Sympathetic Nerve Sprouting After Orthotopic Heart Transplantation

2 downloads 0 Views 573KB Size Report
(nerves/mm2) with respect to time after transplantation in the endocardium; ... As in normal hearts, all 29 allografts had epicardial nerve trunks that extended.
Sympathetic Nerve Sprouting After Orthotopic Heart Transplantation Dave T. Kim, MD,a Daniel J. Luthringer, MD,a Angela C. Lai, BA,a Gina Suh, BA,a Lawrence Czer, MD,a Lan S. Chen, MD,b Peng-Sheng Chen, MD,a and Michael C. Fishbein, MDc Background: Although many studies have documented sympathetic re-innervation in transplanted hearts (allografts) using chemical, imaging, and electrophysiologic methods, little histopathologic proof of this process exists. Methods and We used immunohistochemical techniques with antibodies to S-100 protein, to growth-associated Results: protein 43 (GAP43), and to tyrosine hydroxylase (TH) to detect nerves in the left ventricles in allografts from 29 consecutive recipients. Reasons for transplantation included ischemic heart disease (IHD, n ⫽ 16), non-ischemic dilated cardiomyopathy (DCM, n ⫽ 12), and both (n ⫽ 1). We assessed nerve densities (nerves/mm2) with respect to time after transplantation in the endocardium; in the mid-myocardium; and around intramyocardial blood vessels, scars, foci of rejection, and Quilty lesions. Six normal hearts were used for comparison. As in normal hearts, all 29 allografts had epicardial nerve trunks that extended into the mid-myocardium around blood vessels. Although the total number of nerves (S100-positive) progressively decreased over time, GAP43-positive nerves around the blood vessels increased with time (p ⬍ 0.005). We also observed abundant TH-positive nerves. The density of S100-positive nerves around blood vessels was greater in those undergoing transplantation for IHD (113 ⫾ 88) than in those with prior DCM (54 ⫾ 49, p ⬍ 0.05). Nerve density in each area varied greatly. Conclusions: Heterogeneous sympathetic nerve sprouting and re-innervation occurred around blood vessels in the allografts. The magnitude of nerve sprouting increased with time and varied greatly from patient to patient. Patients with IHD had greater nerve sprouting and re-innervation than did those with DCM. J Heart Lung Transplant 2004;23:1349-58. Copyright © 2004 by the International Society for Heart and Lung Transplantation.

Peripheral nerve injury results in denervation and is followed by neural regeneration through nerve sprouting.1 Because cardiac transplantation results in severed cardiac nerves, we hypothesized that nerve sprouting occurs after transplantation, resulting in re-innervation of transplanted allograft hearts. Indeed, nuclear imaging and norepinephrine assay showed evidence of func-

From the aDivision of Cardiology, Department of Medicine, CedarsSinai Medical Center, bDepartment of Neurology, Childrens Hospital and University of Southern California, and cDepartment of Pathology, University of California at Los Angeles School of Medicine, Los Angeles, California. Submitted April 24, 2003; revised October 10, 2003; accepted October 10, 2003. This study was supported by a Pauline and Harold Price Endowment (Dr. Chen), a Piansky endowment (Dr. Fishbein), and by National Institutes of Health Grants P50 HL52319, R01 HL66389, and R01 HL71140, and by an American Heart Association National Center grant-in-aid (9950464N). Reprint requests: Dr. Michael C. Fishbein, Department of Pathology and Laboratory Medicine, UCLA School of Medicine, 10833 Le Conte Avenue, Los Angeles, California 90095-1732. Telephone: 310-8259731. Fax: 310-794-4161. E-mail: [email protected] Copyright © 2004 by the International Society for Heart and Lung Transplantation. 1053-2498/04/$–see front matter. doi:10.1016/ j.healun.2003.10.005

tional cardiac sympathetic re-innervation in some patients who underwent cardiac transplantation.2– 4 Successful sympathetic re-innervation is associated with improved peak oxygen uptake, heart rate, and contractile function during exercise.3,5 Despite the importance of cardiac re-innervation in the functional performance of transplanted hearts, little histologic data exist on cardiac re-innervation in transplanted hearts. Wharton et al.6 studied cardiac nerves in 9 allografts using immunohistochemical methods. Although they found that all cardiac allografts contained nerve fibers, none displayed tyrosine hydroxylase (TH) immunoreactive nerves, which does not support the presence of sympathetic re-innervation. These investigators also found that the number and distribution of nerves in the cardiac allografts showed no apparent correlation with rejection or longevity of the graft. In contrast, Bengel et al.5 studied 29 cardiac transplant recipients and reported that the interval between transplantation and the time of study is important in demonstrating sympathetic re-innervation by positron emission tomography using [11C]hydroxyephedrine. Because of this apparent discrepancy, whether true sympathetic re-innervation occurs in the allografts remains in dispute.7 In the current study, we sought to investigate the hypothesis 1349

1350

Kim et al.

that active sympathetic nerve sprouting persists for at least 1 year after orthotopic heart transplantation. A second purpose was to determine whether the duration of allograft survival, rejection, or the pre-existing disease status influences the magnitude and distribution of nerve sprouting and sympathetic re-innervation. METHODS Patient Selection We studied samples from 29 consecutive autopsies of human allograft transplanted hearts between 1989 and 1997. As controls, we studied samples from 6 hearts without structural disease. All hearts were preserved in 10% formalin. A portion of the mid-anterior left ventricular wall was excised. The samples were processed routinely and embedded in paraffin. Serial, 5-␮m-thick transmural sections were cut, using a microtome, and placed on glass slides for staining. Immunohistochemical Staining Immunohistochemical staining was performed with primary antibodies against S100-protein (DAKO Corporation; Carpinteria, CA; 1:1000), growth associated protein 43 (GAP43) (Boehringer Mannheim Biochemica, 1:50 dilution), and TH (Accurate Chemical, 1:100 dilution). We used S-100 protein staining to identify cardiac nerves, according to methods previously reported.8 For GAP43 staining, we used a technique described by Shi et al9 to enhance antigen expression. The deparaffinized tissue specimens were placed in 10 mmol/liter citrate buffer solution and heated in a microwave oven before staining. To inactivate endogenous peroxidase and to decrease non-specific staining, we incubated specimens in 3% hydrogen peroxide and methanol for 15 minutes and in serum-free protein block (DAKO) for 30 minutes, respectively. Specimens were incubated with primary antibody for 1.5 hours and then with biotinylated secondary antibodies (DAKO) for 30 minutes. The staining procedure was then resumed with ABComplex (DAKO) for GAP43 staining and with streptavidin (DAKO) for TH staining, followed by visualization of immunoreactive products, counterstaining with hematoxylin, acid and alcohol rinse, dehydration in graded alcohol series, and mounting. We also stained selective specimens with monoclonal mouse anti-human neurofilament protein (DAKO, 1:500) and neuron-specific enolase (DAKO, 1:500). Because the data showed colocalization of S100-protein–positive nerves with those stained positive for neurofilament and neuron-specific enolase, we reported only the results of S100-protein staining. Data Analyses We analyzed the nerves in the endocardium, in the mid-myocardium, and around the intramyocardial blood

The Journal of Heart and Lung Transplantation December 2004

vessels. In addition, we determined nerve density in areas of rejection, scarring, and Quilty lesions. The nerves were quantified by visually counting the number of nerves or nerve twigs per 5 high-power fields (⫻40 objective) in areas of each section that appeared to have the most nerves. The density of nerves was expressed as the number of nerves per mm2. We then analyzed data in the areas of interest in relation to the time after transplantation. At least 2 observers reviewed all sections (DK, MCF). All statistical analyses were performed with commercially available software. The data are presented as the mean ⫾ standard deviation. We used a 2-tailed Student’s t-test to compare the means. For multiple comparisons, we used analyses of variance with Newman-Keuls tests. A p value ⱕ 0.05 was considered significant. RESULTS Patient Characteristics Table 1 summarizes the patient characteristics. The donors weighed 76 ⫾ 20 kg, which is not different from the weight of recipients (77 ⫾ 16 kg). The mean age was 57.4 ⫾ 10.9 years old. Sixteen patients had ischemic cardiomyopathy, 12 had non-ischemic dilated cardiomyopathy, and 1 had both. In 5 transplantations, the donor and recipient blood types were not the same. The blood types in these 5 transplantations were O⫹ and A⫹, B⫺ and AB⫹, O⫹ and A⫹, O⫹ and B⫹, and A⫹ and AB⫹ for donor and recipient, respectively. The donor and recipient blood types were the same in the other transplantations. The surgical techniques were bi-atrial (Shumway-Lower technique) in 15 patients, caval (total orthotopic technique) in 11 patients, and mixed (bicaval anastomosis for the right atrium and a single left atrial anastomosis for the left atrium) in 3 patients. The ischemic time and pump time were 141 ⫾ 30 minutes and 118 ⫾ 22 minutes, respectively. Three patients had post-operative atrial flutter or fibrillation. Four patients needed pacemakers. The time of death from transplantation ranged from 3 days to 75 months. Death occurred from cardiac causes (arrhythmias, myocardial infarction, acute rejection, congestive heart failure) in 23 cases and of non-cardiac causes (pneumonia, sepsis, and possible adverse effect to an antidepressant) in 6 cases. Ten patients have evidence of transplant arteriopathy. We divided the allografts into 3 groups with respect to time after transplant as follows: Group 1 or early (n ⫽ 8, ⬍6 months), Group 2 or intermediate (n ⫽ 7, 6 –12 months), and Group 3 or late (n ⫽ 14, ⬎12 months) after transplantation. Documented primary ventricular fibrillation and tachycardia as causes of death were present only in the late group (2 of 14, 14.3%). Patient 22 was presumed to have died of an adverse reaction to an anti-depressant. Twenty hearts

The Journal of Heart and Lung Transplantation Volume 23, Number 12

Kim et al.

1351

Table 1. Patient Characteristics Death after transplant Group 1 (⬍6 mo.)

Group 2 (6–12 mo.)

Group 3 (⬎12 mo.)

Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

Age 58 65 49 65 42 32 58 51 65 64 64 68 67 70 49 56 67 65 57 68 58 60 62 71 59 54 45 47 28

Sex M M F M M M M M M M M M M M F M M M M M M M M M M M F M F

Grade of rejection 0 0 0 0 3B 4 1A 3 1A hum.rej 4 1A 3B 1A 3A 1A 1B 0 0 0 1A 1A 1A 1A 1A 0 0 1B 1B

Time since transplant 3 days 4 days 11 days 0.5 mo. 1 mo. 1 mo. 2.25 mo. 2.5 mo. 6 mo. 6 mo. 7 mo. 8 mo. 9 mo. 11 mo. 12 mo. 14 mo. 15 mo. 15 mo. 17 mo. 19 mo. 23 mo. 25 mo. 25 mo. 26 mo. 28 mo. 30 mo. 35 mo. 60 mo. 75 mo.

Cause of death Pseudomonas pneumonia Sepsis Post-operative VT and bleeding Sepsis Acute rejection Acute rejection Cardiogenic shock Acute rejection TX CAD/MI Lymphoma Acute rejection CAD/MI Acute rejection CAD/MI Acute rejection bradyarrhythmia CAD/MI CAD/MI CAD/MI Sepsis/HCV SCD/VF Hypotension/acidosis CHF EMD Arrest in cath lab CAD/MI VT SCD/CAD/MI Acute rejection/CHF

Reason for transplant IHD/DCM IHD DCM/Sarcoid IHD DCM DCM IHD DCM IHD IHD IHD IHD DCM DCM DCM DCM IHD IHD DCM IHD IHD IHD IHD DCM DCM IHD IHD IHD Post-partum cardiomyopathy

CAD, coronary artery disease; CHF, congestive heart failure; DCM, dilated cardiomyopathy; EMD, electromechanical dissociation; HCV, hepatitis C virus infection; IHD, ischemic heart disease; MI, myocardial infarction; SCD, sudden cardiac death; TX CAD, transplant coronary artery disease; VF, ventricular fibrillation; VT, ventricular tachycardia.

had acute rejection, 4 had scars, and 16 had Quilty lesions. Scarred areas were few in each group. All specimens in the intermediate group had areas of rejection and Quilty lesions. Time-Dependent Changes of Cardiac Nerve Sprouting Nerve density diminished with time compared with that of controls in all areas, including endocardium, myocardium, and around blood vessels (Table 2). The greatest decrease occurred in the myocardium in an exponential fashion (64%, 97%, and 93% decreases with time compared with control). We found smaller but consistent decreases in the endocardium, which stabilized after 12 months (37%, 65%, and 61% decreases with time compared with control). We observed great variability in nerve density in all areas at all time intervals, as reflected by the large standard deviation in each group. Figure 1A through 1F shows comparisons between normal hearts and allografts. In the control group, the nerve densities did not differ

significantly in the endocardium, the myocardium, and around the intramyocardial blood vessels. Areas of myocardial scarring were evident in 2 normal control hearts, and nerve density was similar to that of the endocardium and myocardium. However, we found significant decreases in total nerve density in the allografts. We also observed transmural heterogeneity of nerve distribution. Growth-Associated Protein 43 The GAP43 staining in the control and transplant groups differed remarkably (Table 2). No GAP43 staining was evident anywhere in the control specimens, including the epicardial nerve bundles. This finding indicates the absence of nerve sprouting in the normal control hearts. In contrast, all except for 2 epicardial nerve bundles stained positively in the allograft hearts (Figure 2). In the transplant group, GAP43-positive nerves were evident in the endocardium, the myocardium, and around the blood vessels

1352

Kim et al.

The Journal of Heart and Lung Transplantation December 2004

Table 2. Nerve Density in Normal Controls and in the Study Groups Nerves/mm2 Stain S100 GAP43 TH§

EC 152.0 ⫾ 126 0 24.7 ⫾ 16.7

Group 1 n⫽8

S100 GAP43 TH§

127.3 ⫾ 90.7 6.0 ⫾ 14 51.4 ⫾ 66.7

Group 2 n⫽7

S100 GAP43 TH§

Group 3 n ⫽ 14

S100 GAP43 TH§

Controls n⫽6

Myo 137.3 ⫾ 134 0 34.7 ⫾ 8.0

BV 208.0 ⫾ 109.3 0 113.4 ⫾ 12.7

SC (n) 136.6 ⫾ 136.6 (2) 0 (2) 46.7 ⫾ 66.0 (2)

REJ (n)

QL (n)

N/A (0)

N/A (0)

49.3 ⫾ 110* 14.0 ⫾ 16.7 46.7 ⫾ 46.0

130.7 ⫾ 103.3 30.7 ⫾ 20.0 78.0 ⫾ 88.0

60.0 (1) 0 (1) 0 (1)

81.3 ⫾ 80.0 (4) 13.4 ⫾ 16.4 (4) 0 (4)

224.7 ⫾ 341.3 (4) 0 (4) 0 (4)

86.7 ⫾ 58.7 8.7 ⫾ 14.7 0

4.7 ⫾ 6.7* 12.7 ⫾ 16 0

72.7 ⫾ 30* 49.3 ⫾ 28.7† 93.4

26.7 (1) 0 (1) 0 (1)

40.0 ⫾ 42 (7) 15.3 ⫾ 26 (7) 0 (7)

191.3 ⫾ 130 (7) 17.3 ⫾ 45.3 (7) 0 (7)

44.7 ⫾ 41.3* 30.0 ⫾ 46.7 0

9.3 ⫾ 10.7* 12.7 ⫾ 17.3 20.0 ⫾ 24.0

82.0 ⫾ 82* 57.3 ⫾ 51.3‡ 75.4 ⫾ 62.0

93.3 ⫾ 75.3 (3) 2.0 ⫾ 4 (3) 0 (3)

71.3 ⫾ 55.3 (9) 11.9 ⫾ 23.3 (9) 0 (9)

90.7 ⫾ 118 (5) 0 (5) 0 (5)

*p ⬍ 0.05 compared with controls. † p ⬍ 0.05 compared with controls and Group 1. ‡ p ⬍ 0.01 compared with controls and Group 1. § TH was positive only in 8 specimens (3 in Group 1, 1 in Group 2, and 4 in Group 3). The nerve density is the mean of those specimens that stained positively for TH. BV, blood vessel; EC, endocardium; GAP43, growth-associated protein 43; Myo, mid-myocardium; N/A, not applicable; QL, Quilty lesion; Rej, area of rejection; SC, scar; TH, tyrosine hydroxylase.

within 6 months (Group 1), but were most significant around the blood vessels (p ⬍ 0.01) in each group. In the mid-myocardium, GAP43-stained nerve density remained small in all 3 groups. In the endocardium, we observed a 3.5-fold increase in the GAP43-stained nerve density from Group 2 to Group 3; however, this did not reach statistical significance. Density of nerve sprouts in the areas of scarring, rejection, or Quilty lesions were no greater than 17.3 nerves/mm2. Nerve sprouting in areas of rejection and in Quilty lesions was greatest in Group 2, 15.3 and 17.3 nerves/ mm2, respectively. Tyrosine Hydroxylase Staining Although hearts with TH-positive staining were few (8 of 29 specimens), positive cases were distributed in all 3 groups. Figure 3 shows examples of TH staining in a control heart and in an allograft 24 months after transplantation. Table 2 shows that TH staining showed significant heterogeneity. Although we saw TH-positive nerves in the endocardium, the myocardium, and around the blood vessels in Group 1, we saw them only around the blood vessels in Group 2 and in both the myocardial layer and around the blood vessels in Group 3. Nerves in Areas of Scarring, Rejection, and Quilty Lesions Nerves also were present in areas of acute rejection, scarring, and were particularly abundant in Quilty lesions (Figure 1G through 1I). The Quilty lesions are focal areas of lymphocytic infiltrates along the endocardium seen in transplant recipients who are receiv-

ing immunosuppressants, especially cyclosporine.10 The nerve densities in areas of scarring, rejection, or Quilty lesions varied in the 3 groups. In Groups 1 and 2, the nerve densities in the areas of scarring and rejection were intermediate to densities of the endocardium and the myocardium. In the same groups, the nerve densities in the Quilty lesion were markedly greater. However, we observed a wide range in density; 1 specimen in Group 1 (Patient 5) showed no nerves in this area, but another (Patient 7) had a density of 734 nerves/mm2. In Group 3, the nerve densities in the areas of scarring, rejection, and Quilty lesions were comparable to those around blood vessels. Significant heterogeneity in nerve densities occurred in these areas as well. Less Nerve Sprouting in Those Transplanted for Nonischemic Cardiomyopathy Table 3 shows the data separated according to patients who required transplantation for ischemic cardiomyopathy (n ⫽ 16) and those requiring transplantation for non-ischemic cardiomyopathy (n ⫽ 12). We excluded Patient 1 from this analysis because pre-transplant diagnosis included both causes. The mean age of those who underwent transplantation for idiopathic dilated cardiomyopathy was significantly younger than that of patients with prior ischemic cardiomyopathy (52 ⫾ 14 vs 61 ⫾ 7 years, p ⫽ 0.046). We found a significant difference between these 2 groups in S100-positive nerves around blood vessels (p ⬍ 0.05). Additionally, we found significantly more nerve sprouting around areas of rejection in those who had ischemic cardiomyopathy than we found

The Journal of Heart and Lung Transplantation Volume 23, Number 12

Kim et al.

1353

Figure 1. S100-positive nerve sprouts (black thin arrows) in normal endocardium (A), myocardium (B), and perivascular area (C); and in allograft endocardium (D), myocardium (E), perivascular area (F), area of rejection (G), scar (H), and Quilty lesion (I) ⬎1 year after transplantation. A nerve fiber is present (large block arrow) around the blood vessel in the allograft heart (F). All figures shown under ⫻40 objective lens.

in those with dilated cardiomyopathy (p ⬍ 0.01). Tyrosine hydroxylase staining was positive in specimens from 5 patients who underwent transplantation for ischemic cardiomyopathy and in specimens from 2 recipients with prior non-ischemic cardiomyopathy. The TH nerve density seemed to differ between the 2 groups, but this was not statistically significant (Table 3). DISCUSSION In this study, we found a time-dependent decrease in total (S100-positive) nerves. The greatest decrease occurred in the mid-myocardium between the first and the second 6 months after transplantation. Afterward, the nerve counts stabilized because of increased nervesprouting activity in various areas of the mid-anterior

left ventricular wall, especially around intramyocardial blood vessels. Nerve-sprouting activity varied greatly within the same patient and among different patients, resulting in a heterogeneous distribution of cardiac nerves in the allografts. We found abundant TH-positive nerves, proving that sympathetic nerves are present in the allografts. In some patients, we found high nerve density in the Quilty lesions, especially within 1 year after transplantation. Patients with ischemic heart disease had a greater magnitude of nerve sprouting and re-innervation than did those with dilated cardiomyopathy. These findings provide the 1st histopathologic proof of cardiac allograft re-innervation through sympathetic nerve sprouting.

1354

Kim et al.

The Journal of Heart and Lung Transplantation December 2004

Figure 2. Regenerating neurons (growth-associated-protein-43–positive) in allograft endocardium (A), myocardium (B), perivascular area (C), scar (D), rejection (E), and Quilty lesion (F) after 1 year. Nerves are marked with black arrows (⫻40 objective).

The Journal of Heart and Lung Transplantation Volume 23, Number 12

Kim et al.

1355

Figure 3. Sympathetic nerves (tyrosine hydroxylase positive) in normal heart endocardium (A), myocardium (B), perivascular area (C); and in allograft myocardium (D) and perivascular area (E) after 2 years. Nerves are marked with black arrows (⫻40 objective).

1356

Kim et al.

The Journal of Heart and Lung Transplantation December 2004

Table 3. Nerve Density in Allografts From Patients With Ischemic Vs Non-ischemic Cardiomyopathy Nerves/mm2 Cardiomyopathy Ischemic

Non-ischemic

Stain S100 GAP43 TH§ S100 GAP43 TH§

EC 94.7 ⫾ 81.3 26.0 ⫾ 44 5.3 ⫾ 12.0 46.0 ⫾ 37.3 8.0 ⫾ 21.3 0

MC 8.0 ⫾ 10 16.7 ⫾ 17.3 14.7 ⫾ 14.7 9.3 ⫾ 9.3 6.0 ⫾ 12.7 23.3 ⫾ 32.7

BV 112.7 ⫾ 88* 57.3 ⫾ 38.7 46.7 ⫾ 34.0 54.0 ⫾ 49.3 38.7 ⫾ 44 110.1 ⫾ 80.0

SC 76.7 ⫾ 70 1.7 ⫾ 3.3 0 60.0 0 N/A

REJ 82.0 ⫾ 74.7 19.3 ⫾ 27.3† 0 42.7 ⫾ 14.7 2.0 ⫾ 4.7† 0

QL 189.3 ⫾ 234.7 0 0 140.7 ⫾ 130.7 17.3 ⫾ 45.3 N/A

*p ⬍ 0.05 compared with controls. † p ⬍ 0.05 compared with controls and Group 1. § Only 5 specimens in the ischemic group and 2 specimens in the non-ischemic group stained positively for TH. The nerve density is the mean of only those samples that stained for TH. The difference in TH nerve density among the groups was not statistically significant. BV, blood vessel; EC, endocardium; GAP43, growth-associated protein 43; Myo, mid-myocardium; N/A, not applicable; QL, Quilty lesion; Rej, area of rejection; SC, scar; TH, tyrosine hydroxylase.

Growth-Associated Protein 43 and Nerve Sprouting in Human Hearts Growth-associated protein 43 is a polypeptide induced in neurons when they grow axons.11 The ability to up-regulate GAP43 is important in determining whether axons in the central nervous system can regenerate after injury.12 In cardiac tissues, GAP43 staining can be used to identify growing nerves in dogs with myocardial infarction or with chronic, rapid atrial pacing.13,14 However, no reports describe GAP43 staining in the human myocardium. In this study, we demonstrated for the 1st time positive GAP43-stained nerves in the human heart. The presence of GAP43-positive nerves strongly supports the conclusion that cardiac nerve sprouting occurs in transplanted human hearts. This active nerve-sprouting activity may underlie the mechanism of cardiac re-innervation after transplantation. Tyrosine Hydroxylase Staining Another novel finding of this study is the presence of TH-positive nerves in the transplant ventricles. Previous histopathologic studies of allograft hearts either failed to detect TH-positive nerves6 or did not attempt to stain for this specific sympathetic nerve marker.15 In our study, although few in number, TH-stained sympathetic nerves occurred in all 3 groups. Our ability to document TH staining may be explained by a larger number of allografts we had available for study, better preservation of tissue for immunohistochemistry, or perhaps more effective antigen retrieval strategies the use of microwave antigen retrieval).9 The demonstration of TH-positive nerves provides the 1st histopathologic proof of sympathetic re-innervation of allograft hearts. The positive cells in Group 1 are most likely retained sympathetic nerves whose demise is not yet evident, especially in the 4 patients who died within 14 days. The distribution of TH-positive nerves in Group 2 occurred primarily around blood vessels, and in

Group 3 distribution occurred around both blood vessels and the myocardial layer. Patterns of Nerve Sprouting Our study reveals that nerve density is more prominent in areas in close proximity to blood vessels. The nerve densities in the endocardium and around the blood vessels are greater than densities in the myocardium in all 3 groups. This finding is not unexpected because similar patterns of sympathetic nerve sprouting are observed elsewhere in the body. For example, invasion of the central nervous system by perivascular sympathetic axons often is used as an example for neuronal plasticity.16 In our own experience, large cardiac nerves often are found along the blood vessels in dogs and in humans.8,13 Areas of rejection also demonstrated increased nerve density compared with those of the myocardium. Various extracellular matrix proteins, endogenous neighboring cells such as fibroblasts, or exogenous cells such as infiltrating macrophages regulate nerve growth and facilitate nerve sprouting.17 Moreover, evidence for marked nerve sprouting in Quilty lesions corroborates an association between inflammation and nerve sprouting. Quilty lesions have been described as focal, endocardial, inflammatory infiltrates, predominantly lymphocytic with associated plasma cells and macrophages, found in transplanted hearts.18 They occur with the use of cyclosporine or related immunosuppressive agents. These focal areas contain venules that course through the lesions. Nerve densities are remarkably greater in these areas when they are present compared with other areas of the tissue that we have analyzed. The wellvascularized nature of Quilty lesions and the abundant inflammatory infiltrates may provide an environment conducive to nerve sprouting. We found less nerve sprouting in those who underwent transplantation for non-ischemic cardiomyopathy than in those who underwent transplantation for ischemic cardio-

The Journal of Heart and Lung Transplantation Volume 23, Number 12

myopathy. This observation was also noted in a nuclear imaging study by De Marco et al19 of transplant recipients. However, conflicting observations have been reported, and the mechanisms by which this difference occurs remain unclear. Auto-antibodies to ␤1-receptors and adenosine diphosphate and adenosine triphosphate carriers in the inner mitochondrial membrane have been described in those with idiopathic dilated cardiomyopathy.20 The authors postulated that anti-autonomic neuronal antibodies could be present in those with dilated cardiomyopathy and could destroy regenerating nerves. In contrast, Bengal et al21 showed that the maximal [11C]hydroxyephedrine retention was greater in patients undergoing transplantation because of idiopathic cardiomyopathy than in those with prior ischemic cardiomyopathy (9.4% ⫾ 3.3% per minute vs 7.2% ⫾ 4.0% per minute, p ⫽ 0.016). The results of this latter study are not necessarily incompatible with our results. Patients with ischemic heart disease may have greater regional nerve-sprouting activity but have less global [11C]hydroxyephedrine retention. Ischemic heart disease often is associated with hypercholesterolemia. Our recent study showed that hypercholesterolemia can trigger cardiac nerve sprouting in rabbits before the development of atherosclerosis or of myocardial infarction.22 These results suggest that dyslipedemia may be responsible for the greater nerve-sprouting activity in patients with ischemic heart diseases than that in patients with cardiomyopathy. Clinical Implications Many functional studies have shown that cardiac reinnervation after transplantation is time dependent. Although little evidence suggests functionally significant cardiac re-innervation early after transplantation,23 the neurochemical markers in some (but not all) patients showed significant re-innervation many years later.5,24 An important clinical implication of the current study is that we provided histologic evidence of initial denervation followed by later re-innervation in some patients after transplantation. These findings provided histologic confirmation of the chronologic sequence of functional re-innervation. Limitations This is a retrospective study of tissue from a limited, albeit significant, portion of the transplanted heart. It would be of interest to study nerve densities in specific regions, such as the roots of the great vessels, or around the atria. However, tissue from these sites was not available in our archival tissue blocks. Because very few patients have documented arrhythmia immediately before death, we could not perform meaningful statistical tests to determine whether an association exists between arrhythmia and nerve

Kim et al.

1357

density. In addition, because of the relatively small number of cases, we could not correlate the degree of nerve sprouting with the numerous potential patient variables that could affect this process. Future prospective studies will allow for more comprehensive morphologic evaluation, including the distribution of nerve sprouting in the transplanted heart and correlation with patient variables and clinical events such as angina, arrhythmias, and autonomic reflexes. The authors thank Dr. Jeffrey Saffitz for assistance with immunohistochemical staining protocols and also thank Catherine Wack and Elaine Lebowitz for their assistance. REFERENCES 1. Guth L. Regeneration in the mammalian peripheral nervous system. Physiol Rev 1956;36:441–78. 2. Wilson RF, Laxson DD, Christensen BV, et al. Regional differences in sympathetic reinnervation after human orthotopic cardiac transplantation. Circulation 1993;88: 165–71. 3. Schwaiblmair M, von Scheidt W, Uberfuhr P, et al. Functional significance of cardiac reinnervation in heart transplant recipients. J Heart Lung Transplant 1999;18: 838 –45. 4. Bengel FM, Ueberfuhr P, Schiepel N, et al. Effect of sympathetic reinnervation on cardiac performance after heart transplantation. N Engl J Med 2001;345:731–8. 5. Bengel FM, Ueberfuhr P, Schiepel N, et al. Myocardial efficiency and sympathetic reinnervation after orthotopic heart transplantation: a noninvasive study with positron emission tomography. Circulation 2001;103:1881–6. 6. Wharton J, Polak JM, Gordon L, et al. Immunohistochemical demonstration of human cardiac innervation before and after transplantation. Circ Res 1990;66:900 –12. 7. Huang MH, Ewy GA. Sympathetic reinnervation of the transplanted heart. N Engl J Med 2001;345:1914 –5. 8. Cao JM, Fishbein MC, Han JB, et al. Relationship between regional cardiac hyperinnervation and ventricular arrhythmia. Circulation 2000;101:1960 –9. 9. Shi SR, Key ME, Kalra KL. Antigen retrieval in formalinfixed, paraffin-embedded tissues: an enhancement method for immunohistochemical staining based on microwave oven heating of tissue sections. J Histochem Cytochem 1991;39:741–8. 10. Joshi A, Masek MA, Brown BW Jr, Weiss LM, Billingham ME. “Quilty” revisited: a 10-year perspective. Hum Pathol 1995;26:547–57. 11. Meiri KF, Pfenninger KH, Willard MB. Growth-associated protein, GAP-43, a polypeptide that is induced when neurons extend axons, is a component of growth cones and corresponds to pp46, a major polypeptide of a subcellular fraction enriched in growth cones (published erratum appears in Proc Natl Acad Sci U S A 1986;83: 9274). Proc Natl Acad Sci U S A 1986;83:3537–41. 12. Anderson PN, Campbell G, Zhang Y, et al. Cellular and molecular correlates of the regeneration of adult mamma-

1358

13. 14.

15.

16.

17.

18.

Kim et al.

lian CNS axons into peripheral nerve grafts. Prog Brain Res 1998;117:211–32. Cao JM, Chen LS, KenKnight BH, et al. Nerve sprouting and sudden cardiac death. Circ Res 2000;86:816 –21. Chang CM, Wu TJ, Zhou SM, et al. Nerve sprouting and sympathetic hyperinnervation in a canine model of atrial fibrillation produced by prolonged right atrial pacing. Circulation 2001;103:22–5. Schwaiger M, Hutchins GD, Kalff V, et al. Evidence for regional catecholamine uptake and storage sites in the transplanted human heart by positron emission tomography. J Clin Invest 1991;87:1681–90. Crutcher KA. Sympathetic sprouting in the central nervous system: a model for studies of axonal growth in the mature mammalian brain. Brain Res 1987;434:203–33. Heumann R, Lindholm D, Bandtlow C, et al. Differential regulation of mRNA encoding nerve growth factor and its receptor in rat sciatic nerve during development, degeneration, and regeneration: role of macrophages. Neurobiology 1987;84:8735–9. Dong C, Winters GL, Wilson JE, et al. Enhanced lymphocyte longevity and absence of proliferation and lymphocyte apoptosis in Quilty effects of human heart allografts. Am J Pathol 1997;151:121–30.

The Journal of Heart and Lung Transplantation December 2004

19. De Marco T, Dae M, Yuen-Green MS, et al. Iodine-123 metaiodobenzylguanidine scintigraphic assessment of the transplanted human heart: evidence for late reinnervation. J Am Coll Cardiol 1995;25:927–31. 20. Schulze K, Becker BF, Schultheiss HP. Antibodies to the ADP/ATP carrier, an autoantigen in myocarditis and dilated cardiomyopathy, penetrate into myocardial cells and disturb energy metabolism in vivo. Circ Res 1989;64: 179 –92. 21. Bengel FM, Ueberfuhr P, Hesse T, et al. Clinical determinants of ventricular sympathetic reinnervation after orthotopic heart transplantation. Circulation 2002;106:831–5. 22. Liu YB, Wu CC, Lu LS, et al. Sympathetic nerve sprouting, electrical remodeling, and increased vulnerability to ventricular fibrillation in hypercholesterolemic rabbits. Circ Res 2003;92:1145–52. 23. Rundqvist B, Casale R, Bergmann-Sverrisdottir Y, et al. Rapid fall in sympathetic nerve hyperactivity in patients with heart failure after cardiac transplantation. J Card Fail 1997;3:21–6. 24. Kaye DM, Esler M, Kingwell B, et al. Functional and neurochemical evidence for partial cardiac sympathetic reinnervation after cardiac transplantation in humans. Circulation 1993;88:1110 –8.