Author + information
- Received December 20, 1999
- Revision received May 15, 2001
- Accepted July 11, 2001
- Published online November 1, 2001.
- Ali H.M Hassan, MDa,
- Irene M Lang, MD*,a (, )
- Mihaela Ignatescu, MDa,
- Robert Ullrich, MDb,
- Diana Bonderman, MDa,
- Paul Wexberg, MDa,
- Franz Weidinger, MDa and
- Helmut D Glogar, MDa
- ↵*Reprint requests and correspondence: Dr. Irene M. Lang, Division of Cardiology, Allgemeines Krankenhaus, University of Vienna, Währinger Gürtel 18-20, A-1090 Vienna, Austria
In a histopathologic study, we assessed the balance of cell proliferation and apoptosis by counting the number of apoptotic and proliferating cell nuclear antigen-positive cells in freshly harvested atherectomy specimens from 34 patients.
Remodeling of human coronary arteries is an adaptive process that alters vascular lumen size.
Intravascular ultrasound was performed prior to atherectomy. Total vessel area (area within the external elastic lamina [EEL]), lumen area and plaque area were measured at the region of interest (ROI), and at a proximal and distal reference segment, utilizing the formula Δ(%)=100×(ROI−reference segment)/reference segment. Positive arterial remodeling (R+) resulting in luminal expansion was defined as ΔEEL >10%. Absence of remodeling (0 < ΔEEL <10%) and constrictive arterial remodeling (ΔEEL <0) were considered as neutral remodeling (R0) and negative remodeling (R−), respectively.
In R− lesions, apoptotic indices (APO) were significantly elevated (17.17 ± 2.19%) compared with R+ lesions (4.89 ± 1.7%; p = 0.0007). In a rabbit iliac percutaneous transluminal coronary angioplasty model intimal apoptosis was increased four weeks after balloon angioplasty injury (APO 8.8 ± 0.03%) compared with contralateral untreated segments (APO 3.0 ± 0.04%, n = 6). Lesions with an EEL/intimal area <3.0 showed significantly more intimal apoptosis than untreated lesions (p = 0.02).
The data indicate that constrictive remodeling of atherosclerotic coronary lesions is associated with increased apoptosis of intimal cells. We speculate that increased apoptosis is due to extensive plaque healing after episodes of symptomatic or asymptomatic plaque rupture.
The development of atheroma (i.e., type IV lesions with thick layers of fibrous connective tissue, calcifications, intramural hematoma and thrombus) results in lumen encroachment and vessel stenosis. Recent studies (2,3)have shown that the biologic processes in the course of atheroma development, generally addressed as arterial remodeling, are bidirectional and transmural. Histopathologic studies (4,5)and intravascular ultrasound (IVUS) data (6,7)have indicated that vascular wall thickening may be associated with a compensatory increase in the size of the external boundary of the artery. Although compensatory enlargement prevails in about two thirds of lesions, inadequate compensatory enlargement or arterial wall shrinkage at the lesion site accounts for a substantial amount of lumen area reduction (8). According to several models of arterial injury, it is believed that myofibroblasts migrate from the adventitial layer toward the intimal layer and continue to proliferate in the neointima (9,10). The present study was designed to examine the in vivo balance between apoptosis and cell proliferation in primary obstructive human coronary atherosclerotic lesions and to relate the data to a rabbit model of arterial injury.
Apoptosisis a form of cell death characterized by cell shrinkage, pyknosis of the nucleus, cytoplasmic budding and formation of dense, rounded masses, called “apoptotic bodies.” Biochemically, DNA is broken into segments that are multiples of ∼180 bp, due to specific cleavage between nucleosomes. Apoptosis can occur within minutes to hours and has been found responsible for the maintenance of stable cell numbers in tissues with various degrees of proliferative activity (11). Apoptosis has been recognized as a biologic mechanism contributing to remodeling of fetal arteries in animal models (12)as well as to intimal thickening after balloon injury (13)in native atherosclerosis (14,15), and in saphenous vein grafts (16). In our study, numbers of apoptotic cells and proliferating cell nuclear antigen (PCNA)-positive cells in the intimal and medial layers of freshly excised human coronary lesions were analyzed in relation to regional vascular remodeling. One of the main new implications of the present study is that primary lesions within vessel segments lacking compensatory enlargement may represent lesions with extensive healing from spontaneous plaque rupture, thus adopting biological features of restenotic lesions.
Materials and methods
We studied atherectomy specimens from 34 patients undergoing directional coronary atherectomy of de novo stenoses at the University of Vienna General Hospital. Patient inclusion was based on the suitability of the lesion for directional coronary atherectomy (DCA)—that is, a lesion of <20 mm length in a vessel with a reference segment ≥3 mm, no tortuosity and no ostial location. Quantitative coronary angiography (QCA) and IVUS were performed both prior to and immediately after the intervention for optimal debulking. Samples were obtained after patients’ informed, signed consent under the auspices of a protocol approved by the Institutional Review Board of the University of Vienna, Austria (EKZ 96/294).
The CVIS/Boston Scientific IVUS (Natick, Massachusetts) was used. Prior to the imaging procedure, a standardized intracoronary dose of 100 to 200 μg of nitroglycerin was administered. The IVUS catheter was advanced to a distal position in the target artery. Imaging was performed using an automated withdrawal catheter device at a pullback velocity of 0.5 mm/s. The IVUS runs were recorded on 1/2-in. (1.27 cm) high-resolution SVHS videotapes for off-line analysis.
Reference segments were defined as segments with a cross-sectional area within 10 mm proximal or distal to the region of interest (ROI) or within 20 mm if there was angiographic evidence for pre- or post-stenotic dilation. The external elastic lamina (EEL) is recognized as the outer border of the sonoluscent medial layer of the vessel wall. The EEL area, lumen area and plaque area were measured in mm2by manual tracing.
Differences in EEL (ΔEEL) were measured at the lesion site (ROI = region of interest), and at a proximal (px) and distal reference segment utilizing the formula:Furthermore, the same formula was applied to the means (m) of EEL measurements at the proximal and distal reference segments (8).
Compensatory arterial enlargement (R+) was defined as a ΔEEL at the stenosis site to be at least 10% greater than the proximal or mean reference site. Inadequate arterial remodeling (constrictive vascular remodeling or vessel shrinkage/contraction, R−) was considered when ΔEEL was smaller by 10% than the proximal or mean reference site. Neutral remodeling (R0) was considered when ΔEEL differed by <10% from the reference site.
The DCA specimens were flushed out of the Simpson Atherocath (Guidant, Zaventem, Belgium) instrument chamber with saline solution and divided into two equivalent parts. The first half was immediately immersed in liquid nitrogen, and the second half was fixed in 7.5% buffered formalin for 24 h. Serial 3-μm sections were stained with hematoxylin and eosin to assess general histology. Thrombus was identified by a modified trichrome stain (17,18). Both frozen sections and paraffin sections were analyzed. All sections of all specimens were analyzed.
In situ detection of apoptotic cells
Sections were washed in phosphate-buffered saline (PBS) containing 2% H2O2to inactivate endogenous peroxidase, and they were then incubated with 20 μg/ml of proteinase K in PBS. The DNA fragments in the tissue sections were determined using an in situ apoptosis detection kit (ApopTag, Oncor, Gaithersburg, Maryland) as described (19). Sections were counterstained with hematoxylin.
Histologic analysis was performed without knowledge of the clinical data. For each specimen the total number of intimal and medial cells was manually counted on several sections. The projection of the microscopic image (Olympus microscope, Sony DigitalCam, Sony Video projector, Tokyo, Japan) and the simultaneous assessment of cell numbers by four blinded observers were chosen for reliable assessment of cell numbers. Portions of the arterial media were encountered in 16 patients. Because vascular remodeling is a transmural process, PCNA and TdT-mediated dUTP biotin nick end labeling (TUNEL) analyses were extended to include these areas.
Counting of TUNEL-positive nuclei
For apoptosis quantitation, areas of necrosis and atheromatous gruel were excluded from the analysis. In these areas the reliability of positive TUNEL staining is low because of artifacts. All specimens per patient (on an average one to three pieces of tissue) were analyzed, yielding 20 to 30 sections each.
Cells were considered apoptotic when cell nuclei demonstrated positive TUNEL staining and apoptotic morphology. For quantification of TUNEL-positive cells, four fields per section were examined at 200-fold magnification. The apoptotic index was calculated using the formula: Sections of human tonsils were used as positive control.
Immunohistochemistry was performed as previously described (20). Mouse monoclonal anti-human alpha-actin antibody and mouse monoclonal anti-PCNA antibody (Oncogene; all at 10 μg/ml) were utilized. Aminoethyl-carbazole served as chromogenic substrate. The relative number of PCNA-positive cells was calculated using the formula: Parallel sections were stained with smooth muscle alpha-actin to help identify medial layer smooth muscle cells.
Computer-assisted quantitative histologic evaluation of the atherectomy specimens and rabbit iliac arteries
All specimens were digitized in their full size utilizing a slide scanner (Nikon 6.0 35-mm film scanner, LS-20, Nikon, Tokyo, Japan). Images were processed, and the color contrast was enhanced using the Adobe Photoshop 3.0 software package (Adobe Systems, San Jose, California). Planimetric analyses of whole vessel area, intimal and medial areas, as well as thrombus and fibrin in percent (%) of total specimen area were performed on trichrome-stained specimens using computer-based planimetry (NIH Image 1.61/ppc, Bethesda, Maryland).
Seven male New Zealand white rabbits, weighing between 2.5 and 3.3 kg, were studied. All rabbits underwent balloon angioplasty of the left iliac artery. Four weeks after the procedure, all animals were sacrificed and bilateral iliac arteries were collected.
Balloon injury of the iliac artery was performed according to published procedures (21). In brief, all animals were anesthetized with intravenous thiopental (10 mg/kg) followed by intravenous sodium pentobarbital (20 mg/kg). A 3-mm balloon catheter (Advanced Cardiovascular Systems, Santa Clara, California) was introduced through the right carotid artery and advanced over a 0.014-in. (0.04 cm) guide wire under fluoroscopic guidance into the left ilical artery. The balloon was inflated in the distal portion of the external iliac artery to 8 atm for three 30-s intervals. Additionally, three inflations were performed with the balloon slightly pulled back. The procedure causes complete endothelial denudation in the treated vessel segment over a length of approximately 2 cm. Both the method and the experimental protocol were approved by the Animal Subjects Committee of the University of Vienna.
Analysis of variance was employed for all analyses. A p value <0.05 was considered significant.
Characterization of patients and lesions
A total of 34 lesions were examined in 34 male patients. Mean age was 62 ± 15 years in R+, 62 ± 11 years in R0 and 61 ± 15 years in R− patients (p > 0.05). All but 10 patients (3 in R+, 4 in R0 and 3 in R−) had known coronary risk factors. Twenty-four patients had undergone an acute coronary syndrome/acute myocardial infarction within 20 ± 15 days prior to the intervention, with 11 ± 7 days (n = 6, R+), 18 ± 13 days (n = 12, R0) and 28 ± 19 days (n = 6, R−) past the event (p < 0.05).
Conventional coronary angiography and QCA were performed prior to and after DCA. Reference luminal diameter, minimal luminal diameter and percent diameter stenosis are listed in Table 1. The mean lesion length was 14.1 ± 4.2 mm. Lumen area (LA) and plaque area (PA) and, consequently, local shear stress were not different between groups (22).
Assessment of stenoses by IVUS
Values for EEL, LA and PA were not different in the proximal and distal reference segments of R+, R− and R0 lesions (Table 2). Because lumen area reduction at the lesion site could theoretically result from a lesional wall area increase or a lesional vessel cross-sectional area reduction compared with the reference site, vascular remodeling was assessed by relative area calculations. Relative to the proximal (ΔEEL/px) and mean references segments (ΔEEL/m) significant differences of ΔEEL were detected (Table 3).
Relative LAs and relative PAs were not different among the three groups regardless of the reference used. Criteria of compensatory enlargement were fulfilled in 10 lesions (29.5%). Twenty-four lesions (70.5%) did not display compensatory enlargement (Table 2). Eighteen lesions (53%) demonstrated ΔEEL between 0% and 10%, thus qualifying as R0 lesions.
Because modified trichrome stains identify thrombus at different stages of organization (18), this technique was utilized for the assessment of thrombus area. The R− lesions displayed more lesional thrombus than the R+ lesions (21.13 ± 3.7% vs. 6.17 ± 2.09%; and 12.68 ± 1.68% of total specimen area in R0; p = 0.0003). Eleven samples (7 R0, and 4 R−) demonstrated more than 50% purple-stained thrombus area, representing granulation tissue of organizing thrombus (18).
Comparative quantitative analysis of apoptosis and proliferation
To understand the biological basis of the two different extremes of vascular remodeling, atherectomy specimens were analyzed utilizing TUNEL and PCNA immunohistochemistry. Deep vessel wall components (media and adventitia) were observed in 12 specimens (35.2%), which is in accord with previous data (23). In 16 instances portions of the medial layer were identified by trichrome and smooth muscle alpha-actin stains. No differences in total cell numbers/field among the three groups of lesions were observed (p = 0.13).
Similarly, when the specimen areas devoid of cell nuclei were compared among the groups, no significant differences were evident (p = 0.78). However, higher apoptotic indices of intimal cells were observed in R− lesions (4.89 ± 1.7 in R+ vs. 17.17 ± 2.19 in R− and 10.39 ± 1.56 in R0 lesions; p = 0.0007, Fig. 1A). The majority of cells undergoing apoptosis were intimal smooth muscle cells and myofibroblasts as assessed by alpha-actin and vimentin stains (data not shown). Mean apoptotic indices in the medial layers were 0.4 ± 0.49 in R− lesions, 1.26 ± 2.4 in R+ lesions and 0 in R0 lesions (p > 0.05).
No differences in numbers of PCNA-positive cell nuclei were observed (3.16 ± 1.26 in R+ vs. 2.27 ± 1.55 in R− and 3.59 ± 0.98 in R0 lesions; p = 0.7741, Fig. 1B). No PCNA-positive nuclei were detected in the medial layers. Figure 2shows two representative specimens.
Quantitative analysis of apoptosis and cell proliferation in a rabbit model of arterial injury
Constrictive remodeling of an artery has been shown to be the major process involved in restenosis after angioplasty (3,24,25). Therefore, vessel-wall apoptosis was investigated in a balloon angioplasty model (n = 6). Four weeks after balloon angioplasty apoptotic indices (APO) were significantly elevated in all angioplasted (treated with a ballon) artery segments (8.8 ± 3.7%) compared with the contralateral intact segments of the same animal (3.0 ± 4.0%; p = 0.0001). In contrast, APO in the medial layer (0.003 ± 0.005% vs. 0%; p = 0.7) and the adventitia (0.14 ± 0.07% vs. 0.3 ± 1.3%; p = 0.7) did not differ. Planimetry of untreated rabbit iliac arteries demonstrated a 3.3 ± 0.3 ratio of the EEL/intimal area. When lesions with an EEL/intimal area <3.0 were regarded as R−, and lesions with an EEL/intimal area >3.5 as R+, then R− showed significantly more intimal APO than did untreated lesions (p = 0.0157).
Analysis of PCNA-positive nuclei revealed 9.4 ± 4.8/100 nuclei in PTAd lesions and 0.8 ± 0.6 PCNA-positive nuclei in untreated coronary segments (p = 0.0018). The media were devoid of PCNA-positive nuclei. In no instance was thrombus present at the lesion site.
The present study is unique in combining the in vivo assessment of vascular remodeling by IVUS of primary atherosclerotic lesions with an analysis of freshly harvested intimal-medial cuts by two simple techniques assessing histologic markers of the control of cell numbers.
A limitation of the present study is that the assessment of remodeling by IVUS involving all layers of the arterial wall was statistically related to histopathologic data derived from a small fragmented intimal-medial sample. However, apoptotic cells were significantly more frequent in specimens from R− lesions. On an average, four apoptotic cells were found per 100 intimal cells, and 0.8 per 100 medial smooth muscle cells of a typical R+ lesion (Fig. 1).
Increased intimal apoptosis and vessel shrinkage in primary atherosclerotic coronary lesions may identify lesions that have healed from spontaneous plaque rupture
For several reasons the data would lead to speculate that R− lesions represent “healed” primary lesions, where the same mechanism that causes restenosis after vascular injury (2,26)is set off, and is associated with increased apoptosis. First, previous studies of atherectomy specimens from primary lesions typically yielded <2% of apoptotic nuclei (27), whereas restenotic lesions displayed a greater degree of apoptosis. Second, intimal APO in the untreated animals of the arterial injury model were 0.3% to 3%, and as high as 8.8% four weeks after angioplasty. The time span of 28 days between injury and specimen recovery in the animal model compares well with 20 days between acute coronary syndrome and DCA in the patients. Therefore, R− and R0 lesions could result from different degrees of plaque healing.
Third, alternating layers of fibrous collagen-rich tissue (Fig. 3)were identified in the specimens, suggesting repeated episodes of plaque rupture and healing. Apoptosis has been found to participate in the transition between granulation tissue and the development of a definitive scar. Recently, enhanced expression of caspase-3 in hypertrophied scars and keloid was found to induce apoptosis of fibroblasts, which may play a role in the process of pathological scarring (28). The IVUS assessment of negative remodeling may identify these lesions as lesions that have gone through at least one cycle of plaque rupture and healing, thus acquiring a more stable architecture. This would explain our clinical observation that constrictive remodeling is associated with stable lesion characteristics (29)and confers protection against coronary events (30).
Because of the different duration of apoptosis and PCNA expression early in the cell cycle (<2 h vs. 14 to 18 h), a net balance over time between apoptosis and proliferation was not assessed in the present study. In light of equal cell numbers per field, the apoptotic cells may be those cells that have migrated into the intimal layer in the course of vascular remodeling (10).
Thrombus usually occurs where the plaque cap has ruptured (31). Increased thrombus area adjacent to the intimal layer was found in R− lesions compared with R+ lesions. Thrombus quantitation was based on histology and not on IVUS, which bears a low sensitivity and specificity for thrombus detection (32). Both apoptotic endothelial cells and apoptotic vascular smooth muscle cells can act as substrates for thrombin generation (33,34), with plaque smooth muscle cells having the same potency to generate thrombin as platelets (34). Apoptotic cells expose phosphatidylserine on the surface early in the process (35). In the presence of tissue factor, thrombin generation ensues (36). Although thrombus could theoretically lead to increased apoptosis, apoptosis in the animal model occurred in the absence of thrombus.
Furthermore, the presence of thrombi after a coronary event may be highly variable, reflecting multiple factors, including the pharmacologic treatments rendered and the patient’s individual plasmatic procoagulant tendency. The biological significance of more or less thrombus in front of a clinical background of recent unstable angina remains uncertain. However, organized thrombus was most common in lesions lacking compensatory enlargement, supporting the concept of healing as an underlying mechanism of nonexpansive remodeling.
- apoptotic indices
- directional coronary atherectomy
- external elastic lamina
- intravascular ultrasound
- lumen area
- quantitative coronary angiography
- region of interest
- plaque area
- phosphate-buffered saline
- proliferating cell nuclear antigen
- proximal reference segment
- TdT-mediated dUTP biotin nick end labeling
- Received December 20, 1999.
- Revision received May 15, 2001.
- Accepted July 11, 2001.
- American College of Cardiology
- Stary H.C.,
- Chandler A.B.,
- Dinsmore R.E.,
- et al.
- Kakuta T.,
- Currier J.W.,
- Haudenschild C.C.,
- Ryan T.J.,
- Faxon D.P.
- Post M.J.,
- Borst C.,
- Kuntz R.E.
- Mintz G.S.,
- Kent K.M.,
- Pichard A.D.,
- et al.
- Nishioka T.,
- Luo H.,
- Eigler N.L.,
- et al.
- ↵Li G, Chen S-J, Chen Y-F, et al. Adventitial fibroblasts contribute to neointima formation in balloon-injured rat carotid arteries (abstr). Circulation 1998;98 Suppl:I182.
- Cho A.,
- Courtman D.W.,
- Langille B.L.
- Kockx M.M.,
- De Meyer G.R.,
- Bortier H.,
- et al.
- Lang I.M.,
- Moser K.M.,
- Schleef R.R.
- Bonderman D.,
- Gharehbaghi-Schnell E.,
- Wollenek G.,
- et al.
- Lang I.M.,
- Marsh J.J.,
- Olman M.A.,
- et al.
- Weidinger F.F.,
- McLenachan J.N.,
- Cybulsky M.I.,
- et al.
- Andersen H.R.,
- Maeng M.,
- Thorwest M.,
- Falk E.
- Wilensky R.L.,
- March K.L.,
- Gradus-Pizlo I.,
- et al.
- Isner J.M.,
- Kearney M.,
- Bortman S.,
- Passeri J.
- Schoenhagen P.,
- Ziada K.M.,
- Kapadia S.R.,
- et al.
- Wexberg P.,
- Gyongyosi M.,
- Sperker W.,
- et al.
- Feld S.,
- Ganim M.,
- Carell E.S.,
- et al.
- Bombeli T.,
- Karsan A.,
- Tait J.F.,
- Harlan J.M.
- Flynn P.D.,
- Byrne C.D.,
- Baglin T.P.,
- Weissberg P.L.,
- Bennett M.R.
- Mallat Z.,
- Hugel B.,
- Ohan J.,
- et al.