Suboptimal Control of Small Dense Low‐Density Lipoprotein Cholesterol Is Associated With Coronary Plaque Progression: An Intravascular Ultrasound Study

Abstract

Background

Plaque progression (PP) is critical between subclinical atherosclerosis and plaque rupture. Small dense low‐density lipoprotein cholesterol (sdLDL‐C) is considered as the most atherogenic lipoprotein. This study aims to investigate the relationship between sdLDL‐C level and PP in patients with stable coronary artery disease.

Methods

We conducted a retrospective analysis of 146 lesions in 86 patients by repeat intravascular ultrasound examinations from January 2020 to May 2023. PP was determined by increases in percent atheroma volume, defined as the atheroma volume in proportion to the volume occupied by the entire vascular wall, ≥5% during follow‐up. Time‐averaged values were calculated for all cardiometabolic parameters including sdLDL‐C. Multivariate logistic regression analysis was performed to interrogate the association between time‐averaged sdLDL‐C and PP.

Results

During a median follow‐up of 12.6 months, PP was found in 65 lesions (44.5%), and mean changes in percent atheroma volume were 4.1%±10.2%. A positive correlation was observed between time‐averaged sdLDL‐C and changes in total atheroma volume (Pearson r=0.29, P=0.006), especially in diabetic patients (Pearson r=0.58, P<0.001). After multivariate adjustment, every 0.1‐mmol/L increase in time‐averaged sdLDL‐C conferred a 1.2‐fold increased risk of PP.

Conclusions

Our findings suggest that sdLDL‐C is an independent risk factor of PP in patients with coronary artery disease. Intensive control of sdLDL‐C along with other risk factors should be considered to mitigate PP and improve cardiovascular outcomes.

Nonstandard Abbreviations and Acronyms

EEM CSA

external elastic membrane cross‐sectional area

JNC 8

Eighth Joint National Committee

MACE

major adverse cardiovascular events

NCEP ATP III

Third Report of the National Cholesterol Education Program Adult Treatment Panel III

PARADIGM

Progression of Atherosclerotic Plaque Determined by Computed Tomographic Angiography Imaging

PAV

percent atheroma volume

PP

plaque progression

sdLDL

small dense low‐density lipoprotein

sdLDL‐C

small dense low‐density lipoprotein cholesterol

TAV

total atheroma volume

Clinical Perspective.

What Is New?

• Small dense low‐density lipoprotein cholesterol level is related to changes in total atheroma volume evaluated by repeat intravascular ultrasound examinations.

• Small dense low‐density lipoprotein cholesterol is an independent predictor of plaque progression after adjusting for conventional risk factors and lesion characteristics.

What Are the Clinical Implications?

• Assessment of small dense low‐density lipoprotein cholesterol aids in a more precise prediction of coronary plaque progression; intensive control of small dense low‐density lipoprotein cholesterol along with other risk factors should be considered to mitigate plaque progression and improve cardiovascular outcomes.

Plaque progression (PP), a critical step between subclinical atherosclerosis and plaque rupture, is one of the major pathological processes driving the onset of acute coronary events. ¹ , ² In recent years, PP is commonly found in 21% to 28% of patients with coronary artery disease (CAD) assessed by invasive and noninvasive coronary imaging techniques, ³ , ⁴ varying according to different PP criteria, follow‐up periods, and the specific imaging modality used. A number of conventional risk factors, such as diabetes, ⁵ hypertension,⁶ and dyslipidemia, ⁷ are related to PP. Low‐density lipoprotein (LDL) in particular has been regarded as the primary proatherogenic lipoprotein possessing a strong relationship with PP and other adverse cardiovascular events. ⁸ However, the residual risk of PP and cardiovascular events remains despite intensive control of LDL cholesterol (LDL‐C) levels. ⁹ In fact, LDL particles represent a heterogeneous lipoprotein group that can be classified into several subclasses based on their size, density, and lipid and apolipoprotein composition, including large buoyant LDL, intermediate LDL, and small dense LDL (sdLDL). ¹⁰

As a subfraction of LDL particles, sdLDL has been recognized as one of the most atherogenic lipoproteins. It is characterized by a higher density of 1.044 g/mL to 1.063 g/mL and a smaller particle size with a diameter <25.5 nm, compared with other LDL subclasses. ¹⁰ , ¹¹ A prospective case‐cohort study showed that small dense LDL‐C (sdLDL‐C) was related to a significantly increased risk of major adverse cardiovascular events (MACE) independent of LDL‐C in statin‐treated patients with CAD. ¹² Schaefer et al found that patients with sdLDL‐C levels ≥50 mg/dL experienced considerably higher atherosclerotic cardiovascular disease risk after adjustment for all standard risk factors. ¹³

However, the relationship between sdLDL‐C and PP is still unclear. In this study, we performed repeat intravascular ultrasound (IVUS) examinations on untreated marginal coronary lesions in patients with CAD and analyzed the relationship between time‐averaged sdLDL‐C levels and PP determined by changes in atheroma volume.

METHODS

Data Availability Statement

The data sets analyzed within the study are available from the corresponding author on reasonable request.

Study Design

We retrospectively analyzed 86 patients with stable CAD and untreated lesions who underwent IVUS examinations at baseline and during follow‐up after ≈12 months in Ruijin Hospital, Shanghai Jiao‐Tong University School of Medicine, between January 2020 and May 2023. The cohort comprised both patients with marginal lesions without percutaneous coronary intervention, as well as those with multivessel disease and untreated marginal lesions after percutaneous coronary intervention (Figure 1).

Figure 1. Flowchart of patient enrollment.

CABG indicates coronary artery bypass graft; CAD, coronary artery disease; IVUS, intravascular ultrasound; PAV, percent atheroma volume; and sdLDL‐C, small dense low‐density lipoprotein cholesterol.

Inclusion criteria were: (1) age ≥18 and ≤85 years; (2) patients with stable CAD undergoing coronary angiography; and (3) marginal lesions with minimal lumen area stenosis between 30% and 70% by IVUS assessment without interventional treatment. Exclusion criteria were: (1) history of coronary artery bypass graft; (2) poor IVUS image quality that may affect the accurate assessment of lesions; (3) history of familial hypercholesterolemia; and (4) <2 tests of sdLDL‐C during follow‐up.

In addition, the post hoc power analysis yielded an effect size of 0.487 and a power of 0.828 with a significant level of 0.05, and the sample sizes of nonprogressed and progressed lesions were 81 and 65, respectively.

This study complies with the Declaration of Helsinki. The study protocol was approved by Shanghai Ruijin Hospital's ethics committee, and written informed consent was obtained from all participants.

Clinical Assessment

Detailed information on medical history and lifestyles was obtained using a standard questionnaire by trained physicians. Body mass index was calculated using the formula of weight/height² (kilograms per square meter). ¹⁴ The diagnosis of type 2 diabetes was made according to the criteria of the American Diabetes Association (symptoms of diabetes with casual plasma glucose concentration ≥200 mg/dL [11.1 mmol/L] or fasting plasma glucose ≥126 mg/dL [7.0 mmol/L], 2 hours postprandial glucose ≥200 mg/dL [11.1 mmol/L] during an oral glucose tolerance test, and currently or previously treated with insulin and/or oral hypoglycemic agents). ¹⁵ Hypertension was diagnosed according to the Eighth Joint National Committee (JNC 8) on the prevention, detection, evaluation, and treatment of high blood pressure. ¹⁶ Dyslipidemia was diagnosed as total cholesterol ≥200 mg/dL, or triglycerides ≥150 mg/dL, or LDL‐C ≥130 mg/dL, or high‐density lipoprotein cholesterol (HDL‐C) <40 mg/dL according to the Third Report of the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III). ¹⁷ A current smoker was defined as a person who smokes at least 1 cigarette per day for at least the past 6 months.

Biochemical Assessment

Blood samples were obtained on the day of angiography in all patients after overnight fasting. Assessment of levels of sdLDL‐C, using a standard peroxidase method, ¹⁸ and other lipid parameters including total cholesterol, LDL‐C, HDL‐C, triglycerides, apolipoprotein A‐I, apolipoprotein B (apoB), and apolipoprotein E was performed on an AU5800 clinical chemistry analyzer (Beckman Coulter). Non–HDL‐C was determined as the difference between total cholesterol and HDL‐C. ¹⁹ , ²⁰ Remnant lipoprotein particle cholesterol (RLP‐C) was derived by subtracting direct LDL‐C from non–HDL‐C. ²¹ Estimated glomerular filtration rate was computed using the Chronic Kidney Disease Epidemiology Collaboration equation. ²²

Evaluation of Time‐Averaged Cardiometabolic Parameters During Follow‐Up

To better reflect control of cardiometabolic factors during follow‐up, time‐averaged parameters were calculated by computing the area under the curve of serial measurements from baseline to follow‐up using the trapezoidal rule. ²³ This involves calculating the area for each time interval, summing these areas, and then dividing by total follow‐up durations.

Acquisition and Analysis of IVUS Images

Acquired baseline and follow‐up IVUS image data were retrospectively analyzed. These data were acquired through the same IVUS imaging system (60 MHz mechanical rotary IVUS catheter, Boston Scientific). IVUS was performed with the catheter reaching the distal epicardial coronary artery and motorized pullback at 0.5 mm/s following intracoronary injection of 0.2 mg nitroglycerin. Anonymous IVUS data were assessed by 2 experienced cardiologists in the Cardiac Catheterization Laboratory, Department of Cardiovascular Medicine, Ruijin Hospital, without any knowledge of patient information. All of the cross‐sectional images and quantitative measurements were analyzed using QIvus Research Edition 3.1 (Medis Medical Imaging System BV). The matched segments between baseline and follow‐up were identified using side branches, as well as a comparison of angiographic images, and then analyzed through consensus standards. ²⁴ , ²⁵ The target lesion is defined as lesions with minimal lumen area stenosis ≥30% and ≤70% by IVUS. The 5‐mm segments proximal and distal to the target lesion were set for reference. External elastic membrane cross‐sectional area (EEM CSA) and lumen area were analyzed every 1 mm in the target lesion. ²⁶ Total atheroma area was defined by the equation: total atheroma area=EEM CSA−lumen area. Total atheroma volume (TAV) was determined according to the pullback speed during image acquisition, and was normalized to account for variations in segment lengths among patients: TAV norm=[∑ (EEM CSA−lumen area)/segment length]×median segment length in the population. ²⁷ Change in TAV norm was calculated as TAV_follow‐up−TAV_baseline. The percent atheroma volume (PAV), defined as the atheroma volume in proportion to the volume occupied by the entire vascular wall, was calculated using the following equation ⁸ : PAV=[∑ (EEM CSA−lumen area)/∑EEM CSA]·100. Change in PAV was calculated as PAV_follow‐up−PAV_baseline. PP was defined as an increase in PAV by ≥5%. ³ For per‐patient analysis, a patient who had PP in any of the analyzed lesions was defined as PP.

SYNTAX Score and Gensini Score

The methodologies for calculating the SYNTAX score and Gensini score were previously described. ²⁸ , ²⁹ Two experienced cardiologists evaluated each patient's SYNTAX score and Gensini score based on coronary angiography.

Statistical Analysis

Continuous variables are presented as median (interquartile range [IQR]) or mean±SD, and categorical data are summarized as numbers (percentages). Normal distribution of continuous variables was evaluated by Shapiro–Wilk test. Differences in normally distributed variables were assessed using Student t test. For nonnormally distributed continuous variables, differences were analyzed by the Mann–Whitney U test or Kruskal–Wallis test. Differences in categorical variables were analyzed by χ ² test. Fisher exact test was used when χ² test was not suitable and the expected values in any of the cells of a contingency table were <5. Correlation between time‐averaged sdLDL and changes in TAV was determined by Spearman correlation test. Univariate logistic regression analysis was performed to identify univariate determinants of PP. Afterwards, per‐lesion and per‐patient multivariate logistic regression models were constructed to interrogate the association between sdLDL‐C and PP by adjusting for traditional risk factors and predictors with P<0.10 in the univariate analysis. In model 1, sex and age were adjusted. In model 2, further adjustments were performed for hypertension, diabetes, high‐sensitivity C‐reactive protein, time‐averaged level of triglycerides, and use of aspirin based on model 1. In model 3, further adjustments were performed for reference lumen area, minimal lumen area, normalized TAV, and PAV based on model 2. All statistical analyses were performed using the R statistical package version 4.0.3 (R Project for Statistical Computing). A 2‐tailed P<0.05 was considered statistically significant.

RESULTS

Baseline Characteristics of the Cohort

We analyzed 146 lesions in 102 arteries of 86 patients with CAD who had a median IVUS follow‐up of 12.6 months (IQR, 11.4–15.8 months). The mean age was 62.7±8.6 years. There were 26.7% female patients and 37.2% patients with type 2 diabetes. Based on the criteria of PAV increasing ≥5%, we divided the 86 patients into a progression group (n=48, 55.8%) and a nonprogression group (n=38, 44.2%). Patients with PP had a higher prevalence of dyslipidemia than those without PP. There were no significant differences in sex, age, history of hypertension and diabetes, admission blood pressure, smoking status, body mass index, renal function, duration of IVUS follow‐up, and medication therapies between the 2 groups. The distribution of lesions was similar between the 2 groups, with the exception of more left main coronary artery lesions in the progression group. In addition, SYNTAX and Gensini scores were significantly higher in the progression group than in the nonprogression group (Table 1).

Table 1.

Baseline Characteristics of Patients With and Without Plaque Progression

Variables	Nonprogression	Progression	P value
	(n=38)	(n=48)
Male sex	32 (84.2)	31 (64.6)	0.072
Age, y	64.03±9.27	61.73±8.01	0.221
Diabetes	14 (36.8)	18 (37.5)	1.000
Hypertension	22 (57.9)	33 (68.8)	0.415
Dyslipidemia	18 (47.4)	37 (77.1)	0.009
Current smoker	20 (52.6)	17 (35.4)	0.167
BMI, kg/m2	25.18±3.05	25.17±2.57	0.993
Systolic BP, mm Hg	141.79±23.57	142.46±21.14	0.890
Diastolic BP, mm Hg	79.05±12.60	78.65±10.37	0.870
eGFR, mL/min per 1.732 m2	100.57±24.80	98.91±16.78	0.725
hs‐CRP, mg/L	0.81 (0.39–2.30)	1.96 (0.45–3.09)	0.280
Aspirin	33 (86.8)	44 (91.7)	0.500
P2Y12 inhibitor	35 (92.1)	46 (95.8)	0.651
β‐Blocker	20 (52.6)	22 (45.8)	0.682
SGLT2 inhibitor	7 (18.4)	4 (8.3)	0.203
ARNI	6 (15.8)	5 (10.4)	0.526
ACEI/ARB	13 (34.2)	21 (43.8)	0.499
CCB	9 (23.7)	18 (37.5)	0.256
Statin	30 (78.9)	43 (89.6)	0.287
PCSK9 inhibitor	3 (7.9)	6 (12.5)	0.725
Duration of IVUS follow‐up, mo	12.57 (11.36–20.23)	12.77 (11.39–14.91)	0.699
Lesion distribution
LM	5 (6.2)	14 (21.5)	0.013
LAD	55 (67.9)	38 (58.5)	0.315
LCX	4 (4.9)	0 (0.0)	0.129
RCA	17 (21.0)	13 (20.0)	1.000
SYNTAX score	18.78±12.74	25.06±10.83	0.002
Gensini score	27.63±22.01	38.46±18.91	0.002

Data are expressed as mean±SD, median (interquartile range), or number (percentage). ACEI indicates angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor‐neurolysin inhibitor; BMI, body mass index; BP, blood pressure; CCB, calcium channel blocker; eGFR, estimated glomerular filtration rate; hs‐CRP, high‐sensitivity C‐reactive protein; IVUS, intravascular ultrasound; LAD, left anterior descending branch; LCX, left circumflex branch; LM, left main coronary artery; PCSK9, proprotein convertase subtilisin/kexin type 9; RCA, right coronary artery; and SGLT2, sodium‐glucose cotransporter 2.

Cardiometabolic Parameters at Baseline and During Follow‐Up

The levels of cardiometabolic parameters at baseline and during follow‐up were analyzed (Table 2). At baseline, compared with the nonprogression group, the progression group had similar glycemic levels and levels of total cholesterol, LDL‐C, HDL‐C, non–HDL‐C, lipoprotein (a), and apolipoprotein A‐I. By contrast, the levels of triglycerides, remnant lipoprotein particle cholesterol, apoB, apolipoprotein E, and sdLDL‐C were significantly higher in the progression group. Time‐averaged values were then calculated to better reflect the control of these cardiometabolic factors during follow‐up. The time‐averaged levels of non–HDL‐C, remnant lipoprotein particle cholesterol, apoB, and sdLDL‐C (0.74±0.28 mmol/L versus 0.57±0.21 mmol/L, P=0.003) were significantly higher in the progression versus the nonprogression group. In this study population, 63.2% of patients in the nonprogression group and 50% in the progression group achieved LDL‐C control targets (≤1.8 mmol/L, P=0.317). Meanwhile, the SD of sdLDL‐C during follow‐up was also significantly higher in the progression group than in the nonprogression group.

Table 2.

Cardiometabolic Parameters at Baseline and During Follow‐Up

Variables	Baseline	Follow‐up
		Time‐averaged	Mean	SD	CV
Fasting plasma glucose, mmol/L
Nonprogression	6.32±2.29	13.75±32.41	6.35±1.61	1.19±1.70	15.19±16.13
Progression	6.12±1.58	17.96±49.04	6.24±1.36	0.85±0.85	12.48±10.09
P value	0.631	0.649	0.721	0.233	0.348
HbA1c, %
Nonprogression	6.45±1.60	6.25±0.86	6.32±0.96	0.50±0.87	6.85±9.94
Progression	6.43±1.20	6.39±1.11	6.39±1.11	0.40±0.46	5.64±5.53
P value	0.933	0.539	0.757	0.486	0.480
Total cholesterol, mmol/L
Nonprogression	3.77±0.87	3.34±0.72	3.41±0.69	0.47±0.34	14.09±10.78
Progression	4.12±1.25	3.57±0.75	3.62±0.74	0.74±0.60	20.05±15.38
P value	0.143	0.147	0.190	0.015	0.046
Triglycerides, mmol/L
Nonprogression	1.47±0.83	1.33±0.69	1.36±0.68	0.37±0.59	22.00±17.59
Progression	1.82±0.78	1.62±0.67	1.64±0.63	0.43±0.37	25.04±15.36
P value	0.044	0.055	0.051	0.555	0.395
HDL‐C, mmol/L
Nonprogression	1.14±0.31	1.20±0.33	1.19±0.33	0.10±0.05	8.47±4.17
Progression	1.05±0.18	1.11±0.20	1.10±0.19	0.11±0.07	10.17±5.70
P value	0.115	0.135	0.096	0.323	0.127
Non–HDL‐C, mmol/L
Nonprogression	2.63±0.78	2.14±0.58	2.22±0.55	0.49±0.36	22.68±17.90
Progression	3.07±1.21	2.46±0.72	2.52±0.71	0.72±0.60	28.33±21.86
P value	0.056	0.029	0.035	0.040	0.202
LDL‐C, mmol/L
Nonprogression	2.14±0.74	1.70±0.54	1.76±0.51	0.40±0.32	23.25±19.19
Progression	2.42±1.09	1.91±0.63	1.97±0.62	0.64±0.52	31.81±23.25
P value	0.173	0.097	0.109	0.014	0.071
sdLDL‐C, mmol/L
Nonprogression	0.69±0.29	0.57±0.21	0.59±0.20	0.15±0.12	26.34±19.75
Progression	0.87±0.42	0.74±0.28	0.73±0.28	0.24±0.22	32.35±24.08
P value	0.026	0.003	0.007	0.027	0.225
RLP‐C, mmol/L
Nonprogression	0.49±0.29	0.44±0.23	0.45±0.23	0.17±0.18	37.26±23.63
Progression	0.65±0.30	0.55±0.24	0.55±0.23	0.20±0.13	37.44±20.43
P value	0.018	0.040	0.050	0.404	0.970
Lipoprotein (a), g/L
Nonprogression	0.23±0.22	0.23±0.23	0.23±0.23	0.05±0.04	31.36±23.81
Progression	0.29±0.33	0.29±0.33	0.29±0.33	0.05±0.05	28.02±22.03
P value	0.357	0.370	0.402	0.673	0.505
Apolipoprotein A‐I, g/L
Nonprogression	1.25±0.21	1.28±0.26	1.28±0.24	0.12±0.07	8.73±4.88
Progression	1.23±0.17	1.26±0.17	1.26±0.16	0.11±0.07	8.27±5.31
P value	0.727	0.688	0.554	0.560	0.685
Apolipoprotein B, g/L
Nonprogression	0.72±0.17	0.64±0.14	0.65±0.13	0.11±0.07	16.59±11.40
Progression	0.83±0.25	0.73±0.18	0.73±0.17	0.16±0.13	21.49±17.56
P value	0.028	0.016	0.020	0.036	0.145
Apolipoprotein E, mg/L
Nonprogression	3.73±0.82	3.59±0.70	3.64±0.76	0.43±0.34	11.94±9.50
Progression	4.12±0.91	3.78±0.59	3.81±0.60	0.57±0.43	14.88±10.08
P value	0.042	0.183	0.260	0.102	0.175

Data are expressed as mean±SD. CV indicates coefficient of variation; HbA_1c, glycated hemoglobin; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol; RLP‐C, remnant lipoprotein particle cholesterol; and sdLDL‐C, small dense low‐density lipoprotein cholesterol.

IVUS Measurements at Baseline and During Follow‐Up

We analyzed all of the lesions by IVUS at baseline and follow‐up (Table 3). During follow‐up, the mean changes in atheroma volumes were 4.1%±10.2% of PAV and 4.3±11.8 mm³ of TAV. PP was detected in 65 lesions (44.5%), with 12 lesions (18.5%) in 9 patients (18.8%) progressing to require stenting during follow‐up. Time‐averaged sdLDL‐C was significantly correlated to changes in TAV (Pearson r=0.29, P=0.006) (Figure 2A). Furthermore, subgroup analysis (Figure 2B) showed that this correlation was significant in diabetic (Pearson r=0.58, P<0.001) but not in nondiabetic (P=0.393) patients.

Table 3.

IVUS Measurements at Baseline and During Follow‐Up

IVUS parameters	Nonprogression	Progression	P value
	(n=81)	(n=65)
Lesion length, mm	7.07±3.33	6.79±3.07	0.593
Reference lumen area, mm2
BL	10.61±4.43	12.17±5.89	0.070
FU	9.91±3.89	9.62±4.88	0.697
Δ	−0.70±2.00	−2.55±2.66	<0.001
Reference EEM CSA, mm2
BL	15.23±5.26	15.82±6.42	0.536
FU	14.36±4.99	14.23±5.40	0.881
Δ	−0.87±1.61	−1.60±2.34	0.028
Reference lumen diameter, mm
BL	4.35±0.77	4.39±0.94	0.773
FU	4.21±0.74	4.14±0.89	0.591
Δ	−0.14±0.26	−0.26±0.38	0.032
MLA, mm2
BL	6.66±3.00	7.80±4.16	0.056
FU	6.43±3.03	5.74±2.70	0.153
Δ	−0.23±1.18	−2.06±2.20	<0.001
EEM CSA at MLA, mm2
BL	14.28±4.86	14.46±5.95	0.847
FU	13.36±4.87	13.36±4.99	0.992
Δ	−0.93±1.69	−1.09±2.44	0.632
MLD, mm
BL	2.85±0.64	3.09±0.86	0.057
FU	3.67±7.85	2.63±0.62	0.288
Δ	0.82±7.79	−0.46±0.43	0.188
TAV, mm3
BL	50.45±33.70	35.97±23.52	0.004
FU	48.01±34.43	48.57±26.41	0.914
Δ	−2.44±8.90	12.60±9.34	<0.001
Normalized TAV, mm3
BL	43.46±18.31	34.13±18.25	0.003
FU	40.86±19.91	47.29±21.65	0.064
Δ	−2.60±8.02	13.15±10.08	<0.001
PAV, %
BL	44.51±11.56	35.70±13.68	<0.001
FU	41.63±12.26	48.54±13.57	0.002
Δ	−2.88±6.03	12.84±7.13	<0.001
Remodeling index
BL	0.95±0.13	0.92±0.10	0.160
FU	0.93±0.11	0.95±0.12	0.360
Δ	−0.01±0.12	0.03±0.13	0.030

Data are expressed as mean±SD. BL indicates baseline; EEM CSA, external elastic membrane cross‐sectional area; FU, follow‐up; IVUS, intravascular ultrasound; MLA, minimal lumen area; MLD, minimal lumen diameter; PAV, percent atheroma volume; TAV, total atheroma volume; and Δ, changes in corresponding parameters.

Figure 2. Correlation between time‐averaged sdLDL‐C and changes in TAV.

Correlations between time‐averaged sdLDL‐C and changes in TAV during follow‐up in the overall population (A) and in the diabetic and nondiabetic subgroups (B). The dashed lines indicate the predictive changes in TAV by time‐averaged sdLDL‐C using linear regression analysis. sdLDL‐C indicates small dense low‐density lipoprotein cholesterol; and TAV, total atheroma volume.

In the progression lesions, baseline plaque burden was lower and reference lumen area was similar compared with the nonprogression lesions. During follow‐up, progression lesions exhibited more reduced reference lumen area and diameter, as well as reference EEM CSA. The decrease in minimal lumen area (−2.06±2.20 mm² versus −0.23±1.18 mm², P<0.001) was significantly greater in the progression versus the nonprogression lesions. Changes in TAV were −2.44±8.90 mm³ and 12.60±9.34 mm³ (P<0.001) and changes in PAV were −2.88%±6.03% and 12.84%±7.13% (P<0.001) in the nonprogression and progression lesions, respectively. A trend towards positive remodeling was also observed in the progressed lesions (P=0.030).

Univariate and Multivariate Analyses

Univariate analysis (Table S1) demonstrated that PP was associated with lower baseline plaque burden, represented by TAV (odds ratio [OR], 0.982 [95% CI, 0.969–0.994]), normalized TAV (OR, 0.971 [95% CI, 0.951–0.990]) and PAV (OR, 0.946 [95% CI, 0.919–0.972]), and higher time‐averaged levels of apoB (OR, 11.357 [95% CI, 1.406–102.908]) and sdLDL‐C (OR, 1.191 [95% CI, 1.048–1.367]).

Per‐lesion multivariate analysis (Table 4) showed that time‐averaged sdLDL‐C level was associated with PP after adjusting for age and sex. After further adjustments for conventional risk factors, lesion characteristics, and other significant predictors from the univariate analysis, time‐averaged sdLDL‐C remained significantly associated with PP (OR, 1.218 [95% CI, 1.035–1.449], per 0.1 mmol/L increase). This association remained to be significant after additional adjustment for time‐averaged level of LDL‐C, but no longer existed after controlling for time‐averaged level of apoB (Table S2). Per‐patient analysis exhibited similar findings (Table S3). In addition, subgroup analysis showed that this association was not affected by diabetic conditions (P for interaction=0.448).

Table 4.

Per‐Lesion Multivariate Analysis

Models	OR (95% CI)	P value
Model 1	1.184 (1.038–1.363)	0.014
Model 2	1.182 (1.018–1.384)	0.032
Model 3	1.218 (1.035–1.449)	0.021

OR indicates odds ratio. Model 1, after adjusting for age and sex. Model 2, model 1 with additional adjustment for hypertension, diabetes, high‐sensitivity C‐reactive protein, time‐averaged level of triglycerides, and use of aspirin. Model 3, model 2 with additional adjustment for reference lumen area, minimal lumen area, normalized total atheroma volume, and percent atheroma volume.

DISCUSSION

The major finding of our study is that patients with CAD who have higher levels of sdLDL‐C tend to have a higher incidence of PP compared with those who have lower levels of sdLDL‐C during follow‐up, even after controlling for LDL‐C and other conventional risk factors. SdLDL‐C is an independent risk factor of PP in patients with CAD.

Existing evidence indicates that coronary PP is closely related to MACE in patients with CAD. ³⁰ , ³¹ Nicholls et al found that each SD increase in PAV quantified by IVUS was associated with a 1.20‐fold (95% CI, 1.10–1.31; P<0.001) greater risk of MACE. ³⁰ Likewise, in a previous study of 1166 patients with CAD from the multicenter PARADIGM (Progression of Atherosclerotic Plaque Determined by Computed Tomographic Angiography Imaging) registry, every 1‐SD annual increase of PAV measured by coronary computed tomography angiography was independently associated with a 1.23‐fold (95% CI, 1.08–1.39) greater risk of MACE during a median of 8.2 years. ³¹ Another small‐scale coronary computed tomography angiography study also showed that PP was associated with a significantly increased risk of MACE after adjusting for clinical and imaging risk factors. ³² These studies suggest that PP is a surrogate end point of MACE, and therapies aiming at mitigating PP should provide favorable effects on cardiovascular outcomes in patients with CAD.

Among the traditional risk factors, substantial evidence revealed that LDL‐C level has a direct link to volumetric changes in atheroma. ¹ , ³³ Treatment by lowering LDL‐C is considered the most effective measure to halt PP in current clinical practice. However, PP occurs in ≈20% of patients with CAD, even with controlled LDL‐C levels. These findings indicate that residual risk of PP exists despite seemingly optimal control of LDL‐C, which remains to be further investigated. ³

LDL particles represent a heterogeneous group composed of lipoproteins of different size, density, lipid content, and composition. Actually, in many cases, the level of LDL particles is discordant with that of LDL‐C. Otvos et al found that the number of LDL particles can better represent the LDL‐attributable atherosclerotic risk than LDL‐C. ³⁴ In patients with metabolic syndrome or diabetes, this discordance between LDL particles and LDL‐C becomes more apparent, characterized by a higher proportion of sdLDL with smaller particle size and higher density. Basic research and clinical studies have revealed that sdLDL is a more atherogenic subfraction of LDL. ³⁵ , ³⁶ Elevated plasma sdLDL‐C was associated with increased risk of MACE in patients with diabetes and CAD. ³⁷ , ³⁸ Furthermore, in individuals with normal levels of LDL‐C and non–HDL‐C, increased sdLDL‐C was also significantly related to incident carotid plaques. ³⁹ , ⁴⁰ However, the relationship between sdLDL‐C and PP remains unclear.

In this study, we for the first time analyzed sdLDL‐C control level and PP determined by IVUS examination. Time‐averaged cardiometabolic factors during follow‐up were calculated to eliminate errors resulting from heterogeneous follow‐up times. In our cohort, more than half of patients achieved LDL‐C control targets during follow‐up in both groups. We showed that the time‐averaged level of sdLDL‐C in the progression group was significantly higher compared with that in the nonprogression group. The level of time‐averaged sdLDL‐C was significantly correlated with changes in TAV, especially in patients with type 2 diabetes. Furthermore, the time‐averaged level of sdLDL‐C was independently associated with PP, even after adjusting for conventional risk factors including LDL‐C and lesion characteristics. Each 0.1‐mmol/L increase in sdLDL‐C levels was significantly associated with a 1.2‐fold increased risk of PP. Interestingly, this association no longer existed after further adjustment for time‐averaged level of apoB, suggesting that apoB seems to be better reflective of the overall atherogenic potential of the lipid profile, which may overshadow the association between sdLDL‐C and PP. Consistent with previous reports, ³ our univariate analysis showed that lower plaque burden was associated with PP, suggesting that evaluation and control of sdLDL‐C should also be considered even in patients with CAD who have moderate coronary artery stenosis.

The possible mechanisms are as follows. First, sdLDL is of smaller particle size, which allows it to more easily transmigrate across the endothelial monolayer and accumulate in the tunica intima. ⁴¹ Second, sdLDL has a lower affinity for LDL receptors while it has a higher affinity for arterial proteoglycans, thereby promoting its retention and the pathological process of PP. ⁴² Third, sdLDL‐C particles are more susceptible to oxidative modification and more likely to be phagocytosed by macrophages to form foam cells in the atheroma. ⁴³ , ⁴⁴ Furthermore, for patients with diabetes and hypertriglyceridemia, lipid exchanges between lipoproteins may further increase sdLDL‐C levels and thus PP. ⁴⁵

Interestingly, although sdLDL‐C appeared to have a stronger correlation with changes in TAV in diabetic than nondiabetic patients with CAD, the multivariate analysis showed that the interaction term between the diabetic condition and sdLDL‐C levels in association with PP was nonsignificant, suggesting that although diabetic patients might be more susceptible to PP under exposure to sdLDL‐C, the role of sdLDL‐C seems to be equally important in promoting PP in diabetic and nondiabetic patients.

Our findings should be interpreted in the context of the following limitations. First, as a retrospective study, all of the enrolled patients were from a single center. Second, the sample size was relatively small because of the stringent inclusion criteria. Third, the cardiometabolic factors were not measured at the same time intervals during follow‐up; however, we calculated the time‐averaged parameters to eliminate errors resulting from heterogenous follow‐up times to a certain extent.

CONCLUSIONS

Our findings suggest that elevated sdLDL‐C is independently associated with PP in patients with CAD. Intensive control of sdLDL‐C along with other risk factors by lifestyle intervention and optimized medication therapies may maximally inhibit PP, thereby leading to a reduction in cardiovascular events.

Sources of Funding

This work was supported by the National Natural Science Foundation of China (grant number 82170423), the Opening Project Program of the National Research Center for Translational Medicine at Shanghai (grant number NRCTM(SH)‐2023‐13), National Key Research and Development Program of China (grant number 2022YFC2503502), Shanghai Jiao‐Tong University School of Medicine—Gaofeng Clinical Medicine Grant Support (grant number 20240801), and Translational Research of Novel Medical Techniques Seed Program by Shanghai Municipal Health Commission (grant number 2024ZZ2042). CY and XW performed the study design, data analysis, and data interpretation. AA and YL performed the manuscript writing. AA, YL, HY, SF, GT, ST, XW, and LT performed the data collection. RZ, LL, and XW performed the manuscript revision. All authors reviewed the article.

Disclosures

None.

Supporting information

Tables S1–S3