In the present study, we compared the predictive value of different blood lipid parameters combined with cIMT on CAD. We had three major findings: (1) Lp (a) better improved the prediction accuracy of CAD in comparison with other blood lipid parameters; (2) in the absence of high levels of LDL-C, the predictive value of Lp (a) on CAD was highlighted; (3) Lp (a) can also be used to predict multi-vessel CAD.
Atherosclerosis is a systemic vascular disease caused by a lipid metabolism disorder and most likely involves the carotid and coronary arteries [27]. Although coronary angiography is the gold standard for the diagnosis of CAD, it is an invasive examination that is unsuitable for general screening of CAD. In addition to identifying traditional risk factors such as demographic information and healthy behavior to predict CAD, screening for subclinical atherosclerosis is also important. Carotid ultrasound is a non-invasive examination [17]. cIMT is widely recognized as a marker of early-stage atherosclerosis and the severity of CAD, which can be used to indirectly assess coronary artery conditions [18, 28]. The results of a meta-analysis indicated that the sensitivity and specificity of cIMT for the diagnosis of CAD were 0.68 (95% CI: 0.57–0.77) and 0.70 (95% CI: 0.64–0.75), respectively, which indicated that only cIMT has limited effectiveness as a diagnostic tool for CAD screening [19]. In addition, since lipid metabolism disorders cannot be reflected directly, it is of limited value in the use of single cIMT in actual clinical application [20, 29]. Combing cIMT and lipid parameters could further improve the diagnostic efficiency of CAD. In recent years, research progress regarding lipid abnormalities has focused on the blood lipid indices, which can reflect lipid metabolism disorders better than a single one [30–32]. Wu et al found that multivariate logistic regression analysis showed that Apo B/Apo A-I ratio, a composite index, had a larger OR value than a single index (LDL-C or Apo B) and was better than a single lipid index in the prediction of CAD [30]. Similarly, in a cohort study by Rabizadeh et al, the analysis showed that LDL-C/Apo B ratio ≤ 1.2 can independently predict CAD (OR = 1.841, p = 0.002) [31]. In a cohort study conducted by Kappelle et al, Apo B/Apo A-I ratio and TC/HDL-C ratio were able to predict CAD and the first major adverse event during follow-up [32]. In the present study, we compared different lipid parameters between the CAD and non-CAD groups and found that Lp (a), TC/HDL-C, LDL-C/HDL-C, Apo B/Apo A-I, TG/HDL-C and LDL-C/Apo B ratios were significantly correlated with CAD. By comparing the non-CAD and CAD group, we found that the blood lipid parameters were significantly different, which was consistent with previous studies. In terms of effectively predicting CAD, the addition of lipid parameters on traditional risk factors model improved the accuracy of the estimates, and Lp (a) was the best.
Many studies have reported that elevated plasma Lp (a) is an independent risk factor for CAD, which mainly participates in the pathophysiological process of CAD via prothrombotic/anti-fibrinolytic effects and promoting the deposition of cholesterol in the vascular intima [33–35]. On an equimolar basis, Lp (a) is more likely to cause atherosclerosis than LDL because it not only contains all proatherogenic components of LDL-C but also those of Apo (a) [13]. Because the structure of Lp (a) has one more Apo (a) than that of LDL-C, its ability to enhance atherosclerotic thrombosis through other mechanisms including inflammation is stronger than that of LDL-C [34]. Elevated levels of LDL-C is strongly associated with the development of CAD [36]. The 2019 ESC guidelines also recommend using LDL-C as the primary indicator of lipid-lowering therapy [37]. In the baseline information of our study patients, the LDL-C of the CAD group was similar to that of the non-CAD group, possibly because the CAD group took more statins. In such condition, the levels of LDL-C were comparable in both groups, and the additional effect of Lp (a) could be investigated. Previous studies have shown that patients still have a certain risk of CAD after the treatment for reducing LDL-C, and this residual risk was related to elevated Lp (a) [38–42]. Among patients whose target blood lipids are normal or below the target, Lp (a) will play an important role in the prediction of CAD. This may explain our results that the AUC value of Lp (a) was greater than that of LDL-C-associated blood lipid indices in the ROC curve.
In this study, we also observed that elevated Lp (a) levels were correlated with multi-vessel disease. This result is consistent with a recent finding demonstrating that high Lp (a) was significantly associated with increased CAD severity, evaluated using the SYNTAX score [43]. The findings of the present study can help physicians to better assess the risk of CAD in their patients and may be useful for guiding primary prevention decisions.
This study has certain limitations. First, this was a retrospective study and no causal conclusion can be drawn. Second, the patients in study was highly suspected as CAD on admission, so the conclusion cannot be extrapolated to the general population. Third, the severity of CAD was based on the number of coronary artery lesions, and the corresponding degree of stenosis in each branch was not well quantified.