Results from the CAVASIC Study - Semantic Scholar

4 downloads 294941 Views 108KB Size Report
Atherosclerosis 2008;198: 214 –22. 2. Lamina C, Meisinger C, Heid IM, Lowel H, Rant- .... Apple FS, Ler R, Murakami MM. Determination of. 19 cardiac troponin ...
Clinical Chemistry 59:4 692–702 (2013)

Lipids, Lipoproteins, and Cardiovascular Risk Factors

Comparison and Evaluation of Cardiac Biomarkers in Patients with Intermittent Claudication: Results from the CAVASIC Study Barbara Kollerits,1† Gisela Sturm,1† Claudia Lamina,1 Angelika Hammerer-Lercher,2 Barbara Rantner,1,3 Marietta Stadler,4 Tim Ziera,5 Joachim Struck,5 Peter Klein-Weigel,6 Gustav Fraedrich,3 and Florian Kronenberg1

BACKGROUND: Plasma concentrations of the peptides midregional proadrenomedullin (MR-proADM), midregional proatrial natriuretic peptide (MR-proANP), and C-terminal endothelin-1 precursor fragment (CTproET-1) are increased in various cardiovascular conditions. However, there is limited information about the association and comparative performance of these peptides in peripheral arterial disease (PAD). METHODS: The associations of MR-proADM, MRproANP, and CT-proET-1 plasma concentrations with symptomatic PAD were investigated in the CAVASIC (Cardiovascular Disease in Intermittent Claudication) Study. Study participants were a male cohort of 238 patients with a diagnosis of intermittent claudication (IC) and 245 age- and diabetes-matched controls. Results were compared to those for N-terminal pro-B-type natriuretic peptide (NT-proBNP). RESULTS: Each increase of MR-proADM, MR-proANP, and CT-proET-1 by 1 SD was significantly associated with symptomatic PAD: odds ratio (OR) ⫽ 1.78 (95% CI, 1.41–2.25, P ⬍ 0.001), OR ⫽ 1.32 (95% CI, 1.06 – 1.66, P ⫽ 0.014), and OR ⫽ 1.80 (95% CI, 1.43–2.28, P ⬍ 0.001), respectively. The association remained significant for all 3 markers after additional adjustment for log C-reactive protein, serum creatinine, HDL cholesterol, and current smoking. When one adjusts for log NTproBNP and excluding individuals with prevalent cardiovascular disease, MR-proADM and CT-proET-1 still predicted symptomatic PAD. Extended adjustment models

1

Division of Genetic Epidemiology, Department of Medical Genetics, Molecular and Clinical Pharmacology, Innsbruck Medical University, Innsbruck, Austria; 2 Central Institute of Medical and Chemical Laboratory Diagnostics, University Hospital Innsbruck, Innsbruck, Austria; 3 Department of Vascular Surgery, Innsbruck Medical University, Innsbruck, Austria; 4 Third Medical Department of Metabolic Diseases and Nephrology, Hietzing Hospital, Vienna, Austria; 5 Research Department, B.R.A.H.M.S GmbH/Thermo Fisher Scientific, Hennigsdorf, Germany; 6 Clinic for Angiology, HELIOS Kliniken Berlin-Buch, Berlin, Germany. † Barbara Kollerits and Gisela Sturm contributed equally to the work, and both should be considered as first authors. * Address correspondence to this author at: Division of Genetic Epidemiology, Department of Medical Genetics, Molecular and Clinical Pharmacology, Inns-

692

including MR-proADM or CT-proET-1 showed significantly improved model fits compared to models including classical cardiac risk factors or NT-proBNP and led to significant reclassification (P ⬍ 0.05). CONCLUSIONS: This study in a male cohort of patients with IC and age- and diabetes-matched controls indicates a significant association of high MR-proADM, MR-proANP, and CT-proET-1 concentrations with PAD. MR-proADM and CT-proET-1 provide additive information in comparison to NT-proBNP. Moreover, MR-proADM and CT-proET-1 significantly predict PAD in those patients and controls free from prevalent CVD.

© 2012 American Association for Clinical Chemistry

Peripheral arterial disease (PAD)7 is a common manifestation of atherosclerosis and has a prevalence of 5%– 15% in the general population age 55 to 74 years. The clinically symptomatic form of PAD usually presents as intermittent claudication (IC) (1–3 ). Patients with PAD have a high incidence of cardiovascular and cerebrovascular events, and their mortality from cardiovascular disease (CVD) is markedly increased (4 ). Adrenomedullin (ADM) and atrial natriuretic peptide (ANP) are peptides possessing vasodilator, diuretic, and natriuretic properties (5, 6 ). ANP is a peptide hormone secreted from the heart (6 ), whereas ADM was found to be secreted from adrenal medulla, ventricles, lung, kidney, vascular endothelial, and smooth muscle

bruck Medical University, Schöpfstasse 41, A-6020 Innsbruck, Austria. Fax ⫹43-512-9003-73560; e-mail [email protected]. Received July 25, 2012; accepted November 30, 2012. Previously published online at DOI: 10.1373/clinchem.2012.193656 7 Nonstandard abbreviations: PAD, peripheral arterial disease; IC, intermittent claudication; CVD, cardiovascular disease; ADM, adrenomedullin; ANP, atrial natriuretic peptide; T2DM, type 2 diabetes mellitus; IHD, ischemic heart disease; ET-1, endothelin-1; CT-proET-1, C-terminal endothelin-1 precursor fragment; MR-proADM, midregional proadrenomedullin; MR-proANP, midregional proatrial natriuretic peptide; NT-proBNP, N-terminal pro-B-type natriuretic peptide; CAVASIC Study, Cardiovascular Disease in Intermittent Claudication Study; DS, discrimination slope; hs-cTNT, high-sensitivity cardiac troponin T; ABI, ankle-brachial index; IDI, integrated discrimination improvement; CRP, C-reactive protein; OR, odds ratio.

Pro-ADM, Pro-ANP, and Pro-Endothelin-1 in PAD

cells (7 ). Increased levels of ADM have been reported in patients with myocardial infarction (8 ), congestive heart failure (9 ), hypertension (10 ), and various diseases such as chronic kidney disease (11, 12 ) and type 2 diabetes mellitus (T2DM) (13 ). Moreover, in small studies ADM was also increased in patients suffering from ischemic heart disease (IHD) and/or PAD (14, 15 ). ANP has been shown to be increased in patients with diseases like coronary atherosclerosis (16 ), heart failure (17 ), and acute myocardial infarction (18 ). In contrast to ADM and ANP, the active molecule endothelin-1 (ET-1), a 21 amino acid vasoconstrictor peptide, is reported to be produced primarily by vascular endothelial cells, with detected binding sites in blood vessels, heart, and kidney (19, 20 ). In patients with heart failure increased plasma concentrations of the C-terminal ET-1 precursor fragment (CTproET-1) have been related to all-cause mortality and mortality due to cardiovascular causes (21, 22 ). Moreover, CT-proET-1 was reported to be a significant predictor of death or heart failure in patients after acute myocardial infarction (23 ). However, there is limited conclusive information about increased concentrations of the peptides midregional proadrenomedullin (MR-proADM), midregional proatrial natriuretic peptide (MR-proANP), or CT-proET-1 in symptomatic PAD and whether such increases are independent from the established heart failure parameter N-terminal pro-B-type natriuretic peptide (NT-proBNP) and persist when those individuals with prevalent CVD are excluded. The aim of this study was therefore to investigate the role of MRproADM, MR-proANP, and CT-proET-1 plasma concentrations in symptomatic PAD. Materials and Methods STUDY PARTICIPANTS AND STUDY DESIGN

The Cardiovascular Disease in Intermittent Claudication (CAVASIC) Study (24 ) is a case control study with the aim to identify cardiovascular risk factors in patients with IC. Patients and controls were enrolled between 2002 and 2006 in 2 clinical centers: the Department of Vascular Surgery, Medical University Innsbruck, and the Third Medical Department of Metabolic Diseases and Nephrology, Hietzing Hospital, Vienna, Austria. Patients (n ⫽ 249) were consecutively included in the study when they presented with or had a history of IC (PAD IIa or IIb according to the criteria of Fontaine), regardless of whether they had already undergone a treatment procedure (bypass surgery or intervention). Patients were excluded from the study for any of the following reasons: presence of acute or critical limb ischemia (Fontaine III or IV), impaired kidney function with serum creatinine ⬎133 ␮mol/L, malig-

nancy, previous organ transplantation, or therapy with nicotinic acid or corticosteroids. We recruited 251 control individuals from the same geographic region matched for age and T2DM. All members of the control group had volunteered to participate in the study after the publication of an invitation in local newspapers. The same exclusion criteria were used for the patients and controls. Control individuals with symptomatic PAD were excluded, but those with a history of CVD were allowed to participate. A total of 19 of the 251 controls either had a positive cardiovascular history for angina pectoris according to the Rose questionnaire (25 ) or had documented cardiovascular events or procedures, such as myocardial infarction, aortocoronary bypass surgery, percutaneous transluminal coronary angioplasty, and/or coronary angiography or stroke. Patients and controls did not suffer from acute illnesses or clinically detectable inflammatory processes at the time of enrollment. All participants provided written informed consent and the examination protocol was approved by the ethics committee of the participating study centers. To minimize interobserver bias all interviews and examinations at each of the 2 clinical centers were performed by 1 medical doctor who was specially trained in vascular examinations and echocardiography. Owing to inappropriate or missing blood samples, MR-proADM, MR-proANP, and CT-proET-1 were measured in 238 patients and 245 controls. All data and analyses described in this manuscript are based on these 483 participants. BASELINE AND VASCULAR EXAMINATION AND OTHER PHENOTYPIC CHARACTERIZATION

Demographic data, patient histories, and atherosclerosis risk profiles were recorded via a standardized interview. A diagnosis of T2DM was determined in the study participants if their fasting plasma glucose concentration was ⱖ7 mmol/L and/or they were being treated with antidiabetic drugs. Participants were considered hypertensive when the systolic blood pressure was ⱖ140 mmHg and/or if the diastolic blood pressure was ⱖ90 mmHg, and/or if they were being treated with antihypertensive drugs. LABORATORY MEASUREMENTS

Serum and EDTA plasma were obtained after an overnight fasting period and stored in aliquots at ⫺80 °C until laboratory measurements were done. We used 3 novel commercially available fully automated immunometric immunoassays (all ThermoFisher Scientific B.R.A.H.M.S KRYPTOR) for the measurement of MRproADM, MR-proANP, and CT-proET-1 according to the manufacturer’s instruction manuals. NT-proBNP and high-sensitivity cardiac troponin T (hs-cTNT, fifth Clinical Chemistry 59:4 (2013) 693

generation) concentrations were measured with a commercially available assay on an E170 instrument (Modular®, Roche Diagnostics). NT-proBNP and hscTNT, as well as MR-proADM, MR-proANP, and CTproET-1, showed sufficient stability concerning longterm storage (26 –29 ). Details of these assays such as precision data are summarized in Table 1 of the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol59/issue4. All blood samples were analyzed in batches by personnel blinded with regard to the case control status of the study participants and with patients and controls in random order on each plate. The samples were stored for a mean of 5 years until MR-proADM, MR-proANP, and CT-proET-1 were measured. The materials of patients and controls were stored under equivalent conditions of temperature and duration. Further details on the baseline examination, such as the ankle-brachial index (ABI) measurement, phenotypic characterization, and laboratory measurements, can be found in the online Supplemental material. STATISTICAL METHODS

We analyzed categorical and continuous data using the ␹2 test, unpaired t-test, or Wilcoxon rank–sum test in case of nonnormally distributed values. We identified variables associated with PAD (or prevalent CVD in PAD patients) by logistic regression analysis applying different adjustment strategies. A test of deviances on nested models was performed to check whether MR-proADM, MR-proANP, CT-proET-1, or NT-proBNP significantly added to various adjustment models. The discriminative abilities of MR-proADM, MR-proANP, and CT-proET-1 were tested by c-statistics and integrated discrimination improvement (IDI). IDI evaluates reclassification by comparing mean predicted probabilities of events in cases compared to controls when adding new variables [ ⫽ discrimination slope (DS) new model DS (new)] to the mean predicted probabilities of events in cases compared to controls of a baseline model [ ⫽ discrimination slope baseline model DS (base)] as a continuous outcome. The absolute (Abs.) and relative (Rel.) IDI are derived by: Abs.IDI ⫽ DS(new) ⫺ DS(base), Rel.IDI ⫽ [DS(new)/DS(base)] ⫺ 1. A positive IDI refers to an improvement in risk prediction. All analyses were conducted in SPSS version 18.0 (IBM SPSS software) and R 2.14.1. Results BASELINE CHARACTERISTICS

Between 2002 and 2006, we included 249 unrelated male patients (age range, 38 – 69 years) presenting with IC graded as Fontaine stage IIa or IIb. A total of 38 of 694 Clinical Chemistry 59:4 (2013)

these patients had T2DM. The patients and the ageand T2DM-matched controls had similar mean (SD) T2DM durations [11.7 (7.5) vs 11.5 (9.0) years]. MRproADM, MR-proANP, and CT-proET-1 were measured in 238 patients and 245 controls. Table 1 provides an overview of the characteristics of the patients and controls with MR-proADM, MRproANP, and CT-proET-1 plasma concentrations available. The patient group had a significantly higher frequency of smokers and significantly higher triglyceride, C-reactive protein (CRP), and glucose concentrations. HDL cholesterol, creatinine, and albumin concentrations were significantly lower in the patient group compared to the control group. Hypertension and prevalent CVD events were more frequent in the patient group. MR-proADM, MR-proANP, AND CT-proET-1 IN PATIENTS AND CONTROLS

Table 1 shows that patients with PAD had significantly higher MR-proADM, MR-proANP, and CT-proET-1 concentrations than controls. Table 2 lists correlations between MR-proADM, MR-proANP, and CT-proET-1 and various parameters. MR-proADM and MR-proANP as well as MR-proANP and CT-proET-1 correlated moderately (r 2 ⫽ 0.360 and 0.354, respectively, both P ⬍ 0.001). MR-proADM and CT-proET-1 correlated to a higher extent (r 2 ⫽ 0.671, P ⬍ 0.001). MULTIPLE LOGISTIC REGRESSION ANALYSIS

We investigated whether MR-proADM, MR-proANP, or CT-proET-1 concentrations are associated with PAD. Each increase of MR-proADM, MR-proANP, and CT-proET-1 by 1 SD was significantly associated with symptomatic PAD: odds ratio (OR) ⫽ 1.78 (95% CI, 1.41–2.25, P ⬍ 0.001), OR ⫽ 1.32 (95% CI, 1.06 – 1.66, P ⫽ 0.014), and OR ⫽ 1.80 (95% CI, 1.43–2.28, P ⬍ 0.001), respectively (Table 3, model 1). After additional adjustment for log CRP, serum creatinine, HDL cholesterol, and current smoking the association was still highly significant for all 3 parameters: MR-proADM (OR ⫽ 1.60, P ⫽ 0.002), MR-proANP (OR ⫽ 1.65, P ⫽ 0.002), and CTproET-1 (OR ⫽ 1.67, P ⬍ 0.001), respectively (Table 3, model 2). When additionally accounting for log NT-proBNP, MR-proADM was borderline significant (OR ⫽ 1.33, P ⫽ 0.08), whereas CT-proET-1 remained significantly associated (OR ⫽ 1.44, P ⫽ 0.021) (Table 3, model 3). We furthermore analyzed if estimates in each model changed when we excluded individuals with prevalent CVD. MR-proADM and CT-proET-1 did not show major differences in estimates; however, MRproANP was no longer significant (Table 3).

Pro-ADM, Pro-ANP, and Pro-Endothelin-1 in PAD

Table 1. Differences between patients with PAD and control individuals matched by age and T2DM at baseline.a Controls (n ⴝ 245)

Age, years

56.9 (9.5) [48.0; 61.0; 64.0]

Body mass index, kg/m2 Smoking (nonsmokers/former smokers/current smokers), n (%) Diabetes mellitus, n (%) MR-proADM, nmol/L

26.6 (3.9)

26.7 (3.8)

0.483

11/101/123c

(46.1%/42.0%/11.8%)

(4.6%/42.4%/51.7%)

41 (16.7%) 0.51 (0.13) 72.1 (34.2) 56.9 (9.8) 77.8 (103.3) [35.4; 51.4; 82.5]

Hs-cTNT, ng/L

[54.0; 60.0; 63.0]

113/103/29

[50.6; 56.3; 62.3] NT-proBNP, ng/Lb

P

0.893

[24.2; 26.5; 28.7]

[47.5; 65.8; 87.3] CT-proET-1, pmol/L

(n ⴝ 238)

58.3 (6.3)

[24.0; 26.0; 28.7]

[0.43; 0.49; 0.57] MR-proANP, pmol/L

Patients

5.12 (5.26) [3.54; 3.54; 3.54]

38 (16.0%)

⬍0.001

0.819 ⬍0.001

0.58 (0.17) [0.47; 0.56; 0.67] 85.4 (58.5)

0.024

[52.5; 70.7; 102.0] ⬍0.001

63.5 (15.5) [54.1; 60.9; 70.1]

⬍0.001

178.7 (296.3) [35.4; 93.4; 184.4] 6.08 (5.80)

0.003

[3.54; 3.54; 6.47]

Total cholesterol, mg/dLb

207.64 (35.10)

204.63 (41.35)

LDL cholesterol, mg/dLb

135.63 (33.07)

132.71 (37.42)

0.362

HDL cholesterol, mg/dLb

59.43 (16.59)

49.31 (13.69)

⬍0.001

Triglycerides, mg/dLb

132.1 (80.4)

171.5 (123.2)

[79.0; 111.0; 155.0]

[94.0; 134.5; 212.0]

[48; 57; 69]

CRP, mg/L

2.64 (3.37) [1.00; 1.40; 2.80]

Glucose, mg/dLb

105.04 (32.34) [89.0; 96.0; 106.0]

Hemoglobin A1c , %

5.90 (0.86) [5.50; 5.60; 5.90]

Creatinine, mg/dLb

1.03 (0.14) [0.93; 1.02; 1.12]

Albumin, g/dLb Urine albumin/creatinine ratio, mg/g creatinineb Systolic blood pressure, mmHg

Hypertension, n (%)e CVD, n (%)

⬍0.001 ⬍0.001

6.25 (10.63) [2.15; 4.20; 7.00] 109.06 (29.89)

0.002

[92.0; 100.0; 113.0] 6.04 (0.96)

0.001

[5.50; 5.80; 6.18] ⬍0.001

0.98 (0.19) [0.87; 0.97; 1.07] 4.45 (0.45)

18.32 (38.74)

78.98 (243.07)

[5.80; 7.77; 13.62]

[8.26; 13.61; 32.86]

139.2 (16.8)

150.3 (19.5)

82.1 (8.6) [80.0; 80.0; 90.0]

Ejection fraction, %d

[41; 48; 54]

4.57 (0.39)

[125.6; 140.0; 150.0] Diastolic blood pressure, mmHg

0.438

61.4 (5.7)

0.001 ⬍0.001

⬍0.001

[135.0; 150.0; 165.0] 83.4 (9.8)

0.254

[80.0; 80.0; 90.0] ⬍0.001

57.2 (8.7)

[59.0; 61.0; 65.0]

[53.0; 57.0; 62.0]

145 (59.2%)

205 (86.1%)

18 (7.3%)

74 (31.1%)

⬍0.001 ⬍0.001

Continued on page 696

Clinical Chemistry 59:4 (2013) 695

Table 1. Differences between patients with PAD and control individuals matched by age and T2DM at baseline.a (Continued from page 695) Controls

Patients

(n ⴝ 245)

(n ⴝ 238)

P

1.07 (0.12)

0.71 (0.24)

⬍0.001

[1.00; 1.08; 1.15]

[0.54; 0.69; 0.89]

f

ABI

a

Data are presented as n (%) or mean (SD) with [25th; 50th; 75th percentile] in case of nonnormal distribution. To convert NT-proBNP concentrations to picomoles per liter, divide by 8.457; to convert cholesterol to millimoles per liter, multiply by 0.0259; to convert triglycerides to millimoles per liter, multiply by 0.0113; to convert glucose to millimoles per liter, multiply by 0.0555; to convert creatinine to micromoles per liter, multiply by 88.4; to convert serum albumin to grams per liter, multiply by 10; to convert urine albumin/creatinine ratio to grams per mole, divide by 8.84. c Smoking status: 3 missing values. d Ejection fraction: 91 missing values (66 patients and 25 controls). e Hypertension was defined as systolic blood pressure ⱖ140 mmHg and/or diastolic blood pressure ⱖ90 mmHg, and/or receiving antihypertension treatment. f The lowest ABI value from the 4 sites was used for data analysis [see Rantner et al. (24 )]. Individuals with ABI values ⬎1.30 were excluded from this analysis. b

COMPARISONS OF MODELS

To evaluate whether the investigated biomarkers contributed to better discrimination between patients and controls, we applied various statistical concepts such as deviances, c-statistics, and IDI. In a first step we used a test of deviances on nested models to evaluate whether the association with PAD can be explained

by all 4 parameters. The models including MRproADM, MR-proANP, CT-proET-1, or log NTproBNP showed significantly improved model fits compared to model 2, with all P ⱕ 0.003, (Table 4). MR-proADM, MR-proANP, and CT-proET-1 also provided improved model fits compared to a model including age and log NT-proBNP as an established marker for

Table 2. Spearman rank correlation coefficients between MR-proADM, MR-proANP, CT-proET-1, and various clinical and laboratory parameters. Correlation coefficient (r)a MR-proADM

a

Age

0.394***

Body mass index

0.267***

MR-proANP

CT-proET-1

0.507***

0.253***

⫺0.089

0.076

Total cholesterol

⫺0.079

⫺0.155**

⫺0.034

LDL cholesterol

⫺0.125**

⫺0.163***

⫺0.086

HDL cholesterol

⫺0.089

0.143**

⫺0.054

⫺0.198***

Triglycerides

0.170***

C-reactive protein

0.308***

0.017

Glucose

0.333***

0.038

0.149***

Hemoglobin A1c

0.226***

⫺0.025

0.121***

Creatinine

0.247***

0.147** ⫺0.120**

0.104** 0.246***

0.232*** ⫺0.049

Albumin

0.002

Urine albumin/creatinine ratio

0.280***

0.268***

Systolic blood pressure

0.309***

0.171***

Diastolic blood pressure

0.052

NT-proBNP

0.389***

0.704***

0.305***

hs-cTNT

0.341***

0.234***

0.206***

⫺0.058

0.246*** 0.179*** ⫺0.009

Ejection fraction

⫺0.251***

⫺0.108*

⫺0.107**

ABI

⫺0.291***

⫺0.041

⫺0.243***

*P ⬍0.05; **P ⬍0.01; ***P ⬍0.001.

696 Clinical Chemistry 59:4 (2013)

Pro-ADM, Pro-ANP, and Pro-Endothelin-1 in PAD

Table 3. Logistic regression analysis predicting PAD for MR-proADM and MR-proANP as well as CT-proET-1. Model 1a ORb

Model 2a

(95% CI)

ORb

P

(95% CI)

Model 3a P

ORb

(95% CI)

P

MR-proADM, per 1-SD increasec All patients and controls (n ⫽ 483)

1.78

(1.41–2.25)

⬍0.001

1.60

(1.18–2.17)

0.002

1.33

(0.97–1.84)

0.080

Excluding CVDs (n ⫽ 388)

1.80

(1.41–2.30)

⬍0.001

1.72

(1.24–2.38)

0.001

1.51

(1.07–2.14)

0.019

All patients and controls (n ⫽ 483)

1.32

(1.06–1.66)

0.014

1.65

(1.21–2.25)

0.002

0.93

(0.64–1.37)

0.720

Excluding CVDs (n ⫽ 388)

1.09

(0.87–1.35)

0.457

1.27

(0.93–1.72)

0.132

0.79

(0.51–1.22)

0.282

MR-proANP, per 1-SD increasec

CT-proET-1, per 1-SD increasec All patients and controls (n ⫽ 483)

1.80

(1.43–2.28)

⬍0.001

1.67

(1.26–2.22)

⬍0.001

1.44

(1.06–1.95)

0.021

Excluding CVDs (n ⫽ 388)

1.61

(1.26–2.05)

⬍0.001

1.56

(1.15–2.11)

0.004

1.41

(1.02–1.94)

0.035

a

Model 1, adjusted for age; model 2, adjusted for age, log C-reactive protein, creatinine, HDL cholesterol, and current smoking; model 3, adjusted for age, log C-reactive protein, creatinine, HDL cholesterol, current smoking, and log NT-proBNP. b The ORs for PAD status were estimated for each increment of 1 SD of MR-proADM, MR-proANP, and CT-proET-1. The SD was calculated on the basis of the values of the total group or the total group excluding those with prevalent CVDs, respectively. c The SDs for each of the biomarkers were as follows: for MR-proADM, the 1-SD increment was 0.15 nmol/L for all patients and controls and 0.13 nmol/L for all patients and controls excluding CVDs; for MR-proANP, the 1-SD increment was 48 pmol/L for all patients and controls and 35 pmol/L for all patients and controls excluding CVDs; for CT-proET-1, the 1-SD increment was 13 pmol/L for all patients and controls and 12 pmol/L for all patients and controls excluding CVDs.

functional cardiac impairment (Table 5, all P ⱕ 0.003). In the extended model 3, including log NT-proBNP, only CT-proET-1 significantly improved model fit (Table 5). The c-statistics improved significantly for all models containing the 4 cardiac biomarkers (0.685 for log NT-proBNP, 0.564 for MR-proANP, 0.645 for MRproADM, and 0.633 for CT-proET-1, respectively) compared to a model including only age [c-statistic, 0.499 (0.445– 0.550)]. The c-statistics for all 4 biomarkers were not significantly different from the c-statistic of the extended adjustment model 2, as can be seen from the overlapping 95% bootstrap CIs (Table 4). The same holds true for comparisons of the c-statistics of the model containing age and log NT-proBNP or the extended model 3 and log NT-proBNP with the c-statistics with the addition of MR-proADM, MRproANP, or CT-proET-1 (Table 5). However, owing to the fact that measures such as IDI are supposed to be more sensitive than the c-statistic because IDI includes the assessment of risk reclassification, we calculated the IDI for MRproADM, MR-proANP, CT-proET-1, and log NTproBNP. We observed a positive and significant IDI (P ⬍ 0.05) for all 4 markers when we compared them to the model including only age and the extended adjustment model 2 (Table 4). The IDI increase was about twice as high for log NT-proBNP (absolute IDI, 0.053; relative IDI, 0.15; P ⬍ 0.0001) compared to MRproADM, MR-proANP, and CT-proET-1, respectively. Nonetheless, MR-proADM (absolute IDI, 0.02; relative ID, 0.17; P ⫽ 0.002) and CT-proET-1 (absolute

IDI, 0.02; relative IDI, 0.17; P ⫽ 0.002) still provide improved reclassification compared to the model containing age and log NT-proBNP. However, a trend in IDI increase was observed only for CT-proET-1 (absolute IDI, 0.008; relative IDI, 0.02; P ⫽ 0.068) when compared to the extended model 3 including log NTproBNP (Table 5). SENSITIVITY ANALYSES

We performed several sensitivity analyses: a model for MR-proADM, MR-proANP, and CT-proET-1 in which model 2 was additionally adjusted for log urine albumin/ creatinine ratio, hypertension, ejection fraction, log hscTNT, triglycerides, or hemoglobin A1c. These analyses did not reveal substantial differences in estimates. PREVALENT CVD

We applied multiple logistic regression analysis to test which marker (MR-proADM, MR-proANP, or CTproET-1) is associated with prevalent CVD in PAD patients. Each increase of MR-proANP by 1 SD adjusted for log CRP, serum creatinine, HDL cholesterol, and current smoking was highly significantly associated with prevalent CVD (OR, 1.80; 95% CI, 1.23–2.62; P ⫽ 0.002). After further adjustment for log NT-proBNP, the P value of the association was only of borderline significance (P ⫽ 0.063, Table 6). This result, however, has to be viewed with caution because of the high correlation between MR-proANP and log NT-proBNP (r ⫽ 0.76, P ⬍ 0.001). There was a trend only for ageadjusted MR-proADM and CT-proET-1, which comClinical Chemistry 59:4 (2013) 697

698 Clinical Chemistry 59:4 (2013) 1.84 ⫻ 10 0.01 2.97 ⫻ 10⫺7 2.69 ⫻ 10⫺7

2.89 ⫻ 10⫺8 3.65 ⫻ 10⫺4 0.003 3.75 ⫻ 10⫺4

⫺30.78 ⫺12.70 ⫺8.81 ⫺12.66

426.39 444.47 448.36 444.52

⫺13

⫺54.17 ⫺6.49 ⫺26.27 ⫺26.46

P

652.96 598.79 646.47 626.69 626.50 457.17

Difference in deviance

0.867 (0.832–0.897) 0.855 (0.817–0.885) 0.852 (0.814–0.884) 0.854 (0.817–0.883)

0.499 (0.445–0.550) 0.685 (0.632–0.729) 0.564 (0.494–0.615) 0.645 (0.595–0.693) 0.633 (0.579–0.683) 0.846 (0.808–0.879)

c-Statistic (95% CI)a

b

0.15 0.06 0.04 0.05

13.72 1.63 6.84 6.66

Relative IDIb

423.29 426.27 420.93

⫹ MR-proADM (per 1-SD increase)

⫹ MR-proANP (per 1-SD increase)

⫹ CT-proET-1 (per 1-SD increase)

426.39

587.97

⫹ CT-proET-1 (per 1-SD increase)

Model 3 (including age, log NT-proBNP, log C-reactive protein, creatinine, HDL cholesterol, and smoking status)

586.40 589.72

⫹ MR-proANP (per 1-SD increase)

598.79

⫹ MR-proADM (per 1-SD increase)

Model 1⫹ NT-proBNP (including age and log NT-proBNP)

Deviance

⫺5.47

⫺0.13

⫺3.10

⫺10.82

⫺9.07

⫺10.53

Difference in deviance

0.019

0.723

0.078

0.001

0.003

0.001

P

0.872 (0.837 to 0.902)

0.868 (0.833 to 0.897)

0.870 (0.835 to 0.899)

0.867 (0.832 to 0.898)

0.704 (0.651 to 0.747)

0.727 (0.679 to 0.773)

0.706 (0.658 to 0.747)

0.685 (0.633 to 0.733)

c-Statistic (95% CI)a

0.008 (⫺0.001 to 0.016)

0.0003 (⫺0.001 to 0.002)

0.005 (⫺0.002 to 0.010)

0.411c

0.019 (0.007 to 0.033)

0.018 (0.007 to 0.029)

0.020 (0.007 to 0.033)

0.116c

IDI (95% CI)

0.068

0.673

0.145

0.002

0.002

0.002

P, IDI

0.02

0.001

0.01

0.17

0.15

0.17

Relative IDIb

Table 5. Logistic regression analysis comparing deviances, c-statistics, and IDI indices of new cardiac markers to NT-proBNP and/or classical risk factors.

CIs for the c-statistic were derived via the bootstrap method. A relative IDI of, e.g., 0.17 refers to a relative improvement of 17%. c Because the interpretation of the IDI of subsequent models depends on the magnitude of the baseline model, the discrimination slope of the baseline model is provided.

a

b

1.86 ⫻ 10⫺7 0.001 0.020 0.004

4.40 ⫻ 10⫺14 0.015 1.70 ⫻ 10⫺7 3.82 ⫻ 10⫺7

0.008c 0.108 (0.080–0.136) 0.013 (0.003–0.023) 0.054 (0.034–0.074) 0.053 (0.032–0.073) 0.358c 0.053 (0.033–0.073) 0.022 (0.010–0.035) 0.013 (0.002–0.024) 0.019 (0.010–0.032)

P, IDI

IDI (95% CI)

CIs for the c-statistic were derived via the bootstrap method. A relative IDI of, e.g., 13.72 refers to a relative improvement of 1372%. c Because the interpretation of the IDI of subsequent models depends on the magnitude of the baseline model, the discrimination slope of the baseline model is provided.

a

Model 1, including age ⫹ Log NT-proBNP (per 1-SD increase) ⫹ MR-proANP (per 1-SD increase) ⫹ MR-proADM (per 1-SD increase) ⫹ CT-proET-1 (per 1-SD increase) Model 2 (including age, log C-reactive protein, creatinine, HDL cholesterol, smoking status) ⫹ Log NT-proBNP (per 1-SD increase) ⫹ MR-proANP (per 1-SD increase) ⫹ MR-proADM (per 1-SD increase) ⫹ CT-proET-1 (per 1-SD increase)

Deviance

Table 4. Logistic regression analysis comparing deviances, c-statistics, and IDI indices of new cardiac markers to basic risk models.

Pro-ADM, Pro-ANP, and Pro-Endothelin-1 in PAD

Table 6. Logistic regression analysis predicting prevalent CVD in PAD patients for MR-proADM, MR-proANP, and CT-proET-1. Model 1a

Model 2a

Model 3a

OR

(95% CI)

P

OR

(95% CI)

P

OR

(95% CI)

P

1.31

(0.97–1.76)

0.077

1.16

(0.81–1.67)

0.413

1.01

(0.68–1.49)

0.959

1.71

(1.20–2.44)

0.003

1.80

(1.23–2.62)

0.002

1.73

(0.97–3.08)

0.063

1.26

(0.96–1.66)

0.097

1.20

(0.87–1.64)

0.266

1.11

(0.77–1.59)

0.589

b

MR-proADM (per 1-SD increase) All patients (n ⫽ 235)

MR-proANP (per 1-SD increase)c All patients (n ⫽ 235) CT-proET-1 (per 1-SD increase)d All patients (n ⫽ 235) a

Model 1, adjusted for age; model 2, adjusted for age, log C-reactive protein, creatinine, HDL cholesterol, and current smoking; model 3, adjusted for age, log C-reactive protein, creatinine, HDL cholesterol, current smoking, and log NT-proBNP. b For MR-proADM the 1-SD increment was 0.17 nmol/L for all patients. c For MR-proANP the 1-SD increment was 57 pmol/L for all patients. d For CT-proET-1 the 1-SD increment was 15 pmol/L for all patients.

pletely disappeared after further adjustment for other risk factors (see Table 6). Discussion In this study we investigated the 3 peptides MRproADM, MR-proANP, and CT-pro-ET1 in a cohort of patients with symptomatic PAD defined by IC and an age- and diabetes-matched control group. We observed significantly higher concentrations of MRproADM, MR-proANP, and CT-pro-ET1 in PAD patients than in controls. The higher the concentrations of all 3 peptides, the higher was the risk for the presence of PAD. Moreover, MR-proADM and CT-pro-ET1 provided incremental information beyond major risk markers such as NT-proBNP. Additionally, effects for these 2 peptides persisted after we excluded data for individuals with prevalent CVDs. A recent review suggested a protective effect of ADM on CVD (30 ). Plasma ADM concentrations reflect activation of defense mechanisms against vascular damage, besides vasodilation and natriuresis, such as antiinflammation and antioxidation, and exhibit proliferative effects on vascular endothelial cells and antiproliferative effects on vascular smooth muscle cells, supporting the role of ADM as an antiproliferative factor inhibiting the development of atherosclerosis (7 ). Natriuretic peptides such as ANP not only play a role in volume and pressure homeostasis, but also show antiproliferative as well as antifibrotic and antiinflammatory effects (31 ). In addition to its vasoconstricting properties, ET-1 has been reported to mediate vessel remodeling by interacting with growth factors and cytokines. Further actions of ET-1 are the promotion of

vascular inflammation and excessive oxidative stress (32, 33 ). In several studies increased ADM, ANP, and CTproET-1 concentrations have been observed in different CVD end points (8, 9, 17, 18, 21–23 ). Investigations in PAD disease are sparse, and previous studies were limited by study populations with only a small number of individuals. In a study with 35 patients with and 37 patients without PAD, ADM concentrations were negatively correlated with ABI and positively with high-sensitivity CRP (15 ). In the CAVASIC study we also found a significant but moderate negative correlation between MR-proADM and ABI (r ⫽ ⫺0.291) and a significant positive correlation with CRP (r ⫽ 0.308). In a further study, Suzuki et al. observed an increase in plasma ADM concentrations in patients with cardiovascular risk factors related to the presence of vascular complications. Furthermore, these authors reported that plasma ADM was independently associated with IHD or PAD but did not clarify whether or not an isolated PAD was associated with high ADM concentrations (14 ). This issue was clearly addressed in the present study, in which we showed that MR-proADM and CT-proET-1 were associated with symptomatic PAD even in individuals without prevalent CVD: when patients and controls with prevalent CVD were excluded from the analyses, the association of MR-proADM and CT-proET-1 with PAD remained clearly significant. The association remained significant if data were additionally adjusted for main PAD risk factors such as age, CRP, creatinine, HDL cholesterol, hypertension, current smoking, cardiac ejection fraction, or even established markers of cardiac stress such as NT-proBNP and hs-cTNT. Therefore, it can be concluded that in Clinical Chemistry 59:4 (2013) 699

patients with symptomatic PAD the increase in MRproADM and CT-proET-1 might be directly linked to vascular lesions caused by PAD irrespective of cardiac dysfunction and complications, kidney function, and systemic inflammation. The associations for MR-proANP were no longer significant after exclusion of patients and controls with prevalent CVD or adjustment for NT-proBNP. The reason why we found an overall stronger effect of MRproADM and CT-proET-1 might have been because of the production of ADM and CT-proET-1 directly in the vasculature as well as in the smooth muscle cells and more autocrine and paracrine action in this case, whereas ANP is a circulating hormone secreted mainly from the cardiac atria. This hypothesis is in line with the observation that specific receptors for ADM as well as for ET-1 are localized in the vascular endothelium in humans (7, 32 ). Moreover, it was shown that CGRPR-1 (the calcitonin gene–related peptide receptor), a target for ADM, acts via the endothelium and not the coronary vascular smooth muscles (34 ). Therefore, the properties of ADM and ET-1 might be more specifically related to vascular disease. In experimental and clinical studies it has been demonstrated that the infusion or increase of endogenous ADM results in a better prognosis in disorders such as heart failure or myocardial infarction and might be a promising approach for the treatment of atherosclerosis (7 ). In PAD increasing concentrations of ADM may also rise constantly in a counteracting and compensatory manner as a reaction to preexisting vascular damage. It has also been discussed that the production of endothelins by endothelial cells is enhanced as a reaction to vascular tissue injury, a process that might be decelerated by applying selective ET-1 antagonists (32 ). Increasing concentrations of endothelins in PAD can thus be seen as an alarming sign of a disturbed vascular function and inflammatory processes. The presence of PAD itself is associated with an increased risk of cardiovascular or cerebrovascular events and total mortality (35 ). The risk is already increased in patients with asymptomatic or subclinical PAD and increases considerably in PAD patients with clinical symptoms, which is important from a general health perspective because PAD is often diagnosed late (36 ). Therefore, laboratory parameters, such as MR-proADM and CTproET-1 besides NT-proBNP, could help to improve the characterization of patients with PAD. CLINICAL UTILITY OF THE MEASURED PARAMETERS

To better characterize PAD patients, we also investigated whether MR-proANP, MR-proADM, or CTproET-1 is associated with prevalent CVD. MRproANP and not MR-proADM or CT-proET-1 was the marker remaining significantly associated with preva700 Clinical Chemistry 59:4 (2013)

lent CVD after adjustment for classical CVD risk factors. This is in line with the observation that MRproANP is a cardiac marker, whereas MR-proADM and CT-proET-1 reflect a generalized endothelial vascular dysfunction that is more evident in symptomatic PAD but not necessarily in prevalent CVD. These 3 biomarkers might be helpful in patients with T2DM who have a high risk of PAD and limb amputation as well as cardiovascular risk. MR-proADM and CTproET-1 could be considered of diagnostic value in patients in whom ABI measurement is inappropriate, such as patients with amputations as well as patients with media sclerosis. This kind of pronounced arterial stiffness is known to be highly prevalent in diabetic and/or geriatric patients and in those with longstanding hypertension. Media sclerosis results in falsely increased ABI values above 1.30. If at the same time the blood flow is decreased by arterial obstruction, the diagnosis of PAD by ABI is often masked by media sclerosis. Consequently, the ABI measurement may show normal values without pointing to media sclerosis, thereby resulting in vessels incorrectly classified as healthy. MR-proADM and CT-proET-1 should point to these patients with incorrectly normal or increased ABI who require an extended clinical examination. The investigated biomarkers might also be especially helpful for the characterization and intensified management of risk factors in patients with diseases that limit their exercise tolerance (e.g., COPD, orthopedic problems) and who would not be able to undergo exercise tests (treadmill). Furthermore, measurements of ABI and treadmill tests are time-consuming and investigator dependent. MEASURES OF PERFORMANCE OF PREDICTION MODELS

Various methods and measures exist for evaluating the performance of prediction models, and their pros and cons have been discussed. We provide tests of deviances on nested models, c-statistics, and IDI. IDI is reported to be more sensitive than the c-statistic because it evaluates reclassification, can be directly referred to event probabilities, and weights the improvement of sensitivity over all possible cutoffs, whereas the c-statistic weights large sensitivities to a higher extent and provides no information on the amount of change in risk. Moreover, we found that the results of the test on deviances are in line with the results of the IDI analysis for all models. Therefore, the model performance of the analyzed markers in the study at hand is described equally well by these 2 methods. All 3 applied methods indicate that NT-proBNP is the strongest marker for PAD risk determination but also show that MR-proADM and CT-proET-1 provide significant additive information beyond classical cardiac risk factors or NT-proBNP.

Pro-ADM, Pro-ANP, and Pro-Endothelin-1 in PAD

STRENGTHS AND LIMITATIONS

As we discussed recently (11 ), it can be considered to be a strength of our study that we measured the markedly more stable midregional parts of the peptide precursor fragments of ADM and ANP instead of the active hormones, which have a short half life and are associated with other technical difficulties related to measurement (37, 38 ). We also measured the CTproET-1 precursor fragment, which is much more stable than the active ET-1 (39 ). Although the material was stored for a mean of 5 years, we do not expect that this has caused the concentration differences between cases and controls, because materials of both groups were treated identically. Our study is limited by the fact that it is a case control study and therefore no causal conclusions can be drawn. Moreover, our study population comprised only white men. It remains to be seen whether our findings can be extrapolated to women and different ethnic groups. In addition, we have limited data on ejection fraction in both patients and controls and therefore the interpretation of this sensitivity analysis has to be treated with caution. In general, due to the high correlation of MR-proANP and NT-proBNP all regression models including both parameters have to be treated with caution as well.

ciation between MR-proADM, MR-proANP, and CTpro-ET1 concentrations and PAD. The association of MR-proADM and CT-proET-1 with PAD seems to be independent from major risk factors such as NTproBNP and remains significant when excluding those with prevalent CVD.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article. Authors’ Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest:

Conclusions

Employment or Leadership: None declared. Consultant or Advisory Role: None declared. Stock Ownership: None declared. Honoraria: None declared. Research Funding: ThermoFisher Scientific/B.R.A.H.M.S GmbH, ROCHE Diagnostics; B. Kollerits, Austrian Nationalbank, Project 13662; F. Kronenberg, Austrian Nationalbank, Project 9331, the Austrian Heart Fund, and the Genomics of Lipid-associated Disorders—GOLD of the “Austrian Genome Research Programme GENAU.” Expert Testimony: None declared. Patents: None declared.

In summary, our study of PAD patients and age- and diabetes-matched controls indicates a significant asso-

Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.

References 1. Kollerits B, Heinrich J, Pichler M, Rantner B, KleinWeigel P, Wolke G, et al. Intermittent claudication in the Erfurt Male Cohort (ERFORT) Study: its determinants and the impact on mortality. A population-based prospective cohort study with 30 years of follow-up. Atherosclerosis 2008;198: 214 –22. 2. Lamina C, Meisinger C, Heid IM, Lowel H, Rantner B, Koenig W, Kronenberg F. Association of ankle-brachial index and plaques in the carotid and femoral arteries with cardiovascular events and total mortality in a population-based study with 13 years of follow-up. Eur Heart J 2006;27: 2580 –7. 3. Selvin E, Erlinger TP. Prevalence of and risk factors for peripheral arterial disease in the United States: results from the National Health and Nutrition Examination Survey, 1999 –2000. Circulation 2004;110:738 – 43. 4. Mukherjee D, Cho L. Peripheral arterial disease: considerations in risks, diagnosis, and treatment. J Natl Med Assoc 2009;101:999 –1008. 5. Kitamura K, Eto T. Adrenomedullin–physiological regulator of the cardiovascular system or biochemical curiosity? Curr Opin Nephrol Hypertens 1997;6:80 –7. 6. Levin ER, Gardner DG, Samson WK. Natriuretic

peptides. N Engl J Med 1998;339:321– 8. 7. Ishimitsu T, Ono H, Minami J, Matsuoka H. Pathophysiologic and therapeutic implications of adrenomedullin in cardiovascular disorders. Pharmacol Ther 2006;111:909 –27. 8. Kobayashi K, Kitamura K, Hirayama N, Date H, Kashiwagi T, Ikushima I, et al. Increased plasma adrenomedullin in acute myocardial infarction. Am Heart J 1996;131:676 – 80. 9. Jougasaki M, Wei CM, McKinley LJ, Burnett JC Jr. Elevation of circulating and ventricular adrenomedullin in human congestive heart failure. Circulation 1995;92:286 –9. 10. Kohno M, Hanehira T, Kano H, Horio T, Yokokawa K, Ikeda M, et al. Plasma adrenomedullin concentrations in essential hypertension. Hypertension 1996;27:102–7. 11. Dieplinger B, Mueller T, Kollerits B, Struck J, Ritz E, von Eckardstein A, et al. Pro-A-type natriuretic peptide and pro-adrenomedullin predict progression of chronic kidney disease: the MMKD Study. Kidney Int 2009;75:408 –14. 12. Gouya G, Sturm G, Lamina C, Zitt E, Freistatter O, Struck J, et al. The association of mid-regional pro-adrenomedullin and mid-regional pro-atrial natriuretic peptide with mortality in an incident dialysis cohort. PLoS One 2011;6:e17803.

13. Hayashi M, Shimosawa T, Isaka M, Yamada S, Fujita R, Fujita T. Plasma adrenomedullin in diabetes. Lancet 1997;350:1449 –50. 14. Suzuki Y, Horio T, Nonogi H, Hayashi T, Kitamura K, Eto T, et al. Adrenomedullin as a sensitive marker for coronary and peripheral arterial complications in patients with atherosclerotic risks. Peptides 2004;25:1321– 6. 15. Suzuki Y, Horio T, Hayashi T, Nonogi H, Kitamura K, Eto T, et al. Plasma adrenomedullin concentration is increased in patients with peripheral arterial occlusive disease associated with vascular inflammation. Regul Pept 2004;118:99 –104. 16. Arad M, Elazar E, Shotan A, Klein R, Rabinowitz B. Brain and atrial natriuretic peptides in patients with ischemic heart disease with and without heart failure. Cardiology 1996;87:12–7. 17. van den Berg MP, Tjeerdsma G, Jan de Kam P, Boomsma F, Crijns HJ, van Veldhuisen DJ. Longstanding atrial fibrillation causes depletion of atrial natriuretic peptide in patients with advanced congestive heart failure. Eur J Heart Fail 2002;4:255– 62. 18. Horio T, Shimada K, Kohno M, Yoshimura T, Kawarabayashi T, Yasunari K, et al. Serial changes in atrial and brain natriuretic peptides in patients with acute myocardial infarction treated

Clinical Chemistry 59:4 (2013) 701

19.

20.

21.

22.

23.

24.

with early coronary angioplasty. Am Heart J 1993;126:293–9. Yanagisawa M, Kurihara H, Kimura S, Tomobe Y, Kobayashi M, Mitsui Y, et al. A novel potent vasoconstrictor peptide produced by vascular endothelial cells. Nature 1988;332:411–5. Nunez DJ, Brown MJ, Davenport AP, Neylon CB, Schofield JP, Wyse RK. Endothelin-1 mRNA is widely expressed in porcine and human tissues. J Clin Invest 1990;85:1537– 41. Dieplinger B, Gegenhuber A, Struck J, Poelz W, Langsteger W, Haltmayer M, Mueller T. Chromogranin A and C-terminal endothelin-1 precursor fragment add independent prognostic information to amino-terminal proBNP in patients with acute destabilized heart failure. Clin Chim Acta 2009;400:91– 6. Jankowska EA, Filippatos GS, von Haehling S, Papassotiriou J, Morgenthaler NG, Cicoira M, et al. Identification of chronic heart failure patients with a high 12-month mortality risk using biomarkers including plasma C-terminal proendothelin-1. PLoS One 2011;6:e14506. Khan SQ, Dhillon O, Struck J, Quinn P, Morgenthaler NG, Squire IB, et al. C-terminal proendothelin-1 offers additional prognostic information in patients after acute myocardial infarction: Leicester Acute Myocardial Infarction Peptide (LAMP) Study. Am Heart J 2007;154: 736 – 42. Rantner B, Kollerits B, Anderwald-Stadler M, Klein-Weigel P, Gruber I, Gehringer A, et al. Association between the UGT1A1 TA-repeat polymorphism and bilirubin concentration in patients

702 Clinical Chemistry 59:4 (2013)

25.

26.

27.

28.

29.

30.

31.

32.

with intermittent claudication: results from the CAVASIC study. Clin Chem 2008;54:851–7. Rose GA. The diagnosis of ischaemic heart pain and intermittent claudication in field surveys. Bull World Health Organ 1962;27:645–58. Hammerer-Lercher A, Puschendorf B, Mair J. B-type natriuretic peptides as powerful markers in cardiac diseases-analytical and clinical aspects. J Lab Med 2006;30:165– 84. Apple FS, Ler R, Murakami MM. Determination of 19 cardiac troponin I and T assay 99th percentile values from a common presumably healthy population. Clin Chem 2012;58:1574 – 81. Melander O, Newton-Cheh C, Almgren P, Hedblad B, Berglund G, Engstrom G, et al. Novel and conventional biomarkers for prediction of incident cardiovascular events in the community. JAMA 2009;302:49 –57. Drion I, Kleefstra N, Landman GW, Alkhalaf A, Struck J, Groenier KH, et al. Plasma COOHterminal proendothelin-1: a marker of fatal cardiovascular events, all-cause mortality, and newonset albuminuria in type 2 diabetes? (ZODIAC29). Diabetes Care 2012;35:2354 – 8. Kato J, Tsuruda T, Kita T, Kitamura K, Eto T. Adrenomedullin: a protective factor for blood vessels. Arterioscler Thromb Vasc Biol 2005;25: 2480 –7. Silver MA. The natriuretic peptide system: kidney and cardiovascular effects. Curr Opin Nephrol Hypertens 2006;15:14 –21. Abraham D, Dashwood M. Endothelin–role in vascular disease. Rheumatology 2008;47(Suppl 5): v23– 4.

33. Iglarz M, Clozel M. Mechanisms of ET-1-induced endothelial dysfunction. J Cardiovasc Pharmacol 2007;50:621– 8. 34. Hagner S, Haberberger R, Kummer W, Springer J, Fischer A, Bohm S, et al. Immunohistochemical detection of calcitonin gene-related peptide receptor (CGRPR)-1 in the endothelium of human coronary artery and bronchial blood vessels. Neuropeptides 2001;35:58 – 64. 35. Criqui MH, Langer RD, Fronek A, Feigelson HS, Klauber MR, McCann TJ, Browner D. Mortality over a period of 10 years in patients with peripheral arterial disease. N Engl J Med 1992;326: 381– 6. 36. Diehm C, Schuster A, Allenberg JR, Darius H, Haberl R, Lange S, et al. High prevalence of peripheral arterial disease and co-morbidity in 6880 primary care patients: cross-sectional study. Atherosclerosis 2004;172:95–105. 37. Lewis LK, Smith MW, Yandle TG, Richards AM, Nicholls MG. Adrenomedullin(1–52) measured in human plasma by radioimmunoassay: plasma concentration, adsorption, and storage. Clin Chem 1998;44:571–7. 38. Ala-Kopsala M, Magga J, Peuhkurinen K, Leipala J, Ruskoaho H, Leppaluoto J, Vuolteenaho O. Molecular heterogeneity has a major impact on the measurement of circulating N-terminal fragments of A- and B-type natriuretic peptides. Clin Chem 2004;50:1576 – 88. 39. Struck J, Morgenthaler NG, Bergmann A. Proteolytic processing pattern of the endothelin-1 precursor in vivo. Peptides 2005;26:2482– 6.