Introduction

Lipoprotein(a) [Lp(a)] is an independent risk factor for atherosclerotic cardiovascular disease, acting by either accelerating atherosclerosis progression or inducing a prothrombotic/antifibrinolytic systemic milieu [1,2,3]. Several studies support the relevant role of Lp(a) in the occurrence of coronary events, especially in patients with premature coronary artery disease (CAD) [4,5,6]. However, the association of elevated Lp(a) levels with the risk of recurrent cardiovascular events in patients with a prior coronary event optimally treated with statins remains controversial [7]. We sought to assess the impact of Lp(a) levels on the recurrence of cardiovascular events in patients with premature CAD treated with percutaneous coronary intervention (PCI).

Methods

This prospective single-center study enrolled consecutive young patients (aged less than 50 years) undergoing first-ever PCI for stable CAD (SCAD) or acute coronary syndrome (ACS) (both ST-segment elevation myocardial infarction [STEMI] and non-ST-segment elevation acute coronary syndrome [NSTE-ACS]), from 2013 to 2017. As Lp(a) can behave as an acute-phase reactant, patients were screened for Lp(a) in clinically stable conditions at least 8 weeks after PCI, by blood samples collected after 12 h fasting. Lp(a) was measured at a single core laboratory using enzyme-linked immunosorbent assay, as previously reported [8]. Lp(a) concentration was reported in mg/dL, and a serum concentration ≥ 30 mg/dL was considered elevated [9]. Based on Lp(a) concentration, the study population was divided into three groups defined as ‘normal’ (< 30 mg/dL), ‘high’ (≥ 30 and < 60 mg/dL), or ‘very high’ (≥ 60 mg/dL) Lp(a). All patients were followed up with clinical visits or phone contact until June 2018. To evaluate the recurrence of major cardiovascular events at follow-up in the three study groups, a survival analysis was performed. The primary endpoint was a composite of cardiovascular death, non-fatal myocardial infarction, stroke, coronary revascularization, and hospitalization for cardiovascular causes. Secondary endpoints consisted of the individual components of the primary endpoint. We analyzed categorical variables by Chi square test or Fisher test, and continuous data by t test, ANOVA, and Mann–Whitney U test (as appropriate). Survival analysis was performed with Cox regression and log-rank test. Event rates were expressed as event per person-years. p values < 0.05 (two-tailed) were considered significant. Analyses were performed using R (R Foundation for Statistical Computing). The study protocol followed ethical guidelines of the Helsinki Declaration, and informed consent was obtained from all participants.

Results

We prospectively evaluated 63 consecutive patients with premature CAD undergoing PCI, receiving a diagnosis of SCAD (14.3%), STEMI (61.9%), or NSTE-ACS (23.8%). Mean follow-up was 3.4 years, not differing among study groups (p = 0.248). Population baseline and procedural characteristics are detailed in Table 1. None of the patients was on lipid-lowering therapy before the qualifying PCI. At the time of blood samples for Lp(a), all patients were on optimal medical therapy, receiving high-dose statins, ezetimibe, and fibrates in 100%, 11.1% and 6.3% of cases, respectively. In the overall population, mean Lp(a) levels were 31.3 mg/dL (median 22.0 mg/dL; interquartile range: 6.00-54.5), resulting elevated (≥ 30 mg/dL) in 42.9% of patients. Based on Lp(a) concentration, 57.1% of patients were allocated to ‘normal’ Lp(a) (< 30 mg/dL), 23.8% to ‘high’ Lp(a) (≥ 30 mg/dL and < 60 mg/dL), and 19.1% ‘very high’ Lp(a) (≥ 60 mg/dL) groups. The event rates (expressed as event per person-years) during follow-up was 0.023, 0.054, and 0.226 for ‘normal’, ‘high’ Lp(a), and ‘very high’ Lp(a) groups, respectively. Survival analysis showed a significantly higher rate of primary endpoint events in patients with ‘very high’ Lp(a) compared with those with ‘normal’ Lp(a) [hazard ratio (HR) 9.91; 95% CI 2.53–38.84; p < 0.001], but no significant difference between patients with ‘high’ versus ‘normal’ Lp(a) [(HR 2.36, 95% CI 0.47–11.76); p = 0.284)] (p value for log-rank test < 0.001). The Kaplan–Meier estimates showed a 2-year event-free survival rate for primary endpoint of 91.1% in ‘normal’ Lp(a) (95% CI 82.0–100%; 3 follow-up events), 79.4% in ‘high’ Lp(a) (95% CI 61.2–100%; 3 follow-up events), and 45.7% in ‘very high’ Lp(a) (95% CI 23.9–89%; 9 follow-up events) groups (Fig. 1). Results for secondary endpoints are reported in Table 2. An additional survival analysis, including Lp(a) as a continuous variable, confirmed that Lp(a) was an independent predictor of primary endpoint in the overall population (HR 1.03, 95% CI 1.01–1.04; p value = 0.003), and in STEMI patients (HR 1.03, 95% CI 1.01–1.05; p value = 0.006), but not in NSTE-ACS (HR 1.03, 95% CI 0.99–1.06; p value = 0.126) nor SCAD patients (HR 1.00, 95% CI 0.93–1.07; p value = 0.909).

Table 1 Population baseline and procedural characteristics
Full size table
Fig. 1
figure 1

Event-free survival curves for the primary endpoint. The study population was divided into three groups defined as ‘normal’ (< 30 mg/dL), ‘high’ (≥ 30 and < 60), and ‘very high’ (≥ 60 mg/dL) Lp(a) levels

Full size image
Table 2 Event-free survival rates at 2-year follow-up for primary and individual secondary endpoints
Full size table

Discussion

In the present study, we investigated the prognostic impact Lp(a) in patients with premature CAD treated with PCI. We showed that elevated Lp(a) can be found in a not negligible proportion of patients with premature CAD, while very high Lp(a) levels (above the threshold of 60 mg/dL) was associated with a higher recurrence of cardiovascular events compared with lower levels. Our findings suggest the importance of Lp(a) measurement in young ‘high-risk’ patients to improve risk stratification, and potentially influence dedicated therapeutic strategies (i.e., lipoprotein apheresis). Several considerations can be made regarding secondary endpoints analysis. First, no cardiovascular death occurred probably due to the young age of patients; a more extended follow-up might be necessary to detect differences in mortality. Second, ‘hard’ coronary events (namely new myocardial infarction and/or coronary revascularization), exclusively occurred in patients with Lp(a) higher than 60 mg/dL. Patients with lower Lp(a), only experienced hospitalization for cardiovascular cause without receiving any coronary interventions, possibly suggesting a lower risk for CAD progression. Moreover, a sub-group analysis confirmed Lp(a) as an independent predictor of outcomes in STEMI, but not in NSTE-ACS and SCAD patients. This result probably led the low number of patients and events in each sub-group, although a different prognostic impact of Lp(a) per clinical presentation cannot be excluded. Numerous studies indicated Lp(a) as an independent risk factor for atherosclerotic disease [1, 5, 10], and similar to low-density lipoprotein cholesterol (LDL-C), this relation seemed to be continuous [4, 9]. However, evidence on the prognostic role of Lp(a) in patients with previous coronary events treated with optimal medical therapy (mainly statins) remains conflicting [7, 9]. A recent study-level analysis of three large trials suggested that high Lp(a) levels were similarly associated with a higher residual cardiovascular risk in patients on statins and with controlled LDL-C [9]. Conversely, a case–cohort analysis of the dal-Outcomes trial failed to demonstrate any associations between Lp(a) and the risk of recurrent cardiovascular events [7], questioning the use of specific Lp(a)-lowering therapies in patients on statins. In the dal-Outcomes trial [7], mean patients’ age was 63 years, possibly masking (at least in part) the impact of Lp(a) on outcomes. Indeed, in a previous study, elevated Lp(a) (> 50 mg/dL) was associated with a significant threefold increase in the risk of coronary events in patients aged less than 45 years [5]. However, this association was weaker (twofold increased risk) in individuals with 45–60 years and entirely abolished after 60 years, suggesting a negative trend with aging [5]. In line with these results, our data showed a significant impact of Lp(a) on the recurrence of events in patients with a premature CAD, suggesting an ‘age dependency’ in the effect of Lp(a) on cardiovascular outcomes. Recent guidelines highlighted the importance of Lp(a) testing in patients with early cardiovascular disease presentation, to quantify more accurately their risk [11]. Our results confirm the relevance of Lp(a) screening in patients with premature CAD to stratify the risk of future events further and potentially identify candidates to Lp(a)-lowering strategies. In recent years, a growing awareness has emerged in the importance of developing effective Lp(a)-lowering drugs. Oral lipid-lowering medications demonstrated modest effects in lowering the Lp(a) levels [9], and only niacin showed to reduce Lp(a) of ~ 30% [9]. Recent reports also indicated a ~ 30% reduction in Lp(a) levels with the use of proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors [12, 13]. To date, lipoprotein apheresis is the only treatment reporting consistent Lp(a) levels reduction and potential benefits on cardiovascular outcomes [14]. However, in young patients with hyperlipoproteinemia, the decision to undergo apheresis is of great psychological and social impact, and informing patients requires the knowledge of principles of informative counseling. Novel antisense oligonucleotides targeting apolipoprotein(a) showed a potent Lp(a)-lowering effect (60–90% reduction) and this sounded promising for the future [15], but additional data are needed to confirm their efficacy and safety.

Our study has several limitations. First, the modest sample size (due to the difficulties in recruiting these special patients), and single-center design might be potential sources of bias, limiting the generalizability of our results. Second, other risk factors for premature CAD (including homocysteine, and genetic prothrombotic risk factors) were not evaluated, but evidence suggests that they have only a minor contribution in the development of early CAD [4, 5]. Third, Lp(a) measurements were performed in patients on lipid-lowering therapy. Although this approach might have influenced Lp(a) levels, their impact on Lp(a) concentration is relatively modest as discussed.

In conclusion, our findings suggest that elevated levels of Lp(a) can be detected in a consistent proportion of patients with premature CAD undergoing PCI in real-world practice. Moreover, ‘very high’ Lp(a) levels (above the threshold of 60 mg/dL) showed a significant association with recurrent cardiovascular events in these patients, mainly due to new myocardial infarction and coronary revascularization. Hence, systematic screening for elevated plasma Lp(a) can help clinicians in the understanding and management of patients with premature coronary atherosclerosis by (1) improving prognostic assessment; (2) intensifying the control of traditional risk factors; and (3) proposing additional Lp(a)-lowering treatment, including lipoprotein apheresis.