Heart Transplantation–Associated Perioperative Ischemic Myocardial Injury
Morphological Features and Clinical Significance
Background The frequency and clinical significance of perioperative ischemic myocardial injury (PIMI) after heart transplantation and the diagnostic features distinguishing PIMI from rejection are not well defined.
Methods and Results We evaluated PIMI in the first four weekly endomyocardial biopsies and/or autopsy myocardium from 140 consecutive orthotopic heart transplantation recipients (1984 to 1991) by grading the severity of coagulative myocyte necrosis (CMN) as absent, 0; mild-focal, 1; moderate-multifocal, 2; or severe-confluent, 3, and determining the evolution of morphological features of its healing. CMN (often with contraction bands) was noted in 124 patients (89%); 24 patients (17%) had grade 3 CMN, of which 4 died within 30 days of transplantation. Nevertheless, at 1 year after surgery, survival was similar in patients with and without severe injury. Increased cold ischemic time but neither donor age nor intensity of inotropic support correlated with more severe early ischemic injury. PIMI inflammation was characterized by a predominantly polymorphonuclear/histiocytic infiltrate that contained lymphocytes and plasma cells, expanding the interstitium but not encroaching upon and separable from adjacent viable myocytes. Histological features of PIMI developed and resolved more slowly than those of typical myocardial infarct necrosis in nonimmunosuppressed patients; at 4 weeks, CMN persisted in 20% of patients and residual healing in nearly half. Diagnostic rejection was observed concurrently with PIMI in 54 of 533 biopsies (10%).
Conclusions Diagnosed by conventional histological criteria, PIMI is prevalent early after heart transplantation and has a protracted healing phase that can mimic or coexist with rejection. Extensive PIMI has deleterious impact on short-term survival, but the long-term impact of PIMI remains to be established.
Perioperative ischemic myocyte injury (PIMI) has been described in endomyocardial biopsies and autopsy heart specimens examined in the first several weeks after orthotopic heart transplantation.1 2 3 4 It is reasonable to expect that severe PIMI would impact markedly on short-term function and survival after heart transplantation and to hypothesize that lesser degrees of PIMI could contribute to late deterioration of function through myocardial fibrosis and possibly stimulate the development of graft arteriosclerosis. PIMI most likely arises predominantly during the obligatory ischemic time that accompanies procurement and implantation of a donor heart.
The histological features that characterize PIMI and those that specifically distinguish ischemic injury and its repair from other pathological processes, such as acute allograft rejection,5 have yet to be systematically described. Furthermore, the incidence, rate of evolution, and clinical significance of PIMI on immediate or long-term survival after orthotopic heart transplantation have not previously been studied in a large cohort of heart transplant recipients. Increasing awareness of and appreciation for PIMI as a distinct clinicopathological entity distinct from myocyte rejection should decrease augmentation of early postoperative immunosuppression, with its resultant short- and long-term sequelae such as infection, diabetes, renal impairment, osteoporosis, cataracts, and posttransplant lymphoproliferative disorders.
Accordingly, the objectives of this study were (1) to develop a comprehensive and systematic pathological description of PIMI; (2) to compare the effects of donor and procurement-related variables on the incidence and severity of PIMI; and (3) to determine the relationship between PIMI and patient prognosis after cardiac transplantation.
This study was based on a retrospective review of endomyocardial biopsies and clinical data obtained from 140 consecutive patients who underwent orthotopic heart transplantation at Brigham and Women’s Hospital between 1984 and 1991.
All data were retrieved from patient files and donor information sheets obtained at the time of organ procurement. Donor data included age, sex, administration of inotropic or vasopressor agents before organ harvesting, and total ischemic time. For the purpose of one facet of this analysis, we arbitrarily divided our population to four classes with regard to ischemic time: <100 minutes, 100 to 149 minutes, 150 199 minutes, and ≥200 minutes. Recipient data included age, sex, and the primary disease process that necessitated cardiac transplantation. Recipient outcome parameters included number of rejection episodes within the first year after transplantation, actuarial survival after transplantation, and, as appropriate, cause of death and other autopsy findings.
Hearts were procured according to a standard protocol that included administration of 1 L of cold potassium crystalloid cardioplegia (K=30 mEq/L) over 5 to 8 minutes via the aortic root. Hearts were then triple-bagged in iced saline and transported. In rare instances, additional cold crystalloid cardioplegia or terminal warm blood cardioplegia was administered during implantation, according to expected prolonged ischemic time or surgeon preference.
Maintenance immunosuppression consisted of cyclosporine, azathioprine, and prednisone; no patient received lymphocytolytic induction therapy with OKT3 or polyclonal antithymocyte globulin.
Endomyocardial biopsy samples (usually three or four pieces from each procedure) were obtained from the right ventricle of the transplanted heart using conventional endomyocardial biopsy technique.6 Surveillance endomyocardial biopsies (EMBs) during the first posttransplant year were performed routinely according to the following schedule: weekly for the first 4 weeks, biweekly for 4 weeks, monthly for 6 months, then every other month to 1 year. Additional unscheduled biopsies may have been performed due to an adverse change in the recipient’s clinical condition or new findings on objective testing. The tissues were fixed in buffered formalin, paraffin embedded, serially sectioned at 3 to 4 μm, and stained with hematoxylin and eosin by routine techniques. At least three separate slides (levels) were examined for each biopsy. A Masson’s trichrome stain was performed on one level of most biopsies. All biopsies performed during the first 4 weeks after transplantation were examined for each patient. Four patients who died before the first biopsy were included on the basis of heart examination at autopsy. Six additional patients who died during the review period were classified on the basis of their previous biopsies, and comparisons of endomyocardial biopsy findings with those of the heart at autopsy were made when available. In total, 533 endomyocardial biopsies were examined (136 at 1 week, 135 at 2 weeks, 132 at 3 weeks, and 130 at 4 weeks). Care was taken to identify prior biopsy sites to exclude these from the estimation of myocyte damage. The following features were noted and recorded for each biopsy.
Myocyte necrosis. Diagnosis of PIMI required the presence of its most definitive manifestation, coagulative myocyte necrosis (CMN), with or without contraction bands in the necrotic cells. Contraction bands in viable cells were considered to be artifact. The extent of myocyte injury was graded semiquantitatively: grade 0, no evidence of CMN; grade 1, mild (focal) CMN; grade 2, moderate (multifocal) CMN; grade 3, severe (confluent) CMN, and its location and distribution were recorded. The CMN grade used in the data represents the most extensive injury noted in the four biopsies on a particular patient.
Myocyte vacuolization. Vacuolization (myocytolysis) of subendocardial or peri-infarct myocytes was considered to be a marker of sublethal ischemic injury.7
Inflammation. Inflammation was classified by cell type (ie, polymorphonuclear leukocytes, macrophages [including macrophage giant cells], lymphocytes, plasma cells, and eosinophils) and location (ie, endocardial, perivascular, interstitial).
Vascular changes. Intramyocardial blood vessels were examined for endothelial swelling, endothelial cell-leukocyte adhesion, thrombosis, and frank vasculitis.
Rejection. Acute cellular rejection was graded according to the International Society for Heart and Lung Transplantation (ISHLT) criteria8 regardless of the presence or absence of CMN within the same biopsy.
Autopsy material was reviewed from all patients who died during the study. Multiple transmural sections from the left ventricular anterior, lateral, and posterior free walls, septum, and right ventricle were examined in the same manner as the endomyocardial biopsies for presence or absence of ischemic injury. The cause of death was recorded.
Pathological and clinical data were analyzed using the SAS statistical analysis package (SAS/PC version 6.04). Contingency tables were analyzed using Pearson and Mantel-Haenzel χ2 (PROC FREQ). Continuous variables between the classes were compared by ANOVA, Tukey’s method was used for pairwise comparisons (PROC GLM), and the distribution of discrete variables was compared by the Kolmogorov-Smirnov test (PROC NPAR1WAY). Survival data were analyzed using the product-limit method (PROC LIFETEST). A value of P<.05 was considered statistically significant.
The age of the recipient population was 45±12 years (mean±SD; range, 14 to 62 years). There were 106 male subjects and 34 female subjects. Indications for transplantation were ischemic heart disease in 53, idiopathic dilated cardiomyopathy in 71, valvular heart disease in 8, and “other” in 8 patients.
The mean age of the donor population was 27±6 years (range, 11 to 55 years). There were 108 male subjects and 32 female subjects. Eighty-four (60%) were maintained on β-adrenergic agents before cardiac procurement. The mean ischemic time was 157±48 minutes (range, 59 to 310 minutes).
Histological Characteristics of PIMI
PIMI was characterized by the following morphological features: CMN with or without contraction bands (in necrotic tissue) observed within 4 weeks of transplantation, usually followed by a heterogeneous interstitial, perivascular, and/or endocardial inflammatory infiltrate that was primarily composed of polymorphonuclear leukocytes and histiocytes (macrophages). The inflammatory response also included plasma cells and occasional lymphocytes, and myocyte injury was often accompanied by fat necrosis, vacuolization of adjacent viable myocytes, and/or ischemic microvascular injury (Figs 1⇓ and 2⇓ and Table 1⇓).
Myocyte hypereosinophilia, loss of nuclei, smudging, and loss of cytoplasmic detail and scattered polymorphonuclear leukocytes were the most characteristic features of early CMN (see Fig 1⇑, A and B). However, in the very early biopsies (approximately 1 week after surgery), necrotic cells were frequently difficult to distinguish in sections stained with hematoxylin and eosin; they were more prominent on the Masson’s trichrome stain. The highest concentration of necrotic myocytes was often immediately beneath the endocardial surface; indeed, an intervening zone of viable myocardium (presumably usually perfused directly from blood in the ventricular cavity) that characterizes typical myocardial infarction (occurring in a blood-filled heart) was frequently absent. Necrotic cells were either isolated or, more frequently, clustered into small groups. Severe CMN was denoted by large areas of confluent myocardial necrosis.
Later biopsies revealed an extensive macrophage infiltrate clearing necrotic myocytes; lymphocytes were a lesser but regular component of this inflammation (Fig 1⇑, C and D). The inflammatory response was predominantly in the interstitium or in perivascular spaces, not encroaching on and usually with a clear separation from adjacent myocytes. Polymorphonuclear leukocytes and obvious cellular debris were frequently present in the inflammatory foci, but eosinophils, often considered characteristic of rejection-associated inflammatory infiltrates in cyclosporine-treated patients,1 were not frequently noted.
Healing fat necrosis, characterized by empty spaces that resembled adipocytes without nuclei, with a surrounding mononuclear inflammatory infiltrate containing macrophage giant cells, was observed in many biopsies (Fig 2A⇑). There was no correlation between the severity of CMN and the observation of fat necrosis. Myocyte vacuolization was often prominent adjacent to necrotic myocardium (Fig 2B⇑). Small blood vessels with endothelial cell swelling, adherent intraluminal polymorphonuclear or mononuclear leukocytes, and mural or perivascular inflammation were occasionally seen. Microvascular thrombosis was not observed. There was no correlation between the severity of CMN and microvascular changes, but the population exhibiting these changes was very small. All of the biopsies with microvascular injury had obvious CMN.
Prevalence and Progression of PIMI
PIMI was frequently observed on endomyocardial biopsies taken within 4 weeks after surgery. Overall, 124 of 140 (89%) recipients had some degree of CMN diagnosed either on endomyocardial biopsies performed 1 to 4 weeks after transplantation or on autopsy material when patients died before the first biopsy (Fig 3⇓). CMN was most frequently seen initially in the first 2 weeks after transplantation (76% of first biopsies and 72% of second biopsies), and the associated inflammation occurred in a protracted fashion, with maximal prevalence at 2 to 3 weeks after transplantation (76% and 77% at weeks 2 and 3, respectively; Fig 4⇓). The amount and progression of PIMI were remarkably consistent on successive biopsies from an individual patient. The posttransplant interval to the initial histological diagnosis of CMN was significantly shorter than the posttransplant interval to the initial onset of inflammation (P<.0001, Kolmogorov-Smirnov test).
Protracted Healing Phase of PIMI and Concurrent Rejection
The healing phase of PIMI was frequently quite protracted. Persistent CMN was noted at 4 weeks in 26 of 130 biopsies (20%). Nearly half (49%) of the biopsies examined 4 weeks after transplantation had residual inflammation with a morphology appropriate for PIMI and not diagnostic of rejection (Fig 1E⇑). The duration of inflammation tended to be longer in patients with more severe CMN.
Nevertheless, the presence of acute or healing PIMI within a biopsy did not preclude the presence of acute cellular rejection within the same biopsy. Acute cellular rejection was noted in 77 of 533 (14%) of the biopsies examined. In only 23 (4%) was rejection isolated; in 54 (10%), rejection and PIMI coexisted.
Increased CMN grade was associated with longer total ischemic time. This correlation was statistically significant (P=.05 Pearson χ2 test, P=.002 Mantel-Haenszel χ2; Table 2⇓). No patient with ischemic time less than 108 minutes had CMN grade 3, and no patient with ischemic time greater than 194 minutes had CMN grade 0. Comparison of mean ischemic time between patients with different CMN grade also shows significant difference (P=.02 by ANOVA), but pairwise only CMN grades 0 and 3 differ at the .05 level. The ischemic times (mean±SD) corresponding to each CMN grade were 133±39, 156±44, 156±55, and 179±35 for grades 0, 1, 2, and 3, respectively. Higher CMN grade was not associated with increased pressor dosage, advanced age, or sex of the donor. Finally, no correlation was observed between the presence or severity of CMN and the number of treated rejection episodes (EMBs with ≥ISHLT grade 3A) occurring during the first posttransplant year.
Moreover, no difference in 1-year actuarial survival was observed in patients with CMN grade 3 versus patients with grade 0, 1, or 2 combined (P=.41, Fig 5⇓). Since the survival curves have essentially the same slope after the first month, the slight survival difference is due to early mortality. Ten patients died early (within 30 days of transplant); of these, 9 had some CMN, with 44% (4 patients) being grade 3 (P=NS). Indeed, recipients with grade 3 CMN suffered early postoperative death from otherwise unexplainable graft failure more frequently than patients with <grade 3 CMN (4 of 24 versus 2 of 116, P<.05).
The present study documented a high frequency of perioperative ischemic myocyte injury diagnosed by conventional histological criteria on endomyocardial biopsies obtained during the early posttransplant period of our heart transplant patients. The specific morphological characteristics of PIMI included CMN and a cellular infiltrate comprised predominantly of polymorphonuclear leukocytes and macrophages, with some lymphocytes and plasma cells. These features are distinct from those of other entities that occur during this period, including cellular allograft rejection. Prolonged total ischemic time of the allograft was associated with more severe grades of CMN. Finally, severe (grade 3) CMN was associated with an increased rate of early postoperative death, but actuarial 1-year recipient survival was not significantly decreased.
Recognition of PIMI and Differentiation From Rejection
In the vast majority of cases, PIMI can be distinguished from acute cellular rejection using conventional histological criteria (Table 3⇓). The morphology of myocyte necrosis observed in PIMI was largely typical for that of non–transplant-associated ischemic coagulation necrosis, with cellular hypereosinophilia, stretching, thinning, and loss of nuclei. However, in contrast to typical myocardial infarction, myocyte necrosis often extended to the endocardial surface, suggesting that the injury occurred while the heart was not filled with oxygenated blood. Moreover, the inflammation due to healing PIMI is generally more diffuse and rich in neutrophils and macrophages but has less lymphocytes than that of acute cellular rejection. In addition, in early PIMI, the extent of myocyte injury usually seems out of proportion to the associated inflammation. Nevertheless, the most important histological features that distinguish PIMI from acute rejection are CMN and a sharp border between polymorphous inflammation and nearby viable myocytes in PIMI. In contrast, cellular rejection is characterized by a predominantly lymphocytic infiltrate closely apposed to individual myocytes, some of which are frayed, scalloped, and/or vacuolated.
Recognition of PIMI on endomyocardial biopsy will contribute to reducing the long-term clinical sequelae of excessive immunosuppression. On the basis of the data in this study, it is apparent how pathologists could overdiagnose PIMI as acute rejection, particularly during its cellular healing phase. From the pathologist’s perspective, this is the “safe” approach, since underdiagnosing a rejection episode may have serious consequences, yet the short-term risks of augmented immunosuppressive therapy are relatively small. There is little support in the literature for doing otherwise. Moreover, diminution of the infiltrate after augmented immunosuppression does not constitute evidence that an infiltrate represented rejection, since inflammation due to healing ischemic injury is expected to progressively resolve and such inflammation is likely to be suppressed by steroids.
Several factors could affect the reliability of diagnosing PIMI after heart transplantation. First, PIMI and cellular rejection are most difficult to distinguish in later (≥3 weeks) endomyocardial biopsies because CMN frequently is resolved and the healing response of PIMI is protracted due to the anti-inflammatory effects of maintenance immunosuppression.9 Indeed, although not part of this study, we have observed inflammation considered to be healing ischemic damage more than 6 weeks after surgery in some patients. Moreover, concurrent rejection and PIMI were diagnosed in 10% of the biopsies taken during this period. Second, lymphocytolytic induction therapy may influence the appearance of PIMI in the early posttransplant period. In some studies, induction with OKT3 has been shown to delay the onset of the first rejection episode,10 and it also may diminish inflammatory infiltrates involved in repair. Interestingly, such therapy was used in a center at which vascular rejection was widely reported11 12 ; the vascular rejection grading schema used in these studies acknowledges the difficulty in distinguishing PIMI from vascular rejection and attributes endothelial swelling and ischemic myocyte changes seen in early biopsies to PIMI. Moreover, the healing response to PIMI is likely to be depressed and potentially more prolonged in patients receiving induction therapy. The results of this study would suggest, however, that biopsies taken even 4 weeks after transplantation can have characteristics of healing ischemic injury and that the associated inflammation may mimic cellular and/or vascular rejection.
Causes of PIMI
Distant procurement of hearts has been associated with decreased long-term recipient survival,13 and a recent study demonstrated the negative influence of increasing cold ischemic time on graft survival after cardiac transplantation.14 Hearts preserved for more than 4 hours had a significantly lower survival rate than those preserved for less than 4 hours; a trend toward decreased survival time with longer cold ischemia was observed at times less than 4 hours. Three intervals probably are important: (1) the brief ischemic period from donor aortic cross-clamping for cold cardioplegic infusion to immersion of the heart in iced saline, (2) the cold ischemic period that comprises the time of transportation while the heart is immersed in iced saline, and (3) the ischemic period in the operating room during donor heart implantation to the initiation of warm, oxygenated blood reperfusion in the recipient. Efforts to limit the occurrence of PIMI have concentrated on donor selection, preservation media, and reperfusion techniques.15 16
While procurement-related ischemic injury is considered most important in this setting,17 other mechanisms of myocyte injury are possible. Donor-associated injury could be secondary to CNS trauma and brain death with resultant endogenous catecholamine-induced myocardial necrosis and/or the exogenous administration of catecholamines.18 19 The dominant morphological pattern of catecholamine-induced injury is focal myocyte necrosis with contraction bands and histiocytic inflammation.20 Furthermore, the specific type of CNS trauma and the duration from declaration of brain death to organ procurement may affect the expression of these lesions in the donor heart.21 22 The possibility of postoperative ischemic injury is also recognized.
Other Clinical Implications
Although PIMI probably has a multifactorial pathogenesis during transplantation, the strongest measurable determinant of allograft ischemic injury was total ischemic time. Prolonged ischemic times have been associated with decreased survival after transplantation, graft dysfunction, and increased incidence of rejection and infection.14 23 24 This study suggests that the previously reported decreased 1-year actuarial survival associated with increasing ischemic time14 may be due to the effects of ischemic myocyte injury.
Ischemic injury could lead to late allograft dysfunction either by fibrosis or by promoting graft arteriosclerosis. The long-term sequelae of the perioperative ischemic myocyte injury and resultant myocardial scarring remain to be studied. Both abnormal left ventricular diastolic filling patterns at 1 year after transplantation and fibrosis in early posttransplant biopsies have been correlated with ischemic time.25 26 Moreover, ischemic damage could contribute to the development of graft arteriosclerosis through endothelial injury,27 increased susceptibility to cytomegalovirus infection, and/or the promotion of rejection through release of donor alloantigens.28 Ischemic allograft injury caused by distant heart procurement and prolonged ischemic time has been shown to correlate with endothelial cell injury in ultrastructural studies.29 In the present study, ischemic microvascular injury was one of the features of the overall perioperative ischemic insult. Nevertheless, a previous experimental study showed no correlation of graft coronary disease with the severity of PIMI.30
The recognition of PIMI on endomyocardial biopsy by pathologists is important after orthotopic cardiac transplantation. The accurate diagnosis of PIMI and distinction of this entity from cellular and/or vascular allograft rejection will decrease unnecessary augmentation of immunosuppression and its associated short- and long-term sequelae. Modifications of preservation techniques aimed at reduction of PIMI incurred during transplantation may decrease early mortality and perhaps decrease late graft dysfunction through scarring and/or vascular lesions. Finally, widespread recognition and consistent grading of PIMI will aid in the evaluation of multi-institutional studies designed to evaluate potentially new immunosuppressive therapies. Further studies are needed to determine whether the high frequency of perioperative ischemic injury reported in this study exists at other institutions but is diagnosed as rejection, or whether it reflects a large institutional/regional variation in the incidence of this phenomenon.
Dr Loh is the recipient of a Physician Scientist Award, K11-HL-02514, from the National Institutes of Health. Dr Schoen is supported in part by grant HL-43364 from the National Institutes of Health (Peter Libby, PI). We thank the members of the Brigham and Women’s Hospital Cardiac Transplant Service for their efforts in the clinical management of the patients reported in this study. We appreciate the helpful comments of Dr Richard Mitchell on the study and manuscript. We also acknowledge the efforts of Claudia Davis in typing the manuscript. Dr Fyfe is presently with the Department of Pathology, Medical College of Pennsylvania and Hahnemann University, Philadelphia, Pa; Dr Loh is presently with the Department of Medicine, Hospital of the University of Pennsylvania, Philadelphia.
Note Added in Proof
Since the submission of this article, a study by Day et al has examined the role of peritransplant ischemic injury in the pathogenesis of cardiac allograft vasculopathy (Day JD, Rayburn BK, Gaudin PB, Baldwin WM, Lowenstein CJ, Kasper EK, Baughman KL, Baumgartner WA, Hutchins GM, Hruban RH. Cardiac allograft vasculopathy: the central pathogenetic role of ischemia-induced endothelial cell injury. J Heart Lung Transplant. 1995;14:S142-S149.). This study proposes that early ischemic injury to vascular endothelial cells initiates a process that ultimately leads to the development of allograft vasculopathy. These authors not only confirm the presence of ischemic myocardial injury in early posttransplant biopsies but also suggest its diagnostic significance in identifying patients at risk for developing allograft vasculopathy.
Reprint requests to Frederick J. Schoen, MD, PhD, Department of Pathology, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115. E-mail firstname.lastname@example.org.
- Received July 17, 1995.
- Revision received October 16, 1995.
- Accepted October 18, 1995.
- Copyright © 1996 by American Heart Association
Billingham ME. Cardiac transplantation. In: Waller BF, ed. Contemporary Issues in Cardiovascular Pathology. Philadelphia, Pa: FA Davis; 1988:185-199.
Schoen FJ. Interventional and Surgical Cardiovascular Pathology: Clinical Correlations and Basic Principles. Philadelphia, Pa: WB Saunders; 1989:186-206.
Billingham ME. Diagnosis of cardiac rejection by endomyocardial biopsy. Heart Transplant. 1981;1:25-30.
Renlund DG, O’Connell JB, Gilbert EM, Hammond ME, Burton NA, Jones KW, Karwande SV, Doty DB, Menlove RL, Herrick CM, Lee HR, Gay WA Jr, Bristow MR. A prospective comparison of murine monoclonal CD-3 (OKT3) antibody-based and equine antithymocyte globulin-based rejection prophylaxis in cardiac transplantation: decreased rejection and less corticosteroid use with OKT3. Transplantation. 1989;47:599-605.
Hammond EH, Hansen JK, Spencer LS, Jensen A, Riddell D, Craven CM, Yowell RL. Vascular rejection in cardiac transplantation: histologic, immunopathologic, and ultrastructural features. Cardiovasc Pathol. 1993;2:21-34.
McLeod AA, Neil-Dwyer G, Meyer CH, Richardson PL, Cruickshank J, Bartlett J. Cardiac sequelae of acute head injury. Br Heart J. 1982;47:221-226.
Connor RC. Fuchsinophilic degeneration of myocardium in patients with intracranial lesions. Br Heart J. 1970;32:81-84.
Shivalkar B, Van Loon J, Wieland W, Tjandra-Maga TB, Borgers M, Plets C, Flameng W. Variable effects of explosive or gradual increase of intracranial pressure on myocardial structure and function. Circulation. 1993;87:230-239.
Hausdorf G, Banner NR, Mitchell A, Khaghani A, Martin M, Yacoub M. Diastolic function after cardiac and heart-lung transplantation. Br Heart J. 1989;62:123-132.
Pickering JG, Boughner DR. Fibrosis in the transplanted heart and its relation to donor ischemic time: assessment with polarized light microscopy and digital image analysis. Circulation. 1990;81:949-958.
Billingham ME, Baumgartner WA, Watson DC, Reitz BA, Masek MA, Raney AA, Oyer PE, Stinson EB, Shumway NE. Distant heart procurement for human transplantation: ultrastructural studies. Circulation. 1980;62(suppl I):I-11-I-19.
Masetti P, DiSesa VJ, Schoen FJ, Sun S-C, Byrne JG, Appleyard RF, Laurence R, Cohn LH. Ischemic injury before heart transplantation does not cause coronary arteriopathy in experimental isografts. J Heart Transplant. 1991;10:597-599.