Association of Lp(a) with coronary plaque burden and high-risk plaque features: A meta-analysis of imaging studies

Abstract

Background and aims:

Lipoprotein(a) [Lp(a)] is a causal risk factor for cardiovascular disease, but its impact on long-term coronary plaque progression remains unclear. This study synthesizes evidence from CCTA, IVUS, and OCT to clarify the relationship between high-risk Lp(a) and coronary plaque burden and high-risk plaque features.

Methods:

We conducted a comprehensive search of multiple databases up to July 2025 for studies evaluating Lp (a) and atherosclerotic plaque progression. Statistical analysis was performed using a random-effects model in RevMan 5.4, reporting odds ratios (OR) and mean differences (MD) with 95% confidence intervals (CI). The protocol is registered in PROSPERO (CRD420251113955).

Results:

Our final analysis included 16 studies comprising 19,822 participants with a mean age of 62 years and a median imaging follow-up ranging from 10 months to 10.2 years. On analysis, high-risk Lp(a) levels were significantly associated with the presence of coronary plaque (OR 1.53; 95% CI, 1.03–2.29; p = 0.04) compared with low Lp(a) levels. Additionally, patients with elevated Lp(a) exhibited significantly greater progression in percent atheroma volume (ΔPAV) than those with low levels (MD 4.31%; 95% CI, 1.08–7.53; p = 0.009). Subgroup analysis by plaque phenotype revealed a statistically significant increase in low-attenuation plaque (LAP) presence among individuals in the high-risk Lp(a) category (OR 1.92; 95% CI, 1.13–3.27; p = 0.02).

Conclusion:

High-risk Lp(a) is associated with greater coronary plaque prevalence, accelerated progression, and increased LAP. These findings underscore Lp(a) as a driver of high-risk, rupture-prone plaques and a critical biomarker and potential therapeutic target in cardiovascular risk management.

1. Introduction

Lipoprotein(a) [Lp(a)] is a genetically mediated risk factor for atherosclerotic cardiovascular disease. Extensive evidence from genetic and observational studies involving over 500,000 individuals indicates that more than 90% of circulating Lp(a) levels are genetically determined [1]. Elevated Lp(a) has been strongly linked to the development and progression of atherosclerotic plaque. Emerging data associate high Lp(a) with high-risk plaque features, also referred to as vulnerable plaques, such as thin-cap fibroatheroma and increased low-attenuation plaque (LAP), highlighting its role in promoting plaque instability and atherothrombotic events [1].

High-risk plaque features, including large necrotic lipid pools, thin fibrous caps, and high plaque burden, are linked to elevated Lp(a) and strongly predict sudden cardiac death and acute coronary syndromes [2, 3]. Coronary computed tomography angiography (CCTA) and intravascular ultrasound (IVUS) detect HRP features such as positive remodeling, low CT attenuation, thin-cap fibroatheroma, and plaque burden >70%, which are associated with a 2.5- to 8-fold higher risk of major adverse cardiovascular events [2,3]. These findings position high-risk plaque as a key mediator of adverse cardiovascular outcomes and a critical imaging biomarker of residual risk.

Additionally, Lp(a) activates monocytes, amplifying vascular inflammation and thrombosis, suggesting an immunothrombotic mechanism contributing to residual cardiovascular risk, even in patients optimized on lipid-lowering therapy [4]. Given these multifaceted pathogenic effects, early identification of elevated Lp(a) is critical to enable more intensive cardiovascular risk reduction through lifestyle and pharmacologic interventions [5,6]. While Lp(a) has been associated with high-risk plaque phenotypes, its impact on long-term coronary plaque progression remains unclear. Existing imaging studies using modalities such as CCTA, IVUS, and optical coherence tomography (OCT) are relatively small and heterogeneous, underscoring the need for a meta-analysis to synthesize the available evidence and clarify the relationship between Lp(a) levels, plaque progression, and high-risk plaque features, thereby better defining its contribution to residual cardiovascular risk.

2. Methods

This study was performed following the guidelines summarized in the Cochrane Handbook for Systematic Reviews of Interventions and the reporting standards set by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [7,8] and was registered with PROSPERO (CRD420251113955).

2.1. Search strategy

We conducted a comprehensive systematic search across multiple databases, including MEDLINE (via PubMed), Scopus, ScienceDirect, the Cochrane Library, and ClinicalTrials.gov, to identify studies published from January 2000 up to July 2025 examining the association between lipoprotein(a) [Lp(a)], atherosclerotic plaque progression, and plaque phenotype. Reference lists of relevant articles were manually screened to capture additional studies potentially missed in the initial search. The search strategy incorporated a combination of keywords and Medical Subject Headings (MeSH) terms related to “lipoprotein (a)”, “LP (a)”, “coronary artery disease”, “plaque”, “coronary plaque”, atherosclerosis, “plaque phenotype”, “plaque burden”, “high risk plaque”, “low-density plaque”, “high-density plaque”, “thin-cap fibroatheroma”, “plaque progression”, “calcified plaque”, “noncalcified plaque”, “atheroma volume”, “coronary computed tomography angiography”, “optical coherence tomography”, “intravascular ultrasound”, “atherosclerotic cardiovascular disease”, “observational studies”, “meta-analysis”, “randomized controlled trials”, and “RCTs”. The detailed search strategy is provided in Supplementary Table 1.

2.2. Study selection

We included studies evaluating the relationship between Lp(a) levels and coronary artery plaque descriptors, including plaque burden and high-risk features including LAP.

High-risk Lp(a) concentrations were defined as ≥125 nmol/L (≥50 mg/dL), in accordance with existing guidelines [9]. Eligible studies utilized non-invasive CCTA or intravascular imaging modalities (IVUS or OCT) to assess plaque characteristics. Both randomized controlled trials and observational studies were considered. We excluded studies focused solely on carotid or peripheral artery plaque, single-case reports, reviews, conference abstracts, animal studies, studies not published in English, and studies that did not report the predefined outcomes.

A comprehensive search across multiple databases yielded 369 articles: 113 from PubMed, 56 from Scopus, 184 from ScienceDirect, 15 from the Cochrane Library, and 1 from ClinicalTrials.gov. Following removal of duplicates, all records underwent a structured two-stage screening process using Rayyan software. Two independent reviewers (SAS and TB) performed the screening, first by evaluating titles and abstracts, then by reviewing full texts based on predefined inclusion and exclusion criteria. Discrepancies, mainly concerning study eligibility and relevance to the predefined criteria, were resolved through discussion, with arbitration by a senior author (ST) when necessary. Inter-rater reliability was assessed using Cohen's Kappa, which demonstrated almost perfect agreement (κ = 0.93) [10]. After title and abstract screening of 358 unique records, 66 studies were selected for full-text review (Fig. 1).

Fig. 1.

PRISMA flowchart illustrating the study selection process for the meta-analysis. The diagram depicts the number of records identified through database and other sources, the number of records screened, assessed for eligibility, and the final number of studies included in the analysis, with reasons for exclusion at each stage.

2.3. Main outcomes

The primary imaging outcomes included the presence of coronary plaque and change in percent atheroma volume (ΔPAV). Subgroup analyses were conducted according to plaque phenotype, primarily low-attenuation plaque (LAP), and by imaging modality (CCTA vs IVUS and OCT).

2.4. Data extraction

Following the completion of the study selection process, two independent reviewers (SAS and TB) extracted relevant information from the 16 studies included in the review. This included key study characteristics such as authorship, publication year, and study design; demographic data including participant age, sample size, number of patients with high-risk and low Lp(a) levels, median Lp(a) concentration, and coronary artery calcium score (if reported); as well as median follow-up duration between imaging and the type of imaging used. In addition, baseline population criteria and study outcomes were documented.

2.5. Risk of bias

For observational studies, we employed the revised Cochrane Risk of Bias tool for Non-Randomized Studies of Exposures (ROBINS-E) [11]. To evaluate the methodological quality of the included randomized controlled trials, we used the Cochrane Risk of Bias 2 (RoB 2) tool [12], which assesses five domains: (1) bias arising from the randomization process; (2) bias due to deviations from intended interventions; (3) bias due to missing outcome data; (4) bias in the measurement of outcomes; and (5) bias in the selection of the reported result. Each domain was assessed to determine the overall risk of bias for each study, classified as low, some concerns, or high. Two independent reviewers (SAS and TB) conducted the assessments. Discrepancies between reviewers were resolved through discussion with the senior author (ST).

2.6. Data synthesis and statistical analysis

Meta-analyses were performed using Cochrane Review Manager (RevMan, version 5.4) [13]. ¹³ For continuous outcomes, mean differences (MDs) were calculated, while odds ratios (ORs) were used for dichotomous outcomes. All analyses employed an inverse variance random-effects model to account for both within-study and between-study variability. Pooled estimates were reported with corresponding Z-values, 95% confidence intervals (CIs), and p-values. Statistical significance was set at p < 0.05. Between-study heterogeneity was assessed using Higgins's I² statistic and classified as low (<25%), moderate (25–50%), or high (>50%) [14].

We assessed the certainty of evidence for the primary outcomes using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach, which evaluates evidence quality across five domains: risk of bias, inconsistency, indirectness, imprecision, and other considerations (such as potential publication bias) [15]. The certainty of evidence was categorized as high, moderate, low, or very low. Initial GRADE assessments were conducted by one reviewer (SAS) and independently verified by a second reviewer (TB), with discrepancies resolved through discussion and consensus (Supplementary Table 2). Publication bias was not assessed because fewer than 10 studies were included in each analysis. Nevertheless, a comprehensive search strategy was employed to minimize the risk of selective non-publication bias.

3. Results

3.1. Included studies and characteristics

After excluding studies that did not meet our selection criteria, our final review included 16 studies [16–31], comprising 2 randomized controlled trials [14,17,23] observational studies, with a total of 19,822 participants (mean age 62 years) and a median imaging follow-up ranging from 10 months to 10.2 years. A summary of the baseline characteristics of the included studies is provided in Table 1.

Table 1.

Baseline characteristics of the included studies.

Author & Country	Type of study	Sample size [High Lp(a) vs low Lp(a)]	Mean age (Years)	Imaging Modality	Study population criteria	Outcomes
Nurmohamed et al., 2024 [16]	Cohort	267 (61/206)	57.1	CCTA	No history of CAD at baseline. Median Lp(a) 25 nmol/L. Median imaging interval: 10.2 y.	ΔPAV, % noncalcified plaque volume; secondary: calcified plaque volume, LAP, pericoronary attenuation.
Kaiser et al., 2022 DIAMOND [17]	RCT	191 (43/148)	65.9	CCTA	Advanced stable multivessel CAD. Median Lp(a) 15 mg/dL. Imaging interval: 12 months. Mean CAC (high Lp(a)): 378.	LAP progression; total, calcific, noncalcific plaque progression.
Lan et al., 2025 [18]	Cohort	1694	62	CCTA	Diabetes, intermediate CAD risk. Median interval: 2.3 y.	Δ total plaque, noncalcified plaque, LAP, fibro-fatty and fibrotic plaque.
Mszar et al., 2024 Miami Heart [19]	Cross-sectional	1795 (291/1504)	52	CCTA	Adults 40–65 in Miami; excluded statin users/missing Lp(a). Median Lp(a) 35 nmol/L. CAC >0 in 28.2% high Lp(a).	Presence of coronary plaque.
Yu et al., 2024 [20]	Cohort	5607 (194/928)	54.5	CCTA	Stable chest pain, suspected CAD. Median Lp(a) 36.5 mg/dL. Follow-up: 8.2 y. Median CAC (with LAP): 82.3.	Low-attenuation plaque, Risk of MI.
Dai et al., 2022 [21]	Cohort	377 (62/315)	63.6	CCTA	Stable CAD patients. Median Lp(a) 15.1 mg/dL.	Total plaque volume; MACE incidence.
Fathieh et al., 2025 [22]	Cohort	1718	62	CCTA	No history of MI or revascularization. Mean Lp(a) 22.5 nmol/L.	Mixed plaque burden.
O’Toole et al., 2024 PROMISE [23]	RCT	1815 (405/1410)	60.2	CCTA	Men >54 y, women >64 y, or younger with ≥1 ASCVD risk factor. Mean Lp(a) 31.1 mg/dL.	Obstructive CAD (stenosis ≥50%, ≥ 70%) and high-risk plaque.
Leistner et al., 2024 [24]	Cross-sectional	975 (264/711)	69.5	Coronary angiography	Adults ≥21 y undergoing angiography. Median Lp(a) 19.3 nmol/L.	Total plaque volume; MI prevalence.
Kato et al., 2022 [25]	Cross-sectional	185 (49/136)	70	OCT	ACS patients undergoing OCT. Median Lp (a) 15 mg/dL.	Prevalence of TCFA.
Wang et al., 2025 [26]	Cohort	177 (51/126)	68.3	OCT	ACS patients with OCT-guided PCI; follow-up OCT non-culprit lesions.	Total atheroma volume; TCFA incidence.
Niccoli et al., 2016 [27]	Cohort	51 (21/30)	65	OCT	ACS patients with obstructive CAD at angiography.	Lipidic plaque and TCFA prevalence.
Matsushita et al., 2020 Yokohama-ACS [28]	Cohort	102 (27/49)	65	IVUS	ACS patients ≥20 y with dyslipidemia after PCI. Median Lp(a) 15 mg/dL. Imaging interval: 10 months.	Plaque progression; MACE (death, MI, revascularization).
Erlinge et al., 2025 PROSPECT II substudy [29]	Cohort	865 (21/760)	61.3	IVUS	Patients with h/o recent MI. Mean Lp(a) 79.2 nmol/L.	Pancoronary plaque volume, lipid core burden, focal vulnerable plaques.
Huded et al., 2020 [30]	Cohort	3943 (683/3260)	58	IVUS	Median Lp(a) 14.8 mg/dL.	Baseline % atheroma volume.
Hartmann et al., 2006 [31]	Cohort	60	58	IVUS	Patients with established CAD. Median Lp (a) 25 ± 23 mg/dL. IVUS follow-up: 18 months.	Plaque progression, adverse cardiovascular events.

^aSample size is presented as total participants or as high Lp(a)/low Lp(a) groups when available.

^bLp(a) indicates lipoprotein(a); CAD, coronary artery disease; CCTA, coronary computed tomography angiography; IVUS, intravascular ultrasound; OCT, optical coherence tomography; ACS, acute coronary syndrome; MI, myocardial infarction; PCI, percutaneous coronary intervention; CAC, coronary artery calcium; LAP, low-attenuation plaque; TCFA, thin-cap fibroatheroma; ΔPAV, change in percent atheroma volume; MACE, major adverse cardiovascular events.

3.2. Quality assessment

The risk of bias assessment for the included studies predominantly indicated some concerns, with variability across different risk levels (Fig. 2). Most studies were rated as having some concerns, while a smaller number were classified as either high or low risk of bias. Confounding bias was an inherent limitation of the predominantly observational designs. Frequent concerns included exposure and outcome measurement and recruitment-related biases, whereas the principal issues in the randomized controlled trials were bias in outcome assessment.

Fig. 2.

Risk-of-bias assessment of included studies. Randomized controlled trials (RCTs) were evaluated using the RoB 2 tool, and observational studies were assessed using ROBINS-E. The figure summarizes the proportion of studies rated as low, some concerns, or high risk of bias across the relevant domains, providing an overview of the methodological quality of the evidence included in the meta-analysis.

3.3. Analysis of outcomes

In the meta-analysis, high-risk Lp(a) was significantly associated with the presence of coronary plaque, with 53% higher odds compared to low Lp(a) levels (OR 1.53; 95% CI, 1.03–2.29; p = 0.04), however, there was substantial heterogeneity across studies (I² = 96%) [Fig. 3]. Subgroup analysis by plaque phenotype showed high-risk Lp(a) to be significantly associated with LAP, one marker of high-risk coronary plaque (OR 1.92; 95% CI, 1.13–3.27; p = 0.02), with moderate heterogeneity (I² = 95%) [Fig. 3]. Furthermore, stratification by imaging modality revealed that studies utilizing non-invasive (CCTA) demonstrated a stronger and statistically significant association between high Lp(a) levels and the presence of coronary plaque (OR 1.63; 95% CI, 1.12–2.37; p = 0.01; I² = 94%), whereas invasive imaging modalities (coronary angiography with IVUS or OCT) showed a weaker and non-significant association (OR 1.31; 95% CI, 0.97–1.78; p = 0.08; I² = 54%) [Fig. 4]. However, the invasive imaging subgroup included only two studies, limiting statistical power and potentially influencing the precision of the pooled estimate. This limited power should be considered when interpreting modality-specific differences [Fig. 4]. Additionally, substantial heterogeneity observed among CCTA studies warrants cautious interpretation of these findings.

Fig. 3.

Forest plot of the association between high Lp(a) levels and coronary plaque. High Lp(a) was associated with increased odds of coronary plaque (OR 1.53; 95% CI, 1.03–2.29; p = 0.04; I² = 96%). Subgroup analysis by plaque phenotype demonstrated that high Lp(a) was significantly associated with low-attenuation plaque (LAP) (OR 1.92; 95% CI, 1.13–3.27; p = 0.02; I² = 95%). OR: Odds ratio, IV: Inverse variance; SE: Standard error.

Fig. 4.

Forest plot of coronary plaque presence stratified by imaging modality. Non-invasive imaging (CCTA) showed a significant association with high Lp(a) (OR 1.63; 95% CI, 1.12–2.37; p = 0.01; I² = 94%), whereas invasive imaging (coronary angiography with IVUS/OCT) showed a non-significant association (OR 1.31; 95% CI, 0.97–1.78; p = 0.08; I² = 54%). OR: Odds ratio, IV: Inverse variance; SE: Standard error.

In the analysis of ΔPAV progression, high-risk Lp(a) was significantly associated with increased plaque progression over follow-up periods ranging from 10 months to 10.2 years, compared to low Lp(a) levels, with a mean difference (MD) of 4.31% (95% CI, 1.08–7.53; p = 0.009) [Fig. 5]. Moderate heterogeneity was observed among studies (I² = 71%), reflecting variability in effect estimates. However, the considerable heterogeneity observed across studies (I² = 96%) necessitates cautious interpretation of these findings.

Fig. 5.

Forest plot showing the effect of high Lp(a) on change in percent atheroma volume (ΔPAV). High Lp(a) was associated with a mean increase of 4.31% (MD 4.31; 95% CI, 1.08–7.53; p = 0.009; I² = 71%). MD: Mean difference; IV: Inverse variance; SE: Standard error.

3.4. Sensitivity analysis

To address heterogeneity, a sensitivity analysis using the leave-one-out method was conducted. In the analysis of five studies assessing the presence of coronary plaque, exclusion of a single study [20] reduced heterogeneity from 96% to 0% (I²), identifying this study as the primary source of variability. Following exclusion, the recalculated OR was 1.26 (95% CI, 1.14–1.39; p < 0.001), confirming the robustness of the association (Supplementary Fig. 1).

Yu et al. [20] may have contributed disproportionately to heterogeneity due to distinct cohort characteristics, including a unique ethnic composition with differing Lp(a) distributions and a focus on patients with stable chest pain rather than broader clinical populations. These factors may influence baseline plaque prevalence and the magnitude of association between Lp(a) and plaque burden, underscoring the need for cautious interpretation when pooling heterogeneous populations. Despite substantial statistical heterogeneity, the association between elevated Lp(a) levels and coronary plaque remained directionally consistent across studies, supporting the validity of the observed relationship.

Similarly, in the analysis of ΔPAV progression, exclusion of one study [30] reduced heterogeneity from 71% to 10% (Supplementary Fig. 2), further suggesting that individual study characteristics contributed substantially to the observed variability.

4. Discussion

Our findings demonstrate that high-risk Lp(a) levels are significantly associated with increased odds of coronary plaque presence compared to low Lp(a) levels. Subgroup analyses by plaque phenotype further reveal a significant association between high-risk Lp(a) and LAP. Furthermore, high-risk Lp(a) is significantly correlated with accelerated plaque progression. Together, these results indicate that elevated Lp(a) contributes to both increased overall coronary plaque burden and a higher prevalence of vulnerable plaque characteristics [Fig. 6].

Fig. 6.

Graphical abstract summarizing the association between elevated Lp(a) levels and coronary atherosclerosis, including coronary plaque presence, high-risk plaque features, and plaque progression. OR: Odds ratio; MD: Mean difference.

While previous meta-analyses have focused on Lp(a) and coronary artery calcium or premature atherosclerotic cardiovascular disease risk [32,33], no prior study has systematically synthesized imaging data on high-risk Lp(a) in relation to coronary plaque progression and high-risk plaque features. Our meta-analysis addresses this gap, integrating evidence from multiple imaging modalities to provide a comprehensive overview of how elevated Lp(a) relates to plaque burden and HRP features. Notably, none of the studies included in the present analysis overlapped with those included in previous meta-analyses, as earlier studies primarily evaluated different endpoints rather than detailed coronary plaque characteristics.

To contextualize our meta-analysis findings, individual studies have provided detailed insights into how elevated Lp(a) affects coronary plaque phenotypes. PROMISE trial demonstrated that Lp(a) ≥50 mg/dL has been linked to accelerated coronary plaque burden, increased low-density noncalcified plaque, and peri-coronary adipose tissue inflammation. Observational data from PROSPECT II study indicate that elevated Lp(a) is associated with vulnerable plaques, pancoronary atherosclerosis, and higher lipid burden, including focal plaques with high plaque burden and lipid core index [29]. Consistent with these observations, Nurmohamed et al. reported that each doubling of Lp(a) was associated with the presence of low-density plaque at baseline (OR 1.23; p = 0.05) [16]. At follow-up, high Lp(a) was linked to low-density plaque (OR 1.22; p = 0.01) and increased pericoronary adipose tissue inflammation (OR 1.24; p = 0.004) [16]. In a recent cohort study, Yu et al. further demonstrated that elevated Lp(a) significantly increased myocardial infarction risk, largely mediated by the presence of LAP detected on CCTA [20].

In terms of imaging, OCT has better resolution than CCTA for the detection of plaque microstructure and morphology, but is limited by lesion-specific use, whereas CCTA can provide whole coronary tree assessment of disease, quantify plaque volume, and provide plaque phenotyping [34–37]. High risk plaque features on CCTA have been standardized as LAP (LAP ≤30 hounsfield units), positive remodeling (≥1.1 remodeling index), spotty calcification ≤ 3 mm, and napkin-ring sign, which enables consistent reporting of clinically relevant plaque modifiers [38]. Across multiple validation studies, CCTA-derived high-risk plaque features show strong correlation with invasive markers of lesion vulnerability, including necrotic core, lipid-rich plaque, and thin-cap fibroatheroma [39–41]. These phenotypes have shown prognostic signals in large cohorts, with a fourfold higher risk of major adverse cardiovascular events in patients with at least one high-risk plaque feature compared to those with none [36]. Importantly, some features correlate with Lp(a) and lesion-specific metrics at greater rates. In one large cohort study, LAP appeared to mediate a substantial proportion (~73%) of the Lp(a)-myocardial infarction relationship and LAP-positive individuals had a three-fold greater risk of MI [20]. CCTA's ability to track plaque progression longitudinally, linking polygenic risk scoring to accelerated growth in total plaque and LAP volumes, as well as good correlation to newer forms of invasive lipid burden assessment via near infrared spectroscopy (NIRS-IVUS) reinforces its ability to flag lesions linked to biologic vulnerability [16,42]. Further integration of perivascular inflammatory metrics and other modalities such as FDG PET may be able to further stratify HRP-related risk in the future [43]. The studies included in our meta-analysis indicate that both CCTA and invasive modalities can provide information on plaque characteristics. Each modality offers distinct perspectives, but a direct comparison of their accuracy or correlation was beyond the scope of this review.

These results underscore the contributory role of Lp(a) in the pathogenesis and progression of high-risk atherosclerotic plaque phenotypes, highlighting its importance as a biomarker for cardiovascular risk. Lp(a) levels vary by ethnicity, and larger studies are needed to further characterize its relationship with plaque characteristics. Incorporating Lp(a) measurements alongside CCTA-derived high-risk plaque metrics could enable more precise risk assessment and guide individualized management. Integrating Lp(a) into existing atherosclerotic cardiovascular disease risk calculators may improve identification of patients at heightened risk who might be underestimated by traditional scores. Furthermore, as novel Lp(a)-lowering therapies become available, prospective studies are warranted to evaluate whether targeted reduction of Lp(a) translates into plaque regression and reduced major adverse cardiovascular events, bridging the gap between biomarker discovery and actionable therapeutic intervention.

A key consideration is that most included studies involved clinically selected cohorts rather than population-based samples. Participants ranged from individuals without prior CAD to patients with stable CAD, acute coronary syndromes, or high-risk subgroups undergoing imaging substudies. Follow-up and imaging intervals varied widely (10 months to 10.2 years), and median Lp(a) levels differed across studies. Because imaging indication, disease stage, and baseline cardiovascular risk strongly influence plaque prevalence and phenotype, the pooled estimates may not fully generalize to asymptomatic or broader populations. While the observed associations between elevated Lp(a) and high-risk plaque features are robust in clinically selected cohorts, caution is warranted when extrapolating these findings to general or asymptomatic populations.

As most prior imaging studies did not selectively enroll patients with very high Lp(a), the associations identified in this analysis represent a wide spectrum of Lp(a) concentrations. Future dedicated studies in cohorts with markedly elevated Lp(a), particularly within randomized trials of potent Lp(a)-lowering therapies, are needed to determine whether reduction of Lp(a) translates into attenuation of high-risk plaque features and progression, thereby establishing its role as a modifiable therapeutic target.

5. Limitations

This meta-analysis has several limitations. First, most included studies were observational, introducing the potential for residual confounding. Second, heterogeneity was high for some outcomes, likely reflecting differences in study populations, Lp(a) thresholds, imaging modalities, and methodological approaches. In particular, follow-up duration for plaque progression analyses varied substantially, which may have contributed to variability in effect estimates and may partially explain the observed between-study heterogeneity. However, sensitivity analyses excluding one study substantially reduced heterogeneity with consistent results, suggesting that the overall findings are directionally robust despite this limitation.

Third, definitions of high versus low Lp(a) varied across studies, with some reporting levels in nmol/L and others in mg/dL, potentially affecting comparability. Additional limitations include limited ethnic and geographic diversity among study populations, variability in CCTA acquisition protocols and plaque quantification methods, a small number of studies for certain plaque phenotypes, and the possibility of overlapping cohorts in multi-center registries. Variability in findings may also reflect differences in imaging modality indications, appropriate use criteria, center availability, and operator preference.

The findings should also be interpreted in light of potential reporting and publication biases. Studies demonstrating associations between elevated Lp(a) levels and adverse plaque characteristics may be more likely to be published, whereas studies with neutral findings or selectively unpublished results may be underrepresented. In addition, coronary imaging studies frequently evaluate multiple plaque characteristics, increasing the possibility of selective outcome reporting. Although formal statistical assessment of publication bias was not feasible due to the limited number of studies in each analysis, small-study effects cannot be excluded and may have influenced the magnitude of the pooled estimates.

Another limitation is that most included studies evaluated low-attenuation plaque (LAP) as the primary high-risk plaque feature, while other established markers of plaque vulnerability, such as the napkin-ring sign, spotty calcifications, and positive remodeling, were not systematically assessed, thereby limiting comprehensive evaluation of Lp(a)-related plaque vulnerability. Although stratification by imaging modality suggested a stronger association between elevated Lp(a) levels and coronary plaque on CCTA compared with IVUS or OCT, these comparisons were not adequately powered and should be interpreted cautiously. Finally, as most included cohorts were clinically selected rather than population-based, generalizability to asymptomatic populations remains limited.

6. Conclusions

Our meta-analysis demonstrates that elevated Lp(a) levels are significantly associated with greater coronary plaque burden, accelerated plaque progression, and a higher prevalence of high-risk low-attenuation plaque. These findings support Lp(a) as a key pathogenic factor in atherogenesis and plaque vulnerability, reinforcing its role as a clinically relevant biomarker for primary or primordial cardiovascular risk assessment. Standardization of Lp(a) measurement and thresholds, along with prospective studies and therapeutic intervention trials, are needed to further clarify its prognostic value and potential as a modifiable target in atherosclerotic cardiovascular disease prevention.

HIGHLIGHTS.

• Elevated Lp(a) is associated with greater coronary plaque burden.

• High Lp(a) is linked to more rapid progression of coronary atherosclerosis.

• Elevated Lp(a) is associated with high-risk coronary plaque features.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.atherosclerosis.2026.120706.

Data availability

The authors confirm that the data supporting the findings of this study are available within the manuscript and the supplementary materials.