Intraplaque Hemorrhage in Cardiac Allograft Vasculopathy



This article is corrected by:

  1. Errata: Erratum Volume 14, Issue 3, 739, Article first published online: 25 February 2014


Plaque hemorrhage, inflammation and microvessel density are key determinants of plaque vulnerability in native coronary atherosclerosis (ATS). This study investigates the role of intraplaque hemorrhage (IPH) and its relation with inflammation and microvessels in cardiac allograft vasculopathy (CAV) in posttransplanted patients. Seventy coronary plaques were obtained from 12 patients who died because of CAV. For each patient we collected both native heart and the allograft, at the time of transplantation and autopsy, respectively. Intralesion inflammation, microvessels and IPH were assessed semi-quantitatively. IPH was observed in 21/35 (60%) CAV lesions and in 8/35 (22.9%) native ATS plaques, with a strong association between fibrocellular lesions and IPH (p = 0.0142). Microvessels were detected in 26/35 (74.3%) of CAV lesions with perivascular leakage as sign of endothelial damage in 18/26 (69.2%). IPH was strongly associated with microvessels (p < 0.0001). Inflammation was present in 31/35 (88.6%) of CAV lesions. CAV IPH+ lesions were characterized by presence of both fresh and old hemorrhage in 12/21 (57.1%). IPH, associated with microvessel damage and inflammation, is an important feature of CAV. Fresh and old intralesion hemorrhage suggests ongoing remodeling processes promoting the lesion progression and vulnerability.


American Heart Association




cardiac allograft vasculopathy


dilated cardiomyopathy


glycophorin A


ischemic cardiomyopathy


intraplaque hemorrhage


smooth muscle cells


von Willebrand factor


Cardiac allograft vasculopathy (CAV) still represents the Achilles' heel of transplant procedures, accounting for its increasing incidence in approximately 10% every year in the long-term clinical follow-up [1]. CAV is recognized as a concentric thickening of the coronary artery wall due to diffuse proliferation of smooth muscle cells (SMCs) in the intima layer (neointima) [2]. Its pathogenesis is complex and can be related to different risk factors [3]. A tight correlation between acute rejection and the development of CAV has been extensively reported [3]. However, the exact pathophysiological origin of CAV is not fully understood [4, 5]. Previous studies have shown that inflammation and oxidative stress in native coronary atherosclerosis (ATS) contribute to rapidly progressive plaque destabilization [6-8].

Recently several groups have also demonstrated that intraplaque hemorrhage (IPH) is associated with the development of ATS lesions and plaque instability [9-11]. Moreover, there is evidence in patients who died suddenly from acute coronary syndromes that IPH may represent a potent atherogenic stimulus both biochemically and mechanically [12].

Red blood cells are rich in phospholipids and free cholesterol, and their accumulation within the plaque plays a key role in promoting lesion instability through necrotic core expansion and inflammatory cell infiltration [12-14].

IPH correlates positively with both inflammation and the presence of microvessels [15]. Pathological examination of unstable lesions has demonstrated that IPH is closely associated with neoangiogenesis at site of microvessel leakage [16].

For the first time we had the possibility to study and compare in the same patients, CAV from explanted hearts at autopsy with native coronary ATS from hearts removed at transplantation.

In this setting, the aim of our study was to investigate whether and to what extent IPH plays a role in the progression of CAV in patients transplanted for dilated cardiomyopathy (DCM) or ischemic cardiomyopathy (ICM).

Materials and Methods

Selection of patients, inclusion criteria

We selected 12 patients from the heart transplant registry at Padua University who died because of CAV, on whom autopsy was performed, and from whom both the native heart and the graft removed at autopsy were available for histopathology. Patients' written consent for keeping the native heart and further use for research purposes was obtained at transplantation. The hearts retrieved at autopsy were kept for diagnostic and further research purposes according to the permission of the family and by ethical committee approval about research on archived materials (Padova Hospital, Ethical Committee, June 2011).

Four of these patients had been transplanted because of DCM while the other eight were transplanted for ICM. Overall we were able to evaluate 24 hearts. All major epicardial arteries were assessed macroscopically by transverse sections every 0.5 cm and the most severe lesions from the three main epicardial coronary arteries were identified. Among the most severe lesions we included the culprit lesions responsible for the cardiac death. A total of 70 lesions were included in the study, of which 35 were from native hearts and 35 from posttransplant hearts.

Plaque histomorphological classification

Different types of lesion classification have been proposed in the literature by Virmani et al [17]. However at present, there is no internationally acknowledged histological classification for the different types of CAV lesions, but only a clinical and angiographic classification [18]. Therefore, we used the classification of native ATS as proposed by the American Heart Association (AHA) [19, 20] but with some modifications to make it suitable for CAV. Classification was carried out by two pathologists blindly.

Plaque morphologies are classified as follows:

  1. Early lesions including fatty streaks and intimal thickening (corresponding to AHA classification I–III lesions).
  2. Fibrocellular lesions: concentric or eccentric fibromuscular intimal hyperplasia with more than 50% of its area occupied by nucleated cells (not included in AHA classification, and considered peculiar to CAV).
  3. Fibrolipid lesions: concentric or eccentric lesions containing a necrotic core with cholesterol clefts and extracellular lipid deposits covered by a fibrous cap (AHA type IV).
  4. Fibrotic lesions: concentric or eccentric lesions composed of collagen rich tissue, cellularity less than 50% and without significant lipid deposition (AHA class Vc).
  5. Complicated lesions: lesions with intraplaque hematoma and/or superimposed mural thrombus (AHA class VI).

Four groups have been identified: (1) native hearts from DCM patients (12 lesions); (2) transplanted hearts in DCM patients (12 CAV lesions); (3) native hearts from ICM patients (23 lesions); (4) transplanted hearts in ICM patients (23 CAV lesions).

Histological and immunohistochemical methods

Formalin-fixed coronary segments were embedded in paraffin and 4- to 5-µm sections were stained with hematoxylin and eosin, Azan-Mallory staining for histomorphologic evaluation of the lesions.

Immunohistochemical single and double stains were performed on serial sections as previously described [21], using mouse monoclonal anti CD68, anti-mouse monoclonal SMA-1 (reactive with vascular SMCs), CD31 (reactive with endothelial cells and platelets), anti-rabbit von Willebrand factor (vWF; reactive with vWF), anti-LCA (reactive with leucocytes) and anti-glycophorin-A (reactive with erythrocytes and erythrocyte membrane remnants). All antibodies were purchased from Dakopatts (Glostrup, Denmark).

(Immuno) histological assessment of plaque inflammation and plaque hemorrhage

  1. IPH: IPH was evaluated using glycophorin A (GFA) immunostain [22] and Perls iron stain. IPH was classified as either recent-onset (extravascular located clusters of GFA+ intact erythrocytes), old (deposits of GFA+ erythrocyte membrane fragments or iron stain) or ongoing (lesions with simultaneous presence of recent and old IPH). Plaque hemorrhage was scored as 0 (absent) and 1 (present) in all cases.
  2. Inflammation: The topographic localization in lesions, of the CD68+ and LCA+ areas, was scored as superficial (adjacent to the lumen) or deep (inside the lesion). Inflammation was semi-quantitatively scored as 0: absent, 1: present. The ratio between immunopositive staining macrophages and SMCs was scored as CD68/SMA < 10%, CD68/SMA = 10–50% and CD68/SMA >50% in all plaques.
  3. Presence of microvessels: CD31 positive microvessels were scored as follows: 0: absent, 1: <10% vessels and 2: >10% vessels.
  4. Microvascular dilatation and endothelial leakage: Dilatation of vessels was evaluated as 0: absent, 1: present. Leakage of microvessels was assessed by presence of a rim of perivascular staining with anti-vWF as reported previously [12] and was scored as either 0: absent (only staining of endothelial cells), 1: present (additional staining of perivascular tissue).

Statistical analysis

Data are presented as counts and percentages. Association of morphologic plaque characteristics was analyzed with chi-square or Fisher's exact test. The statistical significance was set at the 5% level. The analyses were conducted with SAS 9.2 (SAS Institute, Inc., Cary, NC) for Windows.


Clinical profile of patients

Clinical characteristics of all patients who were transplanted for either DCM or ICM are presented in Table 1.

Table 1. Baseline clinical characteristics
ptAge at HT (yrs)Indication to HTSurvival (yrs)HypertensionDiabetesSmokingDonor age (yrs)PRA (%)Mismatch (AB and DR)Ischemic time (min)Total rejection scoreSevere rejection score
  1. DCM, dilated cardiomyopathy; HT, heart transplantation; ICM, ischemic cardiomyopathy; min, minutes; n, no; na, not available; PRA, panel reactive antibodies; pt, patient; y, yes; yrs, years.
121DCM13ynn11na5/12 (42%)12010.18
255DCM18yyn20na2/12 (17%)1201.30.09
345DCM17nnn2804/12 (33%)1800.410.06
427DCM17nnn16na4/12 (33%)1200.460.04
559ICM1.3nnn41na3/12 (25%)601.40.13
663ICM6yyn1904/12 (33%)1800.60.27
740ICM3.3nnn2704/12 (33%)1201.570.19
847ICM1.5nnn5806/12 (50%)601.190.19
951ICM1.1ynn2702/12 (17%)1801.650.3
1053ICM10yyn1603/12 (25%)1202.60.36
1259ICM3.6nnn3004/12 (33%)12010.73

Histomorphological classification of lesions in native hearts and in transplanted hearts

Histomorphological classification of the lesions in native ATS and CAV is reported in Table 2.

Table 2. Plaque classification scheme and morphology of ATS and CAV plaques
Plaque classificationNative ATS (n = 35), n (%)CAV (n = 35), n (%)Total (n = 70), n (%)
  1. ATS, native atherosclerotic plaques in explant hearts; CAV, cardiac allograft vasculopathy plaques in graft hearts at autopsy.
1. No lesions3 (8.6)0 (0)3 (4.3)
2. Early lesions7 (20)3 (8.6)10 (14.3)
3. Fibrocellular lesions3 (8.6)17 (48.6)20 (28.6)
4. Fibrotic lesions8 (22.9)3 (8.6)11 (15.7)
5. Fibrolipid lesions13 (37.1)8 (22.9)21 (30)
6. Complicated lesions1 (2.9)4 (11.4)5 (7.1)

In the DCM hearts, only rarely were ATS plaques detected (9 lesions in 12 samples, of which 7 were early lesions). In ICM hearts, 23 lesions were found, which were all advanced plaques (including 1 lesion complicated by thrombosis). By contrast, nearly all lesions studied from the 12 autopsy hearts (CAV lesions) were advanced plaques, of which 4 were complicated by thrombosis. See Table 3 for details.

Table 3. Plaque classification scheme and morphology according to the native pathology leading to transplantation
Plaque classificationDCM (n = 12)ICM (n = 23)
Native ATS, n (%)CAV, n (%)Native ATS, n (%)CAV, n (%)
  1. CAV, cardiac allograft vasculopathy lesions in allograft at autopsy; DCM, lesions from patients transplanted because of dilated cardiomyopathy; ICM, lesions from patients transplanted because of ischemic cardiomyopathy; native ATS, native atherosclerotic plaques in explanted hearts.
1. No lesions3 (25)0 (0)0 (0)0
2. Early lesions7 (58.3)0 (0)0 (0)3 (13)
3. Fibrocellular lesions0 (0)8 (66.7)3 (13)9 (39.1)
4. Fibrotic lesions0 (0)1 (8.3)8 (34.8)2 (8.7)
5. Fibrolipid lesions2 (16.7)2 (16.7)11 (47.8)6 (26.1)
6. Complicated lesions0 (0)1 (8.3)1 (4.3)3 (13)

IPH in CAV lesions versus native plaques

In the allografts from DCM and ICM patients, altogether 21 out of 35 (60%) CAV lesions showed signs of either fresh (presence of extracellular intact erythrocytes) or old (GFA immunopositive erythrocyte remnants, membrane fragments and/or iron deposits) IPH, while in the native ATS, which were derived from the explant hearts of the same patients, IPH was present in 8 of 35 (22.9%) lesions (Figure 1A–C, p = 0.0016). Fresh IPH was present in 4/35 (11.4%) of CAV lesions and in 3/35 (8.6%) in ATS plaques (Figure 1D). Old hemorrhages were present in 5/35 (14.3%) of CAV lesions versus 2/35 (5.7%) in ATS plaques (Figure 1D–F). Fresh and old IPH, as ongoing IPH, were present in 12/21 (57.1%) CAV lesions compared to 3/8 (37.5%, Figure 1D) ATS.

Figure 1.

Analysis of intraplaque hemorrhage (IPH). (A) The graph shows the percentage (%) of IPH in native atherosclerotic (ATS) lesions and in cardiac allograft vasculopathy (CAV). IPH is present in 21/35 (60%) of CAV lesions compared to ATS 8/35 (22.9%, p = 0.0016); (B) graph shows the percentage of IPH in CAV and ATS lesions separated in fresh, old and for both fresh and old staining. Fresh IPH was present in 4/35 (11.4%) of CAV lesions and in 3/5 (8.6%) in ATS plaques. Old hemorrhage was present in 5/35 (14.3%) of CAV lesions versus 2/35 (5.7%) in ATS plaque lesions. Fresh and old IPH, as ongoing IPH phenomenon, were present in 12/21 (57.1%) CAV lesions compared to ATS with 3/8 (37.5%); (C) immunostaining for glycophorin A antibody that identifies fresh staining for IPH. Original magnification 40×; (D) immunostaining for glycophorin A, antibody that identify staining for IPH. Original magnification 40×.

In CAV lesions IPH co-localized in the necrotic core with extracellular lipid material, close to numerous cholesterol clefts. Fresh and old IPH were localized in different sites in 12/21 (57.1%) plaques.

The incidence of IPH in CAV lesions did not differ significantly from culprit (55.6%) and nonculprit (78.6%, p = 0.2656) lesions.

IPH, microvessels, inflammation and endothelial leakage in CAV lesions versus native plaques

Intraplaque CD31-immunopositive microvessels were present in 23/35 (65.7%) of the native ATS. Only in 8/23 (34.8%) of these cases the microvessels showed a rim of perivascular immunostaining with anti-vWF antibody, indicating endothelial leakage as previously reported [9, 23].

In CAV lesions, 26 out of 35 (74.3%, p = 0.4339 vs. ATS microvessels) showed presence of intimal microvessels, and 18/26 (69.2%) of them presented perivascular anti-vWF immmunostaining, indicating vessel leakage (Table 4).

Table 4. Plaque classification scheme and morphology
Plaque classification CAV (n = 35), n (%)Total (n = 35)
No lesionsEarly lesionsFibro cellular lesionsFibrotic lesionsFibrolipid lesionsComplicated lesions
0 (0)3 (8.6)17 (48.6)3 (8.6)8 (22.9)4 (11.4)
  1. CAV, cardiac allograft vasculopathy; IPH, intraplaque hemorrhage.
IPH, n (%)Positive0 (0)0 (0)11 (64.7)1 (33.3)6 (75)3 (75)21 (60)
 Negative0 (0)3 (0)6 (35.3)2 (66.6)2 (25)1 (25)14 (40)
Inflammation, n (%)Positive0 (0)1 (33.3)15 (88.2)3 (100)8 (100)4 (100)31 (88.6)
 Negative0 (0)2 (66.6)2 (11.8)0 (0)0 (0)0 (0)4 (11.4)
Microvessels, n (%)Positive0 (0)0 (0)13 (76.5)3 (100)6 (75)4 (100)26 (74.3)
 Negative0 (0)3 (100)4 (23.5)0 (0)2 (25)0 (0)9 (25.7)
Vessel leakage, n (%)Positive0 (0)0 (0)11 (64.7)3 (100)3 (37.5)3 (75)20 (57.1)
 Negative0 (0)3 (100)6 (35.3)0 (0)5 (62.5)1 (25)15 (42.9)

Inflammation was identified as areas with CD68+ and LCA+ cells, indicating presence of macrophages and leukocytes. Inflammation was present in 31/35 (88.6%) of the CAV lesions and in 25/35 (71.4%) of the ATS lesions (Table 4, p = 0.1102 vs. ATS lesions). In the ATS plaques, it was not possible to define the presence of inflammation in one early lesion. A superficially located inflammatory infiltrate was more commonly found in CAV lesions (26/35, 74.3%) than in ATS lesions (14/35, 40%, p = 0.0047). Inflammatory infiltrates that were located deep inside the lesions were found both in CAV (25/35, 71.4%) and in ATS lesions (19/35, 54.3%) with no statistically significant differences between them (Table 4). Superficial and deep inflammation together were found in 20/35 (57.1%) CAV lesions compared to 6/35 (17.1%) ATS plaques (p = 0.0121). Moreover, in CAV lesions, the IPH events were strongly associated with presence of microvessels, 20/21 (95.2%) IPH positive plaques (p = 0.0009) as in ATS group (p = 0.0316).

In CAV lesions areas with IPH frequently co-localized with microvessels and perivascular anti-vWF immunostaining and with inflammatory cells at the same sites within the plaques (Figure 2).

Figure 2.

Overview of IPH, microvessels and endothelial leakage in CAV lesions. (A) Hematoxylin and eosin (H&E) of a cross-section of a cardiac allograft vasculopathy (CAV) lesion. The bar indicates 500 µm. (B) Detail of the black insert in (A). Immunostaining for glycophorin A, showing extravasation of intact erythrocytes, indicating fresh hemorrhage. The bar indicates 100 µm. (C) Detail of the black insert in (A). Immunostaining for CD31 indicates presence of microvessels in the plaque. Note as areas with intraplaque hemorrhage (IPH) in CAV lesions frequently co-localized with presence of microvessels. Bar indicates 100 µm. (D) Detail of the black insert in (A). Immunostaining for von Willebrand factor (vWF): leakage of microvessels is identified by the presence of diffuse perivascular vWF deposits. Note as areas with IPH in CAV lesions co-localized also with microvessel and vessel leakage. Bar indicates 100 µm.

IPH, microvessels, inflammation and vessel leakage in relation to plaque morphology

Presence of IPH, neoangiogenesis, inflammation and vessel leakage in CAV lesions were also evaluated in relation to the specific type of plaque morphology (Table 4).

IPH was found in 11/17 (64.7%) of fibrocellular lesions, the lesion type specific for CAV. Three out 11 of these lesions (27.3%) had fresh IPH, 2/11 (18.2%) had old IPH while 6/11 (54.5%) had features of both fresh and old IPH in the same lesion. Moreover, 13/17 (76.5%) fibrocellular lesions also showed presence of microvessels, 15/17 (88.2%) had inflammatory infiltrates, and 11/17 (64.7%) had perivascular anti-vWF immunostaining (Figure 3).

Figure 3.

Intraplaque hemorrhage localization. (A) Hematoxylin and eosin (H&E) stain of a cross-section of a cardiac allograft vasculopathy lesion (2×); insert shows intraplaque hemorrhage. (B) Details of black insert in (A), immunostained for glycophorin A, showing erythrocytes fragments, which indicates old hemorrhage. Bar indicates 100 µm. (C) Perls staining for detection of iron (black deposits), indicating old hemorrhage that co-localized with glycophorin A, in erythrocyte membrane. Bar indicates 100 µm. (D) Adjacent tissue section immunostained for von Willebrand factor showing leakage of vessels. Bar indicates 100 µm. (E) Immunostaining for SMA/CD68, showing the presence of inflammation around necrotic core. Bar indicates 100 µm.

Moreover there was a strong association between fibrocellular lesions and IPH (p = 0.0142).

Relation of CAV lesions phenotype and the native pathology leading to transplantation

IPH was detected in more than half of CAV lesions (52.2%) from ICM patients. CAV lesions from ICM and DCM presented both fresh and old hemorrhage with a mean percentage of 40% and 16%, respectively. Hemorrhage co-localized with lipid and necrotic core in these lesions. Inflammation was a common pattern in most of the CAV lesions (78.3% and 50% from ICM and DCM). No differences were identified between superficial and deep inflammation (Table 5).

Table 5. Plaque characteristics in CAV, according to the native pathology leading to transplantation
Plaque classification CAV from ICM (n = 23), n (%)Total, n = 23
No lesionsEarly lesionsFibrocellular lesionsFibrotic lesionsFibrolipid lesionsComplicated lesions
0 (0)3 (13)9 (39.1)2 (8.7)6 (26)3 (13)
IPH, n (%)Positive0 (0)0 (0)5 (55.6)1 (50)4 (66.7)2 (66.7)12 (52.2)
 Negative0 (0)0 (0)4 (44.4)1 (50)2 (33.3)1 (33.3)8 (34.8)
Inflammation, n (%)Positive0 (0)1 (33.3)7 (77.8)2 (100)6 (100)3 (100)19 (82.6)
 Negative0 (0)2 (66.7)2 (22.2)0 (0)0 (0)0 (0)4 (17.4)
Microvessels, n (%)Positive0 (0)0 (0)5 (55.6)2 (100)4 (66.7)3 (100)14 (60.9)
 Negative0 (0)3 (100)4 (44.4)0 (0)2 (33.3)0 (0)9 (39.1)
Vessel leakage, n (%))Positive0 (0)0 (0)5 (55.6)2 (100)1 (16.7)2 (66.7)10 (43.5)
 Negative0 (0)3 (100)4 (44.4)0 (0)5 (83.3)1 (33.3)13 (56.5)
CD68/SMC<10%0 (0)1 (33.3)7 (77.8)0 (0)0 (0)0 (0)8 (34.8)
 10–50%0 (0)2 (66.7)2 (22.2)1 (50)2 (33.3)1 (33.3)8 (34.8)
 >50%0 (0)0 (0)0 (0)1 (50)4 (66.7)2 (66.7)7 (30.4)
Plaque classification CAV from DCM (n = 12), n (%}Total, n = 12
No lesionsEarly lesionsFibrocellular lesionsFibrotic lesionsFibrolipid lesionsComplicated lesions
0 (0)0 (0)8 (66.7)1 (8.3)2 (16.7)1 (8.3)
  1. CAV, cardiac allograft vasculopathy; DCM, dilated cardiomyopathy; ICM, ischemic cardiomyopathy, IPH, intraplaque hemorrhage.
IPH, n (%)Positive0 (0)0 (0)6 (75)0 (0)2 (100)1 (100)9 (75)
 Negative0 (0)0 (0)2 (25)1 (100)0 (0)0 (0)3 (25)
Inflammation, n (%)Positive0 (0)0 (0)8 (100)1 (100)2 (100)1 (100)12 (100)
 Negative0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)
Microvessels, n (%)Positive0 (0)0 (0)8 (100)1 (100)2 (100)1 (100)12 (100)
 Negative0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)
Vessel leakage, n (%)Positive0 (0)0 (0)6 (75)1 (100)2 (100)1 (100)10 (83.3)
 Negative0 (0)0 (0)2 (25)0 (0)0 (0)0 (0)2 (16.7)
CD68/SMC10%0 (0)0 (0)4 (50)1 (100)0 (0)0 (0)5 (41.7)
 10–50%0 (0)0 (0)1 (12.5)0 (0)0 (0)1 (100)2 (16.7)
 >50%0 (0)0 (0)3 (37.5)0 (0)2 (100)0 (0)5 (41.7)


The main result of our study is that IPH, which is regularly observed in native ATS and considered responsible for episodes of sudden plaque growth [9-11], is also an important feature in CAV. Such IPH could be observed in our study both in CAV lesions (60%) and in native ATS (22.9%) in a high percentage. In both conditions, and with the use of specific staining for erythrocyte membrane remnants [9, 15], these IPH could be identified as being either of recent (fresh) or old onset, supporting the concept of repetitive ongoing IPH as an important mechanism of plaque progression [15].

The contribution of IPH to the progression of coronary atheroma was already reported in native ATS by Kolodgie et al [9], but, as far as we know, no previous articles have reported contribution of IPH to CAV lesions remodeling. Several other aspects like angiogenesis, microvessel leakage and inflammation, which are known to be related to IPH in native ATS, were also frequently present in CAV [15, 16]. Similar to the situation described in detail in native coronary ATS [9, 15], ongoing hemorrhages can be involved in the progressive growth of CAV lesions.

Such IPH could result from either incorporation of mural thrombus by means of fibrocellular organization of the thrombus mass [24], or leakage or rupture of intralesional microvessels [9].

Independently from the diseases leading to heart transplantation, all CAV lesions seem to have similar features undergoing similar remodeling process. In this setting genetic predisposition does not seem to play a role in the progression of lesion hemorrhage.

Based on meticulous quantitative studies on native ATS, Kolodgie et al [9] postulated that erythrocyte membranes, which have a very high lipid content, derived from hemorrhages, could lead to an abrupt increase in the level of free cholesterol in plaques, resulting in expansion of the necrotic core and the potential for plaque destabilization.

Our finding that deposits of erythrocyte membrane–derived material were detected in necrotic core close to the cholesterol cleft and at the level of superficial fibrous layer is in keeping with Kolodgie et al's findings [9]. IPH may stimulate CAV by being a source of free cholesterol and macrophage activation. Our lesions are representative of the heterogeneity of CAV lesions. Lipid could result from lipid insudation from blood or from disrupted erythrocytes membranes.

IPH frequently occurs in ATS, and pathological studies have demonstrated that IPH and plaque rupture are associated with increased microvessel density [9]. In our study, IPH was strongly associated with presence of microvessels.

IPH derived from microvascular disruption or leakage is considered to contribute to necrotic core expansion, which precedes plaque instability and rupture [9, 25]. Repeated hemorrhage events, in our study, occurred in at least one-third of patients with previous IPH.

Iron accumulation due to hemoglobin breakdown can act as a catalyst in the formation of free radicals that modify LDL cholesterol and foam cells [26]. Thus, iron accumulation deposits are not only a histological marker of previous hemorrhage but also a possible source of reactive oxygen species, and could activate the redox state inducing inflammatory reaction and activating matrix metalloproteases, leading to coronary plaque instability via reactive oxygen species generation [26].

CAV is a process that negatively affects survival after heart transplantation. The concentric growth is generally believed to be caused by the influx and growth of SMCs originating in the tunica media of the coronary arteries [6].

In this work, we also identified infiltrating inflammatory cells both superficial and deep in the plaque. There was no difference in plaque inflammation phenotype between ATS and CAV. Moreover in CAV lesions, both superficial and deep inflammation were found in 57.1% compared to ATS lesions with only 19.4%.

Limitation of the Study

For the first time we had the possibility to study and compare within the same group of patients CAV from explanted hearts at autopsy with native coronary ATS from the heart removed at transplantation. The inclusion criteria adopted in this study led to a relatively constrained data set, with an overall number of lesions analyzed around 70. Even though the limited amount of data available suggests caution in extrapolating general patterns from this study, our results provide new insight on the mechanisms underlying the formation of CAV lesions. Further work will be needed to expand the current database and better assess the statistical significance of these results.


Findings from this study suggest the importance of repeated hemorrhage into the plaque as an additional mechanism of CAV progression.

Neoangiogenesis combined with endothelial leakage was a constant feature in CAV lesions. Incorporation of luminal thrombi cannot be entirely excluded. However, microvessels were constantly present in association with hemorrhage while plaque fissure was absent.

Repeated IPH and the resulting consequences related to trapped erythrocytes and activated macrophages may be important factors in the rapid progression of CAV. Understanding the mechanism by which plaque angiogenesis, vessel growth and IPH occur in CAV may help to define a better therapeutic and preventive strategy.

Imaging modalities such as contrast-enhanced ultrasound and dynamic MRI have the potential to evaluate intraplaque neovascularization and IPH noninvasively, improving our stratification of “high-risk” patients [10, 27, 28]. IPH may serve as surrogate for events that are stronger than intima-media thickness [29]. Using microbubbles as vehicles for targeted drug therapy could be foreseen as future therapeutic intervention [30].


This work was supported by a research grant from University of Padua (nos. CPDR099073, CPDA108809, 60A07-5074 and 60A07-8587).


The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.