Extreme lipoprotein(a) levels and the risk of acute myocardial infarction by standard modifiable cardiovascular risk factors among US adults

Abstract

Background/Synopsis

Elevated lipoprotein(a) (Lp[a]) levels have been associated with increased acute myocardial infarction (AMI) risk, but no approved treatment options directly targeting Lp(a) have been approved yet. In addition, risk modification by standard modifiable cardiovascular risk factors (SMuRFs) (hypertension, dyslipidemia, diabetes, chronic kidney disease, alcohol use disorder, smoking [former/current], and BMI [underweight/overweight/obese]) is not well described. Consequently, there is a need to characterize patients with extremely high (XHI) vs low (LO) Lp(a) levels by number of SMuRFs in the real-world setting.

Objective/Purpose

To describe patients with XHI vs LO Lp(a) levels by number of SMuRFs and evaluate their risk of AMI.

Methods

Using natural language processing-enhanced data from the Veradigm Network EHR linked to Komodo Health closed claims, we identified patients with ≥ 1 valid Lp(a) lab test between 01/01/2016 - 01/31/2023 (earliest lab +30 days=index date). Inclusion criteria consisted of EHR/claims activity ≥ 13 months pre- and ≥ 12 months post-index with no severe kidney/liver dysfunction or malignant neoplasm during the study period, and no ischemic stroke, MI, or coronary revascularization at baseline. Selected patients were categorized into LO (<50th percentile) and XHI (>90th percentile) Lp(a) levels and stratified by baseline number of SMuRFs (0, 1, 2, 3, 4+). Inverse probability treatment weighting was used to balance baseline characteristics between Lp(a) cohorts.

Results

Of 22,289 included patients, final weighted cohorts had 2,233 XHI and 11,023 LO patients with a mean (SD) Lp(a) (nmol/L) of 303.9 (78.0) vs 21.1 (10.6) (p<0.0001), respectively. At baseline, total cholesterol, LDL-C, HDL-C, and triglycerides significantly differed between the Lp(a) cohorts (all p<0.0001) (Table 1). During a mean (SD) overall follow-up of 1,239 (642) days, AMI was rare and did not vary by Lp(a) cohorts (XHI: 1.6% vs LO: 1.5%). However, AMI incidence (0: 0.1%, 1: 0.2%, 2: 1.0%, 3: 2.0%, 4+: 4.4%) and proportion of patients with certain risk factors increased with the number of SMuRFs: (hypertension [1: 5.3%, 2: 44.7%, 3: 82.6%, 4+: 97.1%], dyslipidemia [1: 72.2%, 2: 91.2%, 3: 96.2%, 4+: 99.2%], and diabetes [1: 1.1%, 2: 7.8%, 3: 33.3%, 4+: 74.6%]) (Table 2).

Conclusions

Despite the lack of significant difference in incident AMI between the XHI and LO Lp(a) cohorts, the positive association between AMI incidence and baseline number of SMuRFs suggests that XHI Lp(a) patients have other risk factors for AMI and should be followed as a high-risk population to identify candidates for new Lp(a) targeted therapies.