Patients and Methods

This prospective, single-center study recruited consecutive patients with a primary diagnosis of stable CAD who underwent percutaneous revascularization between March 1, 2019, and February 29, 2020, at Rajaie Cardiovascular Medical and Research Center, Tehran, Iran. Patients were eligible for inclusion if they exhibited stable angina with normal cardiac troponin I (cTnI) levels before the procedure. Patients with recent myocardial infarction or incomplete lipid profile data were excluded.

Experienced interventional cardiologists performed all PCI procedures in accordance with the latest guidelines from the American College of Cardiology and the American Heart Association, which recommend aspirin, clopidogrel, and moderate- to high-intensity statin therapy.¹⁴ Before the procedure, all patients received a loading dose of 300 mg of aspirin irrespective of whether they were already on daily aspirin therapy. The day before the procedure, patients on daily clopidogrel received a clopidogrel loading dose of 300 mg; patients not on daily clopidogrel received a 600-mg loading dose. At the beginning of angioplasty, a 100-U/kg bolus of intravenous unfractionated heparin was injected. Additional boluses of 2,000 to 3,000 U were given every hour if the procedure lasted for more than 1 hour, up to a maximum dose of 10,000 U. All participants underwent continuous electrocardiographic monitoring before, during, and after PCI to detect possible ischemic events.

Baseline samples for the serum lipid profile, including LDL, HDL, and TG, were collected before PCI after 12 to 14 hours of fasting. Lipid levels were measured using the electroluminescence method with a Hitachi 917 rack chemistry analyzer (Roche). Postprocedural samples for cTnI analysis were collected 1, 6, and 12 hours after PCI and were analyzed using the chemiluminescence immunoassay method with an Abbott ARCHITECT analyzer.

Diagnoses of procedure-related myocardial injury were based on either (1) the development of new pathological Q waves in at least 2 contiguous electrocardiogram leads or (2) cTnI elevation greater than or equal to 5× the upper limit of normal (ULN), according to the Fourth Universal Definition of Myocardial Infarction.⁶

The study was performed in accordance with the ethical standards of the Declaration of Helsinki and approved by the Review Board and the Human Ethics Committee of Rajaie Cardiovascular Medical and Research Center. Informed consent was obtained from all participants.

Results

During the study period, 1,024 patients at Rajaie Cardiovascular Medical and Research Center had a primary diagnosis of stable CAD and underwent percutaneous revascularization. Of these, 20 patients were excluded because of an elevated baseline cTnI level (indicating previous myocardial infarction) and 27 were excluded because of incomplete lipid profile tests or missing laboratory data. The remaining 977 patients were enrolled, including 697 men (mean [SD] age, 59.8 [10.5] years) and 280 women (mean [SD] age, 62.4 [9.0] years).

Low-Density Lipoprotein

Table I depicts baseline patient characteristics and angiographic and lab data by LDL subgroup. In terms of distribution, LDL less than 70 mg/dL was reported in 681 patients (69.7%), LDL 70 to 99 mg/dL in 224 patients (22.9%), and LDL greater than or equal to 100 mg/dL in 72 patients (7.4%). A higher proportion of patients in all LDL categories were male, and patients in all 3 LDL categories tended to have hypertension (P = .048). Patients with LDL greater than or equal to 100 mg/dL were younger than those in the other groups (P = .01). Other clinical characteristics did not differ significantly by LDL group, nor were there any statistically significant differences in medical therapy by group.

TABLE I Baseline Clinical and Procedural Characteristics by LDL Cholesterol

View Fullscreen

Angiography characteristics did not differ by LDL category, except for chronic total occlusion (P = .04). Preprocedural platelet and hemoglobin levels differed significantly by LDL group (P = .03 and P = .02, respectively). Although patients with LDL greater than 100 mg/dL had a higher platelet level, the level was within the normal range and the difference was not clinically meaningful.

High-Density Lipoprotein

Baseline clinical and procedural characteristics by HDL subgroup are detailed in Table II. In terms of distribution, HDL less than 40 mg/dL was reported in 933 patients (95.5%) and HDL greater than or equal to 40 mg/dL in 44 (4.5%). Patients in either group tended to be nonsmokers (P = .04). Patients with HDL less than 40 mg/dL were more likely to be younger (P = .03) and to be male than were patients with HDL greater than or equal to 40 mg/dL (P = .001). No other statistically significant differences were noted. No correlation was found between HDL and periprocedural myocardial injury in patients with LDL less than 70 mg/dL (β = .018; 95 CI, −.037 to .059; P = .64; R² = −0.001).

TABLE II Baseline Clinical and Procedural Characteristics by HDL Cholesterol Group

View Fullscreen

Triglycerides

Demographic, clinical, and angiographic characteristics by TG subgroup are shown in Table III. In terms of distribution, TG less than 150 mg/dL was reported in 855 patients (87.5%), TG 150 to 199 mg/dL in 71 (7.3%), and TG greater than or equal to 200 mg/dL in 51 (5.2%). Patients with TG greater than or equal to 200 mg/dL were more often younger (P < .001), had a higher mean fasting blood sugar level (P = .005), and had a higher frequency of diabetes mellitus (P = .04) than the other groups. Patients in the group with TG less than 150 mg/dL were more likely to be taking aspirin (P = .045). There were no substantial differences in angiography data among these subgroups. Remarkably, we found a positive correlation between TG and LDL levels (β = .311; 95% CI, .056–.090; P < .001; R² = 0.095), despite the absence of an association between TG and cTnI elevation.

TABLE III Baseline Clinical and Procedural Characteristics by TG Group

View Fullscreen

Postoperative Coronary Troponin I Elevation

Table IV compares the baseline clinical and procedural characteristics of the patients with and without elevated cTnI (≥5× the ULN) after the procedure. Most of the patients had a cTnI level less than 5× the ULN (n = 698 vs 279). The patients with elevated cTnI were less likely to have diabetes mellitus (P = .01) but more likely to have had a previous myocardial infarction (P = .002). The remaining clinical characteristics did not differ significantly between the 2 groups. Nevertheless, there were significant differences vis-à-vis coronary angiography results, the number of vessels subjected to PCI, and use of predilation. Patients with double- and triple-vessel disease were more likely to have elevated cTnI (P < .001). All but 1 patient underwent 1 intervention only. One patient exhibited acute symptoms and electrocardiogram changes 6 hours after PCI, at which time cTnI was 3× the ULN. The changes were attributed to acute stent thrombosis, and the patient underwent immediate angioplasty.

TABLE IV Baseline Clinical and Procedural Characteristics by cTnI Elevation Group

View Fullscreen

Preoperatively, in-stent restenosis was reported in 61 patients and bifurcation in 156. Bifurcation PCI techniques included provisional stenting in 120/977 patients (12.3%), double-kissing crush in 4/977 (0.4%), mini-crush in 9/977 (1%), T-stenting in 2/977 (0.2%), and culotte in 3/977 (0.3%). Patients with elevated cTnI were more likely to have had predilation (P = .009) or bifurcation (P = .003) and had significantly lower LVEF (P = .01) and higher hemoglobin levels (P = .02). Still, mean hemoglobin levels were within the normal range for physiologic blood concentrations. No statistically significant differences were detected in the LDL, HDL, and TG categories by cTnI group (P = .98, .62, and .20, respectively).

Logistic Regression Analyses

Table V shows the logistic regression analysis results for LDL, HDL, and TG according to elevated post-PCI cTnI (cTnI ≥5× the ULN). Neither the unadjusted nor adjusted logistic regression models found an association between elevated postprocedural cTnI and the various LDL, HDL, and TG categories.

TABLE V Logistic Regression Analysis of the Impact of Clinical Parameters on Elevated Postprocedural cTnI (≥5× ULN)

View Fullscreen

As a sensitivity analysis, the relationships between cTnI elevation in the 12 hours after PCI as a continuous variable (in contrast to a dichotomized variable) and the LDL, HDL, and TG categories were separately explored via multiple linear regression analyses (see Table VI). At 1 and 6 hours after PCI, similar results were obtained for the LDL, HDL, and TG categories, except for a difference in LDL and cTnI in patients without periprocedural myocardial injury at 1 hour post-PCI.

TABLE VI Multiple Linear Regression Analysis of the Impact of Lipid Profile on Elevated cTnI (≥5× ULN) at the 12th Postprocedural Hour ^a

View Fullscreen

Discussion

Periprocedural myocardial injury is a frequent complication during PCI and is strongly associated with postprocedural cardiovascular morbidity and death.¹⁷ Major risk factors for periprocedural myocardial injury after PCI can be categorized as patient related, lesion related, or procedure related, all of which contribute equally.¹⁸

Lipid profile was classified as a patient-related factor. Elevated serum LDL is a major cause of CAD and is a key component of coronary plaque development and rupture.^19,20 Nevertheless, the existing literature lacks data on the relationship between lipid levels and periprocedural myocardial injury, especially in the present era of statin use. In the present study, the value of the preprocedural lipid profile (LDL, HDL, and TG) in predicting the incidence of periprocedural myocardial injury (evidenced by elevated cTnI) was examined in patients with stable CAD who underwent PCI. This evaluation was performed in a population with a high prevalence of statin intake and better-controlled LDL. No meaningful association between periprocedural myocardial injury and preprocedural lipid levels were found.

Li et al¹⁶ investigated the association between preprocedural LDL and periprocedural myocardial injury in 2,529 patients with CAD who underwent elective PCI and found a log-linear relationship between LDL and cTnI levels. These authors concluded that a 1-SD increment in preprocedural LDL increased the risk of postprocedural cTnI elevation by 12% to 20%. Similarly, Buturak et al²¹ observed a direct connection between preprocedural LDL levels and periprocedural myocardial injury in 195 patients with stable angina pectoris who underwent elective PCI: In that study, elevated LDL or non-HDL levels were positively correlated with postprocedural cTnI levels. Another study found that non-HDL was more valuable than LDL in predicting periprocedural myocardial injury in patients with type 2 diabetes mellitus.²² In the present study, after adjusting for confounding factors in both the logistic and linear regression analyses, the authors found no association between LDL and postprocedural cTnI elevation.

The researchers also found no relationship between HDL and cTnI elevation, even after adjusting for confounding factors. Silbernagel et al²³ concluded that HDL was inversely associated with cardiovascular mortality in individuals without CAD but not in patients with CAD. Li et al¹⁵ assessed 2,529 patients undergoing elective PCI and found that HDL level was not predictive of periprocedural myocardial injury. However, among patients with LDL less than 70 mg/dL, a 1-mg/dL increase in HDL was associated with a 3% reduction in risk of postprocedural cTnI greater than 3× the ULN. Li et al¹⁶ also reported that patients with LDL greater than or equal to 70 mg/dL had higher TG and C-reactive protein levels, which can lead to HDL dysfunction. Barter et al²⁴ evaluated 2,661 participants with LDL less than 70 mg/dL and concluded that higher HDL levels were associated with a lower risk of major cardiovascular events. Moreover, recent data which investigated the atheroprotective potential of HDL particles elucidated that the large HDL particles are linked to a lower number of circulating LDL particles.^25,26 The present analysis yielded no correlation between HDL and periprocedural myocardial injury in patients with LDL less than 70 mg/dL (β = .018; 95% CI, −.037 to .059; P = .64; R² = −0.001).

Jiao et al²⁷ found that plasma TG and HDL exerted a synergistic effect on the occurrence of CAD. When the plasma LDL concentration was less than 130 mg/dL, the risk of CAD was 10 times as high in a population with high TG levels (>263.93 mg/dL) and low HDL concentrations (<25.90 mg/dL) as it was in a population with low TG levels (<59.34 mg/dL) and high HDL concentrations (>64.19 mg/dL). Li et al¹⁶ found that HDL levels rose in tandem with a substantial fall in TG. Two other studies on the relationship between pre-procedural LDL and periprocedural myocardial injury reported substantially lower TG levels in a population with lower LDL.^16,28 Likewise, the researchers found a positive correlation between TG and LDL (β = .311; 95% CI, .056–.090; P < .001; R² = 0.095), despite the absence of an association between TG and cTnI elevation.

Evidence showing an association between lipid profile and periprocedural myocardial injury comes from studies on lipid-lowering therapy that compare the effects of different doses of statins on periprocedural myocardial injury. Of note, because many included patients adhered to their long-term statin therapy throughout the perioperative period, the authors studied the baseline lipid profile.

Takano et al²⁹ found that high-dose rosuvastatin reduced the incidence of periprocedural myocardial injury more than low-dose rosuvastatin in statin-naive patients but not in patients who were already taking statins. Herrmann et al³⁰ showed that the incidence of creatine kinase elevation greater than 3× the ULN was more than 90% lower in statin-treated patients (0.4% vs 6.0%) and that statin therapy was the only factor independently associated with a lower risk of creatine kinase elevation greater than 3× the ULN; no substantial differences between statin-treated patients and controls were found in terms of TG, LDL, HDL, total cholesterol, and lipoprotein(a). What remains unclear, however, is whether HDL alterations during statin therapy contribute to considerable cardiovascular risks. In the JUPITER trial, HDL and vascular events were inversely associated in patients on rosuvastatin (20 mg/d), whereas patients receiving placebo exhibited no such association.³¹ Moreover, in a meta-analysis of 8 statin trials, HDL increase during statin therapy was not associated with reduced risk of major cardiovascular events.³²

In the current study, no substantial interaction between statin therapy and cTnI was found, even after adjustments were made for confounders (Table V). More than 85% of included patients had statins in their medication history. Therefore, this study's findings can be explained by intensive and optimal medical treatment before the procedure, which might have eliminated the effects of lipids on periprocedural myocardial injury.

As for other patient-related factors, results showed a substantial difference pertaining to LVEF between patients with and without post-PCI cTnI elevation. Patients with cTnI greater than or equal to 5× the ULN had a markedly lower mean LVEF (43.3% [10.5%] vs 45.2% [9.7%]; P = .01). Previous research has indicated that patients with increased cTnI have substantially lower LVEF, higher clinical grading of heart failure, and higher mortality rates.^33,34 Gili et al³⁵ sought to determine the predictors of periprocedural myocardial injury after PCI and concluded that LVEF was a univariate predictor of major adverse cardiovascular events and death. Mancini et al³⁶ found that despite optimal medical treatment and PCI, LVEF and disease burden at baseline were prognostic of residual risk of secondary ischemic events.

Among procedure-related risk factors, the researchers detected a significant association between cTnI elevation and PCI for bifurcation, one of the more complex PCI procedures. Bifurcation correlated with higher levels of postprocedural cTnI, whereas the authors found no significant association between in-stent restenosis and cTnI elevation. Zhang et al³⁷ found that a true bifurcation lesion was a predictor of occlusion in a small side branch and that patients with small branch occlusions had a substantially higher incidence of periprocedural myocardial injury. Ojeda et al³⁸ concluded that the presence of a bifurcation lesion in the context of chronic total occlusion was linked to a higher incidence of periprocedural myocardial injury. Multivessel PCI and predilation during the procedure also may prompt cTnI elevation and periprocedural myocardial injury: Qadir et al³⁹ showed that elevated cTnI was correlated with the greater severity and larger extent of myocardial ischemic territory during non–ST-segment elevation myocardial infarction.

Conclusion

Lipid levels were not associated with an increased incidence of periprocedural myocardial injury in a sample of patients undergoing elective PCI, possibly because these patients received extensive, optimal medical therapy before the procedure. However, angiographic factors such as a lower LVEF, previous myocardial infarction, and the number of vessels subjected to PCI were associated with cTnI elevation and thus with periprocedural myocardial injury.

Although these findings suggest that, in the current era of intervention with fully optimized medical therapy, lipid profile exerts no influence on periprocedural myocardial injury, this does not agree with the findings of previous investigations. It may be that complex vessel anatomy, procedures, and low LVEF may have more influence on postprocedural outcomes based on cTnI elevation. It is essential to evaluate these data in future multicenter studies with a wider range of patients and a wider distribution of lipid values in both elective and nonelective procedures.