Clinica Chimica Acta 412 (2011) 963–969
Contents lists available at ScienceDirect
Clinica Chimica Acta j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c l i n c h i m
Serum tartrate-resistant acid phosphatase isoform 5a (TRACP5a) as a potential risk marker in cardiovascular disease Anthony J. Janckila a,b,⁎, Hseun-Fu Lin c, Yi-Ying Wu c, Chih-Hung Ku c, Shih-Ping Yang c, Wei-Shiang Lin c, Su-Huei Lee c, Lung T. Yam a,b, Tsu-Yi Chao c,d,⁎⁎ a
Department of Veterans Affairs Medical Center, Louisville, KY, USA University of Louisville School of Medicine, Louisville, KY, USA National Defense Medical Center and TriService General Hospital, Taipei, Taiwan, ROC d Shuang Ho Hospital, Taipei Medical University, Taipei, Taiwan, ROC b c
a r t i c l e
i n f o
Article history: Received 24 December 2010 Received in revised form 24 January 2011 Accepted 29 January 2011 Available online 4 February 2011 Keywords: Tartrate-resistant acid phosphatase Macrophage Inflammation Atherosclerosis Cardiovascular risk Myocardial infarction
a b s t r a c t Objective: This study was undertaken to determine the association between serum tartrate-resistant acid phosphatase 5a (TRACP5a) and cardiovascular disease (CVD) risk. Methods: Four hundred patients were enrolled including, 291 asymptomatic subjects grouped by the number of traditional risk factors, 36 patients undergoing cardiac arteriography, 34 undergoing percutaneous cardiac intervention, and 39 with acute myocardial infarction. Serum was collected at baseline and, in arteriograpy and intervention groups, periodically for 1 week afterward. In addition to laboratory and clinical evaluation for risk assessment, serum TRACP5a, C-reactive protein (CRP) and interleukin-6 (IL-6) were determined. Results: All biomarkers rose with increasing CVD risk. Only serum TRACP5a, logCRP and cholesterol were elevated in symptomatic patients. Serum TRACP5a was higher in men and correlated with age, logCRP, logIL-6 and log-triglycerides, and in symptomatic patients, with the number of diseased coronary arteries. IL-6 and CRP showed acute phase responses, whereas TRACP5a did not change over 1 week after arteriography or intervention. After adjustment for all other variables and risk factors, TRACP5a and logCRP were the only biomarkers to associate with symptomatic disease. TRACP5a was more specific than CRP to predict myocardial infarction among all subjects. Conclusions: Serum TRACP5a is a macrophage-derived inflammation marker associated with CVD risk, and with coronary vessel disease and its severity and may be a useful marker for screening and assessment of CVD risk. Published by Elsevier B.V.
1. Introduction Cardiovascular disease (CVD) due to atherosclerosis is common in the Western hemisphere and is a growing concern in Asia. Lifestyle, metabolic risk factors and chronic inflammation combine to initiate and drive atherosclerosis involving macrophage (MΦ) and endothelial activation, vascular injury and CVD [1,2]. Traditional risk factors for CVD include male gender, age, dyslipidemias, diabetes, hypertension, obesity, and smoking. Recently, biomarkers of inflammation, endothelial injury and oxidative stress have been investigated for their contribution to CVD risk [3,4]. C-reactive protein (CRP) is the most thoroughly documented inflammatory biomarker to assess general risk for CVD and to predict acute vascular adverse events [5].
⁎ Correspondence to: A.J. Janckila, VA Medical Center, 800 Zorn Ave., Louisville, KY, 40206 USA. ⁎⁎ Correspondence to: T-Y Chao, Shuang Ho Hospital, Taipei Medical University, No 291 Chung-Cheng Rd. Chung Ho Dist. Taipei, Taiwan, ROC. E-mail addresses:
[email protected] (A.J. Janckila),
[email protected] (T.-Y. Chao). 0009-8981/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.cca.2011.01.035
However, in some studies CRP contributed little more to risk after correction for traditional factors [6,7]. Among other markers, interleukin-6 (IL-6) [8], homocysteine [9], fibrinogen [10], B-type natriuretic peptide [11] and lipoprotein-associated phospholipase A2 (Lp-PLA2) [12] also have impact on CVD risk. Tartrate-resistant acid phosphatase (TRACP) is a marker of differentiation and activation of monocyte-derived cells [13]. In human blood, TRACP circulates as two isoforms, TRACP5a and 5b, each having unique properties and clinical significance [14]. TRACP5b is a proteolytically processed isoform released by bone resorbing osteoclasts [15] reflecting systemic osteoclast number [16] and correlating closely with bone resorption [17]. TRACP5a is an intact polypeptide selectively secreted by macrophages and dendritic cells in vitro [18], and represents the vast majority of circulating TRACP protein in humans [19]. Immunohistochemical studies [20,21] show that macrophages and dendritic cells, associated with inflammatory pathology contain abundant TRACP5a, which could be the source of serum TRACP5a. Based on our studies in rheumatoid arthritis [22], end-stage renal disease [23] and childhood obesity [24], we propose that serum TRACP5a signifies the extent of systemic macrophage
964
A.J. Janckila et al. / Clinica Chimica Acta 412 (2011) 963–969
burden or severity of chronic inflammatory diseases. Atherosclerotic plaque lesions have much in common with rheumatoid nodules including large numbers of activated, TRACP-positive foam cells in plaque or palisading macrophages in nodules, and a Th1-type cytokine profile [25]. Acknowledging that atherogenesis, progression and resolution are complex processes, it has become convenient to distinguish “vulnerable” plaque prone to rupture and thrombosis from “stable” plaque with thick fibrous caps [26]. Our hypothesis is that elevated serum TRACP5a may be an indicator of advanced or systemic inflammatory disease associated with inflammatory vulnerable plaque and a risk marker for adverse events in CVD. This study was undertaken to determine the relationship between TRACP5a and other inflammatory and lipid markers and traditional risk factors for CVD in both asymptomatic and symptomatic patients and to assess whether TRACP5a was an independent risk marker for acute myocardial infarction. 2. Methods and patients 2.1. Patients Two hundred ninety-one apparently healthy adult subjects without CVD symptoms were enrolled from the well patient clinic of Tri Service General Hospital (TSGH), Taipei, Taiwan. A single blood specimen was drawn for biomarker determinations at the time of enrollment. Subjects were grouped according to the number of cardiovascular risk factors as determined by routine screening, regardless of what those factors were. The asymptomatic groups (groups 0–3) consisted of 54 with no risk factors, 119 with only one factor, 80 with two factors and 39 with 3 or more factors. Risk factors included: 1) dyslipidemia (serum triglyceride N200 mg/dL, total cholesterol N200 mg/dL, serum LDL cholesterol of N100 mg/dL, or serum HDL cholesterol b65 mg/dL), 2) obesity (N120% of ideal body weight or abdominal circumference of N90 cm for men and N80 cm for women), 3) hypertension (N130/85 mmHg), 4) diabetes mellitus (fasting blood glucose of N110 mg/dL) and 5) cigarette smoking status (currently or previously a smoker or never having smoked). The demographics and risk status of these patients are summarized in Table 1. Seventy symptomatic subjects were enrolled for a short-term longitudinal study at the time of admission for elective diagnostic coronary arteriography due to chest pain. Coronary artery disease (CAD) was established by i) angina with positive thallium-201 myocardial scan or treadmill stress test; ii) effort angina with electrocardiogram changes; iii) high CAD risk with Framingham score N20% with CAD risk within 10 years and symptoms of progressive heart failure; iv) prior history of CAD with multiple vessel disease; v) previous history of AMI with recurrent angina or
Table 1 Age, gender and CVD risk factor distributions for asymptomatic subjects. Risk
Med. age
Gender
Dyslipid
DM
Obesity
Smoker
Group
range
M/F
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
0 RF
41 22–77 42 23–72 49 22–75 45 23–84 nsa
14/39
0
54
0
54
0
54
0
54
0
54
64/55
35
60
3
88
0
97
17
82
23
75
48/32
47
30
6
70
7
68
32
46
20
60
23/84
33
5
8
27
1
35
31
7
20
18
b0.001a
nsb
1 RF 2 RF N 2 RF P
HTN
0.003b
b 0.001b
0.021b
b0.001b
RF = risk factor; M/F = number male/female; Dyslipid = any dyslipidemia; DM = diabetes mellitus; HTN = hypertension; ns = not significant. a ANOVA; b Chi-square, AMI vs all asymptomatic groups.
previous history of CAD status post percutaneous cardiac intervention. Thirty-six patients had either no significant luminal narrowing (b50%) in any coronary artery or had significant narrowing (between 50 and 70%) in any artery, but did not receive interventions (coronary arteriography group). Thirty-four had significant luminal narrowing of N70% in any coronary artery requiring balloon angioplasty or stenting at the time of study (intervention group). The number and type of traditional risk factors were recorded for each patient according to the criteria above. In addition to lipid profile, use of cholesterol-lowering medication was included as a criterion for dyslipidemia. The number of coronary arteries with significant stenosis was included in the analyses. A blood specimen was drawn for baseline biomarker determinations before the procedure. Additional blood specimens were drawn at time points of 1 h, 1 day, and 1 week after the procedure. The demographics and risk status of these patients are summarized in Table 2. Serum specimens were also obtained from 39 subjects at the time of admission who were determined to have acute myocardial infarction based on clinical symptoms, electrocardiogram abnormalities and elevated serum troponin. The number of diseased coronary arteries was recorded. Traditional risk factors and medications were also recorded and included for adjustment. The demographics and risk status for this cohort are summarized in Table 2. Use of cholesterol-lowering drugs, non-steroidal anti-inflammatory drugs (NSAID), anti-platelet medications, including aspirin, antihypertensive drugs, or anti-diabetic medications by symptomatic patients was included for adjustment. The distribution of medication use among symptomatic groups is summarized in Table 3. All subjects gave signed informed consent prior to enrollment. The study was conducted according to guidelines of the Declaration of Helsinki and approved by the Human Studies Protection Committee of Tri-Service General Hospital. 2.2. Biochemical markers Serum TRACP5a protein was determined by two-site immunoassay as previously published [22,27]. Briefly, streptavidin wells (Pierce Chemical Co.) were coated with 1 microgram biotinlylated mab220 specific for serum TRACP5a. After a brief wash, 10 μl of serum samples were added with 90 μl dilution buffer in duplicate and incubated overnight at 4 °C. Wells were washed and 100 μl of a 1:1000 dilution of horseradish peroxidase (HRP)-conjugated anti-TRACP mab162 were added. After 1 h incubation wells were washed again and HRP detected with a solution of o-phenylene diamine and H2O2 at pH 5.0. After 15 min, the reaction was stopped with 50 μl 2 M H2SO4. TRACP5a assay was calibrated with 2-fold serial dilutions of partially purified serum TRACP5a adjusted to concentrations from 5 μg/L to 0.08 μg/L. The intraassay CV is 3.9% and interassay CV is 7.3%. Serum CRP was estimated by an in-house, high-sensitivity two-site immunoassay, similar to that for TRACP5a, constructed from purified and HRPconjugated polyclonal antibodies to human CRP and purified human CRP as standard (Dako Corp.) [22]. The sensitivity of this microwellbased immunoassay was 0.1 mg/dL with an effective dynamic range of 0.1 to 50 mg/dL, intraassay CV of 5.2% and interassay CV of 15.6%. Serum IL-6 was determined using a commercial kit immunoassay (RayBiotech, Inc.) with an intraassay CV of 10% and interassay CV of 12%. Serum triglycerides (upper limit 200 mg/dL), total cholesterol (upper limit 200 mg/dL), LDL cholesterol (upper limit 100 mg/dL) and HDL cholesterol (lower limit 65 mg/dL) were determined in the Laboratory Service of TSGH by enzymatic colorimetric methods. 2.3. Statistical analyses A power analysis was conducted at a level of 0.8 with alpha at 0.05 based on our previously published studies of rheumatoid arthritis [22] and end stage renal disease [23] estimating the difference between
965
A.J. Janckila et al. / Clinica Chimica Acta 412 (2011) 963–969
Table 2 Age, gender and CVD risk factor distributions for symptomatic subjects undergoing cardiac arteriography alone (CAG) or percutaneous cardiac intervention (PCI) or having acute myocardial infarction (AMI). Group
CAG PCI AMI P
Med age range
Gender M/F
Dyslipid Yes
No
DM Yes
No
HTN Yes
No
Yes
No
Yes
No
63 35–83 66 26–84 68 44–84 nsa
30/6
18
18
6
30
12
24
16
20
16
20
0.842± 1.027
24/10
20
14
12
22
26
8
21
12
18
14
2.000± 0.853
30/9
25
14
9
30
26
13
22
13
26
13
2.077± 0.807
nsa
nsb
nsb
Obesity
nsb
Smoker
nsb
# Disease vessels
nsb
b0.001a
M/F = male-to-female numbers; Dyslipid = any dyslipidemia; DM = diabetes mellitus; HTN = hypertension; # disease vessels = mean number of diseased vessels determined at time of CAG or PCI procedure or AMI evaluation. a ANOVA. b Chi-square, AMI vs CAG + PCI.
means for TRACP5a protein in control and disease cohorts. This analysis suggested that a group size of 22 to 27 was needed to detect a significant difference between groups. Chi-square test was used to compare the proportions of traditional risk factors in acute myocardial infarction patients to those in asymptomatic subjects and to those in symptomatic patients having arteriography and or interventions. Significant differences in the prevalence of 4 of 5 traditional risk factors were noted between myocardial infarction patients and asymptomatic subjects. Asymptomatic subjects were significantly younger than symptomatic patients. Frequency of risk factors and age were no different among AMI, CAG and PCI groups. However, significantly fewer patients in the AMI group were taking antiplatelet medications compared to the combined CAG and PCI groups. For these reasons, some analyses were conducted on asymptomatic and symptomatic groups separately. Mean differences in baseline data were compared among asymptomatic and symptomatic groups separately by one-way ANOVA followed by group-wise t-tests. Mann–Whitney rank sum tests were done to compare median levels of biomarkers between all asymptomatic subjects and all symptomatic patients. Pearson's correlation coefficients were calculated for all pairs of baseline markers in asymptomatic subjects and in symptomatic patients separately. Multiple logistic regressions with stepwise selection were used to assess the associations of interests, with adjustment of all covariates including age, gender, status for each risk factor, inflammatory markers, lipid markers and, in the cases of symptomatic patients, the number of affected vessels and current medication use were also adjusted. Group comparisons included: i) all asymptomatic subjects (groups 0–3) vs all symptomatic subjects (arteriography + intervention + myocardial infarction); ii) all asymptomatic subjects (groups 0–3) vs myocardial infarction; iii) arteriography vs myocardial infarction; and iv) intervention vs myocardial infarction. Receiver operating characteristic curves (ROC) were calculated for TRACP5a and CRP to determine cut-off values and to estimate the sensitivity and specificity of association with myocardial infarction among all subjects and among symptomatic patients separately. Non-Gaussian data for CRP, IL-6 and triglycerides were log-transformed prior to ANOVA, univariate or multivariate analyses;
all other markers were normally distributed and not transformed. For longitudinal studies in arteriography and intervention groups, biomarkers at each time point were normalized to baseline. Using student's t-test, mean percent change was compared between the arteriography and intervention groups. Statistical analyses were conducted or verified by C.-H.K. using SAS version 9.2 and SigmaPlot version 11.2. 3. Results 3.1. Biomarkers as a function of the number of traditional risk factors in asymptomatic subjects Mean levels of all biomarkers, including inflammatory and lipid markers, rose significantly as the number of risk factors increased in asymptomatic subject groups (Table 4). Univariate analysis including all asymptomatic subjects revealed significant correlations between TRACP5a and logCRP (r = 0.239; p b 0.0001), logIL-6 (r = 0.126; p = 0.037), log triglycerides (r = 0.302; p b 0.0001) and total cholesterol (r = 0.168; p = 0.008). 3.2. Biomarkers in symptomatic patients Mean TRACP5a, logCRP and cholesterol were the only biomarkers which were significantly increased in all symptomatic patients compared to all asymptomatic subjects (p b 0.001). Marker levels in patients who underwent arteriography or intervention were no higher at baseline than levels in asymptomatic patients with N2 risk factors. However, myocardial infarction patients had significantly
Table 4 Biomarkers in asymptomatic study population. Variable
0 RF
1 RF
2 RF
≥ 2 RF
n TRACP 5ab CRPc
54 7.23 ± 1.68 0.25 0.04–12.28 2.53 0.11–410 81 38–191 165 ± 24.1
119 8.38± 2.19 0.46 0.67–14.21 3.34 0.14–1016 92 43–327 186± 38.7
80 8.74 ±2.54 0.81 0.49–28.14 3.78 0.14–314 114 32–1166 206 ±43.6
39 9.44 ±2.63 1.03 0.09–34.13 5.04 0.88–179 175 78–567 209 ± 37.2
IL-6c Table 3 Use of cardiovascular disease medications by symptomatic patients.
TGc
Group
No.
Statin
NSAID
Anti-platelet
Anti-HTN
Anti-diabetic
None
CAG PCI AMI Pa
32 28 40
6 9 3 ns
2 0 2 ns
25 23 13 0.028
18 20 14 ns
2 3 5 ns
4 4 17 0.019
a
Chi-square; AMI vs CAG + PCI.
Tot Cholb
Pa
b0.001 b0.001 0.0017 b0.001 b0.001
RF = risk factor; CRP = C-reactive protein; IL-6 = interleukin 6; TG = triglycerides; Tot Chol = total cholesterol. a ANOVA among asymptomatic RF groups (log transformed CRP, IL-6, TG). b Mean ± standard deviation. c Median and range.
966
A.J. Janckila et al. / Clinica Chimica Acta 412 (2011) 963–969
20
Table 5 Biomarkers in symptomatic study populations. Variable
CAG
PCI
AMI
n TRACP 5ab CRPc
36 8.34 ± 1.71 0.58 0.11–7.79 2.16 0.14–350 109 38–450 164± 32.9 107± 29.1 48.5 ± 12.7
34 9.17 ±2.36 0.85 0.09–31.0 2.98 0.11–64 126 49–333 176 ±31.3 105 ±19.2 44.9 ±15.0
39 12.55 ± 2.98 1.92 0.32–34.1 10.17 0.17–253 125 56–425 180 ± 50.9 120 ± 43.0 40.8 ±10.5
IL-6c TGc Tot Cholb LDLb HDLb
P
b 0.001 b 0.001 ns ns ns ns ns
Fold change (mean + SE)
A a
CAG = cardiac angiography; PCI = percutaneous cardiac intervention; AMI = acute myocardial infarction; CRP = C-reactive protein; IL-6 = interleukin-6; TG = triglycerides; Tot Chol = total cholesterol; LDL = low density lipoprotein; HDL = high density lipoprotein; ns = not significant. a ANOVA (log-transformed CRP, IL-6, TG). b Mean ± standard deviation. c Median and range.
15
10
5
0 Baseline
1 hour
1 day
1 week
Time after arteriography 20
B Fold change (mean + SE)
higher mean TRACP5a and mean logCRP than all other symptomatic groups (Table 5). All other inflammatory and lipid markers were no different in the myocardial infarction group than in arteriography or intervention groups. Univariate analyses of biomarkers in symptomatic patients showed significant correlations between TRACP5a and logCRP (r = 0.375; p b 0.0001), logIL-6 (r = 0.193; 0.045), and HDL (r = −0.266; p = 0.034). TRACP5a (r = 0.293; p = 0.002) and logCRP (r = 0.189; p = 0.049) also correlated significantly with the number of diseased coronary arteries. Eighty-eight percent and 86% of CAG and PCI patients, respectively were taking medications for CVD; only 58% of AMI patients were being treated for CVD (p b 0.001) (Table 3). In a short-term longitudinal study of patients undergoing arteriography, IL-6 rose significantly above baseline within 1 h after procedure, peaking after 1 day and falling after 1 week. (Fig. 1A). CRP rose above baseline one day after procedure. In contrast to these acute responses, TRACP5a levels did not change significantly for up to a week after procedures. In patients who also underwent interventions, IL-6 and CRP levels responded similarly to that in arteriography group (Fig. 1B). As with arteriography subjects, TRACP5a levels did not show any significant immediate changes as a result of percutaneous cardiac interventions.
IL-6 CRP TRACP 5a
IL-6 CRP TRACP 5a
15
10
5
0 Baseline
1 hour
1 day
1 week
Time after intervention Fig. 1. Time course of inflammatory biomarker response after cardiac arteriography (CAG) (A) or percutaneous cardiac intervention (PCI) (B). Serum IL-6 elevated after 1 h. Serum CRP elevated after 1 day. Serum TRACP5a did not respond to either treatment for up to 1 week afterward.
comparing myocardial infarction patients to non-infarction patients including either all subjects (Fig. 2A) or only symptomatic subjects (Fig. 2B). TRACP5a had higher area under curve (AUC), superior specificity and higher positive predictive value, whereas logCRP had slightly greater sensitivity whether all subjects or only symptomatic
3.3. TRACP5a associated with acute myocardial infarction To determine if serum TRACP5a was associated with more severe, symptomatic disease or the specific outcome of acute myocardial infarction, multiple logistic regression analyses were done comparing symptomatic groups combined (arteriography, intervention and infarction) to asymptomatic groups (group 0–3) or myocardial infarction group to separate symptomatic groups (Table 6). When all subjects are included, age, male gender, TRACP5a and logCRP were significantly associated with symptomatic disease after correcting for all other interests. As the severity of CVD increased from asymptomatic risk to a need for arteriography to a need for intervention, the association between TRACP5a and myocardial infarction remained strongest, while the significance of other effects diminished. We emphasize that the significant association between AMI and use of anti-platelet medications in our study does not mean that these drugs increases risk for AMI, only that there is association between AMI patients and this treatment regimen after all other adjustments. Since TRACP5a and logCRP were the only biomarkers associated with myocardial infarction, ROC curves were calculated for these variables to estimate their individual sensitivity and specificity,
Table 6 Risk markers associated with acute myocardial infarction at increasing stages of cardiovascular disease.a Group comparison
Effect
OR
95% Wald C.I.
P
C; AUCb
Asymptomatic vs Symptomatic
Age Genderc TRACP5a log CRP Age Genderc TRACP5a log CRP TRACP5a log CRP Anti-platelet TRACP5a Anti-platelet
1.14 0.25 1.20 2.00 1.09 0.24 1.33 4.13 2.04 6.96 9.37 1.50 7.47
1.11–1.17 0.13–0.52 1.06–1.35 1.11–3.53 1.06–1.12 0.12–0.47 1.18–1.50 2.32–7.34 1.26–3.31 1.12–43.25 1.28–68.30 1.15–1.94 2.02–27.60
b0.0001 b0.001 0.003 0.021 b 0.0001 b0.0001 b0.0001 b0.0001 0.004 0.037 0.027 0.002 0.003
0.898
Asymptomatic vs AMI
Arteriography vs AMI
PCI vs AMI
0.871
0.958
0.856
a Multiple logistic regression after adjustment for possible confounders including age, gender, inflammation markers, lipid markers, risk factors and medication use. b C; AUC = C statistic; area under curve. c Female = 1; male = 2.
A.J. Janckila et al. / Clinica Chimica Acta 412 (2011) 963–969
967
4. Discussion
Fig. 2. Receiver-operating characteristic curves describing sensitivity and specificity of TRACP5a and CRP to associate with AMI compared to all other study subjects (A) and compared to only CAG and PCI patients (B). A significant difference between TRACP5a and CRP curves was achieved only for comparison between AMI and CAG plus PCI patients (p = 0.035).
patients were included. The AUC for TRACP 5a was significantly larger than AUC for logCRP (p = 0.035) when only symptomatic patients were included.
We investigated the relationships and interactions of a novel macrophage product, TRACP5a, with other inflammatory and lipid biomarkers and traditional risk factors in the context of CVD. Because age and gender distributions were different between asymptomatic subjects and symptomatic patients, and because information on medication use in asymptomatic subjects is unavailable, these groups were subjected to some analyses separately. Previous studies of TRACP isoform expression by macrophages in vitro have shown that TRACP5a is selectively secreted by macrophages while TRACP5b is retained intracellularly [19]. Circulating monocytes do not express TRACP; only after activation and differentiation does TRACP manifest. Previous immunohistochemical studies have demonstrated that macrophages and osteoclasts are the principal cells to express TRACP [21]. Since it is well documented that the source of serum TRACP5b is osteoclasts and that serum TRACP5a has no relationship to bone metabolism [22,23,28], it is most likely that macrophages are the source of serum TRACP5a. Because serum TRACP5a is a systemic marker, much of it will be derived from sources other than the vascular tree. However, in the context of atherosclerosis and clinical CVD, we hypothesize that elevated serum TRACP5a represents increased macrophages due to the pathological process. In the current study relating biomarkers to CVD risk, serum TRACP5a was higher in men and increased with age and the number of risk factors and correlated with other inflammation markers in asymptomatic subjects. The percentage of men also increased with increasing number of risk factors, perhaps accounting for higher TRACP5a levels in men. Serum TRACP5a was independently associated with symptomatic CVD, and remained strongly associated with acute myocardial infarction as disease severity increased. It was the only biomarker to discriminate patients having infarction from those having a need for percutaneous interventions due to symptoms. These findings corroborate others demonstrating that CVD risk assessment can be improved by considering inflammatory biomarkers along with traditional risk factors and lipid screening, particularly in higher-risk groups [3,29]. We also can advance our hypothesis that the novel macrophage marker, TRACP5a, is a biomarker of chronic inflammatory disease severity and may be an independent risk marker for the CVD end-point of acute myocardial infarction. Atherosclerosis is a chronic inflammatory disease involving progressive endothelial activation, recruitment of inflammatory macrophages and generation of ROS, and can lead to sudden lifethreatening events such as heart attack and stroke [2]. Traditional risk factors for adverse events such as acute myocardial infarction, stroke and death include dyslipidemias, diabetes, hypertension, obesity and smoking. Biochemical markers of inflammation, IL-6 and CRP, are markers with additional impact on CVD risks. Some studies show them to predict the most severe outcomes [30,31]. These markers are proximal components of the acute phase response, reflecting general inflammatory activity at the time of sampling. They may not always be specific to CVD and their incremental impact may not be compelling for individual patient risk stratification [4]. TRACP5a is a secreted product of macrophages and may be a more direct measure of inflammatory macrophage number, which could in turn reflect the degree of systemic atherosclerotic disease. Both CRP and TRACP5a may contribute directly to pathogenesis of disease. CRP is present in plaque [32] and contributes to the perpetuation of macrophage and endothelial activation through FcγR binding and activation [33]. Macrophage TRACP5a may be capable of lipid oxidation through ROS generation [34] and contributes to foam cell generation. Recently a distinction has been made between the pathophysiology and clinical risks of stable and unstable atheromas [26,29]. Stable atheromas bear thick fibrous caps and necrotic lipid cores; they progressively occlude vessels to cause ischemia and chronic heart disease. Unstable atheromas have thin fibrous caps and are laden with
968
A.J. Janckila et al. / Clinica Chimica Acta 412 (2011) 963–969
inflammatory foam cells. Unstable atheromas with “compensatory enlargement” might not cause significant stenosis and chronic symptoms, but are vulnerable to rupture, yielding thromboemboli causing heart attack or stroke [35]. Serum markers associated with vulnerable plaque in addition to Lp-PLA2 [12] would be welcome in assessing those at high risk for acute events and in monitoring risk during treatment. When all subjects were considered, serum TRACP5a significantly increased with age, was higher in men and correlated to logCRP, logIL6 and log triglycerides. These associations are similar to our prior results in rheumatoid arthritis [22] and end-stage renal disease [23] and confirm that TRACP5a is an inflammation biomarker. Considering symptomatic patients alone, only TRACP5a and logCRP were increased in those having suffered heart attack compared to those having less severe manifestations requiring arteriography or angioplasty. None of the lipid markers were further elevated in myocardial infarction patients compared to arteriography or intervention patients. In symptomatic groups, TRACP5a still correlated positively with logCRP and logIL-6. However, the associations among TRACP5a, logCRP and logIL-6 are not always strong, suggesting that macrophage expression of TRACP5a is more distal pathophysiologically from endothelial IL-6 expression and hepatic CRP production. The differential response of these markers to endothelial trauma is evidence that they measure different, but complimentary, aspects of inflammatory disease. The rapid and sequential rise and fall in IL-6 and CRP after arteriography define the acute phase response. In contrast, TRACP5a did not respond in the short-term, consistent with our hypothesis that TRACP5a is a measurement of longer-term accumulation of macrophages. If the number of affected coronary arteries is a reflection of the general atherosclerotic burden, as rheumatoid nodules are a reflection of disease severity and progression in rheumatoid arthritis [36], then the correlation between TRACP5a and the number of affected arteries would support our hypothesis that TRACP5a reflects systemic macrophage burden and is associated with coronary vessel disease and its severity. One goal for biomarkers in CVD assessment is to predict who is at risk for adverse events so that preventative measures can be taken. This is especially important now that a distinction between risk from stable and unstable atheromas can be made. In this study, TRACP5a and logCRP both associated with symptomatic disease. When patients with acute myocardial infarction were compared to other cohorts with increasingly severe disease, the association with TRACP5a remained significant, becoming the only myocardial infarctionassociated biomarker compared to angioplasty patients. Serum TRACP5a assay could provide a new perspective on CVD risk, independent of any ongoing inflammatory activity. Since blood sampling was done in the emergency room, elevated serum TRACP5a in heart attack victims may have been caused by the acute event rather than being a preexisting condition. The fact that TRACP5a did not respond to acute endothelial trauma, while IL-6 and CRP did so, argues against a rapid rise in TRACP5a after infarction and for a protracted elevation over long-term disease progression and systemic macrophage accumulation. One limitation of this study is its cross-sectional design of relatively small cohorts in an ethnically homogeneous group. Nevertheless a power analysis revealed that our group sizes were sufficient in a proof-of-concept study to detect a significant difference in TRACP5a between groups. While we can conclude that TRACP5a is a chronic inflammatory biomarker, and we have generated a reasonable hypothesis for its potential as an independent risk marker in CVD, we cannot yet prove its real clinical value as a predictive marker for adverse events. Another limitation with our cross-sectional design is that we cannot determine the direct effects of each treatment on serum TRACP5a. Moreover, information on medication use is not available for the asymptomatic subjects. Nevertheless, it is interesting that fewer patients in the AMI group were being treated for their
disease compared to other symptomatic cohorts. It may be coincidental that this group also had significantly higher serum TRACP5a, perhaps alerting to unknown risk associated with higher macrophage number in vulnerable plaque. Further prospective, longitudinal study of larger cohorts is required to rigorously test our hypothesis that serum TRACP5a has prognostic value in CVD. Circulating monocytes do not express TRACP, they do so only after activation and differentiation into macrophages. We have not demonstrated whether serum TRACP5a in CVD setting derives solely from atheromas. It probably does not. Serum TRACP5a is a systemic marker, and while atherosclerosis is considered a systemic disease, the pathology is restricted to blood vessels. Patients with atherosclerosis and CVD most likely have systemic inflammation involving other macrophage depots including visceral adipose tissue among other sites. As with the use of hepatic-derived CRP as a systemic risk marker in CVD, it may not matter the precise source of TRACP5a for practical purposes of risk assessment, it may only matter that it is macrophage-derived and increased systemically. In conclusion, serum TRACP5a is associated with chronic inflammation and may be a useful adjunct screening test for CVD risk assessment or as a risk marker for adverse outcome in patients with CVD.
Acknowledgements The authors acknowledge the expert technical assistance of Miss Hsin-Yi Liu. This work was supported in part by the Research Service of the U.S. Department of Veterans Affairs and the Clinical Research Foundation, Inc. (Dr. A. J. Janckila and Dr. L. T. Yam), the Taiwan Department of Defense (DOD97-14-03, Dr. W-S Lin), and the Taiwan Department of Health (DOH-TD-B-111-003, Dr. T-Y Chao). All authors have no conflicts of interest.
References [1] Ross RR. Atherosclerosis — an inflammatory disease. N Engl J Med 1999;340: 115–26. [2] Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med 2005;352:1685–95. [3] Wang TJ, Gona P, Larson MG, et al. Multiple biomarkers for the prediction of first major cardiovascular events and death. N Engl J Med 2006;355:2631–9. [4] Packard RRS, Libby P. Inflammation in atherosclerosis: from vascular biology to biomarker discovery and risk prediction. Clin Chem 2008;54:24–38. [5] Musunuru K, Kral BG, Blumenthal RS, et al. The use of high sensitivity C-reactive protein in clinical practice. Nat Clin Pract Cardiovasc Med 2008;5:621–35. [6] Sukhija R, Fahdi I, Garza L, et al. Inflammatory markers, angiographic severity of coronary artery disease, and patient outcome. Am J Cardiol 2007;99:879–84. [7] Kaptoge S, Di Angelantonio E, Lowe G, Pepys MB, Thompson SG, Collins R. Danesh. J (Writing committee). C-reactive protein concentration and risk of coronary heart disease, stroke, and mortality: an individual participant meta-analysis. Lancet 2010;375:132–40. [8] Ridker PM, Rifai N, Stampfer MJ, Hennekens CH. Plasma concentration of interleukin-6 and the risk of future myocardial infarction among apparently healthy men. Circulation 2000;101:1767–72. [9] Antoniades C, Antonopoulos AS, Tousoulis D, Marinou K, Stefanadis C. Homocysteine and coronary atherosclerosis: from folate fortification to the recent clinical trials. Eur Heart J 2009;30:6–15. [10] Kaptoge S, White IR, Thompson SG, et al. Associations of plasma fibrinogen levels with established cardiovascular disease risk factors, inflammatory markers, and other characteristics: individual participant meta-analysis of 154, 211 adults in 31 prospective studies. Am J Epidemiol 2007;166:867–79. [11] Melander O, Newton-Cheh C, Almgren P, et al. Novel and conventional biomarkers for prediction of incident cardiovascular events in the community. JAMA 2009;302:49–57. [12] Lerman A, McConnell JP. Lipoprotein-A associated phospholipase A2: a risk marker or a risk factor? Am J Cardiol 2008;101(suppl):11F–22F. [13] Moss DW. Changes in enzyme expression related to differentiation and regulatory factors: the acid phosphatase of osteoclasts and other macrophages. Clin Chim Acta 1992;209:131–8. [14] Janckila AJ, Yam LT. Biology and clinical significance of tartrate-resistant acid phosphatases. New perspectives on an old enzyme. Calcif Tiss Int 2010;85: 465–83. [15] Hayman AR, Warburton MJ, Pringle JAS, Coles B, Chambers TJ. Purification and characterization of a tartrate-resistant acid phosphatase from human osteoclastomas. Biochem J 1989;261:601–9.
A.J. Janckila et al. / Clinica Chimica Acta 412 (2011) 963–969 [16] Rissanen JP, Suominen MI, Peng Z, Halleen JM. Secreted tartrate-resistant acid phosphatase 5b is a marker of osteoclast number in human osteoclast cultures and the rat ovariectomy model. Calcif Tiss Int 2008;82:108–15. [17] Halleen JM, Alatalo SL, Suominen H, Cheng S, Janckila AJ, Väänänen HK. Tartrateresistant acid phosphatase 5b: a novel marker of bone resorption. J Bone Miner Res 2000;15:1337–45. [18] Janckila AJ, Neustadt DH, Nakasato YR, Halleen JM, Hentunen T, Yam LT. Serum tartrate-resistant acid phosphatase isoforms in rheumatoid arthritis. Clin Chim Acta 2002;320:49–58. [19] Janckila AJ, Parthasarathy RN, Parthasarathy LK, et al. Properties and expression of human tartrate-resistant acid phosphatase isoform 5a by monocyte-derived cells. J Leukoc Biol 2005;77:209–18. [20] Hayman AR, Macary P, Lehner PJ, Cox TM. Tartrate-resistant acid phosphatase (Acp5): identification in diverse human tissues and dendritic cells. J Histochem Cytochem 2001;49:675–83. [21] Janckila AJ, Slone SP, Lear SC, Martin A, Yam LT. Tartrate-resistant acid phosphatase as an immunohistochemical marker for inflammatory macrophages. Am J Clin Pathol 2007;127:556–66. [22] Janckila AJ, Neustadt DH, Yam LT. Significance of serum TRACP in rheumatoid arthritis. J Bone Miner Res 2008;23:1287–95. [23] Janckila AJ, Lederer ED, Price BA, Yam LT. Tartrate-resistant acid phosphatase isoform 5a as an inflammation marker in end-stage renal disease. Clin Nephrol 2009;71:387–96. [24] Shih KC, Janckila AJ, Kwok CF, Ho LT, Chou YC, Chao TY. Effects of exercise on insulin sensitivity, inflammatory cytokines and serum tartrate-resistant acid phosphatase 5a in obese Chinese male adolescents. Metabolism 2010;59:144–51. [25] Full LE, Ruisanchez C, Monaco C. The inextricable link between atherosclerosis and prototypical inflammatory diseases rheumatoid arthritis and systemic lupus erythematosus. Arthritis Res Ther 2009;11:217–26.
969
[26] Libby P. Atherosclerosis. In: Creager MA, Dzau VJ, Loscalzo J, editors. Vascular Medicine. Philadelphia: Saunders Elsever; 2006. p. 101–18. [27] Chao TY, Lee SH, Chen MM, et al. Development of immunoassays for serum tartrate-resistant acid phosphatase isoform 5a. Clin Chim Acta 2005;359:132–40. [28] Halleen JM, Ylipahkala H, Alatalo SL, et al. Serum tartrate-resistant acid phosphatase 5b, but not 5a, correlates with other markers of bone turnover and bone mineral density. Calcif Tiss Int 2002;71:20–5. [29] Weintraub HS. Identifying the vulnerable patient with rupture-prone plaque. Am J Cardiol 2008;101(suppl):3F–10F. [30] Albert CM, Ma J, Rifai N, Stampfer MJ, Ridker PM. Prospective study of C-reactive protein, homocysteine, and plasma lipid levels as predictors of sudden cardiac death. Circulation 2002;105:2595–9. [31] Sattar N, Murray HM, Welsh P, et al. Are markers of inflammation more strongly associated with risk for fatal than nonfatal vascular events? PLoS Med 2009;6: e1000099, doi:10.1371/journal.pmed.1000099. [32] Sun H, Koike T, Ichikawa T, et al. C-reactive protein in atherosclerotic lesions. Its origin and pathophysiological significance. Am J Pathol 2005;167:1139–48. [33] Devaraj S, Du Clos TW, Jialal I. Binding and internalization of C-reactive protein by Fcgamma receptors on human aortic endothelial cells mediates biological effects. Arteroscler Thromb Vasc Biol 2005;25:1359–63. [34] Hayman AR, Cox TM. Purple acid phosphatase of the human macrophage and osteoclast. Characterization, molecular properties, and crystallization of teh recombinant di-iron-oxo protein secreted by baculovirus-infected insect cells. J Biol Chem 1994;269:1294–300. [35] Libby P, Theroux P. Pathophysiology of coronary artery disease. Circulation 2005;111:3481–8. [36] Turesson C, O'Fallon WM, Crowson CS, Gabriel SE, Matteson EL. Extra-articular manifestations in rheumatoid arthritis: incidence trends and risk factors over 46 years. Ann Rheum Dis 2003;62:722–7.
