Microvascular Thrombosis and Cardiac Allograft Vasculopathy in Rat Heart Transplantation

ALLOGRAFT VASCULOPATHY

Microvascular Thrombosis and Cardiac Allograft Vasculopathy in Rat Heart Transplantation Carlos A. Labarrere, MD,a Miguel A. Ortiz, DVM,a Nargiz Ruzmetov, MD,a Marcelo J. Sosa, BS,a Gonzalo Campana, MD,a Colin Terry, MS,a Lee Ann Baldridge, BS,b Roula Antonopoulos, AAS,a and Hector L. DiCarlo, MDa Background:

The role of a hypercoagulable microvasculature in the development of cardiac allograft vasculopathy (CAV) after heart transplantation in humans is not well understood. The aim of this study was to identify an animal model by which to further evaluate the role of coagulation in the pathogenesis of CAV. Methods: Adult male PVG (RT-1c ) rats were transplanted into ACI (RT-1av1) recipients (n ⫽ 29). ACI donors into ACI recipients (n ⫽ 31) and rats with a sham operation (n ⫽ 33) served as controls. All rats received cyclosporine (10 mg/kg/day) on Days 0 to 9 after surgery. Grafts and native hearts were harvested at 10 days to 3 months after surgery. Hearts were processed for immunohistochemistry and light microscopy. A hypercoagulable microvasculature was defined as presence of microvascular fibrin and capillary antithrombin. CAV was defined as the presence of concentric intimal proliferation and chronic inflammatory infiltrate in the arterial intima, and assessed by computerassisted image analysis. Results: Donor and recipient hearts from PVG–ACI rats showed high levels of fibrin (donors 7.5% to 21.9%, recipients 5.1% to 20.2%) and antithrombin (donors 5.2% to 27.9%, recipients 3.3% to 20.8%) at 10 days to 3 months post-transplant. ACI–ACI donor and recipient hearts had lower deposition of fibrin (donors 0.9% to 9.9%, recipients 0% to 4.0%) and antithrombin (donors 1.4% to 15.2%, recipients 0.8% to 4.5%). Hearts from sham-operated rats had negligible amounts of fibrin (0% to 1.5%) and antithrombin (0% to 2.8%). There was a strong association (p ⬍ 0.001) between presence of fibrin and capillary antithrombin and development of CAV. Conclusions: A hypercoagulable microvasculature in a rat model of heart transplantation was associated with development of CAV, as found in humans. J Heart Lung Transplant 2006;25:1213–22. Copyright © 2006 by the International Society for Heart and Lung Transplantation.

Cardiac allograft vasculopathy (CAV), the principal leading cause of morbidity and mortality among longterm heart transplant recipients,1,2 is characterized by a diffuse process of concentric narrowing of the whole arterial allograft microvasculature. The development of this deleterious disease has been associated with the presence of a hypercoagulable microvasculature in humans.3–7 One of the characteristics of a hypercoagulable microvasculature within the allograft is the presFrom the aDivision of Experimental Pathology, Methodist Research Institute, Clarian Health Partners, Indianapolis, Indiana, and bDepartment of Pathology and Laboratory Medicine, Indiana University School of Medicine, Indianapolis, Indiana. Submitted March 16, 2006; revised May 24, 2006; accepted June 27, 2006. Supported by Grant CEL254 from Roche Pharmaceuticals. Reprint requests: Carlos A. Labarrere, MD, Division of Experimental Pathology, Methodist Research Institute, Clarian Health Partners, 1800 North Capitol Avenue, Noyes Pavilion, Suite E504J, Indianapolis, IN 46202. Telephone: 317-962-3537. Fax: 317-962-9369. E-mail: [email protected] Copyright © 2006 by the International Society for Heart and Lung Transplantation. 1053-2498/06/$–see front matter. doi:10.1016/ j.healun.2006.06.013

ence of fibrin deposits and the development of capillary antithrombin binding after the deposition of fibrin, which is never observed in the microvasculature of normal hearts. Although the development of capillary antithrombin binding is associated with decreased risk of CAV and graft failure, when compared with a persistent loss of vascular antithrombin binding, patients developing capillary antithrombin binding still have significantly worse outcome than patients who never lose vascular antithrombin reactivity after transplantation.3–7 The relevance of a hypercoagulable state for the development of CAV has been demonstrated by several recent studies using rodent heart transplant models. First, it has been demonstrated that tissue factor, the principal activator of the coagulation cascade, is present in the intima of coronary arteries after heterotopic rat heart transplantation,8 and hirudin, a specific and potent thrombin inhibitor, inhibits tissue factor expression and decreases intimal hyperplasia in that model. These data suggest that tissue factor inhibition by hirudin, in addition to its direct effect on thrombin, may attenuate the hypercoagulable state and prevent the development of CAV.9 Second, administration of 1213

1214

Labarrere et al.

antithrombin induces indefinite survival of fully allogeneic grafts in a mouse model of cardiac transplantation, suggesting that administration of high-dose antithrombin may be useful as adjuvant therapy to the currently used immunosuppressive therapies after organ transplantation.10,11 Previous findings in humans and in rodent models of heart transplantation led us to investigate: (a) whether a hypercoagulable microvasculature defined by the presence of microvascular fibrin deposits and capillary antithrombin binding could be identified in a rat model of heterotopic heart transplantation; and (b) whether the presence of a hypercoagulable microvasculature is associated with the development of CAV in a rat model. These investigations are relevant when we consider previous research demonstrating that the presence of a hypercoagulable microvasculature is directly associated with the development and progression of CAV and allograft failure in humans.3–7 METHODS Heterotopic Heart Transplantation Adult male (9 to 12 weeks old, 200 to 250 g) PVG (RT-1c; n ⫽ 29) and ACI (RT-1av1; n ⫽ 122) rats were obtained from Harlan (Indianapolis, IN). The study group consisted of ACI recipients receiving heterotopic heart transplantation from PVG donor rats (n ⫽ 29). Control groups consisted of: (a) a transplanted control group of ACI rats used as donor and recipients (n ⫽ 31); and (b) a sham control group of ACI rats receiving a surgical procedure but not having a transplant (n ⫽ 33). The discrepancy in the number of rats included in each group is due to death before completion of pre-established follow-up time. All animals were maintained at the animal care facilities of the Methodist Research Institute at Clarian Health Partners (Indianapolis, IN). Their environment was maintained at 21 ⫾ 2°C with a time-regulated light period of 12 hours and a relative humidity of 55 ⫾ 10%. Ventilation was handled with 10 to 15 air changes per hour. Rats were provided water and dry food ad libitum. All animals received humane care in compliance with The Principles of Laboratory Animal Care, formulated by the National Society for Medical Research, and Guide for the Care and Use of Laboratory Animals, prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication 86-23, revised 1985). Heterotopic heart transplantation was performed using a modification of the original procedure described by Ono and Lindsey.12 Briefly, animals were anesthetized with isoflurane. To maintain anesthesia, they were intubated and mechanically ventilated (18 cm3/kg/ stroke) with 1% to 2% isoflurane using the “Hallowell EMC” anesthesia workstation. The animals then re-

The Journal of Heart and Lung Transplantation October 2006

ceived 0.01 to 0.05 mg/kg of buprenorphine subcutaneously. Donor hearts were harvested after cardiac arrest by coronary perfusion with Stanford cardioplegia solution by infusion proximal to the aortic cross-clamp. The donor abdomen was opened for injection of 300 U of heparin into the inferior vena cava (IVC). The anterior chest wall was separated from the diaphragm and the rib cage was opened to expose the heart. The IVC was ligated, Stanford cardioplegia solution was injected into the aorta, the heart was depressed inferiorly, and the ascending aorta and main pulmonary artery were transected. Vessels on the lung side of the ligatures were divided and the heart was removed. Immediately after explantation, heart grafts were anastomosed; cold ischemic times did not exceed 10 minutes. The recipient abdomen was opened, and the aorta and IVC were dissected as a unit from surrounding connective tissue. The donor pulmonary artery was anastomosed to the infra-renal IVC in an end-to-side fashion. The donor aorta was anastomosed to the abdominal aorta in an end-to-side fashion and the abdomen was closed. Animals were placed on a heating pad until they became mobile. Post-operative pain medication (Buprenex, 1 mg/ kg, intramuscularly) was given at the end of surgery and as needed. Cardiac allograft survival was monitored by daily palpation of the graft. Recipient and sham-operated rats received cyclosporine (10 mg/kg/day) by gavage on Days 0 to 9 after surgery. A group of rats receiving a sham operation was used as control for both rats receiving either a PVG or an ACI donor heart. All cardiac grafts and recipient hearts as well as hearts from sham-operated rats were then explanted at different time-points (10 days, 1 month, 2 months and 3 months) to evaluate the status of the coronary arteries and to perform immunohistochemical studies. Histologic Examination All explanted hearts were promptly sectioned in five slices: (1) basal (including the atria and the base of the ventricles); (2)–(4) middle (including the mid-portion of the ventricles); and (5) apical (included the ventricular apex), as described elsewhere.8 Slices 1, 2, 4 and 5 were embedded in optimum cutting temperature (OCT) compound (Miles, Elkhart, IN), snap-frozen in liquid nitrogen, and stored at ⫺80°C for future immunohistochemical studies. Slice 3 was fixed in 10% formalin and embedded in paraffin, and hematoxylin– eosin-stained sections (5 ␮m) of paraffin-embedded samples were examined by standard light microscopy to test for the presence of lesions of CAV, as previously described.8,13 The presence of one or more vessels with concentric intimal proliferation and chronic inflammatory infiltrate in the arterial intima was considered positive for CAV.

The Journal of Heart and Lung Transplantation Volume 25, Number 10

Labarrere et al.

1215

Immunohistochemistry

Morphometric Analysis of CAV

Heart tissue samples were examined immunohistochemically for antithrombin with rabbit polyclonal antibody to antithrombin (A0296; DakoCytomation, Carpinteria, CA), and for fibrin with mouse monoclonal antibodies to fibrin (T2G1; Accurate, Westbury, NY, and 350, American Diagnostica, Greenwich, CT). Endothelial cells were identified with monoclonal antibody to rat CD31 (TLD-3A12; Chemicon, Temecula, CA), rabbit polyclonal antibody to von Willebrand factor (N1505; DakoCytomation) or rhodamine-conjugated lectin from Bandeiraea (Griffonia) simplicifolia (BS-1, L5264; Sigma-Aldrich, St. Louis, MO). Secondary antibodies were affinity-purified fluorochrome-labeled F(ab=)2 antibody fragments to mouse or rabbit immunoglobulins (Molecular Probes, Eugene, OR, and Protos ImmunoResearch, Burlingame, CA). Immunofluorescence studies and control tests were performed as previously described.14 Briefly, cryostat sections (4 ␮m) obtained from heart tissue samples were air-dried overnight without chemical fixation. Antibody tests were done with primary antibodies and affinity-purified fluorochrome-labeled F(ab=)2 antibody fragments to mouse or rabbit immunoglobulin (Molecular Probes and Protos ImmunoResearch) secondary antibodies.14 Negative controls were performed using rabbit or mouse immunoglobulins (X0936 and X0931; DakoCytomation). Sections were mounted in gel/ mount (MO1; Biomeda, Foster City, CA). Immunoperoxidase studies were performed with slides obtained from paraffin blocks that were dried for 30 minutes at 60°C. They were deparaffinized, hydrated and rinsed with Tris-buffered saline (Dako). All antibodies were antigen retrieved using pH 6.0 target retrieval solution (Dako) to expose formalin-masked antigens, and then cooled and rinsed with Tris-buffered saline. Slides were processed in an autostainer (Dako). Endogenous biotin was blocked using an avidin– biotin blocking system (Dako). Endogenous peroxidase activity was blocked using 3% hydrogen peroxide. Primary antibodies were applied for 60 minutes at room temperature. Slides were developed using an EnVision⫹ Dual Link HRP Kit (Dako). Double-antibody immunohistochemistry was performed using the EnVision double-stain system. Two investigators (C.A.L., H.L.D.), unaware of any specific treatment or time-point evaluated, assessed the immunohistochemical data. The percentage of capillaries with antithrombin reactivity in two sections each of slices 1, 2, 4 and 5 or vessels with fibrin reactivity in two sections of slice 3 from each heart (donor, recipient and sham), using double-staining techniques with antithrombin and BS-1 lectin or CD31 and fibrin and von Willebrand factor, respectively, was calculated to quantify immunohistochemical reactivity.

Transplanted hearts procured at 10 days and 1 to 3 months as well as hearts obtained from recipients and sham-operated rats were sectioned transversely, including both ventricles and the interventricular septum as described earlier (see Histologic Examination). Thin (5-␮m) sections from slice 3, which was previously fixed in 10% formalin and embedded in paraffin, were stained with elastin–van Gieson to delineate the internal elastic lamina. Two sections per heart were examined in a blinded fashion and each artery was analyzed on a digitally acquired image by computer-based planimetry using IMAGEPRO v5.1 (Media Cybernetics, Silver Spring, MD) software. Each artery was assessed for percentage of intimal proliferation (percentage of luminal narrowing), defined as the area of the intima divided by the entire area inside the elastic lamina (intimal area plus luminal area), as described elsewhere.15 In addition, we assessed hearts for the presence of high-grade CAV, defined as the identification of at least one artery with intimal proliferation of ⬎50%.15 We calculated the percentage of arteries with high-grade occlusion by dividing the number of highly occluded arteries by the total number of arteries and multiplying the quotient by 100.15 Statistical Analysis All data were summarized as median (minimum–maximum) of the percentage of vessels with fibrin or capillaries with antithrombin reactivity. Comparisons of medians between donor and recipient heart combinations within PVG–ACI and ACI–ACI transplanted rats were performed using Wilcoxon’s matched-pairs test. Comparisons of medians between donor or recipient hearts from different groups or between recipient hearts and hearts obtained from sham-operated rats were performed using Wilcoxon’s rank-sum tests. No adjustment was made for multiple comparisons. Fibrin, antithrombin and intimal area levels were grouped as being low (⬍25th percentile), moderate (25th to 75th percentiles) or elevated (⬎75th percentile). Tests of association between levels of antithrombin and fibrin reactivity and intimal area and high-grade CAV were performed using Mantel–Haenszel chi-square tests. Tests for differences in the percentage of arteries with high-grade CAV across levels of antithrombin and fibrin reactivity were performed using the Kruskal–Wallis test. p ⬍ 0.05 was considered statistically significant. Statistical analysis was performed by one of the investigators (C.T.). RESULTS The results of the immunohistochemical assessment of the presence of fibrin or antithrombin reactivity within the

1216

Labarrere et al.

The Journal of Heart and Lung Transplantation October 2006

Table 1. Percentage of Vessels With Fibrin and Capillaries With Antithrombin Reactivity in Rat Hearts in the Different Study Groups ACI–ACI Sham

Donor

PVG–ACI Recipient

Donor

Recipient

Time post-Tx Fibrin 10 days 1 month 2 months 3 months

n

Median (min–max)

n

Median (min–max)

n

Median (min–max)

n

Median (min–max)

n

Median (min–max)

7 8 8 10

0.8 (0–1.0) 1.0 (0–1.4) 0.4 (0–1.0) 0.9 (0–1.5)

7 7 7 8

5.6 (2.0–9.9) 4.2 (1.5–5.2) 1.0 (0.9–1.6) 1.0 (0.9–3.1)

7 7 7 10

1.8 (1.0–2.3) 3.0 (1.5–4.0) 1.7 (0.8–1.9) 1.3 (0–2.2)

8 7 7 7

14.7 (7.5–21.9) 16.1 (12.6–20.8) 14.0 (12.2–19.0) 15.5 (11.1–18.1)

8 7 7 7

8.5 (5.1–14.3) 8.0 (5.7–11.1) 11.4 (8.9–12.5) 18.3 (17.8–20.2)

Antithrombin 10 days 1 month 2 months 3 months

7 8 8 10

1.4 (0–1.9) 1.8 (0–2.8) 0.8 (0–2.0) 1.0 (0–2.5)

7 7 7 8

7.9 (3.4–15.2) 4.0 (1.6–4.3) 2.0 (1.9–4.5) 2.2 (1.4–5.1)

7 7 7 10

2.0 (1.0–3.1) 3.7 (3.2–4.3) 2.4 (0.8–4.4) 2.6 (0.9–4.5)

8 7 7 7

13.5 (5.2–27.9) 18.4 (9.7–23.1) 12.8 (10.9–16.0) 14.9 (10.5–19.6)

8 7 7 7

5.8 (3.3–14.6) 6.3 (5.2–11.0) 11.4 (5.7–12.3) 19.0 (18.3–20.8)

No adjustment was made for multiple comparisons. All values are expressed as median (minimum–maximum) of the percentage of vessels with fibrin or capillaries with antithrombin reactivity. a PVG–ACI (donor) vs PVG–ACI (recipient). b ACI–ACI (donor) vs ACI–ACI (recipient). c PVG–ACI (donor) vs ACI–ACI (donor). d PVG–ACI (recipient) vs ACI–ACI (recipient). e PVG–ACI (recipient) vs sham. f ACI–ACI (recipient) vs sham. g Wilcoxon’s matched pairs test. h Wilcoxon’s rank sum test. Continued on page xxx.

cardiac microvasculature of the different groups studied are depicted in Table 1. Hearts obtained from rats having a sham operation showed negligible amounts or an absence (Figure 1) of fibrin (0% to 1.5% during 10 days to 3 months after surgery). Donor hearts from PVG–ACI rats showed the highest deposition of fibrin (7.5% to 21.9%) during follow-up (10 days to 2 months), as shown in Figure 1. Interest-

ingly, the recipient hearts obtained from PVG–ACI rats also showed increased deposition of fibrin (5.1% to 20.2%). Comparisons for microvascular fibrin deposition between donor and recipient heart combinations in PVG–ACI rats showed significant differences at 10 days to 3 months after transplantation (Table 1). When we evaluated the deposition of fibrin in ACI– ACI rats, we identified increased reactivity in donor

Figure 1. Detection of myocardial fibrin in rat hearts. Double antibody technique using antibodies to fibrin (brown) and von Willebrand factor (red). Representative photomicrographs showing absence of fibrin in the cardiac microvasculature (left) and presence of microvascular fibrin deposits (right) in the rat myocardium. Original magnification: ⫻640.

The Journal of Heart and Lung Transplantation Volume 25, Number 10

Labarrere et al.

1217

Table 1. Continued.

p-valuea,g

p-valueb,g

p-valuec,h

p-valued,h

p-valuee,h

p-valuef,h

0.016 0.016 0.031 0.031

0.016 0.219 0.078 0.641

0.008 0.009 0.009 0.007

0.007 0.009 0.009 0.004

0.007 0.007 0.006 0.004

0.016 0.007 0.025 0.127

0.039 0.031 0.031 0.063

0.016 0.703 1 0.461

0.062 0.009 0.009 0.007

0.007 0.009 0.009 0.004

0.007 0.007 0.007 0.004

0.097 0.007 0.017 0.013

hearts (0.9% to 9.9%) during follow-up (10 days to 3 months), but the highest reactivity in this group of hearts was found at 10 days after transplantation (2.0% to 9.9%). During the follow-up period of between 1 and 3 months the capillary antithrombin reactivity was reduced (0.9% to 5.2%). Recipient hearts from ACI–ACI rats had a low deposition of fibrin (0% to 4.0%) during the entire follow-up period (10 days to 3 months) after transplantation. Interestingly, statistical comparisons performed between ACI–ACI donor and recipient heart combinations showed differences ( p ⫽ 0.016) only at 10 days post-transplant (Table 1). We evaluated differences in fibrin deposition between donor hearts from PVG–ACI vs ACI–ACI rats, and found that a higher reactivity (7.5% to 21.9% vs 0.9% to 9.9%) was consistently observed in PVG–ACI donor hearts during the entire follow-up period (10 days to 3 months), with differences being statistically significant at all time-points (p ⬍ 0.01). Comparisons performed between recipient hearts in PVG–ACI vs ACI–ACI rats showed a persistently increased reactivity in PVG–ACI rats (5.1% to 20.2% vs 0% to 4.0%), during the whole follow-up period (10 days to 3 months), with differences reaching statistical significance at all time-points (p ⬍ 0.01). Comparisons performed between recipient hearts from PVG–ACI vs hearts obtained from rats having a sham operation showed significant differences in the deposition of fibrin at all time-points (5.1% to 20.2% vs 0% to 1.5%, during the entire follow-up period of 10 days to 3 months), as shown in Table 1. Comparisons between recipient hearts from ACI–ACI vs hearts obtained from sham-operated rats showed significantly more fibrin (p ⬍ 0.03) in ACI–ACI rat hearts at 10 days to 2 months after surgery (Table 1). Similar findings were observed when we compared the presence of capillary antithrombin binding in the myocardium of the different study groups (Table 1).

Hearts obtained from rats having a sham operation showed the presence of antithrombin binding predominantly in arteries and veins, but very little or no capillary reactivity (0% to 2.8%) was identified during follow-up (10 days to 3 months), as shown in Figure 2. An increased number of capillaries showing antithrombin reactivity was identified in ACI–ACI rat hearts, and the highest reactivity was found in PVG–ACI rat hearts (Figure 2). Donor hearts from PVG–ACI rats showed the highest reactivity (5.2% to 27.9% during 10 days to 3 months follow-up) when compared with reactivity of donor (1.4% to 15.2%) or recipient (0.8% to 4.5%) hearts from ACI–ACI rats (Table 1). Interestingly, as described for the deposition of fibrin, recipient hearts from PVG–ACI rats had an elevated percentage of capillaries with antithrombin binding (3.3% to 20.8%) during the whole follow-up period (10 days to 3 months), as shown in Table 1. We subsequently evaluated whether the deposition of fibrin and the presence of capillary antithrombin binding were associated with development of CAV. To determine a possible association, we grouped all rat hearts, irrespective of the groups studied or the characteristics of the hearts (donors, recipients or sham), according to the percentage of vessels with fibrin deposits or capillaries with antithrombin reactivity. We found a significant association between the presence of microvascular fibrin and increased intimal area (intimal thickening) in the coronary arteries, and between fibrin and high-grade occlusion CAV, as shown in Table 2. A significant association was also found between the percentage of capillaries with antithrombin reactivity and the presence of increased coronary artery intimal thickening, and also between capillary antithrombin and high-grade occlusion CAV (Table 3). The association between the percentage of vessels with microvascular fibrin or capillary antithrombin binding and coronary artery intimal area is shown in Figure 3. Interestingly, the presence of increasing fibrin

1218

Labarrere et al.

The Journal of Heart and Lung Transplantation October 2006

Figure 2. Detection of capillary antithrombin binding in rat hearts. Double staining using antibody to antithrombin (green) and Bandeiraea (Griffonia) simplicifolia (BS-1) lectin (red). Representative photomicrographs showing absence of capillary antithrombin (top row) and presence of capillaries (arrow) antithrombin reactive in rat myocardium (bottom row). Original magnification: ⫻640.

deposits and capillaries showing antithrombin reactivity was associated with increased arterial intimal thickening, not only in large arteries but also in small arteries and arterioles (Figure 4). DISCUSSION In this study we have reported that a rat model of heterotopic heart transplantation exhibits microvascular deposition of fibrin and concomitant development of capillary antithrombin binding, and that these changes are associated with development of CAV. The findings obtained are similar to those previously de-

scribed by our own laboratory in human heart transplantation. We were able to demonstrate that, after heart transplantation in humans, allografts demonstrate fibrin deposits within the first 10 days of the procedure.16,17 Deposition of fibrin within the cardiac microvasculature is associated with development of capillary antithrombin binding in the allograft microvasculature. We proposed that the development of capillary antithrombin reactivity occurs as a wound-healing mechanism and could represent vascular remodeling or neovessel formation after the initial deposition of fibrin and loss of vascular antithrombin.3–7 The development

Table 2. Relation Between Fibrin-reactive Vessels, Intimal Area and High-grade Occlusion Cardiac Allograft Vasculopathy (CAV) Percent of fibrin-reactive vessels Low (⬍1.0%) n ⫽ 36

Moderate (1.0% to 11.4%) n ⫽ 77

Elevated (⬎11.4%) n ⫽ 38

a

Intimal area (%) Low (⬍15.6%) Moderate (15.6% to 29.8%) Elevated (⬎29.8%) High-grade CAV (⬎50% occlusion) Hearts with high-grade CAV, n (%)a,b Arteries with high-grade CAV (%)c a

21 (58.3) 15 (41.7) 0 (0.0) 0 0

15 (19.5) 54 (70.1) 8 (10.4) 10 (13.0) 0 (0–76.9)

p value ⬍0.001

2 (5.3) 7 (18.4) 29 (76.3) 28 (73.7) 36.9 (0–100)

⬍0.001 ⬍0.001

Outcomes are presented as frequency (percent), and tested for association using the Mantel–Haenszel chi-square test. b Data represent the number of hearts (percent) with one or more arteries per heart having high-grade CAV. c Data summarized as median (min–max) of the percentage of arteries per heart with high-grade CAV, and tested for association using the Kruskal–Wallis test.

The Journal of Heart and Lung Transplantation Volume 25, Number 10

Labarrere et al.

1219

Table 3. Relation Between Capillary Antithrombin (AT), Intimal Area and High-grade Occlusion Cardiac Allograft Vasculopathy (CAV) Percent of AT-reactive capillaries Low (⬍2.0%) n ⫽ 39 Intimal area (%)a Low (⬍15.6%) Moderate (15.6% to 29.8%) Elevated (⬎29.8%) High-grade CAV (⬎50% occlusion) Hearts with high-grade CAV, n (%)a,b Arteries with high-grade CAV (%)c a

21 (53.8) 18 (46.2) 0 (0.0) 0 0

Moderate (2.0% to 11.0%) n ⫽ 74 15 (20.3) 52 (70.3) 7 (9.5) 10 (13.5) 0 (0–76.9)

Elevated (⬎11.0%) n ⫽ 38

p-value ⬍0.001

2 (5.3) 6 (15.8) 30 (78.9) 28 (73.7) 33.3 (0–100)

⬍0.001 ⬍0.001

Outcomes are presented as frequency (percent), and tested for association using the Mantel–Haenszel chi-square test. b Data represent the number of hearts (percent) with one or more arteries per heart having high-grade CAV. c Data summarized as median (min–max) of the percentage of arteries per heart with high-grade CAV, and tested for association using the Kruskal–Wallis test.

of a hypercoagulable state within the allografts in humans is associated with increased incidence of CAV and allograft failure. Our study has shown that PVG–ACI rat transplants are associated with increased fibrin deposition and development of capillary antithrombin binding in donor hearts, and these events are detected 10 days after the procedure. These findings are almost identical to those described after human heart transplantation and suggest that, as noted in humans, they occur very early after transplantation. We did not include earlier timepoints in this study to compare groups receiving complete cyclosporine treatment up to Day 9. The detection of microvascular fibrin and capillary antithrombin reactivity in donor hearts from PVG–ACI rats early after transplantation—when high-grade occlusion CAV lesions have not yet developed in this model— emphasizes the importance of looking for early events after heart transplantation as being relevant for subsequent outcome. The early development of these changes suggests that early phenomena, such as ischemia–reperfusion, could be determining factors leading to genera-

tion of a hypercoagulable microvasculature within the allografts. The deposition of fibrin and the generation of capillaries able to bind antithrombin is evidence of a hypercoagulable microvasculature in this model. The principal activator of the coagulation cascade is tissue factor and, when expressed at the endothelial surface, immediately promotes coagulation. Interestingly, the presence of tissue factor has been demonstrated in human cardiac allografts, and myocardial tissue factor levels can predict subsequent development of CAV.18 Furthermore, tissue factor expression has also been demonstrated in a rat model of heterotopic heart transplantation, and such expression was significantly associated with the development of CAV.8 A relevant point that emerges from the present investigations is the detection of microvascular changes within the recipient hearts. The identification of such changes suggests that soluble factors, possibly generated within the allografts, could be reaching the circulation and may subsequently affect recipients’ native hearts. One of these factors is vascular endothelial

Figure 3. Relationship between intimal area and microvascular fibrin and capillary antithrombin binding in rat myocardium. A significant association was found between percentage of fibrin-reactive vessels (left) and percentage of capillaries with antithrombin reactivity (right) and development of intimal proliferation (increased intimal area) in arteries and arterioles of rat myocardium.

1220

Labarrere et al.

The Journal of Heart and Lung Transplantation October 2006

Figure 4. Identification of cardiac allograft vasculopathy (CAV) in the rat heart. Representative photomicrographs showing absence of intimal thickening in large arteries (top row, left) and arterioles (bottom row, left), and presence of intimal proliferation in large arteries (top row, right) and arterioles (bottom row, right) of rat myocardium. Original magnification: ⫻640.

growth factor (VEGF), the expression of which is induced under hypoxic conditions.19,20 Hypoxia/ischemia has been shown to increase myocardial cell VEGF expression in vitro and in vivo.21,22 Human myocardial cells also express VEGF23 as well as smooth muscle cells surrounding an infarcted area.22 We were able to demonstrate increased VEGF expression in hearts having fibrin deposition, and that myocardial cells and smooth muscle cells in these hearts demonstrate VEGF mRNA, strongly suggesting that fibrin deposition and subsequent ischemia within the transplanted myocardium could induce synthesis and release of the growth factor.24 Interestingly, increased VEGF expression within the biopsies with fibrin deposits is mirrored in circulation

by the elevation of VEGF levels.24 These findings suggest that the development of microvessels with the ability to bind antithrombin in recipients’ native hearts, in the rat model of heterotopic heart transplantation, could be a consequence of factors released within the allografts. Another possibility that could explain the presence of both fibrin and capillary antithrombin binding in recipients’ native hearts is that the allografts may release other factors capable of inducing activation of coagulation not only in the allografts but also in the native hearts of the recipient rats. An interesting finding that emerges from these investigations is the persistent identification of fibrin deposits and antithrombin capillary reactivity during the entire follow-up period (10 days to 3 months) in the

The Journal of Heart and Lung Transplantation Volume 25, Number 10

donor hearts of PVG–ACI rats, and the predominantly early identification of fibrin deposits and antithrombin capillary reactivity (10 days) with subsequent reduction of these findings in donor hearts from ACI–ACI rats. These findings suggest that ischemia–reperfusion phenomena could be involved early after transplantation, and the persistence of pro-inflammatory and pro-coagulant events remains in allografts but not isografts. The persistence of a hypercoagulable microvasculature in the allografts could be associated with endothelial cell activation.25 Endothelial activation is not only associated with increased expression of major histocompatibility antigens, adhesion molecules and cytokines, but with disruption of the normal anti-coagulant state that leads to the final deposition of fibrin. The role of anti-coagulation with regard to outcome of solid-organ transplantation has been clearly demonstrated. First, antithrombin pre-treatment was shown to decrease tissue damage after ischemia–reperfusion in experimental pancreas transplantation in rats.26 Treatment with antithrombin inhibited T- and B-lymphocyte activation and improved parameters of inflammation in a rat model of lung transplantation,27,28 and induced indefinite survival of fully allogeneic cardiac grafts.10,11 Interestingly, antithrombin can affect inflammatory processes via inhibition of nuclear factor-␬B, and these effects could directly inhibit expression of pro-inflammatory molecules and tissue factor, because these genes are known to be under the control of nuclear factor-␬B.29 Treatment with hirudin in a rat model of cardiac transplantation inhibited tissue factor expression and decreased neo-intimal hyperplasia, suggesting that hirudin, in addition to the direct effects on thrombin inhibition, may attenuate the hypercoagulable state and prevent the development of CAV.9 Other therapies may also affect the hypercoagulable state within the allografts; for example, it was found that statins can inhibit tissue factor expression in inflammatory cells after heart transplantation.30 The present investigation suggests that the rat model of heterotopic heart transplantation can be used to study the local and systemic hypercoagulable state that occurs after heart transplantation, and could also be utilized to introduce and evaluate new therapies that could affect that state and reduce the intimal proliferation characteristic of CAV. The model also offers the possibility of studying mechanisms involved in microvascular antithrombin binding, such as evaluation of the role of heparan sulfate proteoglycans, because these molecules appear to be involved in antithrombin binding in human cardiac allografts.5,7 Furthermore, heparan sulfate proteoglycan expression is induced in capillaries within healing wounds,31 syndecan-4 is induced by hypoxia in ischemic hearts,32 and the inhibitory potency of antithrombin depends on the interaction of

Labarrere et al.

1221

this molecule with heparin-like cell surface glycosaminoglycans.29 REFERENCES 1. Eisen H, Kobashigawa J, Starling RC, Valantine HA, Mancini DM. Improving outcomes in heart transplantation: the potential of proliferation signal inhibitors. Transplant Proc 2005;37(suppl):4S–17S. 2. Taylor DO, Edwards LB, Boucek MM, Trulock EP, Keck BM, Hertz MI. The registry of the International Society for Heart and Lung Transplantation: twenty-first official adult heart transplant report—2004. J Heart Lung Transplant 2004;23:796 – 803. 3. Labarrere CA, Torry RJ, Nelson DR, et al. Vascular antithrombin and clinical outcome in heart transplant patients. Am J Cardiol 2001;87:425–31. 4. Labarrere CA, Nelson DR, Pitts DE, Kirlin PC, Halbrook H. Immunohistochemical model to predict risk for coronary artery disease and failure in heart transplant patients. Am J Transplant 2001;1:251–9. 5. Labarrere CA, Nelson DR, Park JW. Pathologic markers of allograft arteriopathy: insight into the pathophysiology of cardiac allograft chronic rejection. Curr Opin Cardiol 2001;16:110 –7. 6. Labarrere CA, Deng MC. Microvascular prothrombogenicity and transplant coronary artery disease. Transplant Immunol 2002;9: 243–9. 7. Labarrere CA, Nelson DR, Spear KL. Non-immunologic vascular failure of the transplanted heart. J Heart Lung Transplant 2003; 22:236 – 40. 8. Holschermann H, Bohle RM, Zeller H, et al. In situ detection of tissue factor within the coronary intima in rat cardiac allograft vasculopathy. Am J Pathol 1999;154:211–20. 9. Holschermann H, Bohle RM, Schmidt H, et al. Hirudin reduces tissue factor expression and attenuates graft arteriosclerosis in rat cardiac allografts. Circulation 2000;102:357– 63. 10. Aramaki O, Takayama T, Yokoyama T, et al. Intravenous administration of antithrombin III induces indefinite survival of fully allogeneic cardiac grafts. Transplant Proc 2002;34:1409 –10. 11. Aramaki O, Takayama T, Yokoyama T, et al. High dose of antithrombin III induces indefinite survival of fully allogeneic cardiac grafts and generates regulatory cells. Transplantation 2003;75:217–20. 12. Ono K, Lindsey ES. Improved technique of heart transplantation in rats. J Thorac Cardiovasc Surg 1969;57:225–9. 13. Poston RS Jr, Billingham ME, Pollard J, Hoyt EG, Robbins RC. Effects of increased ICAM-1 on reperfusion injury and chronic graft vascular disease. Ann Thorac Surg 1997;64:1004 –12. 14. Labarrere CA, Lee JB, Nelson DR, Al-Hassani M, Miller SJ, Pitts DE. C-reactive protein, arterial endothelial activation, and development of transplant coronary artery disease: a prospective study. Lancet 2002;360:1462–7. 15. Balsam LB, Mokhtari GK, Jones S, et al. Early inhibition of caspase-3 activity lessens the development of graft coronary artery disease. J Heart Lung Transplant 2005;24:827–32. 16. Labarrere CA, Nelson DR, Faulk WP. Myocardial fibrin deposits in the first month after transplantation predict subsequent coronary artery disease and graft failure in cardiac allograft recipients. Am J Med 1998;105:207–13. 17. Labarrere CA, Nelson DR, Cox CJ, Pitts D, Kirlin P, Halbrook H. Cardiac-specific troponin I levels and risk of coronary artery disease and graft failure following heart transplantation. JAMA 2000;284:457– 64. 18. Yen MH, Pilkington G, Starling RC, et al. Increased tissue factor expression predicts development of cardiac allograft vasculopathy. Circulation 2002;106:1379 – 83.

1222

Labarrere et al.

19. Ferrara N. Vascular endothelial growth factor: basic science and clinical progress. Endocr Rev 2004;25:581– 611. 20. Pages G, Pouyssegur J. Transcriptional regulation of the vascular endothelial growth factor gene—a concert of activating factors. Cardiovasc Res 2005;65:564 –73. 21. Hashimoto E, Ogita T, Nakaoka T, Matsuoka R, Takao A, Kira Y. Rapid induction of vascular endothelial growth factor expression by transient ischemia in rat heart. Am J Physiol 1994;267:H1948 –54. 22. Li J, Brown LF, Hibberd MG, Grossman JD, Morgan JP, Simons M. VEGF, flk-1, and flt-1 expression in a rat myocardial infarction model of angiogenesis. Am J Physiol 1996;270:H1803–11. 23. Shinohara K, Shinohara T, Mochizuki N, et al. Expression of vascular endothelial growth factor in human myocardial infarction. Heart Vessels 1996;11:113–22. 24. Torry RJ, Bai L, Miller SJ, Labarrere CA, Nelson D, Torry DS. Increased vascular endothelial growth factor expression in human hearts with microvascular fibrin. J Mol Cell Cardiol 2001;33:175– 84. 25. Salom RN, Maguire JA, Hancock WW. Endothelial activation and cytokine expression in human acute cardiac allograft rejection. Pathology 1998;30:24 –9. 26. Hackert T, Werner J, Uhl W, Gebhard MM, Buchler MW, Schmidt J. Reduction of ischemia/reperfusion injury by antithrombin III

The Journal of Heart and Lung Transplantation October 2006

27.

28.

29.

30.

31.

32.

after experimental pancreas transplantation. Am J Surg 2005;189:92–7. Zuo XJ, Okada Y, Nicolaidou E, et al. Antithrombin III inhibits Tand B-lymphocyte activation in vitro and improves parameters of inflammation in a rat model of acute lung allograft rejection. Transplant Proc 1999;31:816 –7. Okada Y, Zuo XJ, Marchevsky AM, et al. Antithrombin III treatment improves parameters of acute inflammation in a highly histoincompatible model of rat lung allograft rejection. Transplantation 1999;67:526 – 8. Oelschlager C, Romisch J, Staubitz A, et al. Antithrombin III inhibits nuclear factor ␬B activation in human monocytes and vascular endothelial cells. Blood 2002;99:4015–20. Holschermann H, Hilgendorff A, Kemkes-Matthes B, et al. Simvastatin attenuates vascular hypercoagulability in cardiac transplant recipients. Transplantation 2000;69:1830 – 6. Elenius K, Vainio S, Laato M, Salmivirta M, Thesleff I, Jalkanen M. Induced expression of syndecan in healing wounds. J Cell Biol 1991;114:585–95. Kojima T, Takagi A, Maeda M, et al. Plasma levels of syndecan-4 (ryudocan) are elevated in patients with acute myocardial infarction. Thromb Haemost 2001;85:793–9.