Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Mar 1.
Published in final edited form as: Atherosclerosis. 2022 Jan 29;344:31–39. doi: 10.1016/j.atherosclerosis.2022.01.012

Histopathological correlation of near infrared autofluorescence in human cadaver coronary arteries

Mie Kunio a,b,*, Joseph A Gardecki b, Kohei Watanabe a,b, Kensuke Nishimiya b, Sarika Verma a, Farouc A Jaffer b,c, Guillermo J Tearney b,**
PMCID: PMC9106423  NIHMSID: NIHMS1793265  PMID: 35134654

Abstract

Background and aims:

Prior coronary optical coherence tomography (OCT)-near infrared auto-fluorescence (NIRAF) imaging data has shown a correlation between high-risk morphological features and NIRAF signal intensity. This study aims to understand the histopathological origins of NIRAF in human cadaver coronary arteries.

Methods:

Ex vivo intracoronary OCT-NIRAF imaging was performed on coronary arteries prosected from 23 fresh human cadaver hearts. Arteries with elevated NIRAF were formalin-fixed and paraffin-embedded. Microscopic images of immunostained Glycophorin A (indicating intraplaque hemorrhage) and Sudan Black (indicating ceroid after fixation) stained slides were compared with confocal NIRAF images (ex. 635 nm, em. 655–755 nm) from adjacent unstained slides in each section. Different images from the same section were registered via luminal morphology. Confocal NIRAF-positive 45° sectors were compared to immunohistochemistry and colocalization between NIRAF and intraplaque hemorrhage or ceroid was quantified by Manders’ overlap and Dice similarity coefficients.

Results:

Thirty-one coronary arteries from 14 hearts demonstrated ≥1.5 times higher NIRAF signal than background, and 429 sections were created from them, including 54 sections (12.6%) with high-risk plaques. Within 112 confocal NIRAF-positive 45° sectors, 65 sectors (58.0%) showed both Glycophorin A-positive and Sudan Black-positive, while 7 sectors (6.3%) and 40 sectors (33.6%) only showed Glycophorin A-positive or Sudan black-positive, respectively. A two-tailed McNemar’s test showed that Sudan Black more closely corresponded to confocal NIRAF than Glycophorin A (p < 1.0 × 10−6). NIRAF was also found to spatially associate with both Glycophorin A and Sudan Black, with stronger colocalization between Sudan Black and NIRAF (Manders: 0.19 ± 0.15 vs. 0.13 ± 0.14, p < 0.005; Dice: 0.072 ± 0.096 vs. 0.060 ± 0.090, p < 0.01).

Conclusions:

As ceroid associates with oxidative stress and intraplaque hemorrhage is implicated in rapid lesion progression, these results suggest that NIRAF provides additional, complementary information to morphologic imaging that may aid in identifying high-risk coronary plaques via translatable intracoronary OCT-NIRAF imaging.

Keywords: Atherosclerosis, Plaque progression, Plaque destabilization, Ceroid, Intraplaque hemorrhage, Oxidized lipoprotein, Near-infrared auto-fluorescence

Graphical Abstract

graphic file with name nihms-1793265-f0004.jpg

1. Introduction

Intracoronary imaging technologies may enable the identification non-stenotic atherosclerotic plaques that are at high risk of progression [1–6]. Studies have shown that the presence of a thin-cap fibroatheroma (TCFA) increases the risk of major adverse cardiovascular events by 3.9-fold [7]. TCFAs may rupture, which may lead to an acute coronary syndrome due to sudden luminal thrombosis [8]. Optical coherence tomography (OCT), one of the intravascular imaging modalities that provides intra-luminal morphologic information within coronary arteries, is an in vivo standard for identifying TCFA, made possible by its high in-plane resolution (12–15 μm) [9–14]. OCT can also accurately detect plaque rupture and erosion, thrombus, and calcified nodules, features also implicated in acute coronary syndromes. Owing to its capability to characterize the arterial wall and plaque and clearly visualize lumen borders and implanted stent-struts [9–15], intravascular imaging with OCT during PCI may improve procedural success and clinical outcomes [1–4].

Atherosclerosis is a chronic inflammatory condition, and multiple studies have described a number of complex multi-factorial mechanisms that play a key role in coronary atherosclerotic plaque formation (atherogenesis) and its progression, such as oxidation, intraplaque hemorrhage and accumulation of lipids in the artery wall [16–22]. While OCT is capable of identifying certain high-risk morphological features of plaques, it cannot readily assess other chemical or molecular characteristics. The possibility of acquiring additional complementary information that could increase the predictive capacity of OCT motivated combining near-infrared auto-fluorescence (NIRAF) detection to the OCT catheter device [23,24]. Previous preclinical and clinical research with this multimodality OCT-NIRAF intracoronary imaging platform showed that high-risk morphological features (e.g., TCFA) were more likely to harbor high levels of NIRAF [23,24]. NIRAF hotspots in these lesions were focal, suggesting the detection of pathobiological information that could not be identified by structural imaging alone. These findings motivated the investigation of the relationships between coronary NIRAF and mechanisms of plaque progression and destabilization. In a recent study by Htun et al. conducted in a mouse model and human endarterectomy specimens, the authors provided evidence that heme degradation products arising from intraplaque hemorrhage could be one source of NIRAF [25]. Other studies have shown that lipid oxidation byproducts (e.g., ceroid) also provide a strong NIRAF signal in human plaque [26–29]. Most recently, a study in excised human carotid arteries and macrophages in vitro found that NIRAF overlaps with ceroid [29].

In this study specifically investigating coronary atherosclerosis, we examine the spatial correlation between human coronary NIRAF and histopathological markers of intraplaque hemorrhage and ceroid. Results provide insight into the origin of coronary NIRAF and its potential clinical relevance.

2. Materials and methods

2.1. OCT-NIRAF imaging system and catheter

A prototype multi-modality OCT-NIRAF imaging system and catheters, further described in Supplemental Methods, were provided by Canon U.S.A., Inc (Cambridge, MA, USA).

2.2. Human cadaver hearts

Explanted hearts were obtained from deceased donors identified by the National Disease Research Interchange (NDRI, Philadelphia, PA). Inclusion criteria for donor selection included a medical history of hypertension, hyperlipidemia, age (>55 y/o) and known cardiovascular disease. Hearts were recovered with 24 h postmortem, packed in phosphate buffered saline solution with antibiotics, and transported on wet ice for delivery in less than 48 h.

2.3. Selection of coronary arteries and tissue processing

The three major epicardial coronary arteries were dissected from the heart. Totally occluded arteries were excluded from the study as they would not permit intravascular imaging. The resected arteries (approx. 5 cm length) were helically scanned and imaged with an OCT-NIRAF system and catheter at room temperature to ascertain the presence of atherosclerosis and NIRAF as previously demonstrated [24]. After OCT-NIRAF imaging, the arteries were perfusion fixed at 100 mm Hg with 10% neutral buffered formalin (Thermo Scientific, Kalamazoo, MI) for greater than 1 h with continued fixation in formalin for greater than 48 h. Formalin-fixed arteries were cut into 4 mm long segments and placed in decalcification solution (Protocol B, Thermo Scientific, Kalamazoo, MI) for 2–4 h depending on the degree of calcification. After decalcification, the arteries were washed in water, placed back into 10% formalin, and submitted for standard formalin-fixed paraffin-embedded (FFPE) histology.

Six serial 5-μm sections were taken from each segment and stained with hematoxylin and eosin (H&E), Masson’s trichrome, Sudan Black (Sigma Aldrich, St. Louis, MO, USA), Glycophorin A (BioCareMedical, Concord, CA, USA), and one unstained section respectively using standard protocols. Since Glycophorin A is a membrane protein of the erythrocyte that is specific for anion exchange, Glycophorin A staining can be used as an intraplaque hemorrhage marker [20]. In fixed tissue, Sudan Black can identify ceroid (insoluble lipid), a known oxidative product of lipoproteins [30]. All slides were digitized as described above. The unstained slides were submitted for confocal fluorescence imaging for NIRAF as detailed below. All histology slides were digitized at 40× magnification using a whole slide imaging system (NanoZoomer, Hamamatsu Photonics, Hamamatsu, Shizuoka, Japan).

2.4. Classification of lesions and annotation on histology slides

Histology sections were read by a board-certified pathologist using a modified Virmani classification scheme [8]. Atherosclerotic lesions were classified as ruptured plaque, fibroatheroma, fibrocalcific plaque, pathological intimal thickening, fibrous plaque, and intimal hyperplasia. Slides were randomized and marked with a unique identifier prior to analysis.

2.5. Confocal fluorescence microscopy of unstained paraffin sections

Unstained sections collected from FFPE sections immediately adjacent (5–15 μm) to the histology sections were used for confocal fluorescence microscopy. Histology slides were not deparaffinized prior to scanning. The entire cross-section of each artery was imaged with a confocal fluorescence microscope (FLUOVIEW FV1000, Olympus, Tokyo, Japan) with a 10× objective lens (NA = 0.4, UPLSAPO 10X NA:0.40, Olympus, Tokyo, Japan), resulting in a resolution of 3.314 μm/pixel. Confocal fluorescent images acquired with an excitation at 405 nm, emission at 425–475 nm showed relatively non-specific autofluorescence from the section and served to indicate the whole tissue shape and orientation. Confocal fluorescent images obtained with an excitation at 635 nm, emission at 655–755 nm were used to show the microscopic NIRAF distribution in each section. Multiple Z slice confocal images were acquired to account for cases where the optical axis and the tissue on the glass slide were not completely perpendicular. Stitched images were constructed using the Olympus FLUOVIEW software (FV10-ASW, version 04.02.03.06, Olympus, Tokyo, Japan) and the averaged Z-stack image was constructed using ImageJ/Fiji software (National Institutes of Health, Bethesda, MD, USA) [31].

2.6. Registration of confocal image and histology and immunohistochemistry images

The lumen of the confocal fluorescent image taken with an excitation at 635 nm (i.e., confocal NIRAF image) is the same as that of the confocal fluorescent image taken with an excitation at 405 nm, since these images are acquired from the same section as described above. Therefore, all histology and immunohistochemistry images were registered to the confocal NIRAF image by identifying the same feature points on the lumen seen in the confocal fluorescent image at 405 nm excitation, histology, and immunohistochemistry images. Using these lumen feature points, rigid transformation was applied to H&E, Masson’s trichrome, Glycophorin A, and Sudan Black images independently to match each of them to the confocal NIRAF image (Supplemental Figure 2).

2.7. Determination of threshold for immunohistochemistry stainingpositive and confocal NIRAF-positive

To determine the distribution of intraplaque hemorrhage, ceroid, and NIRAF, all the immunohistochemistry and confocal NIRAF images were first binarized. For Glycophorin A-stained images, a color deconvolution matrix was first applied to each image. The color deconvolution matrix was defined based on the deconvolution matrix for Fast Red, Fast Blue and DAB in ImageJ/Fiji plug-in ‘Color Deconvolution’, and the first color channel was extracted as Glycophorin A-positive channel. The extracted Glycophorin A-positive channel was binarized with the threshold of top 30% of the 8-bit gray-scale level (pixel value of 70). For Sudan Black staining, the image was converted into gray-scale, and binarized with a threshold of the top 30% of the 8-bit gray-scale level (pixel value of 70). For the confocal NIRAF image, after averaging the intensity of the Z-stack (see above section), the image was binarized with a threshold of the top 50% of the 8-bit gray-scale level (pixel value of 128). A subset of thresholded images (10%) were reviewed by a pathologist who deemed these threshold levels to be indicative of the lower bound of staining/fluorescence that was indicative of positivity.

2.8. Assessment of percent area of confocal NIRAF-positive and immunohistochemistry-positive in each plaque

After registration of all histology and immunohistochemistry images to the confocal image and annotation of plaque type on H&E and Masson’s trichrome, the percent area of confocal NIRAF-positive and immunohistochemistry-positive in each of annotated plaque was evaluated as: ANIRAF/ATISSUE or AIHC/ATISSUE, where ANIRAF and AIHC were the areas that were confocal NIRAF-positive and immunohistochemistry-positive, respectively, and ATISSUE was the area of the plaque. The plaques were diagnosed from either H&E or Masson’s trichrome images, depending on the stain for which the plaque type annotation was made.

2.9. Analysis of contribution of histopathological staining to NIRAF signal

To understand the contribution of each stain to the NIRAF signal, colocalization was analyzed from two perspectives: co-occurrence and correlation. Co-occurrence refers to spatial overlap of NIRAF with histopathological markers, and correlation represents both co-existence and co-distribution of NIRAF and histopathological markers. Co-occurrence was evaluated in terms of sensitivity and specificity, and the Manders’ overlap coefficient, while the Dice similarity coefficient was selected as a metric for correlation. To minimize the influence of histology processing on signal local distribution, this analysis was performed 45° sector-basis. The 45° sectors were made by splitting each registered microscopic images into 8, centered at the lumen’s centroid.

First, the sector-based positivity was determined. This determination was performed visually by two independent, blinded observers, and the results were reviewed. Each sector was determined as NIRAF, Glycophorin A, or Sudan Black-positive when more than 10% of the intima in the sector contained NIRAF, Glycophorin A, or Sudan Black staining that was above the 30% (Glycophorin A and Sudan Black) or 50% (NIRAF) intensity threshold.

Sensitivity, specificity, accuracy, positive predictive value, and negative predictive value were evaluated using all 45° sectors (Supplemental Methods). Manders’ overlap coefficient [32,33] was evaluated for 45° -sectors that were confocal NIRAF-positive to determine the fraction of NIRAF that co-existed with immunohistochemistry staining (Supplemental Methods). Dice similarity coefficient [34,35] was evaluated for 45° -sectors where at least one of confocal NIRAF, Glycophorin A, or Sudan Black was positive to determine the correlation between NIRAF and immunohistochemistry staining (Supplemental Methods).

2.10. Image processing and analysis

Averaging Z-stack images and color deconvolution were performed using ImageJ/Fiji software (National Institutes of Health, Bethesda, MD, USA) [31]. All the other image processing and analyses were performed using MATLAB R2018a (MathWorks, Natick, MA, USA) with Computer Vision System Toolbox (version 8.1), Image Processing Toolbox (version 10.2) and Signal Processing Toolbox (version 8.0).

2.11. Statistical analysis

One-way ANOVA and Student’s t-tests were used to compare groups of values. Two-tail McNemar’s test was performed to compare two paired results. Continuous data are expressed as mean ± standard deviation, and p-values <0.05 are considered statistically significant.

3. Results

3.1. Characteristics of human cadaver hearts

A total of 23 fresh human cadaver hearts and 66 coronary arteries, were imaged ex vivo by intracoronary OCT-NIRAF in this study. Out of the 23 hearts, 14 hearts and 31 arteries, presented a NIRAF-positive signature and were selected for analysis. Supplemental Table 1 summarizes patient characteristics of the 14 NIRAF-positive hearts.

3.2. Distribution of lesions in human coronary artery specimens and their correlation with intraplaque hemorrhage and lipoprotein oxidation markers

From the 31 coronary arteries with a positive NIRAF signal, a total of 40 plaques and 429 sections of histology, immunohistochemistry, and unstained paraffin-covered slides were created and analyzed. Ruptured plaques were found in 16 sections (3.73%), fibroatheroma were found in 38 sections (8.86%), fibrocalcific plaques were in 166 sections (38.7%), pathological intimal thickening in 147 sections (34.7%), fibrous plaques in 141 sections (32.9%), and intimal hyperplasia in 84 sections (19.6%).

Ruptured plaques, fibroatheroma, and fibrocalcific plaques showed moderate-high NIRAF, Glycophorin A, or Sudan Black staining; while pathological intimal thickening, fibrous plaques, and intimal hyperplasia showed little or no NIRAF, Glycophorin A, or Sudan Black staining (Figs. 1 and 2). Fig. 1 shows representative NIRAF and histology from a ruptured coronary plaque. Magnified portions of different regions of the plaque and the rupture site are shown in panels 1 E-F. Fig. 1E shows plaque constituents within the ruptured core. Residual necrotic core at the base of the rupture cavity is shown that is high in NIRAF, Glycophorin A, and has focal Sudan Black staining. Surrounding necrotic regions show less Glycophorin A with focal NIRAF and Sudan Black, which appears to be more co-localized. Fig. 1F–H represents a part of fibroatheromas. In the region of Fig. 1F, Glycophorin A and NIRAF showed a similar distribution with focal Sudan Black staining. In Fig. 1G and H, NIRAF and Sudan Black staining were observed in a similar area while Glycophorin A appeared to be less co-localized with NIRAF. Fig. 2 shows representative NIRAF and histology from a fibroatheroma and fibrocalcific plaque, surrounded by fibrous tissue. In Fig. 2E, NIRAF, Glycophorin A, and Sudan Black were observed in the fibrous region that was next to fibroatheroma, and these three seemed to be co-localized. In Fig. 2F and G, Glycophorin A was scarce, while NIRAF and Sudan Black were observed and co-localized. In Fig. 2H, NIRAF, Sudan Black, and less Glycophorin A were observed, at similar locations. Typically, wherever NIRAF was observed, Glycophorin A or Sudan Black was observed. Glycophorin A and Sudan Black usually did not overlap but were located adjacent to each other (Fig. 1E–H). In a few cases, NIRAF, Glycophorin A, and Sudan Black were observed in the same locations (Fig. 2E and H).

Fig. 1.

Fig. 1.

Open in a new tab

Comparison of histology, immunohistochemistry, and confocal NIRAF images.

(A) H&E, (B) Glycophorin A, (C) confocal NIRAF, (D) Sudan Black, and (E–H) magnified images of Glycophorin A, confocal NIRAF, and Sudan Black in the area specified in B-D. Confocal NIRAF was observed in residual necrotic core at the base of the rupture cavity (E) and fibroatheroma (F–H), and colocalized with Glycophorin A or Sudan Black. The positive areas of Glycophorin A and Sudan Black were often located adjacent to each other, and infrequently overlapped.

Fig. 2.

Fig. 2.

Open in a new tab

Comparison of histology, immunohistochemistry, and confocal NIRAF images.

(A) H&E, (B) Glycophorin A, (C) confocal NIRAF, (D) Sudan Black, and (E–H) magnified images of Glycophorin A, confocal NIRAF, and Sudan Black in the area specified in C. Confocal NIRAF-positive areas in fibroatheromas (E–H) colocalized with Glycophorin A and/or Sudan Black. In E and H, the positive areas of Glycophorin A and Sudan Black were observed in the same locations.

On average, high-risk lesions (ruptured plaques and fibroatheroma) contained more NIRAF, Glycophorin A, and Sudan Black-positive regions than lower risk plaques (fibrocalcific plaques, pathological intimal thickening, fibrous plaque, and intimal hyperplasia) (p < 0.05, Fig. 3). Within the lower-risk plaques, fibrocalcific plaques contained more NIRAF signal, Glycophorin A, and Sudan Black-positive areas than the other low risk plaques (p < 0.05, Fig. 3).

Fig. 3.

Fig. 3.

Open in a new tab

Relationship between plaque type and percent area of confocal NIRAF/Glycophorin A/Sudan Black positivity.

High-risk plaques contained not only more NIRAF-positive areas, but also Glycophorin A-positive and Sudan Black-positive areas, than low-risk plaques, with statistical significance (*p < 0.05, **p < 0.01, Supplemental Table 2). Error bars show standard error of mean.

3.3. Contribution of intraplaque hemorrhage and lipoprotein oxidation to NIRAF signal

From 429 distinct coronary locations from 31 arteries from 14 hearts, a total of 3432 (429 sections × 8 sectors/section) 45° -sectors/stain were analyzed. One hundred and one (101) sectors were excluded due to damage to the sample that occurred during histology processing and/or poor staining quality.

From visual 45° sector-based assessment, 112 sectors (3.4%) were positive for confocal NIRAF. For the 112 confocal NIRAF-positive sectors, 72 sectors (64.3%) were Glycophorin A-positive, while 105 sectors (93.8%) were Sudan Black-positive (Supplemental Table 3A). For the 3219 confocal NIRAF-negative sectors, 136 sectors (4.2%) were Glycophorin A-positive, while 7 sectors (0.22%) were Sudan black-positive (Supplemental Table 3B). Thus, using NIRAF as the gold standard, the sensitivity and specificity of Glycophorin A were 64.3% (95% CI: 54.7%–73.1%) and 95.8% (95% CI: 95.0%–96.4%), respectively, while those of Sudan black were 93.8% (95% CI: 87.6%–97.5%) and 99.8% (95% CI: 99.6%–99.9%), respectively (Table 1). In addition, positive predictive value (PPV) of Sudan black was higher than that of Glycophorin A (93.8% vs. 34.6%), while negative predictive value (PPV) was similar between Glycophorin A and Sudan Black (98.7% and 99.8%, respectively; Table 1).

Table 1.

Sensitivity, specificity, and accuracy analysis result using confocal NIRAF as the ground truth.

Glycophorin A Sudan Black
Sensitivity 64.3% (95% CI: 54.7%–73.1%) 93.8% (95% CI: 87.6%–97.5%)
Specificity 95.8% (95% CI: 95.0%–96.4%) 99.8% (95% CI: 99.6%–99.9%)
Accuracy 94.7% (95% CI: 93.9%–95.5%) 98.6% (95% CI: 99.3%–99.8%)
Positive predictive value 34.6% (95% CI: 29.9%–39.6%) 93.8% (95% CI: 87.7%–96.9%)
Negative predictive value 98.7% (95% CI: 98.3%–99.0%) 99.8% (95% CI: 99.6%–99.9%)
Open in a new tab

Values in parentheses are 95% confidence intervals

Within the 112 confocal NIRAF-positive sectors, 65 sectors (58.0%) observed both Glycophorin A-positive and Sudan black-positive staining, while 7 sectors (6.3%) only showed Glycophorin A-positive staining and 40 sectors (33.6%) only showed Sudan black-positive staining (Supplemental Table 4A). A two-tailed McNemar’s test for confocal NIRAF-positive sectors and confocal NIRAF-negative sectors, respectively (Supplemental Table 4C), showed that Sudan Black more closely corresponded to confocal NIRAF than Glycophorin A (p < 1.0 × 10−6).

Within 3331 sectors, thrombus was observed in 45 sectors. The sensitivity of Sudan black staining in sectors without thrombus was higher (96.6%) than that in sectors with thrombus (84.0%), while that of Glycophorin A was lower in sectors without thrombus (56.3% vs. 92.0%) (Table 2 and Supplemental Table 5). The positive predictive value of Sudan black staining was higher in both sectors with and without thrombus (Table 2 and Supplemental Table 5).

Table 2.

Sensitivity, specificity, and accuracy analysis result using confocal NIRAF as the ground truth in sectors with thrombus and without thrombus.

Sectors with thrombus
 Sectors without thrombus
Glycophorin A Sudan Black Glycophorin A Sudan Black
Sensitivity 92.0% (74.0%–99.0%) 84.0% (63.9%–95.5%) 56.3% (45.3%–66.9%) 96.6% (90.3%–99.3%)
Specificity 55.0% (31.5%–76.9%) 100.0% (83.2%–100.0%) 96.0% (95.3%–96.7%) 99.8% (99.6%–99.9%)
Accuracy 75.6% (60.5%–87.1%) 91.1% (78.8%–97.5%) 95.0% (94.2%–95.7%) 99.7% (99.4%–99.9%)
Positive predictive value 71.9% (60.8%–80.8%) 100.0% 27.8% (23.1%–33.2%) 92.3% (85.1%–96.2%)
Negative predictive value 84.6% (57.9%–95.7%) 83.3% (67.1%–92.5%) 98.8% (98.5%–99.0%) 99.9% (99.7%–100.0%)
Open in a new tab

Values in parentheses are 95% confidence intervals

In addition to sensitivity and specificity analyses, the contribution of Glycophorin A and Sudan Black to the NIRAF signal was evaluated quantitatively via Manders’ overlap and Dice similarity coefficients. For 101 of confocal NIRAF-positive 45° -sectors (the sectors with tissue folding, air bubbles, or dust were removed), the Manders’ overlap coefficient between confocal NIRAF and Sudan Black was higher than that between confocal NIRAF and Glycophorin A (0.19 ± 0.15 vs. 0.13 ± 0.13, p = 0.0014). Within these 101 sectors, the maximum Manders’ overlap coefficient for Sudan Black was 0.73, while that for Glycophorin A was 0.56. In addition, in 81 (80.2%) of confocal NIRAF-positive 45° -sectors, the Manders’ overlap coefficient between confocal NIRAF and Sudan Black was higher than that between confocal NIRAF and Glycophorin A. For 221 of confocal NIRAF or Glycophorin A or Sudan Black-positive 45° -sectors (the sectors with tissue folding, air bubbles, or dust were removed), the Dice coefficient between confocal NIRAF and Sudan Black was higher than that between confocal NIRAF and Glycophorin A (0.072 ± 0.096 vs. 0.060 ± 0.090, p = 0.0074). The Dice coefficient between confocal NIRAF and Sudan Black was higher in 168 (76.0%) of 45° -sectors than that between confocal NIRAF and Glycophorin A.

These results showed that NIRAF was spatially associated with both intraplaque hemorrhage (Glycophorin A-positive) and insoluble lipid, or ceroid (Sudan Black-positive), and that NIRAF was significantly more correlated with ceroid than intraplaque hemorrhage.

4. Discussion

In this study, we evaluated the correlation between NIRAF signal and histopathological features of intraplaque hemorrhage (Glycophorin A staining) and ceroid (Sudan Black staining) in human cadaver coronary arteries. These two histopathologic features were selected since they are well-known factors for plaque progression and destabilization [8,36–42] and have previously been shown to be associated with NIRAF [25,29, 43]. Findings showed that NIRAF correlated with both intraplaque hemorrhage and ceroid, and the correlation between NIRAF and ceroid was stronger than that between NIRAF and intraplaque hemorrhage.

Accumulation of ceroid in atherosclerotic plaque has been reported in multiple studies [38–40,44]. Although ceroid is an undegradable catabolism product of macrophages and foam cells, it is a product of direct oxidization, and not related to aging [27,40]. Direct oxidation can exacerbate the inflammatory process, suggesting that ceroid could be found accumulated in destabilized and advanced plaques, which is supported by our findings and those of Michinson et al. [39] In this paper, ceroid showed a yellow-orange fluorescence spectrum [28]; several polyenic compounds that are responsible for ceroid formation are known to exhibit maximum fluorescence in 500–630 nm (yellow-red) wavelength range [27]. Our finding of a spatial association between NIRAF and ceroid could be explained by the NIRAF signal in part representing the tail of the autofluorescence emission from ceroid.

The NIRAF characteristics of another important factor of plaque progression, intraplaque hemorrhage, has been reported by Htun et al. [25]. Using human carotid endarterectomy specimens and an atherosclerotic mouse model, they suggested that heme degradation products, particularly bilirubin and protoporphyrin IX could serve as a source of NIRAF at 740 nm excitation and at 800–880 nm emission [25]. Our findings of spatial association between NIRAF and intraplaque hemorrhage is also supported by this study, although the correlation between NIRAF and ceroid was significantly higher. The major role of ceroid as a contributor to NIRAF here could be explained by our excitation wavelength of 635 nm, whereas the excitation wavelength for Htun et al. was 740 nm [25]. Future studies should be conducted to confirm whether 740 NIRAF is dominated by intraplaque hemorrhage and 650 by ceroid.

Oxidative stress not only associates with the inflammatory process, as described above, but also triggers neovascularization in the plaque [45]. Neovasculature is often leaky, potentially leading to intraplaque hemorrhage [46]. Intraplaque hemorrhage promotes more oxidization and angiogenesis; thus causing plaque progression to accelerate in a positive feedback loop [8,21,37,42,47,48]. The existence of these mechanisms in advanced human coronary atherosclerosis is supported by our finding of increased ceroid and intraplaque hemorrhage in high-risk plaques, such as fibroatheromas and plaque rupture. This study revealed that intraplaque hemorrhage and ceroid accumulation occurred in adjacent, but frequently non-overlapping locations. This finding highlights the interplay between two important pathways that lead to advanced atherosclerosis, plaque irreversibility/destabilization due to accumulation of necrotic cell debris and ceroid and plaque progression caused by intraplaque hemorrhage, both potentially associated with oxidative stress.

There are several limitations of the present study. The study was conducted ex vivo from cadaver arteries that were previously frozen and thawed. Additionally, the NIRAF images were obtained from FFPE slides. It is unclear as to whether these processing steps had a significant effect on NIRAF signal intensity or distribution [29]. Results of colocalization analysis (evaluation of Manders’ overlap coefficient and Dice similarity coefficient) may be affected in part by the fact that the all the histology, immunohistochemistry, and unstained slides were created from serial slides (5–10 μm away) and/or potential shrinkage during the process, and thus the correspondence between the two sections were close but not one-to-one. Finally, even though intraplaque hemorrhage and ceroid were found to be clear associates of NIRAF in this study, other potential contributors to NIRAF may be present.

4.1. Conclusions

The importance of this study is not only in the finding of increased ceroid and intraplaque hemorrhage in high-risk coronary plaques, along with increased NIRAF, but also the translational impact of intracoronary OCT-NIRAF that has already been demonstrated in a previous clinical study [24]. A key question in this field is the clinical significance of intracoronary NIRAF and ultimately how it can be used to better discriminate plaques at risk for a future event. The data shown here demonstrates a strong relationship between coronary NIRAF and both ceroid and intraplaque hemorrhage. Thus, it is reasonable to assert that NIRAF when seen intravascularly denotes advanced plaques that contain byproducts of oxidative stress and/or intraplaque hemorrhage. Since the two plaque mechanisms are known to contribute to plaque progression [20,38,49], future human studies should focus on the natural history of high NIRAF regions in coronary arteries to determine the clinical significance of this signal.

Supplementary Material

Supplemental Materials

Acknowledgements

The authors thank Dr. Jenny Zhao and Wellman Center’s Photopathology Lab, at the Massachusetts General Hospital for creating all histology slides.

Some of the figures were created with BioRender.com.

Financial support

This study was funded by Canon USA, Inc. and NIH grants (R01HL137913 and R01HL150538).

Footnotes

Declaration of competing interest

M.K., K.W., and K.N. does not have any conflict of interests. G.T. and J.A.G. has a patent and a patent application on OCT-NIRAF technology. F.A.J. sponsored research from Canon USA, Inc, Siemens, Teleflex, Shockwave, Mercator, and Boston Scientific; consultant for Boston Scientific, Abbott Vascular, Siemens, Magenta Medical, Asahi Intec, and IMDS. Equity interest, Intravascular Imaging Inc., DurVena. Massachusetts General Hospital has a patent licensing arrangement with Terumo, Canon, and Spectrawave; FAJ has the right to receive royalties. G.T. has a financial/fiduciary interest in SpectraWave, a company developing an OCT-NIRS intracoronary imaging system and catheter. His financial/fiduciary interest was reviewed and is managed by the Massachusetts General Hospital and Partners HealthCare in accordance with their conflict of interest policies. G.T. also has a consulting relationship with SpectraWave. G.T. receives sponsored research funding from Astrazeneca, VivoLight, Verdure Biotech, and Canon U.S.A., Inc.

CRediT authorship contribution statement

Mie Kunio: Conceptualization, Formal analysis, Writing – original draft. Joseph A. Gardecki: Conceptualization, Formal analysis. Kohei Watanabe: Conceptualization, Formal analysis. Kensuke Nishimiya: Conceptualization. Sarika Verma: Conceptualization. Farouc A. Jaffer: Conceptualization. Guillermo J. Tearney: Conceptualization, Formal analysis.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.atherosclerosis.2022.01.012.

References

  • [1].Dowe DA, Fioranelli M, Pavone P, Imaging Croronary Arteries, second ed., Springer, 2013. [Google Scholar]
  • [2].Thompson CA, Textbook of Cardiovascular Intervention, first ed., Springer, 2014. 10.1007/978-1-4471-4528-8. [DOI] [Google Scholar]
  • [3].Ali ZA, Maehara A, Généreux P, Shlofmitz RA, Fabbiocchi F, Nazif TM, Guagliumi G, Meraj PM, Alfonso F, Samady H, Akasaka T, Carlson EB, Leesar MA, Matsumura M, Ozan MO, Mintz GS, Ben-Yehuda O, Stone GW, Optical coherence tomography compared with intravascular ultrasound and with angiography to guide coronary stent implantation (ILUMIEN III: optimize PCI): a randomised controlled trial, Lancet 388 (2016) 2618–2628, 10.1016/S0140-6736(16)31922-5. [DOI] [PubMed] [Google Scholar]
  • [4].Kubo T, Shinke T, Okamura T, Hibi K, Nakazawa G, Morino Y, Shite J, Fusazaki T, Otake H, Kozuma K, Ioji T, Kaneda H, Serikawa T, Kataoka T, Okada H, Akasaka T, Optical frequency domain imaging vs. intravascular ultrasound in percutaneous coronary intervention (OPINION trial): one-year angiographic and clinical results, Eur. Heart J 38 (2017) 3139–3147, 10.1093/eurheartj/ehx351. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [5].Fitzgerald PJ, Oshima A, Hayase M, Metz JA, Bailey SR, Baim DS, Cleman MW, Deutsch E, Diver DJ, Leon MB, Moses JW, Oesterle SN, Overlie PA, Pepine CJ, Safian RD, Shani J, Simonton CA, Smalling RW, Teirstein PS, Zidar JP, Yeung AC, Kuntz RE, Yock PG, For the C Investigators, final results of the can routine ultrasound influence stent expansion (CRUISE) study, Circulation 102 (2000) 523–530, 10.1161/01.cir.102.5.523. [DOI] [PubMed] [Google Scholar]
  • [6].Steinvil A, Zhang Y-J, Lee SY, Pang S, Waksman R, Chen S-L, Garcia-Garcia HM, Intravascular ultrasound-guided drug-eluting stent implantation: an updated meta-analysis of randomized control trials and observational studies, Int. J. Cardiol 216 (2016) 133–139, 10.1016/j.ijcard.2016.04.154. [DOI] [PubMed] [Google Scholar]
  • [7].Stone GW, Maehara A, Lansky AJ, de Bruyne B, Cristea E, Mintz GS, Mehran R, McPherson J, Farhat N, Marso SP, Parise H, Templin B, White R, Zhang Z, Serruys PW, A prospective natural-history study of coronary atherosclerosis, N. Engl. J. Med 364 (2011) 226–235, 10.1056/NEJMoa1002358. [DOI] [PubMed] [Google Scholar]
  • [8].Virmani R, Burke AP, Farb A, Kolodgie FD, Pathology of the vulnerable plaque, J. Am. Coll. Cardiol 47 (2006) C13–C18, 10.1016/j.jacc.2005.10.065. [DOI] [PubMed] [Google Scholar]
  • [9].Bezerra HG, Costa MA, Guagliumi G, Rollins AM, Simon DI, Intracoronary optical coherence tomography: a comprehensive review, JACC Cardiovasc. Interv 2 (2009) 1035–1046, 10.1016/j.jcin.2009.06.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Gonzalo N, Garcia-Garcia HM, Serruys PW, Commissaris KH, Bezerra H, Gobbens P, Costa M, Regar E, Reproducibility of quantitative optical coherence tomography for stent analysis, Eurointervention 5 (2009) 224–232. [DOI] [PubMed] [Google Scholar]
  • [11].Murata A, Wallace-Bradley D, Tellez A, Alviar C, Aboodi M, Sheehy A, Coleman L, Perkins L, Nakazawa G, Mintz G, Kaluza GL, Virmani R, Granada JF, Accuracy of optical coherence tomography in the evaluation of neointimal coverage after stent implantation, JACC Cardiovasc Imaging 3 (2010) 76–84, 10.1016/j.jcmg.2009.09.018. [DOI] [PubMed] [Google Scholar]
  • [12].Suzuki Y, Ikeno F, Koizumi T, Tio F, Yeung AC, Yock PG, Fitzgerald PJ, Fearon WF, In vivo comparison between optical coherence tomography and intravascular ultrasound for detecting small degrees of in-stent neointima after stent implantation, JACC Cardiovasc. Interv 1 (2008) 168–173, 10.1016/j.jcin.2007.12.007. [DOI] [PubMed] [Google Scholar]
  • [13].Templin C, Meyer M, Muller MF, Djonov V, Hlushchuk R, Dimova I, Flueckiger S, Kronen P, Sidler M, Klein K, Nicholls F, Ghadri JR, Weber K, Paunovic D, Corti R, Hoerstrup SP, Luscher TF, Landmesser U, Coronary optical frequency domain imaging (OFDI) for in vivo evaluation of stent healing: comparison with light and electron microscopy, Eur. Heart J 31 (2010) 1792–1801, 10.1093/eurheartj/ehq168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [14].Guagliumi G, Bezerra HG, Sirbu V, Ikejima H, Musumeci G, Biondi-Zoccai G, Lortkipanidze N, Fiocca L, Capodanno D, Wang W, Tahara S, Vassileva A, Matiashvili A, Valsecchi O, Costa MA, Serial assessment of coronary artery response to paclitaxel-eluting stents using optical coherence tomography, Circulation-Cardiovascular Interventions 5 (2012) 30–38, 10.1161/circinterventions.111.965582. [DOI] [PubMed] [Google Scholar]
  • [15].Russo M, Fracassi F, Kurihara O, Kim HO, Thondapu V, Araki M, Shinohara H, Sugiyama T, Yamamoto E, Lee H, Vergallo R, Crea F, Biasucci LM, Yonetsu T, Minami Y, Soeda T, Fuster V, Jang I-K, Healed plaques in patients with stable Angina pectoris, Arterioscler. Thromb. Vasc. Biol (2020), ATVBAHA120314298, 10.1161/ATVBAHA.120.314298. [DOI] [PubMed] [Google Scholar]
  • [16].Kumar V, Abbas AK, Fausto N, Aster J, Robbins & Cotran Pathologic Basis of Disease, eighth ed., Saunders/Elsevier, Philadelphia, PA, 2010. [Google Scholar]
  • [17].Lilly LS, Pathophysiology of Heart Disease: a Collaborative Project of Medical Students and Faculty, fifth ed., Lippincott Williams & Wilkins, Philadelphia, 2011. [Google Scholar]
  • [18].Takumi T, Yang EH, Mathew V, Rihal CS, Gulati R, Lerman LO, Lerman A, Coronary endothelial dysfunction is associated with a reduction in coronary artery compliance and an increase in wall shear stress, Heart 96 (2010) 773–778, 10.1136/hrt.2009.187898. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Stone PH, Saito S, Takahashi S, Makita Y, Nakamura SS, Kawasaki T, Takahashi A, Katsuki T, Nakamura SS, Namiki A, Hirohata A, Matsumura T, Yamazaki S, Yokoi H, Tanaka S, Otsuji S, Yoshimachi F, Honye J, Harwood D, Reitman M, Coskun AU, Papafaklis MI, Feldman CL, Investigators P, Prediction of progression of coronary artery disease and clinical outcomes using vascular profiling of endothelial shear stress and arterial plaque characteristics: the PREDICTION Study, Circulation 126 (2012) 172–181, 10.1161/CIRCULATIONAHA.112.096438. [DOI] [PubMed] [Google Scholar]
  • [20].Kolodgie FD, Gold HK, Burke AP, Fowler DR, Kruth HS, Weber DK, Farb A, Guerrero LJ, Hayase M, Kutys R, Narula J, Finn A.v., Virmani R Intraplaque hemorrhage and progression of coronary atheroma, N. Engl. J. Med 349 (2003) 2316–2325, 10.1056/NEJMoa035655. [DOI] [PubMed] [Google Scholar]
  • [21].Kolodgie FD, Virmani R, Burke AP, Farb A, Weber DK, Kutys R, V Finn A, Gold HK, Pathologic Assessment of the Vulnerable Human Coronary plaque., Heart, vol. 90, British Cardiac Society, 2004, pp. 1385–1391, 10.1136/hrt.2004.041798. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [22].Cornelissen A, Guo L, Sakamoto A, Virmani R, Finn AV, New insights into the role of iron in inflammation and atherosclerosis, EBioMedicine 47 (2019) 598–606, 10.1016/J.EBIOM.2019.08.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [23].Wang H, Gardecki JA, Ughi GJ, Jacques PV, Hamidi E, Tearney GJ, Ex vivo catheter-based imaging of coronary atherosclerosis using multimodality OCT and NIRAF excited at 633 nm, Biomed. Opt Express 6 (2015) 1363, 10.1364/BOE.6.001363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [24].Ughi GJ, Wang H, Gerbaud E, Gardecki JA, Fard AM, Hamidi E, Vacas-Jacques P, Rosenberg M, Jaffer FA, Tearney GJ, Clinical characterization of coronary atherosclerosis with dual-modality OCT and near-infrared autofluorescence imaging, JACC (J. Am. Coll. Cardiol.): Cardiovascular Imaging 9 (2016) 1304–1314, 10.1016/j.jcmg.2015.11.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [25].Htun NM, Chen YC, Lim B, Schiller T, Maghzal GJ, Huang AL, Elgass KD, Rivera J, Schneider HG, Wood BR, Stocker R, Peter K, Near-infrared autofluorescence induced by intraplaque hemorrhage and heme degradation as marker for high-risk atherosclerotic plaques, Nat. Commun 8 (2017) 75, 10.1038/s41467-017-00138-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [26].van de Poll SWE, Bakker Schut TC, van der Laarse A, Puppels GJ, In situ investigation of the chemical composition of ceroid in human atherosclerosis by Raman spectroscopy, J. Raman Spectrosc 33 (2002) 544–551, 10.1002/jrs.865. [DOI] [Google Scholar]
  • [27].Yin D, Biochemical basis of lipofuscin, ceroid, and age pigment-like fluorophores, Free Radic. Biol. Med 21 (1996) 871–888, 10.1016/0891-5849-(96)00175-X. [DOI] [PubMed] [Google Scholar]
  • [28].Haka AS, Kramer JR, Dasari RR, Fitzmaurice M, Mechanism of ceroid formation in atherosclerotic plaque: in situ studies using a combination of Raman and fluorescence spectroscopy, J. Biomed. Opt 16 (2011), 011011, 10.1117/1.3524304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [29].Albaghdadi MS, Ikegami R, Kassab MB, Gardecki JA, Kunio M, Chowdhury MM, Khamis R, Libby P, Tearney GJ, Jaffer FA, Near-infrared autofluorescence in atherosclerosis associates with ceroid and is generated by oxidized lipid-induced oxidative stress, Arterioscler. Thromb. Vasc. Biol (2021), 10.1161/ATVBAHA.120.315612. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [30].Georgakopoulou EA, Tsimaratou K, Evangelou K, Fernandez Marcos PJ, Zoumpourlis V, Trougakos IP, Kletsas D, Bartek J, Serrano M, Gorgoulis VG, Specific lipofuscin staining as a novel biomarker to detect replicative and stress-induced senescence. A method applicable in cryo-preserved and archival tissues, Aging 5 (2013) 37–50, 10.18632/aging.100527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [31].Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B, Tinevez J-Y, White DJ, Hartenstein V, Eliceiri K, Tomancak P, Cardona A, Fiji: an open-source platform for biological-image analysis, Nat. Methods 9 (2012) 676–682, 10.1038/nmeth.2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [32].Crum WR, Camara O, Hill DLG, Generalized overlap measures for evaluation and validation in medical image analysis, IEEE Trans. Med. Imag 25 (2006) 1451–1461, 10.1109/TMI.2006.880587. [DOI] [PubMed] [Google Scholar]
  • [33].Dunn KW, Kamocka MM, McDonald JH, A practical guide to evaluating colocalization in biological microscopy, Am. J. Physiol. Cell Physiol 300 (2011) C723–C742, 10.1152/ajpcell.00462.2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Dice LR, Measures of the amount of ecologic association between species, Ecology 26 (1945) 297–302, 10.2307/1932409. [DOI] [Google Scholar]
  • [35].Pereira S, Pinto A, Alves V, Silva CA, Brain tumor segmentation using convolutional neural networks in MRI images, IEEE Trans. Med. Imag 35 (2016) 1240–1251, 10.1109/TMI.2016.2538465. [DOI] [PubMed] [Google Scholar]
  • [36].Sakakura K, Nakano M, Otsuka F, Ladich E, Kolodgie FD, Virmani R, Pathophysiology of atherosclerosis plaque progression, Heart Lung Circ. 22 (2013) 399–411, 10.1016/j.hlc.2013.03.001. [DOI] [PubMed] [Google Scholar]
  • [37].Michel J-B, Virmani R, Arbustini E, Pasterkamp G, Intraplaque haemorrhages as the trigger of plaque vulnerability, Eur. Heart J 32 (2011), 10.1093/eurheartj/ehr054, 1977–85, 1985a, 1985b, 1985c. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [38].Mitchinson MJ, Hothersall DC, Brooks PN, de Burbure CY, The distribution of ceroid in human atherosclerosis, J. Pathol 145 (1985) 177–183, 10.1002/path.1711450205. [DOI] [PubMed] [Google Scholar]
  • [39].Mitchinson MJ, Insoluble lipids in human atherosclerotic plaques, Atherosclerosis 45 (1982) 11–15, 10.1016/0021-9150(82)90167-8. [DOI] [PubMed] [Google Scholar]
  • [40].Perrotta I, Occurrence and characterization of lipofuscin and ceroid in human atherosclerotic plaque, Ultrastruct. Pathol 42 (2018) 477–488, 10.1080/01913123.2018.1544953. [DOI] [PubMed] [Google Scholar]
  • [41].Marchio P, Guerra-Ojeda S, Vila JM, Aldasoro M, Victor VM, Mauricio MD, Targeting early atherosclerosis: a focus on oxidative stress and inflammation, Oxid. Med. Cell. Longev (2019) 8563845, 10.1155/2019/8563845, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Vinchi F, Muckenthaler MU, Da Silva MC, Balla G, Balla J, Jeney V, Atherogenesis and iron: from epidemiology to cellular level, Front. Pharmacol 5 (2014) 94, 10.3389/fphar.2014.00094. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [43].Gardner CM, Tan H, Hull EL, Lisauskas JB, Sum ST, Meese TM, Jiang C, Madden SP, Caplan JD, Burke AP, Virmani R, Goldstein J, Muller JE, Detection of lipid core coronary plaques in autopsy specimens with a novel catheter-based near-infrared spectroscopy system, JACC (J. Am. Coll. Cardiol.): Cardiovascular Imaging 1 (2008) 638–648, 10.1016/J.JCMG.2008.06.001. [DOI] [PubMed] [Google Scholar]
  • [44].Kakimoto Y, Okada C, Kawabe N, Sasaki A, Tsukamoto H, Nagao R, Osawa M, Myocardial lipofuscin accumulation in ageing and sudden cardiac death, Sci. Rep 9 (2019) 3304, 10.1038/s41598-019-40250-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [45].Mury EP, Chirico EN, Mura M, Millon A, Canet-Soulas E, Pialoux V, Oxidative stress and inflammation, key targets of atherosclerotic plaque progression and vulnerability: potential impact of physical activity, Sports Med 48 (12) (2018) 2725–2741, 10.1007/s40279-018-0996-z. [DOI] [PubMed] [Google Scholar]
  • [46].Camaré C, Pucelle M, Négre-Salvayre A, Salvayre R, Angiogenesis in the atherosclerotic plaque, Redox Biol 12 (2017) 18–34, 10.1016/J.REDOX.2017.01.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [47].Virmani R, Kolodgie FD, Burke AP, et al. , Atherosclerotic plaque progression and vulnerability to rupture: angiogenesis as a source of intraplaque hemorrhage, Arterioscler. Thromb. Vasc. Biol 25 (10) (2005) 2054–2061, 10.1161/01.ATV.0000178991.71605.18. [DOI] [PubMed] [Google Scholar]
  • [48].Cornelissen A, Guo L, Sakamoto A, Virmani R, Finn AV, New insights into the role of iron in inflammation and atherosclerosis, EBioMedicine 47 (2019) 598–606, 10.1016/J.EBIOM.2019.08.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [49].Takaya N, Yuan C, Chu B, Saam T, Polissar NL, Jarvik GP, Isaac C, McDonough J, Natiello C, Small R, Ferguson MS, Hatsukami TS, Presence of intraplaque hemorrhage stimulates progression of carotid atherosclerotic plaques: a high-resolution magnetic resonance imaging study, Circulation 111 (2005) 2768–2775, 10.1161/CIRCULATIONAHA.104.504167. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Materials
NIHMS1793265-supplement-Supplemental_Materials.docx (2.8MB, docx)

RESOURCES