MRS examination (ie, positioning of the heart in the magnet and data acquisition) did not exceed 8 minutes in the present study and did not delay surgical procedures.
For each 31P spectrum, 32 or 64 free induction decays were averaged, with the use of adiabatic 90° pulses with a repetition time of 5 seconds (see Fig 1⇓), resulting in total acquisition times of 160 or 320 seconds, respectively. All 31P spectra were quantified by means of a time-domain–fitting routine without any user interaction.9 10 This routine also yields an estimate of the accuracy of the fit (Crame´r-Rao SDs). PCr/ATP, Pi/ATP, PME/ATP, and PCr/Pi ratios were calculated and used as an index for myocardial energy state (see Table 1⇑). Intracellular pH at 0°C was obtained from the equation: pH=6.88+log([δ−3.35]/[5.60−δ]), in which δ is the chemical shift in parts per million of the Pi signal with respect to the PCr signal in the 31P MR spectra.6
Right heart hemodynamics were measured with a Swan-Ganz thermodilution catheter during the first hours after the transplantation procedure at the intensive care unit and ≈1 week after reperfusion at the time of the first transvenous endomyocardial biopsy for monitoring cardiac allograft rejection. Heart rate and blood pressure were recorded and the MAP was calculated. The CI, in liters of blood per minute per square meter of body surface area (L·min−1·m−2), was calculated and used as a measure of cardiac performance.
Statistics and Calculations
At the time of the 31P MRS measurements, the donor hearts had experienced different cold ischemic times (see Fig 2⇓). In addition, ischemic times at the time of reperfusion were different. To obtain PCr/ATP, Pi/ATP, and PME/ATP ratios and intracellular pH of the donor hearts at the time of reperfusion, the values of these parameters at the time of the 31P MRS examination were linearly extrapolated to values at the onset of reperfusion. Individual values were extrapolated with respect to their corresponding total ischemic times with the use of the slope of a linear relationship obtained from the pooled data. We obtained PCr/Pi ratios by dividing the extrapolated PCr/ATP ratios by the extrapolated Pi/ATP ratios.
Correlations between CI and each of the following parameters were analyzed by linear regression: PCr/ATP, Pi/ATP, PME/ATP, PCr/Pi, and pH. Test results with a value of P<.05 were considered significant. All data are presented as mean±SD, unless otherwise indicated.
Cold cardioplegic ischemic times at the time of 31P MRS acquisition and total ischemic times are shown in Table 1⇑. Mean ischemic time of the hearts from the moment of cold cardioplegic arrest to the 31P MRS measurement was 65±22 minutes (n=25) (range, 15 to 106 minutes), and mean total ischemic time on reperfusion was 147±26 minutes (n=25) (range, 90 to 210 minutes). Fig 1⇑ shows 31P MR spectra of two different hearts (donor hearts 2 and 3; see Table 1⇑) after ischemic times of 52 and 56 minutes, respectively. Although these ischemic times are comparable and relatively short, both PCr/ATP and Pi/ATP ratios showed considerable differences. Fig 2 ⇑ shows plots of PCr/ATP, Pi/ATP, PME/ATP, PCr/Pi, and pH versus ischemic time of all transplanted human donor hearts at the time of the 31P MRS examination. Linear regression between PCr/ATP, Pi/ATP, PME/ATP, and pH versus corresponding ischemic times of the hearts yielded the lines shown in Fig 2⇑. These lines were used to calculate individual values at the onset of reperfusion. The hyperbola in Fig 2D⇑ was obtained by dividing the equations describing the linear relationship between PCr/ATP and Pi/ATP ratios versus time.
All hearts showed functional recovery after reperfusion (Table 2⇓) and good long-term functional performance. No significant correlations were observed between the CI of heart transplant patients during the first hours after transplantation and the extrapolated PCr/ATP (r=.13, P=.54), Pi/ATP (r=.26, P=.21), PME/ATP (r=.02, P=.92), or PCr/Pi (r=.16, P=.44) ratio or the intracellular pH (r=.06, P=.78) at the time of reperfusion.
However, 1 week after transplantation, modest but significant correlations between the CI of heart transplant patients and the extrapolated PCr/ATP (r=.49, P=.01), PME/ATP (r=.45, P=.02), and PCr/Pi (r=.40, P=.05) ratios at the time of reperfusion were observed in the present study. In contrast, the extrapolated Pi/ATP (r=.10, P=.63) ratio and pH (r=.31, P=.13) showed poor correlations with the CI 1 week after transplantation. (See Fig 3⇓.)
No correlation was observed between the CI and the actually measured Pi/ATP, PME/ATP, and PCr/Pi ratios or pH, but the PCr/ATP ratio showed a correlation (r=.45, P=.02) with CI, albeit less significant than the extrapolated ratio.
Also, no correlation was observed between cardiac allograft rejection grade and CI or metabolic parameters (data not shown).
The main finding of the present study is the significant correlation between the CI of heart transplant patients 1 week after transplantation and the PCr/ATP, PME/ATP, and PCr/Pi ratios at the time of reperfusion. In contrast, no significant correlations were observed between CI within the first hours of reperfusion and the 31P MRS parameters. Another important finding was the considerable variation in both PCr/ATP and, especially, Pi/ATP ratios of the donor hearts, even when ischemic times were comparable. These differences may be caused by the condition of the donor, the use of pharmacological therapy, and the quality of cardioplegic arrest and hypothermic preservation.
Correlation Between 31P MRS Data and CI
A significant correlation between the CI of heart transplant patients 1 week after transplantation at the time of the first transvenous endomyocardial biopsy and the extrapolated PCr/ATP, PME/ATP, and PCr/Pi ratios at the time of reperfusion was observed in the present study. These energy indexes were extrapolated assuming a linear relationship of PCr/ATP, PME/ATP, and Pi/ATP with respect to ischemic time within the first hours of cold ischemia. Results of another study on long-term preservation of human hearts suggested an exponential behavior of these energy indexes, which certainly for the first 4 hours may be approximated by a linear relationship (C.J.A. Van Echteld, PhD, et al, unpublished data, 1996).
Our findings in the present study suggest that cardiac functional performance is linked to myocardial phosphate metabolism. Recently, several studies2 3 4 have supported the hypothesis that cardiac contractility is regulated by levels of myocardial high-energy phosphates (PCr/ATP and PCr/Pi) in the underperfused animal heart. However, rather than comparing present functioning and energy metabolism from a prior period, function and energy metabolism were evaluated simultaneously in these studies.2 3 4 Unfortunately, only limited 31P MRS data on cardioplegic preservation6 8 and subsequent functional recovery are available.5 11 12 To the best of our knowledge, the present study presents the first 31P MRS data concerning the relationship of the energy metabolism of the cardioplegia-arrested and subsequently cold-preserved human donor heart and its functional performance postoperatively.
In the present study, significant correlations between both the PCr/ATP and the PCr/Pi ratios at the onset of reperfusion and functional performance after transplantation were observed, whereas no such correlation was found for the Pi/ATP ratio, suggesting an important role for PCr. The PCr content of a cardioplegic heart may be highly affected on reperfusion. Recently, Bøtker et al13 observed no significant change in total creatine content in cold ischemic pig hearts 90 minutes after the onset of cardioplegic arrest, whereas a significant decrease of ≈42% in total creatine content was observed 2 hours after reperfusion. During preservation, PCr decreases and free creatine levels concomitantly increase in the cardioplegic heart. This results in a rising creatine gradient over the myocardial cell membrane. Consequently, the larger the gradient, the larger the loss of creatine on reperfusion may be. Moreover, a decrease in total creatine may lead to decreased cardiac PCr/ATP ratios after reperfusion, which may compromise functional performance.14 It has been established in postischemic animal hearts that a decrease in creatine kinase and PCr content, alone and in combination, leads to a decrease in myocardial performance.14 15
The total creatine content and hence PCr/ATP ratios are not only reduced early after reperfusion subsequent to preservation but are likely to be depressed for a longer period than a few hours after reperfusion. In a preliminary study,16 we have shown that the cardiac PCr/ATP ratio in heart transplant patients is reduced during the first months after transplantation. This lower PCr/ATP ratio is observed irrespective of rejection and recovers to normal values following an exponential association curve with a time constant of 39 days. The temporarily lower PCr/ATP ratio in heart transplant patients can be explained by creatine loss during preservation and reperfusion and a slow uptake of creatine from the blood during the first months after transplantation.
The relationship between PMEs before reperfusion and function after transplantation is less clear. The PME signal could originate from glycolytic intermediates, such as glucose-6-phosphate and fructose-6-phosphate, but may also represent AMP. Interestingly, in a recent 31P MRS study on the ischemic rat myocardium,5 Jeffrey et al showed that the PME signal was the only 31P MRS parameter that predicted poor functional recovery. They found that the most prominent components of the PMEs were glucose-6-phosphate, α-glycerol-phosphate, and AMP. Spectral resolution, unfortunately, was not sufficient to discriminate between these compounds in the present study, but the chemical shift of the PME signals of ≈6.80 ppm corresponds well with the chemical shifts of these compounds. Jeffrey et al5 interpret the accumulation of PMEs and the concomitant occurrence of contracture as an inhibition of glycolysis followed by calcium overload. Such a sequence of events could also contribute to temporary diminished performance of the transplanted human heart. In addition, if indeed AMP is involved, then a similar reasoning as for creatine may apply. On reperfusion, the heart could lose adenine nucleosides. However, the decreased myocardial PCr/ATP ratio in heart transplant patients we have observed early after transplantation suggests that the extent of adenosine leakage is less than that of creatine or that replenishment of adenine nucleotides is faster.16
The present study showed no significant relationship between the Pi/ATP ratio or pH and the CI 1 week after transplantation. Other studies have indicated that contractile dysfunction may be explained by increased intracellular Pi levels3 17 18 and by acidosis.3 17 18 19 20 It is possible that high Pi and H+ levels contributed to reduced myocardial contractility during early reperfusion in the present study, whereas these levels were normalized 1 week postoperatively due to either washout or redistribution. Also, Jeffrey et al5 found no correlation between intracellular pH, ATP, and Pi measured before reperfusion and the extent of recovery. On the other hand, one might argue that the increase in Pi levels during hypothermic storage reflects the decrease in PCr and that therefore Pi/ATP should show a similar correlation with CI as PCr/ATP. However, based on the fair assumption that ATP levels remain largely unchanged (C.J.A. Van Echteld, PhD, et al, unpublished results, 1996), part of the Pi from PCr must end up as PME and not as free Pi, thereby reducing free Pi levels and their correlation with CI. In line with this reasoning is the lower degree of correlation of CI with PCr/Pi versus PCr/ATP.
The poor correlation between the CI within the first hours after reperfusion and the 31P MRS data may have several causes. At this short time after reperfusion, all heart transplant patients were still generally anesthetized and ventilated and their hearts were inotropically supported and stimulated by a pacemaker. All these features may have their influence on the functional recovery and performance of the transplanted reperfused heart. Moreover, myocardial stunning is probably the most important phenomenon that may influence the hemodynamics of a transplanted heart during the first days after transplantation.21 22 23 Because myocardial stunning is expected to be significantly decreased or even absent 1 week after transplantation, the CI obtained at the time of the first endomyocardial biopsy was considered a more reliable measure for functional recovery, although many other factors may still influence cardiac performance.
We have chosen CI as a measure of myocardial performance because it is closely related to cardiac work and consequently to cardiac energy metabolism. CI is dependent on loading conditions and therapies, but after 1 week after heart transplantation, external pacing and inotropic support had been ceased. We have compared other indexes of myocardial performance and found generally decreasing degrees of correlation with the metabolic parameters in the following sequence: stroke volume<cardiac output<CI. Because a cardiac work index (CI×MAP) may be an even better measure of cardiac work, we have also calculated correlations of this parameter with the metabolic parameters but found equal or slightly less degrees of correlation. Unfortunately, our measurements of MAP were recorded the same day but not simultaneously with the CI measurements.
The 31P MRS data of the donor hearts were not corrected for partial saturation. However, all 31P MRS spectra of the donor hearts were obtained with the same repetition time of 5 seconds and therefore can be compared with each other. Limited examination time precluded sufficient measurements to determine the actual T1 relaxation times of the different metabolites involved. In addition, T1 relaxation times of 31P metabolites from literature could not be used for the correction of partial saturation because all published values were determined in vivo at 37°C. In a preliminary study on explanted diseased human hearts preserved at 0°C, we observed that the T1 relaxation time of PCr was dramatically decreased with respect to its value at 37°C and was comparable to the T1 relaxation times of ATP (data not shown). From the linear relationship of the PCr/ATP ratio in the present study, the PCr/ATP ratio at the onset of cardioplegia can be extrapolated and is found to be 1.31. This value corresponds closely with the value of 1.23±0.17 in the human heart in vivo24 and shows that correction for partial saturation of the PCr/ATP ratio is not necessary in the present study.
As with most other MRS studies that use surface coils, data have to be presented as ratios rather than absolute metabolite content. Theoretically, absolute measurement by use of surface coils is possible but requires a more elaborate and more time-consuming protocol. Alternatively, one might consider using a volume coil, but the bulky plastic bags containing the heart dramatically reduce the signal-to-noise ratio of the 31P MRS measurement.
To obtain 31P MRS parameters of the donor hearts at the time of reperfusion, the values of these parameters at the 31P MRS examination were extrapolated to values at the onset of reperfusion, assuming a linear relationship between these parameters and time. As mentioned before, support for this assumption has been found in a previous study on long-term preservation of human hearts. However, this approach does not allow for individual variation in metabolic deterioration. Nevertheless, the absent or reduced correlation between CI and the directly measured metabolic parameters emphasizes the necessity of extrapolation and the significance of ongoing metabolic deterioration during the implantation procedure.
The present study clearly demonstrates that 31P MRS examination of a human donor heart before implantation is feasible within the short period of time available during the transplantation procedure.
In conclusion, functional recovery after human heart transplantation is related to the metabolic condition of the hypothermic donor heart. Improved preservation of the energy metabolism may result in improved functional recovery of the donor heart, which may allow a shorter postoperative stay in the hospital. In addition, 31P MRS may even offer a selection criterion for the transplantation of a donor heart.
Selected Abbreviations and Acronyms
|MAP||=||mean arterial pressure|
|MRS||=||magnetic resonance spectroscopy|
Dr Van Echteld was supported as an established investigator by The Netherlands Heart Foundation (grant No. D88001). The authors would like to thank Prof Dr W.P.Th.M. Mali for providing access to the magnetic resonance system, Prof Dr E.O. Robles de Medina and Prof Dr T.J.C. Ruigrok for helpful discussions, and M.C.H. de Groot for assistance during the measurements.
- Received February 19, 1996.
- Revision received May 22, 1996.
- Accepted July 11, 1996.
- Copyright © 1996 by American Heart Association
Ingwall JS, Weiss RG. 31P NMR spectroscopy: the noninvasive tool for the study of the biochemistry of the cardiovascular system. Trends Cardiovasc Med. 1993;3:29-37.
Schaefer S, Schwartz GG, Gober JR, Wong AK, Camacho SA, Massie B, Weiner MW. Relationship between myocardial metabolites and contractile abnormalities during graded regional ischemia: phosphorus-31 nuclear magnetic resonance studies of porcine myocardium in vivo. J Clin Invest. 1990;85:706-713.
Schwartz GG, Schaefer S, Meyerhoff DJ, Gober JR, Fochler P, Massie BM, Weiner MW. Dynamic relation between myocardial contractility and energy metabolism during and following brief coronary occlusion in the pig. Circ Res. 1990;67:490-500.
Van Echteld CJA, Kirkels JH, Jambroes G, Kodde JH, Lahpor JR, Van Dobbenburgh JO, Ruigrok TJC. Preservation of human donor hearts. Proc 9th Annu Meet SMRM. 1990;1:247. Abstract.
De Beer R, Van Ormondt D. Analysis of NMR data using time domain fitting procedures. In: Diehl P, Fluck E, Gunther H, Kosfeld D, Seelig J, eds. NMR Basic Principles and Progress. Berlin, Germany: Springer-Verlag; 1992;26:201-248.
Pernot A-C, Ingwall JS, Menasche´ P, Grousset C, Bercot M, Piwnica A, Fossel ET. Evaluation of high-energy phosphate metabolism during cardioplegic arrest and reperfusion: a phosphorus-31 nuclear magnetic resonance study. Circulation. 1983;67:1296-1303.
Neubauer S, Hamman BL, Perry SB, Bittl JA, Ingwall JS. Velocity of the creatine kinase reaction decreases in postischemic myocardium: a 31P-NMR magnetization transfer study of the isolated ferret heart. Circ Res. 1988;63:1-15.
Van Dobbenburgh JO, De Jonge N, Klo¨pping C, Lahpor JR, Woolley SR, Van Echteld CJA. Postischemic alterations in myocardial energy metabolism in heart transplant patients. Circulation. 1993;88(suppl I):I-378. Abstract.
Aussedat J, Ray A, Lortet S, Reutenauer H, Grably S, Rossi A. Phosphorylated compounds and function in isolated hearts: a 31P-NMR study. Am J Physiol. 1991;260:H110-H117.
Orchard CH, Kentish JC. Effects of changes of pH on the contractile function of cardiac muscle. Am J Physiol. 1990;258:C967-C981.
Bolli R. Mechanism of myocardial ‘stunning.’ Circulation. 1990;82:723-738.
Braunwald E, Kloner RA. The stunned myocardium: prolonged, postischemic ventricular dysfunction. Circulation. 1982;66:1146-1149.