Lipoprotein-Associated Phospholipase A2: An Independent Predictor of Coronary Artery Disease Events in Primary and Secondary Prevention

Assessment of Risk Using Lipoprotein–Associated Phospholipase A2 in Primary Prevention Studies 

The key clinical trials that have evaluated the relative risk and the hazard or odds ratios for Lp-PLA2 in primary and secondary prevention studies are shown in Table 1 and Table 2.1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25 Hazard and odds ratios generally express the risk of a cardiovascular event based on being in the top quantile compared with the bottom quantile of the Lp-PLA2 level (eg, top tertile vs bottom tertile or top quartile vs bottom quartile). Relative risk in these trials is generally defined as the relative risk per 1 standard deviation increase in Lp-PLA2 level as a continuous variable. These trials demonstrate generally consistent and statistically significant correlations between elevated levels of Lp-PLA2 and increased risks for cardiovascular events. It should be noted that the results reflect a full multivariable adjustment for traditional risk factors, indicating the independent contribution of Lp-PLA2 beyond traditional risk factors in risk assessment. Overall, elevated Lp-PLA2 is associated with roughly a doubling of risk, with some attenuation in the elderly population.

Table 1. Elevated lipoprotein-associated phospholipase A2 (Lp-PLA2) as a predictor of first coronary events in primary prevention in 12 studies⁎

	Study	CV End Point	Year	Population	#Cases/#Controls	Relative Risk (95% CI)†	Hazard/Odds Ratio (95% CI)†
	WOSCOPS1	Coronary events	2000	Men with hypercholesterolemia	580/1,160	1.18(1.05–1.33)	1.80(1.3–2.6)
	WHS2	Coronary events	2001	Healthy women	123/123	1.17(0.5–3.0)	—
	ARIC4	Coronary events	2004	Healthy subjects	608/740	—	1.15(0.8–1.6)
				LDL <130 mg/dL‡	203/507		2.08(1.2–3.6)
	Winkler et al6	Severe CAD	2004	Patients with type 2 diabetes mellitus	23/66	—	2.09(1.0–4.2)
	MONICA5	Coronary events	2004	Men with moderate hypercholesterolemia	97/837	1.21(1.01–1.5)	—
	Rotterdam Study8	Coronary events	2005	Healthy subjects aged >55 years	308/1,820	—	1.97(1.13–3.0)
	PROSPER13	Coronary events	2006	Elderly subjects	856/4,801	—	1.25(1.02–1.5)
	Cardiovascular Health Study14	MI	2006	Elderly subjects	504/4,318	—	1.26(1.03–1.6)
	Rancho Bernardo Study15	CAD events	2006	Elderly subjects	1,077	—	1.64(1.1–2.6)
	Malmo Study24	MI and stroke	2007	Subjects without diabetes	262/4,480	1.16(1.01–1.33)	1.54(1.1–2.2)
	Bruneck Study21	CV events	2007	Healthy subjects	82/765	1.4(1.1–1.4)	—
	Nurse's Health Study25	MI	2007	Healthy women	431/858	—	1.76(1.1–2.8)

ARIC = Atherosclerosis Risk in Communities; CAD = coronary artery disease; CI = confidence interval; CV = cardiovascular; LDL = low-density lipoprotein; MI = myocardial infarction; MONICA = Monitoring of Trends and Determinants in Cardiovascular Disease; PROSPER = Prospective Study of Pravastatin in the Elderly at Risk; WHS = Women's Health Study; WOSCOPS = West of Scotland Coronary Prevention Study.

⁎ All models were multivariate adjusted for all traditional risk factors, lipids and C-reactive protein, and sometimes for other inflammatory markers. † Risk ratios and 95% CIs are expressed as relative risk per 1 standard deviation or as hazard/odds ratios based on levels in the highest versus lowest quantile (tertile, quartile, etc.) of the Lp-PLA2 biomarker. ‡ 1 mg/dL = 0.02586 mmol/L.

Table 2. Elevated lipoprotein-associated phospholipase A2 (Lp-PLA2) as a predictor of incident cardiovascular (CV) events or severe coronary artery disease (CAD) in 13 prospective secondary prevention studies⁎

	Study	CV End Point	Year	Population	#Cases/#Controls	Relative Risk (95% CI)†	Hazard/Odds Ratio (95% CI)†
	AtheroGENE3	ACS and angina	2003	CAD	496/477	—	1.80(1.01–3.20)
	Mayo Heart Study7	Coronary events	2005	CAD	61/466	1.3(1.06–1.59)	2.29(1.12–4.68)
	LURIC9	Severe CAD	2005	Angio patients	2,454/694	—	1.85(1.23–2.78)
	HELICOR10	Severe CAD	2005	Angio patients	312/479	—	1.91(1.12–3.28)
	KAROLA11	Recurrent CV	2005	S/P ACS or Revascularization	95/1,051	—	2.09(1.10–3.96)
	Intermountain Heart Study12	CAD	2006	Angio patients	475/1,012	—	2.44(1.58–3.79)
	THROMBO17	Recurrent MI	2006	Post MI	766	—	1.90(1.31–2.75)
	Mayo (Olmsted) Study18	Death after MI	2006	Acute MI	42/229	—	4.93(2.10–11.6)
	PROVE-IT30	Recurrent CV events	2006	ACS	3,265	—	1.33(1.01–1.74)
	VA-HIT20	CV events	2006	Stable CAD	927	1.13(p = 0.018)	—
	GUSTO & FRISC21	Recurrent CV events	2007	ACS	435/2,266	—	1.40(0.77–2.5)
	NOBIS-II22	Coronary events	2007	Chest pain	56/429	—	2.60(1.1–6.6)
	PEACE19	MI and stroke	2007	Stable CAD	1,108/3,766	—	1.41(1.17–1.70)

ACS = acute coronary syndromes; angio = angiography; CI = confidence interval; FRISC = Fragmin During Instability in Coronary Artery Disease; GUSTO = Global Use of Strategies to Open Occluded Coronary Arteries; KAROLA = Langzeiterfolge der Kardiologischen Anschlussheil-Behandlung; LURIC = Ludwigshafen Risk and Cardiovascular Health Study; MI = myocardial infarction; NOBIS-II = North Wuerttemberg and Berlin Infarction Study-II; PROVE-IT = Pravastatin or Atorvastatin and Infection Therapy; PEACE = Prevention of Events with Angiotensin-Converting Enzyme Inhibition; S/P = status post; THROMBO = Thrombogenic Factors and Recurrent Coronary Events; VA-HIT = Veterans Affairs HDL Intervention Trial.

⁎ All models were multivariate adjusted for all traditional risk factors, lipids and C-reactive protein, and sometimes for other inflammatory markers. † Risk ratios and 95% CIs are expressed as relative risk per 1 standard deviation or as hazard/odds ratios based on levels in the top quantile to bottom quantile of the Lp-PLA2 biomarker. Major population studies evaluating Lp-PLA2 as an independent predictor of future CV events have shown that Lp- PLA2 generally indicates an approximate 2-fold increased risk.

Prospective Epidemiologic Studies of Elevated Lipoprotein-Associated Phospholipase A2 as a Predictor of Coronary Events in Primary Prevention 

As listed in Table 1, there have been several studies of Lp-PLA2 elevation as a predictor of coronary events in primary prevention. The first major study evaluating Lp-PLA2 as a risk marker was the West of Scotland Coronary Prevention Study (WOSCOPS), a trial evaluating the role of pravastatin in the prevention of coronary events that was published in 2000.1 WOSCOPS provided the first evidence that Lp-PLA2 could be a potential biomarker for cardiovascular risk assessment. As shown in Figure 1, in WOSCOPS, both high-sensitivity CRP (hs-CRP) and Lp-PLA2 are correlated by quintiles with an increase in cardiovascular risk, with a statistically significant risk increase above the 40th percentile (2nd quintile). In the third, fourth, and fifth quintiles of Lp-PLA2 elevation, there is almost a doubling of risk for cardiovascular events. This risk was not confounded by traditional risk factors or hs-CRP or leukocyte count, as demonstrated by multivariable adjustment. Although there were weak-to-modest but significant associations of Lp-PLA2 levels with total and LDL cholesterol levels, the Lp-PLA2 level was independent of age, body mass index, systolic and diastolic blood pressure, and even smoking status. As the risk ratio within each quintile is adjusted for progressively more traditional risk factors, Lp-PLA2 continues to demonstrate a consistent relation with risk, whereas the risk associated with elevated hs-CRP is markedly attenuated as more traditional risk factors are added into the model.1

View Large Image Download to PowerPoint

Figure 1. Association of C-reactive protein (CRP) and lipoprotein-associated phospholipase A2 (Lp-PLA2) levels with the risk of a coronary event. In the case-cohort West of Scotland Coronary Prevention Study (WOSCOPS), both high-sensitivity CRP (hs-CRP) and Lp-PLA2 were associated with coronary events in univariate analysis. After full adjustment for age, blood pressure, lipids, and other inflammatory markers, hs-CRP was no longer a significant predictor of coronary events. However, the significant association of Lp-PLA2 with risk persisted in the fully adjusted model. BP = blood pressure. (Adapted from N Engl J Med.1)

Another large prospective study, the Atherosclerosis Risk in Communities (ARIC) trial, found that Lp-PLA2 levels were higher in middle-aged US men and women who subsequently developed coronary artery disease (CAD) events than in those who remained free of CAD.4 Individuals with incident CAD had a significantly higher body mass index, systolic blood pressure, and levels of total cholesterol, triglycerides, and LDL cholesterol, and significantly lower HDL cholesterol levels than patients without incident CAD. Hypertension, diabetes mellitus, and current smoking were more prevalent in CAD cases. In addition to these differences in traditional risk factors, the weighted mean levels of both Lp-PLA2 and hs-CRP were higher in patients with CAD events than in those without events. Weak-to-modest correlations of Lp-PLA2 levels to lipids were noted, with positive correlations to LDL cholesterol and total cholesterol levels and a negative correlation to HDL cholesterol level, in both men and women. Lp-PLA2 levels also correlated weakly with triglycerides. Importantly, there was no significant correlation between Lp-PLA2 levels and body mass index. This observation suggests an advantage in specificity over CRP because persons with abdominal adiposity have been shown to release cytokines from their mesenteric fat that stimulate hepatic expression of CRP and several other inflammatory markers, potentially weakening the link between the serum levels of these other markers and vascular inflammation. In addition, Lp-PLA2 levels were independent of other traditional risk factors, such as systolic and diastolic blood pressure. Lp-PLA2 levels also were independent of CRP levels.

This is the only study where adjustment for LDL cholesterol attenuated the risk ratio for Lp-PLA2. However, when patients with LDL cholesterol levels below the median (130 mg/dL) (1 mg/dL = 0.02586 mmol/L) were considered separately, elevation of the Lp-PLA2 enzyme was again significantly associated with a doubling of risk for coronary events. Because current guidelines call for an LDL cholesterol level <130 mg/dL in intermediate-risk persons, the intermediate- or moderate-risk patient who has achieved this lipid goal may benefit from the use of Lp-PLA2 to further assess residual risk. The ARIC results support the proposal that Lp-PLA2 and hs-CRP may be useful to identify intermediate-risk patients with an LDL cholesterol target level <130 mg/dL to assess residual risk, which might warrant reclassifying their risk status from intermediate to high risk: a risk doubling of an intermediate-risk person with 10%–20% 10-year risk would put the person into the high-risk >20% 10-year risk category. Interestingly, in the ARIC study when both Lp-PLA2 and hs-CRP were elevated (top tertile), the risk for coronary events increased by 4-fold and for stroke by 11-fold compared with the bottom tertile for each marker.4, 27 The Monitoring of Trends and Determinants in Cardiovascular Disease (MONICA)–Augsburg study also found that risk related to high levels of Lp-PLA2 and CRP were additive for cardiovascular events.5

The National Cholesterol Education Program (NCEP) Adult Treatment Panel III (ATP III) was published in 2001.28 However, at that time, there was only 1 major published study—WOSCOPS—of Lp-PLA2 as a risk marker. Although the ATP III committee did not recommend the routine measurement of inflammatory markers for clinical risk assessment, they did state that proinflammatory markers collectively should be considered as “emerging risk factors” for coronary heart disease (CHD) that “appear to contribute to CHD risk to varying degrees and can have utility in selected persons to guide intensity of risk-reduction therapy. Their presence can modulate clinical judgment when making therapeutic decisions.”28 Thus, the ATP III guidelines recognize the potential clinical utility of inflammatory markers in selected cases, although they do not promote them for screening low-risk individuals. ATP III also acknowledges the imperative to improve on traditional risk factor assessment of cardiovascular risk: “Nonetheless, when major risk factors are present, they account for only about half of the variability in CHD risk in the US population; other factors, yet to be identified, seemingly influence how much the major risk factors affect absolute CHD risk.”29

Building on the ATP III guidelines, an expert panel of the American Heart Association (AHA) and the Centers for Disease Control (CDC) in 2003 reviewed several inflammatory markers and made recommendations on how an inflammatory marker might be used in conjunction with traditional risk factors to refine the assessment of a person's Framingham cardiovascular risk category.30 This guideline also did not recommend the use of inflammatory markers for universal screening, but it did state that “hs-CRP measurement appears to be best employed to detect enhanced absolute risk in persons in whom multiple risk factor scoring projects a 10-year CHD risk in the range of 10% to 20%” (ie, in intermediate-risk persons). Again CRP measurement was not mandated in these individuals but was recommended as optional.30

At that time, the writing group recommended limiting the use of inflammatory markers to CRP because “the laboratory tests to assess inflammation are limited to those that are employable in clinical settings, have commercially available assays that can be standardized, and have adequate precision.” Lp-PLA2 measurement was not standardized nor commercially available at that time, but it subsequently has been standardized and become widely available through national commercial and reference laboratories. In addition, an immunoassay for Lp-PLA2 mass concentration has received US Food and Drug Administration (FDA) clearance for use as an aid in the prediction of coronary events and stroke. It also is worth noting that the AHA/CDC writing group recommended against testing inflammatory markers in high-risk patients, arguing that they already should receive maximum therapy.30

The recommendation to consider inflammatory marker assessment in intermediate-risk patients could have a major epidemiologic impact. Greenland et al31 made important proposals in a seminal article “Improving Coronary Heart Disease Risk Assessment in Asymptomatic People: Role of Traditional Risk Factors and Noninvasive Cardiovascular Tests” in 2001.31 They did not recommend the use of novel noninvasive risk marker tests in the low-risk population, which “constitutes approximately 35% of the US adult population 20 years of age and over. In the short-term, they can be offered general public health recommendations, and they can usually avoid further risk assessments for approximately 5 years.” In contrast, they stated that refining risk assessment with noninvasive testing after traditional risk assessment could make a large impact. Citing the National Health and Nutrition Examination Survey III (NHANES III), they stated, “This is a sizable group, judging from NHANES III data, roughly 40% of the US adult population.”31 Global risk assessment using Framingham risk scoring continues to be recommended, but patients found at intermediate risk might be further stratified (eg, into a higher cardiovascular risk category), based on the now substantial literature identifying Lp-PLA2 as an independent risk marker for cardiovascular events.

Metabolic Syndrome: A Clinical Marker of Intermediate Cardiovascular Risk 

The Neptune survey of 5,000 patients determined that metabolic syndrome was highly correlated with risk in the intermediate category.32 Because Framingham risk scoring is often regarded as cumbersome for use in routine clinical practice, it has been suggested that the recognition of metabolic syndrome might more easily be used to identify an important subset of persons who are at intermediate cardiovascular risk, and therefore, who would be appropriate for further risk assessment with the use of inflammatory biomarkers, such as CRP and/or Lp-PLA2.

As patients with the metabolic syndrome progress toward diabetes, their cardiovascular risk increases substantially. The Nurse's Health Study observed >100,000 female nurses for up to 20 years.33 Subjects who did not have diabetes during the entire follow-up period had the lowest rate of developing CVD. Those with prevalent diabetes at the time of the baseline questionnaire had a 5-fold increased risk of CVD. Nurses with prediabetes who developed diabetes by the end of the study had a 2.8-fold increased risk of CVD relative to subjects who never developed diabetes, reflecting the rapidly increasing cardiovascular risk in insulin-resistant persons who do not yet have diabetes.33 This link among metabolic syndrome, prediabetes, and CVD is becoming increasingly important as the epidemic of obesity continues to grow. In parallel, the prevalence of metabolic syndrome is now approaching 50% for individuals aged >50 years.

Because Lp-PLA2 has been shown to be independent of insulin resistance, it should be additive to the metabolic syndrome as a risk predictor. In a careful analysis of 78 healthy women evaluated for insulin resistance using the insulin suppression test, there was no association between Lp-PLA2 mass concentration levels and increasing insulin resistance.34 Indeed, Lp-PLA2 has been shown to be independent of body mass index in a dozen epidemiologic studies.1, 3, 4, 5, 8, 9, 10, 17, 18, 19, 20, 24 In 2 recent studies, findings showed that an elevated Lp-PLA2 level increases cardiovascular risk beyond the risk of having metabolic syndrome.24, 35 In the Malmö Diet and Cancer study, 4,480 subjects without diabetes without manifest CVD were studied for 10 years. An incident first CVD event included stroke or myocardial infarction (MI).26 As Figure 2 shows, when Lp-PLA2 enzyme activity is low and metabolic syndrome is not present, subjects are at significantly lower risk than persons with either metabolic syndrome or elevated Lp-PLA2 activity. When metabolic syndrome is not present but Lp-PLA2 is high, the CVD risk is similar to those subjects with low Lp-PLA2 and metabolic syndrome (ie, intermediate). However, subjects with a combination of both metabolic syndrome and high Lp-PLA2 activity exhibited the highest risk for cardiovascular events.

View Large Image Download to PowerPoint

Figure 2. High levels of lipoprotein-associated phospholipase A2 (Lp-PLA2) and metabolic syndrome (MS): independent and additive risk factors. In the Malmö study comprising 4,480 patients without diabetes mellitus with 261 hard cardiovascular (CV) events over 10 years, Lp-PLA2 activity was determined to be independent of and additive to insulin resistance–associated cardiovascular risk, a relatively unique characteristic compared with other inflammatory markers. (Adapted from Arterioscler Thromb Vasc Biol.38)

The Malmö Study findings are consistent with a recent report from the Intermountain Heart Collaborative Study, where 1,493 angiographically studied patients were observed for 7.5 years.35 The average age was 63 ± 12 years, 70% were men, 67% had CAD, and 42% had the metabolic syndrome. Lp-PLA2 mass concentration above the median was more predictive of angiographic CAD in metabolic syndrome subjects than in non–metabolic syndrome subjects, and these results remained significant after adjustment for traditional risk factors. Lp-PLA2 predicted CAD death significantly whether or not patients had the metabolic syndrome.

In summary, Lp-PLA2 is independent of insulin resistance, a relatively unique characteristic for an inflammatory marker, and elevated Lp-PLA2 adds to cardiovascular risk prediction in patients with the cardiovascular metabolic syndrome.

Secondary Prevention: A Heterogeneous “High-Risk” Category 

A log-linear relation exists between LDL cholesterol levels and relative risk for CAD, with event rates increasing more steeply with increasing LDL cholesterol concentrations.36 Randomized clinical trials of primary and secondary prevention have studied a variety of lipid-lowering agents at various degrees of intensity with LDL levels in placebo and treated groups ranging from 200 mg/dL to <50 mg/dL. For every 30-mg/dL change in LDL cholesterol, the relative risk for CAD is changed by about 30%. In the 2004 ATP III guidelines, a new risk subcategory beyond high risk called “very high risk” was introduced, whereby lowering the LDL cholesterol target to <70 mg/dL would be a “reasonable clinical strategy.”36 The option to further intensify LDL lowering to <100 mg/dL created a need to understand which high-risk patients with an LDL cholesterol at the “optimal” <100 mg/dL level might benefit from even more intensive lipid modification. Very-high-risk patients were defined as those having established CVD plus 1 of several criteria: (1) multiple major risk factors, especially diabetes; (2) severe and poorly controlled risk factors, especially continued cigarette smoking; (3) multiple risk factors of the metabolic syndrome, especially high levels of triglycerides ≥200 mg/dL plus elevated non–HDL cholesterol ≥130 mg/dL with HDL cholesterol levels <40 mg/dL; and (4) acute coronary syndromes (ACS).

In 2006, the AHA/American College of Cardiology (ACC) updated the guidelines for secondary prevention for patients with coronary and other atherosclerotic vascular diseases, confirming the ATP III panel recommendation in 2004 that it would be reasonable to treat LDL cholesterol to the target of <70 mg/dL in very-high-risk patients.37 Creating this fourth cardiovascular risk category raises the question whether inflammatory markers, such as CRP or Lp-PLA2, could identify which high-risk patients already at the optimal LDL cholesterol goal of <100 mg/dL might still have sufficient residual risk to benefit from an even more aggressive therapy.

Evidence supporting this suggestion has been forthcoming in the Pravastatin or Atorvastatin and Infection Therapy–Thrombolysis in Myocardial Infarction 22 (PROVE-IT–TIMI 22) and other recent studies (see below).16, 38 In PROVE-IT–TIMI 22, elevated CRP was shown to identify those patients with ACS with residual cardiovascular risk despite intensive LDL cholesterol–lowering therapy.39 Despite this suggested use, CRP might not be the ideal marker to substratify high-risk patients because of its vascular nonspecificity. CRP may produce false-positive results in patients with CAD because of infection, extravascular inflammatory diseases, injury, abdominal adiposity, or recent MI.40 It may also have limited utility in postmenopausal women, in whom it can be confounded by hormone replacement therapy, which stimulates hepatic production of CRP.41

A growing body of literature has demonstrated the predictive ability of Lp-PLA2 to identify residual risk in persons in the high cardiovascular risk category (Table 2). In all, 13 secondary prevention studies have tested the hypothesis that Lp-PLA2 might help identify which high-risk persons are in fact at very-high-risk for coronary events. The specific populations studied included patients with angiographically defined CAD, stable CAD, and post-ACS status.

The Intermountain Heart Collaborative Study measured Lp-PLA2 mass concentration and hs-CRP from samples donated by consecutive consenting patients (N = 1,493).12 All patients underwent coronary angiography (1996–1998) for CAD determination and were studied for 6.7 ± 0.5 years (range, 5.7–7.9 years) for cardiovascular events (death [including all-cause, CAD, and non-CAD cardiovascular death], MI, and cerebrovascular accident).12 There was a stepwise progression in CAD death associated with elevated Lp-PLA2 mass concentrations over the follow-up period in these patients, from 4.6% in quartile 1, 6.8%, 8.6%, and finally, 10.9% in quartiles 2–4, respectively, over a follow-up duration of almost 7 years. Lp-PLA2 was confirmed to predict the presence of CAD, even among patients at high pretest probability undergoing coronary angiography. Lp-PLA2 predicted the risk of CAD death, but not all-cause death, MI, or cerebrovascular accident, in this specific patient population referred for angiography. In this study, hs-CRP was also predictive of angiographic CAD. However, it did not attenuate the association of Lp-PLA2 with angiographic CAD. Instead, Lp-PLA2 showed an additive effect with CRP on CAD risk prediction. Compared with high CRP/low Lp-PLA2 levels, high CRP/high Lp-PLA2 levels predicted a significant risk increment for CAD death (p = 0.048).12

An important issue faced by physicians treating high-risk, and even very-high-risk patients is determining when atherosclerosis disease activity has subsided (ie, when has rupture-prone, inflamed, thin fibrous cap plaque stabilized?) In the PROVE-IT–TIMI 22 trial, an average LDL cholesterol level of 62 mg/dL was achieved in the atorvastatin 80-mg/day arm, and yet despite this very aggressive lipid-lowering therapy, 22.4% of post-ACS patients had a recurrent cardiovascular event in just 2 years (Figure 3).38

View Large Image Download to PowerPoint

Figure 3. Kaplan-Meier estimates of the incidence of the primary end point of death from any cause or a major cardiovascular event. Intensive lipid lowering with a 80-mg dose of atorvastatin compared with moderate lipid lowering with a 40-mg dose of pravastatin reduced the hazard ratio for death or a major cardiovascular event by 16%. No. = number. (Reprinted with permission from N Engl J Med.38)

A better predictor of low risk was the combined presence of low levels of LDL (regardless of treatment group) and a low inflammatory index as measured by hs-CRP.39 Thus, although lipids are a well-established treatment target, they should not be perceived as a reliable predictor of low subsequent risk (ie, of whether atherosclerotic plaques have fully stabilized).

Because Lp-PLA2 is involved in a causal pathway of plaque inflammation, is found in thin fibrous cap atheroma, and is lowered by all lipid-modifying medications, it is tempting to hypothesize that lower Lp-PLA2 may be a predictor of stabilized plaque. Indeed, several studies suggest that low Lp-PLA2 may have a particularly useful role as a negative predictor (ie, that it more reliably indicates plaque stabilization with a low risk of subsequent ACS). In the Langzeiterfolge der Kardiologischen Anschlussheil-Behandlung (KAROLA) study (Figure 4), plasma concentrations and activity of Lp-PLA2 were determined at baseline in a cohort of 1,051 patients aged 30–70 years with CAD.11 A total of 95 secondary CVD events occurred, despite fairly intensive treatment and despite baseline LDL cholesterol levels of 100 mg/dL. In multivariable analyses, Lp-PLA2 mass and activity were strongly associated with CVD events after controlling for traditional risk factors, lipids, severity of CAD, statin treatment, cystatin C, and N-terminal pro–brain natriuretic peptides (NT-proBNP), with a hazard ratio (HR) for the top versus bottom tertile of Lp-PLA2 of 2.09 (95% confidence interval [CI], 1.10–3.96). Also, after 4–6 years of follow-up, 95% of high-risk patients remained free of cardiovascular events if their Lp-PLA2 levels were <223 ng/mL. Thus, not only does a high level of Lp-PLA2 predict residual risk in high-risk ACS patients, but a low level of Lp-PLA2 was associated with a remarkably low 5% risk of recurrent cardiovascular events, suggesting that plaque stabilization was achieved in patients in the bottom tertile for Lp-PLA2.

View Large Image Download to PowerPoint

Figure 4. The Langzeiterfolge der Kardiologischen Anschlussheil-Behandlung (KAROLA) study: cardiovascular event-free survival with optimal low-density lipoprotein levels of 100 mg/dL (1 mg/dL = 0.02586 mmol/L) by lipoprotein-associated phospholipase A2 (Lp-PLA2) tertiles. Kaplan-Meier estimate of secondary fatal and nonfatal cardiovascular disease events during follow-up (time = days) according to tertiles of Lp-PLA2 mass at baseline. (Reprinted with permission from Arterioscler Thromb Vasc Biol.11)

Similar results were noted in the Mayo Heart Study of 504 patients with CAD. After 4 years of follow-up, 72 major adverse events occurred in 61 of 466 patients.7 Patients with Lp-PLA2 mass concentrations in the bottom tertile (<200 ng/mL) had only a 5% cardiovascular event rate compared with an event rate >10% for patients in the top tertile for Lp-PLA2.7 In the Ludwigshafen Risk and Cardiovascular Health Study (LURIC), 2,513 patients with angiographic evidence of ≥10% stenosis were observed for 6.5 years.42 Similar to results in the KAROLA and Mayo Heart Studies, the Kaplan-Meier survival curve demonstrated residual rates of only about 5% of cardiac death for persons in the bottom tertile for Lp-PLA2 enzyme activity, whereas risk approximately doubled for other patients (HR, 1.9; 95% CI, 1.4 –2.5; p <0.001). Taken together, these consistent results suggest that low Lp-PLA2 has a high negative predictive value in patients at secondary risk, and this may be hypothesized to indicate successful therapeutic plaque stabilization.

A marker of rupture-prone plaque should intuitively predict near-term cardiovascular event risk. The results of 3 recent secondary prevention studies—the Thrombogenic Factors and Recurrent Coronary Events (THROMBO) study, a Mayo study of the residents of Olmsted County, and the North Wuerttemberg and Berlin Infarction Study–II (NOBIS-II)17, 18, 22—suggest that Lp-PLA2 may represent such a marker. The THROMBO study evaluated the role of Lp-PLA2 activity in the risk of recurrent coronary events after a 2-year follow-up in 766 MI survivors.17 After full multivariate adjustment for a very wide range of 17 different lipid, lipoprotein subfractions, and hemostatic and inflammatory markers, apolipoprotein B was the only significant predictor (HR, 1.66; 95% CI, 1.14–2.40) for the top quartile versus the bottom 3 quartiles together. However, when Lp-PLA2 activity was entered into the model, atherogenic particle concentration (apolipoprotein B) lost significance, and Lp-PLA2 emerged as the single best predictor of recurrent MI, with an HR of 1.90 (95% CI, 1.31–2.75).17

A Mayo Clinic study observed 271 residents of Olmsted County after MI to evaluate the role of Lp-PLA2 as a predictor of death 1 year after acute MI.18 During the first year of follow-up, 42 deaths occurred. The survival estimates (95% CIs) at 1 year were 92% (86%–98%), 85% (78%–93%), and 74% (65%–84%) in the lowest, middle, and upper Lp-PLA2 tertiles, respectively (p = 0.007). After adjustment for age and sex, HRs for death in the middle and upper Lp-PLA2 tertiles were 2.20 (95% CI, 0.88–5.54) and 4.93 (95% CI, 2.10–11.60) compared with the lowest tertile, respectively (p trend <0.001). Further adjustment for other risk indicators resulted in even stronger associations. It is important to note that blood was drawn at 43 ± 39 hours, and that Lp-PLA2 is generally suppressed (as is LDL cholesterol) early post MI (see below). Patients with an Lp-PLA2 mass <218 ng/mL had a 95% 1-year survival rate, and those with <166 ng/mL had approximately a 98% 1-year survival rate, again highlighting the negative predictive value of this biomarker for cardiovascular events when it is low.18

In the NOBIS-II study, 429 consecutive patients were admitted to the emergency room with suspected ACS.22 Classification and regression tree analysis was conducted to examine 4 biomarkers as predictors of near-term, 42-day cardiovascular outcomes. Assessment of NT-proBNP and then troponin levels identified most persons at high cardiovascular risk, but an important subgroup of 35.9% of the patients with borderline NT-proBNP and negative troponin levels were identified; in this group Lp-PLA2 mass concentration provided a significant further risk stratification at a cutoff of 210 ng/mL (event rate with elevated Lp-PLA2, 17.9% vs 6.9%; relative risk, 2.6; 95% CI, 1.1–6.6). None of the clinical or electrocardiographic variables of the thrombolysis in myocardial infarction (TIMI) risk score provided comparable incremental information for risk stratification. The investigators commented that “these 3 biomarkers represent different pathways in the pathophysiology of the ACS, and this represents the rationale behind the application of a panel of biomarkers in the assessment of ACS patients.” CRP did not emerge as a significant independent marker of cardiovascular risk.22

Timing of the blood draw for Lp-PLA2 post ACS may be important because Lp-PLA2 is suppressed as is LDL cholesterol. In the Air Pollution and Inflammatory Response in Myocardial Infarction Survivors: Gene-Environment Interaction in a High-Risk Group (AIRGENE) study, Lp-PLA2 mass concentrations were measured in 204 post-MI survivors monthly for 6 months. Levels did not return to baseline until 60–90 days post MI.43 This is a possible explanation for why in PROVE-IT, Lp-PLA2 in blood drawn 7 days post ACS was not a significant predictor of recurrent cardiovascular events, whereas Lp-PLA2 activity in blood drawn 30 days post ACS in PROVE-IT significantly predicted residual risk (adjusted HR, 1.33; 95% CI, 1.01–1.74; p = 0.002) after full multivariable adjustment, including CRP. In the NOBIS-II study of ACS patients, Lp-PLA2 in blood drawn within 6–7 hours of symptom onset was predictive of future events and suggests that a brief window exists in the first several hours after ACS onset when Lp-PLA2 has not yet been suppressed and may be useful for risk prognostication.22 Lp-PLA2 in blood drawn an average of >1 day post ACS in the Mayo study of Olmsted County residents18 and in the Global Use of Strategies to Open Occluded Coronary Arteries/Fragmin During Instability in Coronary Artery Disease (GUSTO/FRISC) study21 gave mixed results (ie, it was a significant predictor of recurrent cardiovascular events in the first but not in the second case, although the trend was positive). Thus, the impact of timing of Lp-PLA2 measurements post MI on clinical risk prediction is unresolved and is an important area for future investigation.

Conclusion 

Putting these observations and considerations together, the question arises of how Lp-PLA2 should be best used in clinical practice? The evidence seems to support its use in intermediate-risk persons for further primary risk stratification and, more recently, for substratification of risk for secondary prevention (by definition, a high-risk category) in patients in whom a more aggressive strategy would be implemented if Lp-PLA2 is still elevated. These implications also apply to hs-CRP. Lp-PLA2, which measures an independent aspect of vascular inflammation, is a more vascular-specific marker and is also incremental to hs-CRP in risk assessment.4, 5, 12 In contrast to intermediate- and high-risk persons, low-risk individuals are not recommended for testing with Lp-PLA2. Lp-PLA2, as with CRP and other biomarkers, is not recommended for general population screening. However, once a patient is determined to be at intermediate or high cardiovascular risk, elevated Lp-PLA2 might be used to move the patient to the next higher-risk category, and low Lp-PLA2 might increase confidence that the patient is optimally treated or at lower risk.

Among the simplest means of identifying a large and relevant subset of intermediate-risk persons is to evaluate them for the presence of the metabolic syndrome. These patients appear to be promising candidates for inflammatory marker testing, including Lp-PLA2 testing, to further stratify those who are at higher risk. AHA/CDC and ATP III guidelines now also recognize a fourth risk subcategory of patients deemed to be at very high risk, and Lp-PLA2 appears to be a relatively inexpensive and noninvasive way to identify which high-risk or coronary-risk-equivalent patients are at very high residual risk. Whereas intermediate-risk patients have an LDL cholesterol target of <130 mg/dL, high-risk patients have an LDL cholesterol target of <100 mg/dL, and very-high-risk patients may be targeted to an LDL cholesterol level of <70 mg/dL. Other important considerations in patients with high Lp-PLA2 levels include intensification of therapeutic lifestyle changes and combination lipid-lowering therapy. A basic principle of preventive therapies is to match the intensity of treatment to individual patient risk.

Given its role in plaque pathology and consistent results from clinical trials to date as summarized above, Lp-PLA2 may be hypothesized to be unique as a biomarker of patients with rupture-prone plaque, and conversely, of patients whose plaques have been successfully stabilized therapeutically. This interesting hypothesis is an important area for future research into Lp-PLA2 diagnostics as well as therapeutics, including interventions that are targeted to inhibit Lp-PLA2 activity.

Author Disclosures 

The author who contributed to this article has disclosed the following industry relationships.

Jeffrey L. Anderson, MD, has received speaker honoraria from diaDexus, Inc. who has also supported research projects with his institution.