Association Between Aortic Wall Parameters on Multidetector Computed Tomography and Ruptured Plaques By Nonobstructive General Angioscopy

Abstract

Background

Nonobstructive general angioscopy (NOGA) can identify vulnerable plaques in the aortic lumen that serve as potential risk factors for cardiovascular events such as embolism. However, the association between computed tomography (CT) images and vulnerable plaques detected on NOGA remains unknown.

Methods and Results

We investigated 101 patients (67±11 years; women, 13.8%) who underwent NOGA and contrast‐enhanced CT before or after 90 days in our hospital. On CT images, the aortic wall thickness, aortic wall area (AWA), and AWA in the vascular area were measured at the thickest point from the 6th to the 12th thoracic vertebral levels. Furthermore, the association between these measurements and the presence or absence of NOGA‐derived aortic plaque ruptures (PRs) at the same vertebral level was assessed. NOGA detected aortic PRs in the aortic lumens at 145 (22.1%) of the 656 vertebral levels. The presence of PRs was significantly associated with greater aortic wall thickness (3.3±1.7 mm versus 2.1±1.2 mm), AWA (1.33±0.68 cm² versus 0.89±0.49 cm²), and AWA in the vascular area (23.2%±9.3% versus 17.2%±7.6%) (P<0.001 for all) on the CT scans compared with the absence of PRs. The frequency of PRs significantly increased as the aortic wall thickness increased. Notably, a few NOGA‐derived PRs were detected on CT in near‐normal intima.

Conclusions

The presence of NOGA‐derived PRs was strongly associated with increased aortic wall thickness, AWA, and AWA in the vascular area, measured using CT. NOGA can detect PRs in the intima that appear almost normal on CT scans.

Nonstandard Abbreviations and Acronyms

AWA

aortic wall area

AWA/VA

aortic wall area in the vascular area

AWT

aortic wall thickness

NOGA

nonobstructive general angioscopy

NOGA‐AAS

NOGA‐Guided Analysis of Regional Myocardial Perfusion Abnormalities Treated With Intramyocardial Injections of Plasmid Encoding Vascular Endothelial Growth Factor A‐165 in Patients With Chronic Myocardial Ischemia

PR

plaque rupture

VA

vascular area

Clinical Perspective.

What Is New?

• Aortic wall thickness, aortic wall area, and aortic wall area in the vascular area measured on computed tomography were closely associated with the presence of nonobstructive general angioscopy–derived aortic plaque ruptures.

• Nonobstructive general angioscopy detected plaque ruptures in 9.0% of the intima that appear almost normal on computed tomography scans.

What Are the Clinical Implications?

• Nonobstructive general angioscopy may be able to identify patients who could benefit from atherosclerosis treatment by detecting unstable plaques, even in the aorta, that do not appear strongly atherosclerotic on computed tomography.

• Computed tomography and nonobstructive general angioscopy could emerge as potential imaging indications for guiding atherosclerosis treatment.

Monitoring noncoronary atherosclerosis in patients with ischemic heart disease is crucial owing to its association with systemic cardiovascular events. Blood pressure and lipid‐level targets were set according to the systemic disease guidelines of the United States and the European Union. ¹ , ² The aorta is a representative organ that strongly represents the effects of atherosclerosis, ³ often linked to embolisms resulting from aortic plaque ruptures (PRs) (Video S1). ⁴ , ⁵ Contrast‐enhanced computed tomography (CT) is one of the most common modalities used to evaluate the characteristics of the entire aortic wall, which is strongly associated with systemic atherosclerosis and future cardiovascular events. ⁶ , ⁷ However, existing studies have highlighted CT's limitations in detecting atherosclerotic qualities on the basis of the biological activity or vulnerability of plaques. ⁸ , ⁹ , ¹⁰ , ¹¹

Nonobstructive general angioscopy (NOGA) allows for precise, direct, in situ observations of the aortic lumen, enabling the detection of aortic plaques, including aortic PRs. ¹² PRs are an independent predictor of cardiovascular events, renal impairment, and ischemic stroke. ⁴ , ¹³ The accumulation of asymptomatic spontaneous ruptured aortic plaques may lead to vulnerability through systemic “embolic” processes, such as mechanical obstruction and activation of the inflammasome pathway. ¹⁴ , ¹⁵ However, the association between the aortic wall characteristics on CT and the presence of aortic plaques on NOGA has not yet been clarified. The present study aimed to examine the associations between aortic wall thickness (AWT); aortic wall area (AWA); and the percentage of aortic wall area in the vascular area (AWA/VA), as measured on CT; and the PRs evaluated by NOGA.

METHODS

The authors declare that all supporting data are available within the article and its online supplementary files.

Study Patients

From January 2017 to December 2022, 394 consecutive patients scheduled for angioscopy of the aorta following coronary angiography for the assessment of coronary artery disease were deemed eligible for inclusion in the study. Only a limited number of studies have compared the NOGA and CT findings. Numerous studies revealed that the significant changes in aortic wall thickness may take up to 12 months to manifest. ¹⁶ , ¹⁷ , ¹⁸ A previous study reported that changes in atherosclerosis could be only observed on CT after a minimum of 3 months. ¹⁷ Therefore, patients who underwent contrast‐enhanced CT within 90 days before or after undergoing NOGA for aortic screening were included in the study. Patients who developed acute coronary syndrome, received mechanical circulatory support, were hemodynamically unstable (ie, Killip state >2 or cardiogenic shock), underwent aortic dissection, or had aortic aneurysm requiring surgery, after surgical graft replacement or stent graft placement, were excluded. Of the 394 patients, 293 who were unable to undergo CT (n=285) or whose descending aorta was not observable on NOGA (n=8) were excluded. Hence, only 101 patients (67±11 years; women, 13.8%) were included in the final analysis (Figure 1). The study was approved by the institutional review board of Nihon University Itabashi Hospital (RK‐210713‐7) and registered in the University Hospital Medical Information Network Clinical Trials Registry of Japan (UMIN000051008). The institutional review boards of all participating hospitals also approved the study, and all patients provided written informed consent.

Figure 1. Flowchart showing the participant selection process.

CT indicates computed tomography; and NOGA, nonobstructive general angioscopy.

Cardiac Catheterization and Angioscopic System

Following the same procedure as outlined in the previous study, ⁴ cardiac catheterizations were performed in a standard manner via the radial or femoral artery approach. After coronary angiography or percutaneous coronary intervention, NOGA was performed using a 6‐French Ikari Left 3.5 guiding catheter (Goodman, Nagoya, Japan) with a VISIBLE Fiber (FT‐203F, Fiber Tech Co. Ltd., Tokyo, Japan) and console (InterTec Medicals Co. Ltd., Osaka, Japan). Angioscopic images were recorded using a digital video recorder for subsequent offline analysis. Recordings were made during the injection of 10 to 20 mL (1–2 g) of dextran 40 and lactated Ringer's solution (Low Molecular Dextran L Injection; Otsuka Pharmaceutical Factory, Tokushima, Japan) to facilitate the removal of red blood cells from the observation site. The volume of solution used in each examination was ≈50 mL to 80 mL, which was significantly below the threshold associated with renal injury. ¹⁹ , ²⁰

The tips of the fiber, 4‐French probing, and 6‐French guiding catheters were positioned identically. These catheters were systematically rotated at each vertebral level for comprehensive vessel‐wide screening and then pulled back at a constant speed to scan the aortic intima from the ascending aorta to the descending aorta.

Assessment of Nonobstructive General Angioscopy

An aortic plaque was defined as visible atherosclerosis involving the luminal surface of the aorta. Various plaque characteristics were documented, encompassing PRs, thrombi, ulcerations, and intense yellow plaques (Figure 2A through 2E). PRs were defined as puff ruptures or puff‐chandelier ruptures, displaying a scattered “puff‐like” appearance. ²¹ , ²² The yellow grade was defined as a “4‐point‐scale grading,” ²² with the intensive yellow plaque defined as yellow grade 3. Plaque and yellowness were examined at every level of the vertebral body using frontal fluoroscopy (Figure 2F). The use of NOGA for visualizing aortic plaques adhered to a standardized procedure overseen by a committee, ensuring a quantitative evaluation of all aortic plaques as reproducibly as possible. ¹² NOGA analysis also showed good intra‐ and interobserver reproducibility. ⁴ , ²³ , ²⁴

Figure 2. Aortic plaques detected by nonobstructive general angioscopy.

A, Plaque ruptures. B, Ulceration. C, Thrombi. D, Intensive yellow plaque. E, Normal intima. F, Angioscopy findings are recorded according to the level of the evaluated vertebral body. L indicates lumbar vertebra; and Th, thoracic vertebra. *Probing catheter.

Enhanced CT

CT scans were conducted at the discretion of the clinicians. Aortic CT imaging was carried out in a single breath‐hold. Contrast‐enhanced CT was performed using a 320‐section (Aquilion ONE Vision; Canon Medical Systems Corp., Tokyo, Japan), 128‐section (Brilliance CT 128‐channel scanner; Philips Medical Systems, Andover, MA), or 64‐section CT scanner (Brilliance CT 64‐channel scanner; Philips Medical Systems). The choice of CT scanner was primarily determined on the basis of certain factors such as time, whether the patient was an inpatient or outpatient, and equipment availability, rather than being specified by the clinician. In a 320‐section CT scanner, iodinated contrast medium (Iomeron; Bracco‐Eisai Co., Ltd., Tokyo, Japan) was administered to all patients at a dose of 400 mg of iodine/kg with an injector, unless contraindicated. The following scan parameters were used: collimation, 1.00 mm; rotation time, 0.5 second; tube voltage, 120 kV; and tube current (mA) calculated following the automatic exposure control technique. In 128‐ and 64‐section CT scanners, iodinated contrast medium (Iomeron; Bracco‐Eisai Co., Ltd.) was administered to all patients at a dose of 521 mg of iodine/kg with an injector, unless contraindicated. The scan parameters were set as follows: collimation, 0.625 mm; rotation time, 0.5 second; tube voltage, 100 kV; and tube current (mA) calculated following the automatic exposure control technique. CT was performed under the aforementioned conditions and adjusted as required.

Assessment of Enhanced CT

The aortic wall parameters on multidetector CT were evaluated on the basis of axial images obtained at 1‐mm intervals. ⁶ The CT images were evaluated using Ziostation 2 Workstation (Ziosoft Inc., Tokyo, Japan). First, a centerline configuration was created between the distal aortic arch and abdominal aorta on axial CT images using a curved multiplanar reconstruction application (Figure 3A). Next, the curved multiplanar reconstruction image was converted into a straight multiplanar reconstructed image (Figure 3B). Third, grid lines were placed on the straight multiplanar reconstructed image, and every 1‐mm axial image within the aortic analysis range was evaluated. Fourth, we measured the AWT and AWA at the location of the thickest AWT within the largest noncalcified plaque at a single vertebral level (Figure 3C and 3D). An interactive freehand manual drawing tool was used for tracing the aortic lumen and wall contour, allowing the examination of the lumen area and vascular area (VA). AWA was calculated after excluding the lumen area from the VA. Aortic lumen tracing was partially referred to as the automatic measurement function on Ziostation 2. The AWA with calcified plaques measured >150 Hounsfield units. ²⁵ The minimum AWT was defined as the distinct luminal protrusions with a radial thickness of ≥1 mm, visually discernible from the blood signal. ²⁶ , ²⁷ Finally, the AWT, AWA, and AWA/VA were obtained for each vertebra. ²⁷

Figure 3. AWT and AWA were measured on Ziostation 2.

CT parameters were measured on Ziostation 2. A centerline configuration was created between the distal aortic arch and abdominal aorta on axial CT images using a curved multiplanar reconstruction (CPR) application (A). The CPR image was converted into a straight multiplanar reconstructed (sMPR) image (B).The aortic lumen area was measured semiautomatically (2.71 cm²). AWT was measured manually (5.83 mm) (C). AWA was measured manually. The vascular area was measured manually (4.97 cm²), and the lumen area was deducted to calculate the AWA (2.26 cm²) (D).

Range of Aortic Investigations

The CT and NOGA findings obtained at the descending aorta from the 6th to the 12th thoracic vertebral levels were compared (Figure 4). The selection of this observation range observations was driven by 2 factors. First, the descending aorta may be a more indicative site of atherosclerosis compared with the ascending aorta or aortic arch. ²² , ²⁸ Second, the descending aorta is relatively straight and is less influenced by the heartbeat. Therefore, this facilitated the accurate investigation of the aortic parameters using CT.

Figure 4. CT and NOGA findings obtained from the descending aorta from the 6th to the 12th thoracic vertebral levels were compared.

The 2 modalities were compared at the same vertebral level. AWA indicates aortic wall area; AWA/VA, aortic wall area in the vascular area; AWT, aortic wall thickness; CT, computed tomography; NOGA, nonobstructive general angioscopy; PRs, plaque ruptures; and VA, vascular area.

Evaluations and Study Outcomes

The primary end point was the relationship between aortic parameters on CT, such as AWT, AWA, and AWA/VA, and the presence or absence of NOGA‐derived aortic PRs. The secondary end point was the presence of NOGA‐derived PRs in the aortic wall with near‐normal CT findings.

Statistical Analysis

JMP Pro Statistics 16 (SAS Institute, Cary, NC) was used to perform all statistical analyses. Meanwhile, the differences between the receiver operating characteristic (ROC) curves were analyzed using MedCalc version 19.3 Software (MedCalc Software, Ostend, Belgium). The baseline continuous variables were expressed as the mean±SD for normally distributed variables or as the median (interquartile range) for nonnormally distributed variables. The normality was a reference to the Shapiro–Wilk tests. The study population was divided into 2 groups on the basis of the median number of PRs per patient from the 6th to the 12th thoracic vertebral levels. The groups were compared using Student's t test and analysis of variance, or Mann–Whitney U‐test. Categorical variables were expressed as frequencies and analyzed using χ² statistics or Fisher's exact test. Univariate and multivariate logistic regression analyses were employed to determine the factors associated with high PRs. In the multivariate analysis for potential confounders, models were constructed to adjust for the effects of clinically relevant factors, including age, the use of calcium channel blockers, estimated glomerular filtration rate, triglyceride levels, and median AWT from the 6th to the 12th thoracic vertebral levels. These factors had a P value of <0.05 in the univariate analysis. The odds ratios (ORs) and 95% CIs were calculated. All aortic sections from the 6th to the 12th thoracic vertebral levels were divided into 2 groups on the basis of the presence or absence of PRs. The aorta sections were divided into 2 groups using AWT as a cutoff value, determined by ROC analysis. A ROC curve was plotted to determine the cutoff values of the AWT, AWA, and AWA/VA for presence of PRs in each aortic section (n=656). The AWT and AWA were independently assessed by 2 expert cardiologists who were blinded to the patient's clinical information. Two independent observers (M.M. and K.K.) measured the AWT and AWA on 35 aortic CT images from 5 randomly selected patients and assessed the interobserver reproducibility using the analysis of intraclass correlation coefficients. Cohen's κ was used to determine the interobserver variability for the absence or presence of PRs on 35 aortic NOGA images. A P value of <0.05 was considered significant.

RESULTS

Study Patients

The patients' baseline characteristics are summarized in Table 1. Of the 101 patients (67±11 years; women, 13.8%), 78.2% had hypertension, 36.6% had diabetes, 73.2% had dyslipidemia, 74.2% had a history of percutaneous coronary intervention, and 2.9% had undergone coronary artery bypass grafting. Statins were prescribed in 72.0% of the patients. None of the patients who underwent NOGA experienced embolic events within 24 hours. The median number of NOGA‐derived aortic PRs from the 6th to the 12th thoracic vertebral levels per patient was 1 (0–2), while the mean was 1.5±1.7. The patients were divided into 2 groups on the basis of the median number of PRs (Table 1). The high‐PR group had ≥2 PRs (n=45), while the low‐PR group had 1 or no PR (n=56). The patients with high PRs had a higher age (70±9 versus 64±13 years; P=0.006), higher levels of triglycerides (OR, 134 [95% CI, 100–228] versus 115 [74, 147] mg/dL, P=0.041), and a higher prevalence of taking calcium channel blocker (26 [57.8%] versus OR, 20 [95% CI, 35.7%]; P=0.027) and antiplatelet drugs (39 [86.7%] versus OR, 35 [95% CI, 62.5%]; P=0.006) compared with the patients with low PRs. Furthermore, the estimated glomerular filtration rates were lower in the high‐PR group (66.2±17.6 versus 75.8±20.4 mL/min per 1.73 m²; P=0.014). In the univariate analysis, age, the use of calcium channel blockers and antiplatelet drugs, triglyceride levels, median AWT, mean AWA, and mean AWA/VA were predictors of high PRs (Table 2). In the multivariate analysis, median AWT was a predictor of high PRs (OR, 3.13, 95% CI: 1.57–6.24, P<0.001).

Table 1.

Patient Characteristics

Variables	Total (N=101)	High‐PR group (n=45)	Low‐PR group (n=56)	P value
Age, y	67±11	70±9	64±13	0.006
Female sex	14 (13.8)	7 (15.6)	7 (12.5)	0.66
Body mass index, kg/m2	23.8±3.7	23.7±3.4	24.1±4.0	0.58
Past history
Hypertension	79 (78.2)	37 (82.2)	42 (75.0)	0.38
Diabetes	37 (36.6)	17 (37.8)	20 (35.7)	0.83
Dyslipidemia	74 (73.2)	36 (80.0)	38 (67.8)	0.17
Smoking	64 (63.4)	27 (60.0)	37 (66.1)	0.53
History of PCI	75 (74.2)	35 (77.8)	50 (71.4)	0.47
History of CABG	3 (2.9)	2 (4.4)	1 (1.8)	1.00
History of OMI	22 (21.8)	10 (17.0)	22 (28.6)	0.16
History of stroke	9 (8.9)	4 (8.9)	5 (8.9)	1.00
History of PAD	9 (8.9)	5 (11.1)	4 (7.1)	0.51
Medications
β Blockers	45 (44.6)	22 (48.9)	23 (41.1)	0.55
Renin–angiotensin system inhibitors	58 (57.4)	26 (57.8)	32 (57.1)	0.95
Calcium channel blockers	46 (45.4)	26 (57.8)	20 (35.7)	0.027
Statin	72 (72.0)	34 (75.6)	39 (69.6)	0.51
Ezetimibe	18 (17.8)	9 (20.0)	9 (16.1)	0.61
Antiplatelet drugs	74 (73.3)	39 (86.7)	35 (62.5)	0.006
Oral hypoglycemic agent	27 (26.7)	13 (28.9)	14 (25.0)	0.66
Insulin	2 (2.0)	2 (4.4)	0 (0.0)	0.20
Laboratory data
Hemoglobin, g/dL	13.3±1.6	13.0±1.7	13.5±1.6	0.13
Creatinine, mg/dL	0.83 (0.73–0.94)	0.87 (0.77–0.97)	0.80 (0.68–0.91)	0.06
eGFR, mL/min per 1.73 m2	70.8±19.3	66.2±17.6	75.8±20.4	0.014
Glycated hemoglobin, %	6.3±0.9	6.3±0.9	6.3±1.0	0.94
Total cholesterol, mg/dL	171±42	175±42	163±43	0.13
LDL cholesterol, mg/dL	94±36	98±38	90±32	0.27
HDL cholesterol, mg/dL	45±11	46±11	44±13	0.33
Triglyceride, mg/dL	122 (89–175)	134 (100–228)	115 (74–147)	0.041
Non‐HDL cholesterol, mg/dL	121 (96–157)	124 (96–159)	118 (99–145)	0.74
Uric acid, mg/dL	5.6±1.3	5.6±1.4	5.5±1.2	0.55
hs‐CRP, mg/dL	0.100 (0.037–0.300)	0.085 (0.032–0.278)	0.100 (0.04–0.300)	0.74
NT‐proBNP, pg/mL	199 (85–659)	213 (74–820)	199 (99–601)	0.99
CT finding
Median AWT from Th6 to Th12, mm	2.2 (1.4–3.0)	2.8 (2.0–3.5)	1.8 (1.0–2.4)	< 0.001
Mean AWA from Th6 to Th12, cm2	0.98±0.57	1.20±0.64	0.81±0.44	< 0.001
Mean AWA/VA from Th6 to Th12, %	18.4±8.5	21.7±9.3	15.8±6.8	< 0.001
Difference between CT and NOGA, d	17 (11–40)	22 (12–50)	15 (10–32)	0.18

Values are n (%), mean±SD, and median (25th–75th quartile). AWA indicates aortic wall area; AWT, aortic wall thickness; CABG, coronary artery bypass grafting; CT, computed tomography; eGFR, estimated glomerular filtration rate; HDL, high‐density lipoprotein; hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; NOGA, nonobstructive general angioscopy; NT‐proBNP, N‐terminal pro‐B‐type natriuretic peptide; OMI, old myocardial infarction; PAD, peripheral artery disease; PCI, percutaneous coronary intervention; PR, plaque rupture; Th, thoracic vertebra; and VA, vascular area.

Table 2.

Predictors of High PRs

	Univariate		Multivariate
	Odds ratio (95% CI)	P value	Odds ratio (95% CI)	P value
Age, y	1.04 (1.01–1.10)	0.005	1.00 (0.94–1.05)	0.90
Hypertension	1.54 (0.58–4.09)	0.38
Diabetes	1.09 (0.48–2.46)	0.32
Dyslipidemia	1.89 (0.75–4.76)	0.17
Smoking	0.77 (0.34–1.73)	0.53
Calcium channel blockers	2.46 (1.10–5.51)	0.028	1.97 (0.75–5.12)	0.16
Antiplatelet drugs	3.90 (1.41–10.80)	0.009
eGFR, mL/min per 1.73 m2	0.97 (0.95–1.00)	0.012	0.98 (0.96–1.01)	0.15
LDL cholesterol, mg/dL	1.00 (0.99–1.01)	0.96
Triglyceride, mg/dL	1.00 (1.00–1.01)	0.039	1.00 (1.00–1.01)	0.31
Glycated hemoglobin, %	1.02 (0.66–1.56)	0.93
Median AWT from Th6 to Th12, mm	3.31 (1.83–6.01)	<0.001	3.13 (1.57–6.24)	<0.001
Mean AWA from Th6 to Th12, cm2	6.53 (2.19–19.5)	<0.001
Mean AWA/VA from Th6 to Th12, %	1.17 (1.07–1.27)	<0.001

AWA indicates aortic wall area; AWT, aortic wall thickness; eGFR, estimated glomerular filtration rate; LDL, low‐density lipoprotein; PRs, plaque ruptures; Th, thoracic vertebra; and VA, vascular area.

CT Parameters and NOGA Findings

Of the 707 aortic section levels per vertebra, 51 vertebral levels were excluded for the following reasons: beyond the CT image range (n=40), lack of data on NOGA findings due to inadequate blood flow clearance or access limitations arising from severe aortic or arterial angulation (n=8), and difficulty in measuring the AWT due to severe aortic calcification (n=3). The intraclass correlation coefficients of the AWT and AWA on 35 aortic CT images were 0.81 and 0.79, respectively, indicating good reproducibility. Within the observational range of this study, the median AWT, average AWA, and average AWA/VA were 2.2 mm (interquartile range, 1.4–3.0), 0.98±0.57 cm², and 18.4%±8.5%, respectively (Table 3). The frequencies of PRs, ulceration, thrombi, and intense yellow plaque detected by NOGA on all aortic sections were 22.1%, 4.3%, 16.3%, and 25.1%, respectively. No significant difference was observed in the frequency of PRs in any of the thoracic vertebrae. The aortic sections with PRs exhibited higher values of AWT (3.3±1.7 mm versus 2.1±1.2 mm; P<0.001), AWA (1.33±0.68 cm² versus 0.89±0.49 cm²; P<0.001), and AWA/VA (23.2±9.3% versus 17.2±7.6%; P<0.001) compared with those without (Figure 5A through 5C). The Cohen's κ was 0.80 for the absence or presence of PRs, indicating good reproducibility.

Table 3.

CT Parameters and NOGA Findings of the Aortic Atherosclerosis

	CT parameters			NOGA findings
	AWT, mm	AWA, cm2	AWA/VA, %	PRs	Thrombus	Ulceration	Intensive yellow plaque
Total (N=656)	2.2 (1.4–3.0)	0.98±0.57	18.4±8.5	145 (22.1)	107 (15.3)	28 (4.3)	165 (25.2)
Th6 (n=91)	2.2 (1.5–2.9)	1.05±0.51	17.8±7.6	19 (20.9)	10 (11.0)	3 (3.3)	21 (23.1)
Th7 (n=93)	2.2 (1.5–3.0)	1.00±0.54	18.2±8.1	17 (18.3)	13 (14.0)	5 (5.4)	24 (25.8)
Th8 (n=93)	2.1 (1.8–3.2)	1.04±0.58	19.3±8.5	20 (21.5)	16 (17.2)	5 (5.4)	22 (23.7)
Th9 (n=95)	2.1 (1.5–3.0)	1.05±0.78	19.5±11.8	18 (20.0)	12 (12.6)	2 (2.1)	24 (25.3)
Th10 (n=95)	2.1 (1.4–3.0)	0.90±0.56	17.4±8.2	23 (24.2)	13 (13.7)	4 (4.2)	20 (21.1)
Th11 (n=95)	2.2 (1.1–3.0)	0.95±0.53	18.3±7.7	23 (24.2)	22 (23.2)	4 (4.2)	26 (27.4)
Th12 (n=94)	2.2 (1.0–3.1)	0.89±0.43	18.6±7.6	25 (26.6)	21 (22.3)	5 (5.3)	28 (29.8)

Values are n (%), mean±SD, and median (25th–75th quartile). AWA indicates aortic wall area; AWA/VA, aortic wall area in the vascular area; AWT, aortic wall thickness; CT, computed tomography; NOGA, nonobstructive general angioscopy; PRs, plaque ruptures; Th, thoracic vertebra; and VA, vascular area.

Figure 5. Bar graphs showing the AWT, AWA, and AWA/VA.

Bar graphs showing the (A) AWT, (B) AWA, and (C) AWA/VA. Details are in the text. The aortic sections with an absence of PRs had higher AWT (3.3±1.7 vs 2.1±1.2 mm, P<0.001), AWA (1.33±0.68 vs 0.89±0.49 cm², P<0.001), and AWA/VA (23.2±9.3 vs 17.2±7.6%, P<0.001) than those with no PRs. AWA indicates aortic wall area; AWT, aortic wall thickness; PRs, plaque ruptures; and VA, vascular area.

Determinants of the Presence of Aortic PRs

The ROC curve for predicting aortic PRs revealed that the optimal discriminating points for AWT, AWA, and AWA/VA were 3.0 mm, 1.09 cm², and 18.4%, respectively, with the areas under the curve of 0.74, 0.71, and 0.70, respectively (Figure 6). The AWT had a nonsignificantly larger area under the curve than the AWA and AWA/VA (P=0.31 and 0.42, respectively, versus AWT). In the descending aorta with an AWT of 3.0 mm, the specificity for the presence of PRs was 83.1% (sensitivity, 55.7%), corresponding to a positive predictive value of 48.5% and a negative predictive value of 87.7%.

Figure 6. The ROC curve for the maximum AWT, AWA, and AWA/VA value predicting aortic PRs.

The ROC curve for the maximum (A) AWT, (B) AWA, and (C) AWA/VA value predicting aortic PRs. The optimal discriminating values of the maximum AWT, AWA, and AWA/VA value were 3.0 mm, 1.09 cm², and 18.4%, with the area under the curves of 0.74, 0.71, and 0.70, respectively. AWA indicates aortic wall area; AWT, aortic wall thickness; PRs, plaque ruptures; and VA, vascular area.

Relationship Between the AWT and the Frequency of PRs

We investigated the relationship between AWT and the frequency of PRs owing to its ease of measurement compared with AWA and AWA/VA. The frequency of PRs progressively increased with increasing AWT (Figure 7). NOGA could detect PRs at 9.0% (11 of the total 122 aortic section levels even in an AWT of ≤1 mm). However, the frequency of PRs gradually increased at 11.9% (21/177) for an AWT of 1.1 to 2 mm, followed by 26.2% (38/201) for an AWT of 2.1 to 3 mm, and 48.1% (75/156) for an AWT of >3 mm. The NOGA findings at AWT values of >3.0 mm and ≤3.0 mm on CT are shown in Table 4. All NOGA findings, including PRs, thrombi, ulceration, and intensive yellow plaque, were significantly associated with an AWT of >3.0 mm.

Figure 7. The frequency of plaque rupture gradually increased as the AWT increased.

The frequency of plaque rupture gradually increased as the AWT increased. When the AWT values were ≤1 mm, 1.1 to 2 mm, 2.1 to 3 mm, and >3 mm, NOGA detects PRs at probability rates of 9.0%, 11.9%, 18.9%, and 48.1%, respectively. Representative images are presented in the bottom row. The AWT values are <1 mm (image of ≤1 mm), 1.8 mm (image of 1.1 to 2 mm), 2.3 mm (image of 2.1 to 3 mm), and 3.6 mm (image of >3 mm), respectively. AWT indicates aortic wall thicknesses; NOGA, non‐obstructive general angioscopy; and PRs, plaque ruptures.

Table 4.

Relationship Between the AWT and NOGA Findings

	AWT>3 mm (n=156)	AWT ≤3 mm (n=500)	P value
Plaque ruptures	75 (48.1)	79 (14.0)	<0.001
Thrombus	48 (30.8)	59 (11.8)	<0.001
Ulceration	13 (8.3)	15 (3.0)	0.004
Intensive yellow plaque	78 (50.0)	87 (17.4)	< 0.001

Values are n (%). AWT indicates aortic wall thickness; and NOGA, nonobstructive general angioscopy.

DISCUSSION

The present study is the first to report an association between PRs determined by NOGA and aortic wall parameters on CT. The AWT, AWA, and AWA/VA measured on CT were closely associated with the presence of PRs, as evaluated using NOGA. In addition, our study revealed that the frequency of PRs increased as the AWT increased. NOGA could detect PRs in 9.0% of nearly normal intima on CT, suggesting that NOGA may offer a more precise detection of aortic atheroma compared with CT.

Increased AWT is associated with atherosclerotic intimal thickening, chronic renal disease–related arteriopathy, and hypertension‐related medial hypertrophy. ²⁹ , ³⁰ , ³¹ AWA, reflecting AWT, is influenced by several risk factors including age, sex, diabetes, and hypertension. ³² , ³³ , ³⁴ , ³⁵ Moreover, both AWT and AWA are independently associated with incident death or cardiovascular disease. ¹⁴ , ²⁷ , ³⁶ The AWT in the upper descending aorta is also associated with coronary atherosclerosis. ³⁶ In that study, 64% of patients had a history of coronary atherosclerosis, and the mean AWT was 2.29 mm, similar to the findings of the present study. Consequently, AWT serves as a representative marker of systemic atherosclerosis associated with cardiovascular events.

Conversely, NOGA can directly detect plaque vulnerability, such as PRs in the aortic wall. ⁴ , ²² PRs assessed through NOGA have been associated with age, smoking, chronic kidney disease, peripheral arterial disease, and coronary artery disease. ⁴ , ²² , ²⁴ Notably, PRs are associated with the spontaneous release of debris rich in cholesterol crystals and leukocytes. ⁵ This debris consistently contains large cholesterol crystals, which are implicated in the activation of leukocytes involved in innate inflammation. Such events are considered potential causes of embolisms, such as stroke or peripheral arterial disease. ⁵ Recent studies revealed a significant correlation between ischemic stroke and PRs in patients with coronary artery disease. ¹³ , ³⁷ The clinical adverse events related to NOGA‐derived aortic plaques were not only a consequence of the progression of systemic atherosclerosis or inflammation but also resulted from the embolism of thrombi, cholesterol crystals, or other plaque materials from proximally located plaques into distal organs. ⁵ , ³⁸ As mentioned above, PRs detected by NOGA can be regarded as predictors of cardiovascular and embolic events. In our study, 59.4% of the individuals had at least ≥1 PRs as detected by NOGA in the descending aorta from the 6th to 12th thoracic level. This high detection rate aligns with the findings from another institution, which reported that all atherosclerotic plaques were detected in the descending aorta of 76% of patients with stable angina pectoris. ²⁴

Although the patients with high PRs were older or had higher triglyceride levels than the patients with low PRs, no significant differences were observed in the prevalence rates of hypertension, dyslipidemia, and diabetes between the 2 groups. This can be attributed to the fact that many of the study patients had a history of revascularization and therefore had high medication rates and well‐controlled levels of glycated hemoglobin, low‐density lipoprotein, and other parameters. Interestingly, the high‐PR group had a higher rate of calcium channel blocker use, although no difference was found in the prevalence of hypertension or the use of renin–angiotensin system inhibitors or β blockers. These findings may suggest a need for more aggressive antihypertensive treatment in this group.

The progression of aortic atherosclerosis often undergoes a prolonged subclinical phase. ³⁹ Thus, the early diagnosis and treatment of atherosclerosis in the subclinical stage may reduce the occurrence of overt cardiovascular disease. Previous studies reported that lipid‐lowering therapy with simvastatin reduced the AWA and AWT on magnetic resonance imaging in 21 patients with asymptomatic hypercholesterolemia. ⁴⁰ PRs, which are associated with an increased risk of cardiovascular events, can also be stabilized by lipid‐lowering therapy with rosuvastatin. ³⁸ The ROC analysis in this study showed that the frequency of PRs can be predicted by performing a simplified measurement of aortic wall parameters on CT. This study highlights a noteworthy finding: even when AWT does not indicate intense atherosclerotic findings, PRs can still be detected using NOGA. Another study reported that slight aortic arch plaques detected on transesophageal echocardiography may be a possible source of embolism, thus leading to stroke. ⁴¹ The positive predictive value for predicting PRs on the basis of the AWT measured on CT images was not relatively high (48.5%). This suggests that patients who could benefit from atherosclerosis treatment should be identified by detecting unstable plaques using NOGA, even in the aortas that do not appear strongly atherosclerotic on CT. It may be crucial to identify PRs that reflect long‐term atherosclerosis rather than transient laboratory results. Significantly, the decision to perform CT was made by clinicians on the basis of the following indications: (1) screening for systemic vascular disease in case of suspected highly advanced atherosclerosis and (2) evaluation for abdominal bruit or poor arterial palpation of extremities. Additionally, caution was exercised in performing CT scans in patients with severe renal failure, with only 2 of the 101 patients having an estimated glomerular filtration rate of <30 mL/min per 1.73 m².

Limitations

Our study has several limitations. First, it included a small, single‐center sample. In addition, the study focused on evaluating the aortic wall parameters measured on the CT images using the same protocol, potentially limiting the generalizability of the results to other imaging modalities (such as magnetic resonance imaging or ultrasound) or other CT protocols. Second, matching the aortic location between the 2 modalities was not possible, revealing the association of PRs with NOGA and the aorta section with higher AWT or AWA values within each vertebral level. This study could not confirm that the NOGA detected PRs precisely at the “point” of high AWT. This may affect the accuracy of the correlation results. However, considering the observed rotation of NOGA in the horizontal plane per vertebral body and the longitudinal spread of aorta wall thickening, it is reasonable to assume that the same region was captured using these 2 modalities. Third, the significance of PRs detected by NOGA, although identified as a prognostic factor, ⁴ , ⁵ , ¹³ , ¹⁵ remains unclear. It is also uncertain whether NOGA is a more accurate predictor of clinical outcomes compared with other imaging modalities. Further studies, such as the ongoing DREAM‐NOGA study (Detection of Ruptured plaques as Embolic source in Aorta to clarify the Mechanism of systemic organ deterioration by Non‐Obstructive General Angioscopy study) or NOGA‐AAS (NOGA‐Guided Analysis of Regional Myocardial Perfusion Abnormalities Treated With Intramyocardial Injections of Plasmid Encoding Vascular Endothelial Growth Factor A‐165 in Patients With Chronic Myocardial Ischemia) study, are required to address these issues. Finally, we analyzed the images from a few CT scanners. Nonetheless, several studies have shown no difference in the image quality between 64‐ and 320‐section gated or nongated CT, suggesting a limited impact on the results of this study. ⁴² , ⁴³ , ⁴⁴

CONCLUSIONS

Both the AWT and AWA measured on CT images were strongly associated with the presence of NOGA‐derived PRs. Our observations indicate that NOGA can provide a more detailed evaluation of plaque vulnerability compared with CT imaging parameters.

Sources of Funding

This work was supported by JSPS KAKENHI (Grant No. 21K16041) and Nihon University School of Medicine Alumni Association 60th Anniversary Fund Research Grant (2022).

Disclosures

None.

Supporting information

Video S1