Association of Lipoprotein (a) and Standard Modifiable Cardiovascular Risk Factors With Incident Myocardial Infarction: The Mass General Brigham Lp(a) Registry.
May 18, 2024DOI: 10.1161/jaha.123.034493Journal ArticleAHA
Written by: Arthur ShiyovichAdam N BermanStephanie A BesserDavid W BieryGurleen KaurSanjay DivakaranAvinainder SinghDaniel M HuckBrittany WeberJorge PlutzkyMarcelo F Di CarliKhurram NasirChristopher CannonJames L JanuzziDeepak L BhattRon Blankstein
Find these authors on Roon
AS
Dr. Arthur Shiyovich
Advanced Cardiac Imaging · PHS
AB
Dr. Adam Berman
Cardiac Critical Care · NYU Langone
METHODS
The patient population was derived from the Mass General Brigham Lp(a) Registry, as previously described.29, 30 In summary, this is a retrospective cohort study, which included all patients who had undergone Lp(a) testing as part of their routine care between January 2000 and July 2019 and was conducted at 2 large academic medical centers in Boston, Massachusetts (Brigham and Women's Hospital and Massachusetts General Hospital). The Mass General Brigham Lp(a) Registry was granted approval by the Institutional Review Board at Mass General Brigham, and the requirement for informed consent has been waived. All individuals aged ≥18 years with at least 1 Lp(a) result were screened for inclusion in the cohort. Exclusion criteria consisted of 2 factors: (1) severe kidney dysfunction, which was defined as stage 5 chronic kidney disease (estimated glomerular filtration rate <15 mL/min per m2), prior renal transplant, or those receiving renal replacement therapy; or (2) the presence of a diagnostic International Classification of Diseases (ICD) (ICD‐9, ICD‐10) code for a malignant neoplasm during the covariate assessment window, except for nonmelanoma skin cancer. Additionally, for the current study, patients with a history of ASCVD, defined as a previous myocardial infarction, history of coronary revascularization, or a history of ischemic stroke, were also excluded.Because of the sensitive nature of the data collected for this study, requests to access the data set from qualified researchers trained in human subject confidentiality protocols may be sent to the corresponding author.Three primary sources were used to collect the data, including: (1) the RPDR (Research Patient Data Registry)31 at Mass General Brigham, which provides demographic, laboratory, imaging, diagnostic, procedural, medication, vital status (based on the Social Security Administration Death Master File), and clinical documentation for individuals who meet specific search criteria; and (2) ICD‐coded death information from the National Death Index or the Massachusetts Office of Vital Statistics was used to establish the causes of death for each patient who passed away during the study period.To determine the presence of cardiovascular risk factors, validated natural language processing (NLP) modules,32 laboratory data, and diagnostic and procedural ICD‐9, ICD‐10, and Current Procedural Terminology codes were used, as previously described.29
The baseline covariate assessment period was established as the 12‐month period before and 30 days after the Lp(a) measurement. For individuals with >1 Lp(a) test, their covariate assessment period was assessed relative to the first Lp(a) test. Any clinical covariate that occurred or was documented during this time was attributed to baseline characteristics.
The definitions used to define SMuRFs were as follows. Diabetes was classified on the basis of the presence of a combination of diagnostic ICD codes, NLP, and hemoglobin A1c results. Hypertension was defined on the basis of the presence of a combination of diagnostic ICD codes and NLP. Non‐Lp(a) dyslipidemia was determined through NLP, treatment with a cholesterol‐lowering medication, or laboratory values (median) that exceeded any 1 of the following thresholds during the covariate window: (1) total cholesterol ≥240 mg/dL, (2) LDL cholesterol (LDL‐C) ≥160 mg/dL, (3) high‐density lipoprotein cholesterol <40 mg/dL (men), (4) high‐density lipoprotein cholesterol <50 mg/dL (women), or (5) total triglycerides ≥175 mg/dL. Smoking was classified as current, former, or never smoker by an NLP algorithm that was proven to be highly accurate in our institution.33 Further details on the covariate assessment, ascertainment, and adjudication are found in Data S1.
Lp(a) was measured as part of routine medical care using available immunochemical‐based assays with reference ranges <30 mg/dL or <75 nmol/L.34, 35, 36 All Lp(a) laboratory testing was performed at commercial laboratories over the study period using industry‐standard assays. We converted all Lp(a) values to nmol/L using the following conversion formula: Lp(a) nmol=[2.18×Lp(a)‐M]−3.83.37, 38 To avoid potential biases attributable to possible differences in Lp(a) testing techniques over the study period, percentile distributions for each assay were defined separately, as previously also performed by others.37, 39, 40, 41 Subsequently, Lp(a) percentiles were then combined across assay types.
Given the well‐established distribution of Lp(a), a priori percentiles were determined and used in the present study: 1st to 50th, 51st to 70th, 71st to 90th, and 91st to 100th.21, 42, 43 Accordingly, high Lp(a) was defined as a value >90th percentile in the cohort, and low Lp(a) was defined as a value <50th percentile.
The primary outcome for the current study was the occurrence of AMI during the follow‐up period. As previously described,29 AMI was defined by the presence of a diagnostic ICD code in the primary hospital discharge position. This method has been thoroughly validated and is associated with high specificity, high positive predictive value, and reasonable sensitivity.44, 45, 46, 47, 48
Continuous variables are reported as medians, and categorical variables are reported as frequencies and percentages. Ordinal variables were compared using Jonckheere‐Terpstra trend test. Cumulative incidence of AMI was compared according to the study variables using Kaplan‐Meier approach. The relative risk of AMI was assessed with univariate and multivariable levels of Cox proportional hazard models, presented as hazard ratio (HR) and 95% CI. We evaluated multicollinearity among the independent variables using Spearman correlation coefficients. The correlations between variables were all <0.7, indicating no evidence of substantial multicollinearity in the multivariable model.
The proportional hazards assumption was tested using Schoenfeld residuals and found to be valid (P>0.05), indicating that the HRs for the covariates remained constant over time. Incidence rate ratios were calculated when comparing the rates of cardiovascular events between different levels of Lp(a). The χ2 goodness‐of‐fit test for multinominal data was used to calculate the P values for low and high Lp(a). Patients were censored either upon their death or at the conclusion of the query period, which extended until July 2019. We used the Fine‐Gray model to evaluate the potential for competing risks of death. Tests were 2 tailed and considered statistically significant with a P<0.05. All statistical analyses were conducted using STATA MP, version 17 (College Station, TX).
RESULTS
A total of 6238 individuals met the established inclusion and exclusion criteria. The mean age of the patients was 54 years (interquartile range, 43–65 years), and 45% were women. Table 1 provides the baseline characteristics of the cohort. The distribution of the number of SMuRFs in the study cohort is shown in Figure 1. It is noteworthy that 23.7% of the study cohort did not have any SMuRFs, whereas 4.3% had all 4 modifiable risk factors. For subsequent analyses, patients who had 4 SMuRFs were combined with those who had 3 SMuRFs. Figure S1 displays the distribution of Lp(a) percentiles in the study cohort. Of the total cohort, 436 (7%) patients had high Lp(a), whereas 3618 (58%) had low Lp(a). Table S1 presents a comparison of baseline characteristics between patients with no SMuRFs and those with a least 1 SMuRF. Patients without SMuRFs were younger, with comparable rates of women and racial or ethnic distribution, and exhibited a more favorable lipid profile compared with individuals with at least 1 SMuRF. Figure 2 displays the rates of high and low Lp(a) according to the number of SMuRFs. The figure shows a gradual increase in the rates of high Lp(a) and a gradual decrease in the rate of low Lp(a) as the number of SMuRFs increases.
Baseline Characteristics of the Study Population
IQR indicates interquartile range.
At time of lipoprotein (a) test.
Other includes Indian, Middle Eastern, Native American, other, Pacific Islander, and unknown.
*Percentages presented are out of total number of patients in every SMuRF category. High Lp(a) was defined as a value >90th percentile in the cohort, and low Lp(a) as <50th percentile.
Over a median follow‐up of 8.8 years (interquartile range, 4.2–12.8 years; range, 0.02–19.2 years), 234 (3.75%) patients experienced an AMI. Among the entire cohort and when separately examining patients with high and low Lp(a), there was an increase in the risk of incident AMI with a higher number of SMuRFs (Figure 3, Figure S2). However, in all subgroups of SMuRFs, the incidence of AMI was considerably higher among patients with high Lp(a) compared with those with low Lp(a), as shown in Figure 3. Notably, as the number of risk factors increased, the absolute difference in the incidence of AMI between patients with high Lp(a) versus those with low Lp(a) became greater. When evaluating the incidence rates of these subgroups, the risk associated with elevated Lp(a) was similar to having 2 additional SMuRFs. Table S2 shows the hazard ratios of the individual SMuRFs for developing incident myocardial infarction (MI). Notably, all SMuRFs were significantly associated with increased risk of AMI with an increasing risk for AMI with higher number of SMuRFs. Similarly, high Lp(a) levels [versus low Lp(a)] were significantly associated with increased risk of AMI (Table S2, Figure S3). After adjusting for age, sex, and the number of SMuRFs in a multivariable Cox model (Table 2), high Lp(a) remained significantly associated with the first occurrence of AMI, with an HR of 2.9 (95% CI, 2.0–4.3), and a P<0.001. This was similar to the HR associated with having 2 SMuRFs (HR, 2.97; P=0.001). The cumulative incidence function of MI was significantly higher in patients with high Lp(a) than in patients with low Lp(a) after adjusting for other causes of death as competing risk (P<0.001).
Incidence rates are presented with 95% CIs.
Multivariable Cox Model Adjusted for Age and Sex
HR indicates hazard ratio; Lp(a), lipoprotein (a); and SMuRF, standard modifiable risk factor.
Significance was considered when P<0.05.
Increase of 1 year.
DISCUSSION
In this retrospective contemporary cohort study of patients who did not have prior ASCVD, we evaluated the risk of first AMI associated with high Lp(a) relative to SMuRFs. The key findings were that: (1) high Lp(a) was significantly associated with an increased risk of first AMI regardless of the presence or number of SMuRFs, and this association remained highly significant after adjusting for risk factors; and (2) the excess risk imparted by having a high Lp(a) was similar to the risk of having 2 SMuRFs.
The association between Lp(a) and AMI or CAD has been extensively studied through various epidemiologic and genetic investigations in different populations. Indeed, the association between Lp(a) and CAD has been supported by multiple genetic studies, which have provided strong mechanistic evidence for this association.20, 49, 50 In addition, epidemiologic studies have demonstrated a robust and independent graded association between Lp(a) and CAD or AMI.20, 21, 39, 51, 52, 53 Although some of these studies have accounted for baseline characteristics and standard risk factors, the risk associated with elevated Lp(a) in relation to other risk factors has not been fully explored, and prior studies have not focused on modifiable risk factors.
Recently, several studies have examined the cardiovascular risk associated with increased Lp(a) relative to various individual risk factors. For example, elevated Lp(a) was found to be linked to an increased risk for MI, stroke, and cardiovascular mortality in patients with diabetes and prediabetes in a stepwise association.54 Mehta et al55 studied >12 000 participants in the ARIC (Atherosclerosis Risk in Communities) study and found that elevated Lp(a) and family history of coronary heart disease had independent and additive joint associations with the risk of coronary heart disease. In another study, Verbeek et al56 evaluated the prognostic association between Lp(a) and LDL‐C in a primary prevention setting and found that the risk associated with elevated Lp(a) was attenuated among patients with LDL‐C levels <100 mg/dL. However, the justification for the use of statins in prevention: an intervention trial evaluating rosuvastatin57 demonstrated that the association between baseline Lp(a) and first incident cardiovascular event was similar at LDL‐C levels above and below the median (110 mg/dL) and even at low levels (54 mg/dL).
Trinder et al58 investigated whether apolipoprotein B could explain the risk of CAD associated with Lp(a) particles, but found that the concentration of apolipoprotein B could only explain the association of LDL‐C with CAD, not that of Lp(a).58 Hedegaard et al59 conducted a study to contextualize the risk associated with Lp(a) by comparing it with that of familial hypercholesterolemia. According to their findings, the risk associated with Lp(a) becomes clinically significant at a level of 30 mg/dL, and the strength of the association was similar to that of familial hypercholesterolemia at levels which exceeded 70 mg/dL.
Thomas et al21 reported that Lp(a) is a significant contributor to ASCVD and MI regardless of CRP (C‐reactive protein) levels. Furthermore, Lp(a) was found to be an independent prognostic marker among patients with acute coronary syndrome without SMuRFs, potentially explaining their worse outcomes.28 Previous evidence suggests that inflammation and thrombosis may play a significant role in mediating these adverse outcomes.22, 60, 61
The current study builds on and extends previous research in several ways. First, it focuses on the role of SMuRFs and Lp(a) among a primary prevention population who did not have prior ASCVD. Second, it offers a simple estimation of risk that can enable clinicians to understand the interplay between Lp(a) and the number of SMuRFs in assessing the risk of MI. For instance, for any given number of risk factors (ie, SMuRF category), having a high Lp(a) was associated with a nearly 3× higher level of risk of incident MI. Moreover, having a high Lp(a) was associated with a similar risk level as having 2 additional modifiable risk factors.
Given the robust risk of Lp(a) in risk assessment observed in this study, our results reinforce the European and Canadian Dyslipidemia guidelines that advocate for wider Lp(a) screening, wherein Lp(a) is measured at least once in every adult to enhance risk stratification and therapeutic decision‐making.4, 24, 62 Our results support the role of Lp(a) testing even among individuals with minimal or no cardiovascular risk factors, because high Lp(a) levels can be linked to a substantially elevated risk. Additional assessment of other nonstandard risk factors was advised for these patients.17 Although there are currently no effective therapies for lowering Lp(a), patients who are found to have higher risk should be aggressively managed with respect to their modifiable risk factors, especially LDL‐C.63 Indeed, our results show that even in the presence of high Lp(a), the number of SMuRFs is strongly associated with risk (ie, 4× increase in risk between those with ≥3 risk factors versus those with no SMuRFs). These data suggest that controlling modifiable risk factors likely has an important role on risk reduction.
The current study has certain limitations. First, because of its retrospective design, risk factors were not assessed in a standard manner, and there is possible unmeasured confounding. Additionally, the evaluation of risk factors was conducted around the time of Lp(a) measurement, which was a priori defined as the baseline covariate window.29 However, relying on a single baseline period may underestimate the prevalence of SMuRFs, as patients may develop additional risk factors over additional follow‐up time up until the incidence of AMI. Underestimation, and consequently undertreatment, of SMuRFs is considered 1 of the proposed mechanisms contributing to poorer outcomes observed in SMuRFless patients, as reported in numerous other studies.7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 64 Second, our cohort comprised patients who were tested for Lp(a) as part of routine clinical care, which may introduce selection bias. However, we focused on patients without prior ASCVD who were referred for testing, and as such these results remain generalizable to the population of patients in whom Lp(a) is currently tested in the United States. Nevertheless, the generalizability of our findings may be limited in other cohorts and, thus, further investigation is warranted. Moreover, there is a possibility that the risk observed among high‐risk patients [those with a high prevalence of SMuRFs or high Lp(a)] was attenuated through the management of these risk factors. Third, the registry used for this study represents a single geographic region of the United States, which could limit the generalizability of the findings, particularly as our patient population was predominantly White race. Fourth, Lp(a) was measured by different techniques in our cohort; however, we standardized the assays using percentiles, as also done by prior studies. Nevertheless, there is a need for additional validation of conversion methods as well as further data comparing the performance of different types of Lp(a) assays. Fifth, for practical and simplicity reasons, our risk factors were assessed in a binary manner, and thus we did not account for the severity of various SMuRFs (eg, LDL level, blood pressure); however, our approach has been used in the vast majority of observational cohorts. Sixth, we have included low high‐density lipoprotein as a component of modifiable dyslipidemia, which may be controversial.
CONCLUSIONS
This retrospective analysis from a large contemporary Lp(a) registry demonstrated that high Lp(a) levels are significantly associated with an increased risk for first AMI, regardless of the presence or number of SMuRFs. The study also revealed that the incremental risk posed by high Lp(a) was similar to that of having 2 additional SMuRFs. These findings could be useful in risk assessment and in guiding currently available preventive therapies.
Sources of Funding
The study was funded in part by Amgen Inc.
Disclosures
Dr Weber serves on the scientific advisory boards for Kiniksa, Horizon Therapuetics, and Novo Nordisk. Dr Cannon reports in calendar years 2021 to 2023: research grants from Amgen, Better Therapeutics, Boehringer‐Ingelheim (BI), Daiichi Sankyo, Merck, Novo Nordisk, and Pfizer; consulting fees from Amryt/Chiesi, Alnylam, Amarin, Amgen, Applied Therapeutics, Ascendia, Biogen, BI, BMS, CSL Behring, Eli Lilly, Janssen, Lexicon, Merck, Milestone, Pfizer, Rhoshan, and SanofiI; and serving on Data and Safety Monitoring Boards for Applied Therapeutics, Areteia, and Novo Nordisk. Dr Januzzi is a trustee of the American College of Cardiology; is a board member of Imbria Pharmaceuticals; is a director at Jana Care; has received grant support from Abbott, Applied Therapeutics, HeartFlow, Innolife, and Roche Diagnostics; has received consulting income from Abbott, Beckman, Boehringer Ingelheim, Bristol Myers Squibb, Janssen, Merck, Novartis, Pfizer, Roche Diagnostics, and Siemens; and participates in clinical end point committees/data safety monitoring boards for Abbott, AbbVie, CVRx, Intercept, and Takeda. Dr Di Carli reports grant support from Amgen. Dr Bhatt discloses the following relationships: advisory board: Angiowave, Bayer, Boehringer Ingelheim, CellProthera, Cereno Scientific, Elsevier Practice Update Cardiology, High Enroll, Janssen, Level Ex, McKinsey, Medscape Cardiology, Merck, MyoKardia, NirvaMed, Novo Nordisk, PhaseBio, PLx Pharma, and Stasys; board of directors: American Heart Association New York City, Angiowave (stock options), Bristol Myers Squibb (stock), DRS.LINQ (stock options), and High Enroll (stock); consultant: Broadview Ventures, Hims, SFJ, and Youngene; Data Monitoring Committees: Acesion Pharma, Assistance Publique–Hôpitaux de Paris, Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute, for the PORTICO trial, funded by St. Jude Medical, now Abbott), Boston Scientific (Chair, PEITHO trial), Cleveland Clinic, Contego Medical (Chair, PERFORMANCE 2), Duke Clinical Research Institute, Mayo Clinic, Mount Sinai School of Medicine (for the ENVISAGE trial, funded by Daiichi Sankyo; for the ABILITY‐DM trial, funded by Concept Medical; for ALLAY‐HF, funded by Alleviant Medical), Novartis, Population Health Research Institute, and Rutgers University (for the National Institutes of Health–funded MINT Trial); honoraria: American College of Cardiology (Senior Associate Editor, Clinical Trials and News, ACC.org; Chair, ACC Accreditation Oversight Committee), Arnold and Porter law firm (work related to Sanofi/Bristol‐Myers Squibb clopidogrel litigation), Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute; RE‐DUAL PCI clinical trial steering committee funded by Boehringer Ingelheim; AEGIS‐II executive committee funded by CSL Behring), Belvoir Publications (Editor in Chief, Harvard Heart Letter), Canadian Medical and Surgical Knowledge Translation Research Group (clinical trial steering committees), CSL Behring (AHA lecture), Cowen and Company, Duke Clinical Research Institute (clinical trial steering committees, including for the PRONOUNCE trial, funded by Ferring Pharmaceuticals), HMP Global (Editor in Chief, Journal of Invasive Cardiology), Journal of the American College of Cardiology (Guest Editor; Associate Editor), K2P (Co‐Chair, interdisciplinary curriculum), Level Ex, Medtelligence/ReachMD (CME steering committees), MJH Life Sciences, Oakstone CME (Course Director, Comprehensive Review of Interventional Cardiology), Piper Sandler, Population Health Research Institute (for the COMPASS operations committee, publications committee, steering committee, and US national coleader, funded by Bayer), WebMD (CME steering committees), and Wiley (steering committee); other: Clinical Cardiology (Deputy Editor); patent: sotagliflozin (named on a patent for sotagliflozin assigned to Brigham and Women's Hospital, who assigned to Lexicon; neither I nor Brigham and Women's Hospital receive any income from this patent); research funding: Abbott, Acesion Pharma, Afimmune, Aker Biomarine, Alnylam, Amarin, Amgen, AstraZeneca, Bayer, Beren, Boehringer Ingelheim, Boston Scientific, Bristol‐Myers Squibb, Cardax, CellProthera, Cereno Scientific, Chiesi, CinCor, Cleerly, CSL Behring, Eisai, Ethicon, Faraday Pharmaceuticals, Ferring Pharmaceuticals, Forest Laboratories, Fractyl, Garmin, HLS Therapeutics, Idorsia, Ironwood, Ischemix, Janssen, Javelin, Lexicon, Lilly, Medtronic, Merck, Moderna, MyoKardia, NirvaMed, Novartis, Novo Nordisk, Otsuka, Owkin, Pfizer, PhaseBio, PLx Pharma, Recardio, Regeneron, Reid Hoffman Foundation, Roche, Sanofi, Stasys, Synaptic, The Medicines Company, Youngene, and 89Bio; royalties: Elsevier (Editor, Braunwald's Heart Disease); site coinvestigator: Abbott, Biotronik, Boston Scientific, CSI, Endotronix, St. Jude Medical (now Abbott), Philips, SpectraWAVE, Svelte, and Vascular Solutions; trustee: American College of Cardiology; unfunded research: FlowCo. Dr Blankstein has received research support from Amgen Inc, Novartis Inc,, and has served as a consultant/advisory board for Caristo Inc, Elucid Inc, Hearflow Inc, Nanox AI, and Beren Therapeutics. The remaining authors have no disclosures to report.