Abstract

Study Population

This single-center, retrospective study included 929 patients who underwent PCI followed by coronary angiography 1 year after PCI in the Department of Cardiology, Daping Hospital, The Third Military Medical University, from January 2017 to December 2019. Patient data, including cardiovascular risk factors, clinical features, and prognosis, were obtained from the hospital's electronic records. Patients who underwent PCI did not have coronary angiography 1 year after PCI, and these patients did not have an electronic record of blood lipids and were excluded. According to the normal reference range of blood lipids, high triglycerides were defined as a triglyceride concentration exceeding 1.70 mmol/L; high total cholesterol was defined as a total cholesterol concentration exceeding 5.72 mmol/L; and high LDL-C was defined as an LDL-C concentration exceeding 3.12 mmol/L. Coronary artery disease was defined by coronary angiography confirming that at least 1 epicardial main vessel stenosis was more than 50%. Multivessel lesions were defined by coronary angiography confirming that at least 2 epicardial main vessel stenoses were more than 50%. In-stent restenosis was diagnosed by coronary angiography, in which the narrow diameter of the lumen in the implanted segment of the stent and the proximal and distal 5-mm segments of the stent exceeded 50%. Family history of hypertension, diabetes, and CAD refers to immediate family members, including parents, brothers, and sisters. Secondary PCI was defined as PCI undertaken in patients with angina and at least 1 epicardial main vessel stenosis of more than 80% 1 year after PCI. This study was conducted in compliance with human studies guidelines and the Declaration of Helsinki, and this study was approved by the Ethical Committee of Daping Hospital.

Statistical Analysis

Values are presented as mean (SD) or number (%). A nonparametric Kruskal-Wallis test was used to compare continuous variables between groups, and the Bonferroni correction method was used to adjust multiple comparisons. Wilcoxon signed-rank test was used to compare blood lipid levels between the patients undergoing PCI, and then coronary angiography 1 year after PCI. Categorical variables between groups were compared using χ² or Fisher exact test. Receiver operating characteristic curve was used to evaluate the capacity of baseline Lp(a) levels to discriminate secondary PCI or in-stent restenosis. For a diagnostic test to be meaningful, the area under the curve (AUC) must be greater than 0.50. Generally, an AUC of 0.50 to less than 0.60 is considered a fail, 0.60 to less than 0.70 is considered poor, and 0.80 or more is considered acceptable. A multivariate logistic regression model was used to analyze factors potentially related to the incidence of in-stent restenosis and secondary PCI in patients undergoing coronary angiography 1 year after PCI, adjusted for baseline clinical characteristic factors, including age, smoking, gender, hypertension, diabetes, high total cholesterol, high triglycerides, high LDL-C, multivessel lesions, and Lp(a) group, with 95% CIs. P < .05 was considered statistically significant. All statistical analysis was carried out in SPSS, version 26, software (IBM Corp).

Comparison of Clinical Characteristics of Patients With Different Lp(a) Concentrations

The 929 patients in this study were allocated to 1 of 4 groups according to their Lp(a) concentration: group 1 (normal: Lp(a) concentration 0-300 mg/L), group 2 (moderate: Lp(a) concentration 301-500 mg/L), group 3 (high: Lp(a) concentration 501-1000 mg/L), and group 4 (extreme: Lp(a) concentration >1000 mg/dL). The non-parametric Kruskal-Wallis test was used to compare the continuous variables, and the χ² or Fisher exact test was used to compare the categorical variables between Lp(a) groups. Baseline characteristics of the 929 patients are presented in Table I. As shown in Table I, Lp(a) groups did not differ by sex, age, body mass index, hypertension, diabetes, obesity, hyperuricemia, smoking, cardiac insufficiency, acute myocardial infarction, or multivessel lesion. There were no significant differences in the concentrations of total cholesterol (P = .25), high-density lipoprotein cholesterol (P = .24), LDL-C (P = .10), or apo AI (P = .33). There were no significant differences in the incidence of high total cholesterol (P = .51) or high LDL-C (P = .15). There were no statistically significant differences in family history of hypertension (P = .35), diabetes (P = .85), or CAD (P = .28). Furthermore, 66.20% of the patients had Lp(a) concentrations of 0 to 300 mg/L, 15.50% had Lp(a) concentrations of 301 to 500 mg/L, 14.21% had Lp(a) concentrations of 501 to 1000 mg/L, and 4.09% had Lp(a) concentrations greater than 1000 mg/L.

TABLE IBaseline Characteristics of Patients Undergoing Percutaneous Coronary Intervention, by Lp(a) Concentration (n = 929)

View Fullscreen

A multigroup comparison of continuous variables and categorical variables was conducted to assess the differences between Lp(a) groups. As shown in Supplemental Table I, the incidence of hypertriglyceridemia was greater in group 1 (normal) than in group 2 (moderate) (P = .001).

Changes in Blood Lipids During PCI and After 1 Year After PCI

Among 929 patients who received PCI and underwent coronary angiography 1 year after PCI, Wilcoxon signed-rank test was used to compare blood lipid levels between the patients undergoing PCI and 1 year after PCI. As shown in Supplemental Table II, after patients received statins for 1 year after PCI, the concentrations of triglycerides (P < .001), total cholesterol (P < .001), LDL-C (P < .001), and apo B (P < .001) were statistically significantly lower in the patients who underwent coronary angiography 1 year after PCI, although statins had no effect on the Lp(a) concentration in patients undergoing PCI (P = .35).

Comparison of Clinical Characteristics of Patients With Different Lp(a) Concentrations 1 Year After PCI

After grouping according to Lp(a) levels, 1 year after PCI, there were statistically significant differences in the average concentrations of triglycerides (P = .03), total cholesterol (P = .007), LDL-C (P < .001), and apo B (P = .02), whereas there were no statistically significant differences in the average concentrations of apo AI (P = .54) and high-density lipoprotein cholesterol (P = .35) (Table II). There were no statistically significant differences in the incidence of high triglycerides (P = .13), high total cholesterol (P = .52), or high LDL-C (P = .051). There were no statistically significant differences in the incidence of in-stent restenosis (P = .39) or secondary PCI (P = .64) between Lp(a) groups. There were no statistically significant differences in the incidence of stable angina (P = .39), unstable angina (P = .08), chest distress (P = .07), dyspnea (P = .56), acute myocar-dial infarction (P = .40), cardiac insufficiency (P = .76), chronic kidney disease (P = .16), or stroke (P = .999) between Lp(a) groups.

TABLE IIComparison of Patient Clinical Characteristics, by Lp(a) Concentration 1 Year After PCI (N = 929)

View Fullscreen

A multigroup comparison of continuous variables was then conducted that reached statistical differences between Lp(a) groups. As shown in Supplemental Table III, 1 year after PCI, the average apo B concentration was higher in group 3 (high) than in group 1 (normal) (P = .001). The average LDL-C concentration was higher in group 3 (high) than in group 1 (normal) (P = .002).

Capacity of Lp(a) Levels for Discriminating Secondary PCI or In-Stent Restenosis

The receiver operating characteristic curve was used to investigate the capacity of Lp(a) levels to discriminate secondary PCI or in-stent restenosis. As shown in Figure 1, Lp(a) level has no capacity to discriminate secondary PCI (AUC, 0.539 [95% CI, 0.50-0.58]; P = .055). Also, Lp(a) level has no capacity to discriminate in-stent restenosis (AUC, 0.536 [95% CI, 0.48-0.60]; P = .24).

View Figure

• Download as Powerpoint
• Download Figure

Multivariate Logistic Regression Model to Analyze Risk Factors Related to Secondary PCI and In-Stent Restenosis

A multivariate logistic regression model was used to analyze risk factors associated with the incidence of secondary PCI, with 95% CIs. As shown in Table III, after adjustment for the clinical characteristic factors, diabetes (odds ratio [OR], 1.48 [95% CI, 1.07-2.05]; P = .02) and multivessel lesion (OR, 3.12 [95% CI, 2.23-4.52]; P < .001) increased the risk of secondary PCI in patients undergoing coronary angiography 1 year after PCI. Patients in group 2 (moderate) (OR, 1.12 [95% CI, 0.80-1.79]; P = .40), group 3 (high) (OR, 1.27 [95% CI, 0.84-1.92]; P = .26), and group 4 (extreme) (OR, 1.05 [95% CI, 0.51-2.13]; P = .90) have no obvious differences from patients in group 1 (normal) (reference) in risk of secondary PCI.

TABLE IIIMultivariate Logistic Regression Model on the Association of Secondary PCI and In-Stent Restenosis With Cardiovascular Risk Factors

View Fullscreen

A multivariate logistic regression model was used to analyze risk factors associated with the incidence of instent restenosis, with 95% CIs. As shown in Table III, after adjustment for the clinical characteristic factors, diabetes (OR, 1.76 [95% CI, 1.10-2.79]; P = .02) increased the risk of in-stent restenosis in patients undergoing coronary angiography 1 year after PCI. Patients in group 2 (moderate) (OR, 1.50 [95% CI, 0.86-2.61]; P = .16), group 3 (high) (OR, 1.33 [95% CI, 0.73-2.40]; P = .35), and group 4 (extreme) (OR, 0.57 [95% CI, 0.16-2.03]; P = .38) have no obvious differences from patients in group 1 (normal) (reference) in risk of instent restenosis.

Study Limitations

This study has several limitations. First, this was a retrospective study, and it is difficult to rule out the effects of activity and diet, which may introduce bias. Second, this was a single-center study, and various ethnic populations may have different results. Third, the sample size of this study was small, which may have led to some bias. Fourth, an imbalanced sample size is a main limitation of this study, which would probably limit statistical inference of the group effect of Lp(a).