Data for this retrospective analysis were derived from comprehensive angiography databases at Guangdong Provincial People's Hospital. These databases utilize longitudinal follow-up protocols to ensure robust long-term monitoring following ICA. Adult patients (aged ≥18 years) who underwent angiography for suspected angina between January 1, 2012, and December 31, 2022, were screened for eligibility. CAE was defined as a coronary segment dilated to at least 1.5 times the diameter of an adjacent normal segment. Angiographic assessments were performed independently by two cardiovascular imaging specialists, each possessing over 10 years of experience in angiographic imaging. Patients were included in the ICAE cohort if they met the diagnostic criteria for ectasia and exhibited <20% stenosis across all coronary vessels. All patients underwent clinically indicated ICA, performed at the discretion of treating cardiologists. ICAE was diagnosed retrospectively based on angiographic findings and no invasive procedures were performed for research purposes.

Subsequently, to ensure a strict ICAE cohort, we applied a rigorous and clinically pragmatic exclusion strategy. Patients with any angiographic evidence of ≥20% coronary artery stenosis were excluded. Secondary causes of coronary ectasia were excluded primarily through detailed reviews of medical history, clinical presentation, and available medical records, including documented histories of Kawasaki disease, systemic vasculitis, syphilis, obstructive sleep apnea, or prior coronary intervention. Additional paraclinical investigations were not performed systematically in all patients but were obtained selectively when clinically indicated, in the presence of suggestive symptoms, abnormal findings, or at the request of patients (e.g., serology for vasculitis).

To minimize follow-up bias, we established a control group consisting of normal individuals from the same database (19, 20). For each patient in the ICAE group, a control was matched 1:1 based on gender, similar age (within one year difference), and ICA procedure timeframe (21).

Baseline demographic and clinical data included age, sex, body mass index (BMI), smoking status, hypertension, and diabetes mellitus. Laboratory tests assessed: metabolic parameters: hemoglobin A1c (HbA1c), estimated glomerular filtration rate (eGFR); inflammatory markers: leukocyte count, high-sensitivity C-reactive protein (hs-CRP) (22); cardiac biomarkers: α-hydroxybutyrate dehydrogenase (α-HBDH), lactate dehydrogenase (LDH), creatine kinase (CK), creatine kinase-MB (CK-MB), high-sensitivity cardiac troponin T (hs-cTnT), and N-terminal pro–B-type natriuretic peptide (NT-proBNP) (13); and lipid profile: total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) (21).

Echocardiography was performed within one month of angiography to measure left ventricular ejection fraction (LVEF). Standard 12-lead electrocardiograms were reviewed for ST–T abnormalities and pathologic Q waves (23).

Angiographic data included the distribution of ectasia, number of affected vessels, Markis classification, and the presence of thrombus or slow coronary flow (9).

The follow-up period spanned from the initial angiography until an outcome event or the study's conclusion. The primary outcome was all-cause mortality. The secondary outcome was major adverse cardiovascular events (MACE), a composite of sudden cardiac death, non-fatal myocardial infarction (MI), coronary revascularization, and hospitalization for unstable angina (UA) or heart failure (HF) (24). Mortality data were adjudicated using official death certificates and medical records. Cardiovascular events and hospitalizations were ascertained through clinic visits, telephone interviews, and review of hospital records (10, 25, 26).

Continuous normally distributed variables were expressed as mean ± standard deviation (SD) and compared using the Student's t-test. Non-normally distributed variables are presented as medians with interquartile ranges (IQR) and compared using the Mann–Whitney U-test. Categorical variables were expressed as frequencies and percentages and compared using the chi-square test. Survival curves were estimated using the Kaplan–Meier method and compared via the log-rank test. Hazard ratios (HR) and 95% confidence intervals (CI) were calculated using Cox proportional-hazards models. Due to the limited number of outcome events and the biological overlap between certain baseline variables and the ICAE phenotype, we prioritized unadjusted analyses to avoid overfitting and overadjustment bias. Statistical significance was defined as a p-value < 0.05. Analyses were conducted with SPSS software, version 27.0 (15, 17).

The study was conducted in compliance with the principles outlined in the Declaration of Helsinki. Ethical approval for this study was obtained from the Guangdong Provincial People's Hospital Ethics Committee. Since this retrospective study involved the review of existing medical records and did not involve intervention, the requirement for informed consent was waived by the ethics committee.

The study population was derived from 19,144 ICA records in the database. As detailed in Figure 2, initial screening identified 1,546 patients who met the angiographic criteria for CAE. Subsequent application of exclusion criteria resulted in a final analytical cohort of 171 patients with ICAE. These exclusions included 1,263 patients with concomitant coronary artery stenosis of ≥20% and 112 patients with identifiable secondary etiologies for CAE or sever systemic disorders. For comparative analysis, a 1:1 matched control group (n = 171) consisting of patients with angiographically normal coronary arteries were selected from the same database.

Baseline characteristics were well-matched for age (60.7 ± 12.0 years vs. 60.3 ± 11.7 years, p = 0.718) and sex (27.5% female in both groups, p = 1.000) (Table 2). BMI (24.6 ± 3.0 kg/m² vs. 24.5 ± 3.0 kg/m², p = 0.754) was also similar between the two groups, and the prevalence of diabetes mellitus (12.9% vs. 12.3%, p = 0.870) and current smoking (22.8% vs. 18.1%, p = 0.284) were comparable. However, ICAE patients exhibited a significantly higher prevalence of hypertension (56.7% vs. 37.4%, p = 0.001).

Laboratory analysis revealed several differences in key parameters between the groups (Table 2). Some cardiac biomarkers were significantly elevated in ICAE patients, including LDH (175.9 ± 57.6 U/L vs. 162.7 ± 33.6 U/L, p = 0.013), CK-MB (11.3 ± 6.3 U/L vs. 9.8 ± 4.3 U/L, p = 0.016), hs-cTnT [9.4 pg/mL [IQR 7.1–16.3] vs. 7.5 pg/mL [IQR 5.3–9.4], p = 0.001], and NT-proBNP [69.2 pg/mL [IQR 35.2–217.1] vs. 41.2 pg/mL [IQR 19.5–93.6], p = 0.001]. However, no significant differences were observed between the groups in inflammatory markers such as leukocyte count (7.1 ± 1.9 × 10⁹/L vs. 6.9 ± 2.0 × 10⁹/L, p = 0.379) and hs-CRP [1.0 mg/L [IQR 0.3–4.3] vs. 1.4 mg/L [IQR 0.4–3.2], p = 0.225], metabolic parameters such as HbA1c (6.0% ± 0.9% vs. 5.9% ± 0.9%, p = 0.559), eGFR (81.5 ± 18.4 mL/min/1.73 m² vs. 85.0 ± 18.4 mL/min/1.73 m², p = 0.057), and standard lipid profiles (TC, TG, HDL-C, LDL-C; all p ≥ 0.05).

ECG analysis also showed a higher prevalence of pathologic Q-waves (15.8% vs. 7.0%, p = 0.002) and ST-T changes (45.6% vs. 26.9%, p = 0.001) in the ICAE group (Table 2). LVEF was slightly lower in ICAE patients, although remaining within the normal range (63.7% ± 7.5% vs. 65.5% ± 6.2%, p = 0.021).

In addition, for long-term medications, ICAE patients exhibited significantly higher drug utilization, including renin-angiotensin-aldosterone system inhibitors (RAAS-i) (42.7% vs. 28.7%, p = 0.007), calcium channel blockers (CCB) (30.4% vs. 18.7%, p = 0.012), and diuretics (10.5% vs. 4.7%, p = 0.041). Remarkably, more than half of ICAE patients had taken antithrombotic therapy (56.7%) and statins (57.9%), whereas the control group was significantly lower at 15.8% and 32.2% (both p = 0.001).

Among ICAE patients, the left anterior descending artery (LAD) was the most commonly affected vessel (70.8%), followed by the left circumflex artery (LCX, 57.3%) and the right coronary artery (RCA, 56.1%). Left main (LM) involvement was observed in 18.7% of patients (Table 3). Regarding the number of affected vessels, 45.6% had single-vessel ectasia, while 19.9% and 34.5% had two and three affected vessels, respectively. Based on the Markis classification, Type IV was the most common (54.4%), followed by Type I (25.7%), Type II (11.1%), and Type III (8.8%). Additionally, slow blood flow was observed in 26.9% of patients, and coronary thrombus formation was rare (2.3%) (Table 3).

First, its retrospective, single-center design may limit the generalizability of the findings and the ability to draw definitive causal inferences. Although we applied strict angiographic definitions and excluded overt secondary causes, unrecognized subclinical etiologies cannot be entirely ruled out. Second, although multiple clinical, laboratory, and imaging abnormalities were observed, their limited sensitivity and specificity preclude their use as standalone diagnostic markers, and ICA remains the reference standard for diagnosis.

Third, we did not perform formal multivariable adjustment. Some candidate variables, such as hypertension, are biologically intertwined with the ICAE phenotype. These might represent upstream contributors or downstream consequences of the disease rather than independent confounders, making adjustment potentially prone to overadjustment bias. In addition, many disease-specific characteristics and treatment variables were present almost exclusively in ICAE patients, limiting the interpretability of joint multivariable modeling with control groups. Furthermore, the relatively small number of outcome events constrained the feasibility of stable multivariable models (44). Finally, pharmacologic therapy was not randomized or standardized; it was initiated or adjusted based on clinical judgment. Consequently, this likely reflects confounding by indication, whereby patients with more severe disease tend to receive more intensive therapy (9). Therefore, the independent effects of specific medical treatments on long-term outcomes might not be reliably assessed in this study.