This is a cross-sectional study conducted at the Hippokrateion hospital of the National and Kapodistrian University of Athens between 2022 and 2025. Participants were prospectively enrolled and comprised of adult ANOCA patients who underwent invasive coronary functional assessment at the cardiac catheterization laboratory of the First University Department of Cardiology.

The primary endpoint of the study was the comparison of Lp(a) levels between CMD patients and non-CMD controls. Secondary endpoints were (1) to detect a possible association between Lp(a) and coronary flow reserve (CFR)/index of microvascular resistance (IMR) indices and (2) to compare Lp(a) levels between patients with structural versus functional CMD.

The study protocol was approved by the review board of the Hippokrateion hospital (6061/06-04-2022) and was in accordance with the principles of the Declaration of Helsinki. All patients provided written informed consent.

The prospectively enrolled patients were 18 to 85 years old presenting with symptoms of stable ischemic heart disease who were found to have nonobstructive CAD on clinically indicated elective coronary angiography and who also underwent invasive functional assessment for evaluation of CMD. Nonobstructive CAD was defined as the presence of either no coronary stenosis, stenosis <50%, fractional flow reserve (FFR) >0.8 or instantaneous wave-free ratio >0.89.

The exclusion criteria of the study were the following: (1) hemodynamic instability, (2) severe left ventricular systolic dysfunction (left ventricular ejection fraction <35%), (3) severe valvular stenosis or regurgitation, (4) coronary artery bypass graft surgery, (5) known allergy to adenosine, (6) underlying conditions deterring adenosine administration (severe bradyarrhythmia, sick sinus syndrome, second or third degree atrioventricular block, severe asthma, severe chronic obstructive pulmonary disease, and long-QT syndrome), (7) advanced renal or liver disease, (8) hematologic disease (thrombocytopenia <80,000 platelets/mm³, severe anemia with hemoglobulin <8 mg/dl), (9) malignancies under chemo-, immune-, or radiation therapy, and (10) pregnancy, breastfeeding, or postpartum period.

Coronary physiology assessment was performed by 2 experienced cardiologists who were blinded to the patients’ laboratory profile. All patients underwent invasive assessment using a pressure-temperature sensor guidewire (Pressure Wire X, Abbott Vascular) and CoroFlow system (Coroventis Research AB) with the standard bolus thermodilution technique. Before guidewire insertion, the aortic pressure transducer’s sensor instrumentation and guidewire underwent zeroing at atmospheric pressure. The guidewire pressure was then equalized to aortic pressure at the ostium of the guiding catheter. The equalization was considered successful when the mean coronary pressure wire to aortic pressure ratio was equal to 1.00. After being calibrated and equalized to guide catheter pressure, the guidewire was then advanced to the distal two-thirds of the left anterior descending artery. Coronary flow was estimated using thermodilution curves by calculating the mean transit time (Tmn) from bolus injections of room temperature saline (3 ml) at rest and during maximal hyperemia. Tmn was averaged over 3 measurements at rest and 3 at hyperemia. For maximal hyperemia induction, intravenous adenosine infusion (140 mg/kg/min for 2 minutes) was used. Hyperemic state was identified by a decreased Pd/Pa pattern and a left shift in the Tmn. Baseline and hyperemic flow were estimated as the inverse of the average baseline (bTmn) and hyperemic (hTmn) mean transit times, respectively. Continuous recordings of aortic pressure (Pa) and distal pressure (Pd) were obtained during both baseline and hyperemic periods. FFR was calculated as the ratio of Pd/Pa during maximal hyperemia. CFR was defined as the bTmn/hTmn ratio. IMR was calculated by multiplying the mean Pd during maximal hyperemia with the hTmn. Resistive reserve ratio (RRR) was calculated as the ratio between basal resistance index (bTmn × resting Pd) and IMR. CMD was defined as the presence of either CFR ≤2.5 or IMR ≥25. Regarding endotype definition, structural CMD was defined as the presence of CFR ≤2.5 and IMR ≥25, while functional CMD was defined as the presence of CFR ≤2.5 and IMR <25. After measurement completion, the guidewire was pulled back to the guide catheter and possible presence of a pressure drift was inspected. Re-equalization and new measurements were made if the presence of a drift >0.03 in FFR was identified.

Blood samples were collected within 1 month before or after invasive functional assessment. Τhe concentration of Lp(a) was quantified by using a mass assay and was measured in mg/dl. All Lp(a) laboratory testings took place at the same laboratory of the Hippokrateion hospital in Athens using industry-standard assays.

For the primary endpoint, Lp(a) was evaluated both as a continuous and as a categorical variable. In the latter case, patients were grouped into 3 groups based on the Lp(a) level: <30, (30 to 50), and ≥50 mg/dl. Lp(a) was additionally analyzed as a discrete variable with the generation of the “Lp(a) score,” in which a score from 1 to 5 was assigned according to the following Lp(a) level groups, correspondingly: <10, (10 to 20), (20 to 30), (30 to 50), and ≥50 mg/dl.

Categorical variables were expressed as frequencies and proportions and compared with the chi-square or Fisher’s exact test, as appropriate. Continuous variables were reported as mean ± SD or median (IQR) and compared with Student’s t-test, analysis of variance or Pearson correlation coefficient, as appropriate, if the distribution was normal. Otherwise, Mann–Whitney U test, Kruskal–Wallis test, or Spearman rank correlation coefficient were employed. The presence of normal distribution was evaluated with the Shapiro-Wilk test. Equality of CFR and IMR variances across Lp(a) categories was assessed using Levene’s test with Brown–Forsythe modification.

Univariable and multivariable logistic regression was conducted in order to examine predictors of CMD. Univariable and multivariable quantile regression was used to model the relationship between laboratory findings and CFR/IMR at the 50th percentile of the outcome distribution (CFR/IMR distribution). Quantile regression was chosen over linear regression given the highly-skewed distribution of the outcome. All multivariable models were adjusted for age and sex, which were included as covariates in the regression models (age modeled as a continuous variable and sex as a binary variable).

Significance level was set at a = 0.05 level and all reported p values were two-sided. Data analysis was performed using the STATA statistical package for Windows (version 13, StataCorp).