Invited review
New cardiovascular risk markers in the general population and in hypertension. Do they improve risk prediction and influence treatment?

Introduction

Due to increasing life expectancy and an obesity epidemic in Western countries the demand for prevention and treatment of cardiovascular diseases is growing. In order to prioritize limited health resources cardiovascular risk stratification is essential. Several years ago the Framingham Risk Score was developed based on a large American population survey in Framingham [1] and less than a decade ago the HeartScore was developed based on several European population surveys [2]. These risk scores were only based on traditional cardiovascular risk factors because newer risk markers were not measured in these large population surveys. New risk markers, more closely related to cardiovascular disease, have been developed and successfully tested in well defined groups of patients [3]. However, this multiple risk marker approach has only been done systematically in a few general populations with rather disappointing results [4]. Furthermore, the prognostic importance of reducing these new risk markers was until recently unknown.
We wanted to test the additive prognostic value of three relatively new but established cardio-vascular risk markers: N-terminal pro-brain natriu-retic peptide (Nt-proBNP) [5], high sensiti-vity C-reactive protein (hsCRP) [6] and urine albumin/creatinine ratio (UACR) [7]. These three risk markers were chosen because we expected them to have additive prognostic importance as Nt-proBNP was primarily related to haemodynamic cardiovascular risk factors [8, 9], hsCRP was primarily related to metabolic cardiovascular risk factors [10, 11], and UACR was related to haemodynamic as well as metabolic risk factors [12] (Figure 1). Furthermore, in patients with hyper-tension and left ventricular (LV) hypertrophy enrolled in the LIFE study we measured UACR yearly during either losartan- or atenolol-based antihyper-tensive treatment to test whether changes in UACR had prognostic importance independently of changes in blood pressure and LV hypertrophy.
Material and methods
In the clinical setting it is recommended to evaluate hsCRP [13], Nt-proBNP [14, 15] and UACR [16, 17] on at least two occasions separated by a few weeks to reduce the intra-individual variations and to exclude ongoing infection to avoid measuring false elevations of hsCRP and UACR. However, in our population study we had only one measurement and could not exclude subclinical infection.
Results and discussion
In the general population, UACR, Nt-proBNP and hsCRP above or below the gender-adjusted median values had additive predictive value (Figure 2). As logarithmic transformed continuous variables, UACR, Nt-proBNP and hsCRP predicted the composite cardiovascular endpoint (CEP) of cardiovascular death, non-fatal myocardial infarction and non-fatal stroke independently of traditional cardiovascular risk factors including left ventricular mass and pulse wave velocity [18], supporting a recent study [19]. All three risk markers were associated with increased cardiovascular risk even at low values and the risk increased continuously [18] (Figure 3). Due to lack of consensus regarding cut-off values for these apparently continuous risk markers we decided to test different age- and gender-specific cut-off values [20]. Pre-specified 90% specificity, gender-adjusted (men/women) cut-off values of 110/164 pg/ml for Nt-proBNP, 6.0/7.3 mg/l for hsCRP, and 0.73/1.06 mg/mmol for UACR lead to the highest positive predictive values without compromising the necessarily high negative predictive values [20].
High sensitivity C-reactive protein predicted CEP primarily in 41- or 51-year old men, Nt-proBNP in 61- or 71-year old subjects and UACR independently of age and gender [20]. This is probably because hsCRP and UACR are elevated early in the development of atherosclerosis [21, 22] secondarily to metabolic risk factors and endothelial dysfunction [10], and because Nt-proBNP and UACR are elevated later in the process in connection with subclinical cardiovascular damage [8]. Consistent with this we also found: 1) that hsCRP primarily predicted CEP in low-risk subjects without any elements of the metabolic syndrome [23, 24], 2) that Nt-proBNP primarily predicted CEP in high-risk subjects with the metabolic syndrome, diabetes or known cardiovascular disease [23, 24], and 3) that UACR predicted CEP independently of cardiovascular risk assessed by elements of the metabolic syndrome or by HeartScore [23, 24]. In patients with hypertension and LV hypertrophy, Nt-proBNP, but not hsCRP, predicted CEP independently of traditional cardiovascular risk factors and UACR, which supports the idea that hsCRP primarily predicts outcome in low-risk subjects.
In healthy subjects with a 10-year risk of cardiovascular death lower than 5% based on HeartScore [2] and therefore not eligible for primary prevention [25], the actual 10-year risk of cardiovascular death exceeded 5% in a small subgroup of subjects with hsCRP higher than 5.6 mg/l [24], which was close to the pre-specified 90% specificity, gender-adjusted cut-off value of 6.0/7.3 mg/l [20] (Figure 4). As hsCRP ł6.0/7.3 mg/l was found only in 124 subjects predicting only 6 CEPs and as 82% of the subjects in the low-moderate risk group were 41 or 51 years old [24], one could argue for a lower cut-off value accepting intervention at a lower absolute 10-year cardiovascular risk if the relative risk was high. That would be especially relevant in subjects with moderate cardiovascular risk as recommended by the Centers for Disease Control and Prevention and the American Heart Association in 2003 [26]. However, our data also suggested that hsCRP did not add new prognostic information in subjects with low-moderate cardiovascular risk, if younger subjects were regarded as being 60 years of age when calculating cardiovascular risk [24] in order to avoid withholding intervention that would be recommended only if the subjects were older [25]. However, this method almost doubled the number of subjects eligible for primary prevention due to high cardiovascular risk based on HeartScore, which is not rational. The impact of measuring hsCRP is still controversial. Ridker et al. [27] and others [28] have previously found hsCRP to predict cardio-vascular events independently of Framingham risk score and recently claimed that a new risk score using hsCRP as a continuous variable together with traditional cardiovascular risk factors in subjects with moderate cardiovascular risk can reclassify 40-50% of the subjects to either higher or lower CV risk [29], whereas Danesh et al. [30] have questioned the additive predictive value of hsCRP.
In the same low-moderate risk group, the actual 10-year risk of cardiovascular death exceeded 5% for UACR >1.6 mg/mmol (Figure 5), giving indication for primary prevention [25] in a small subgroup of 61 subjects (4.3%) [24]. However, as most of the subjects were 41 or 51 years old with an overrepresentation of women [24], intervention might be relevant at a lower absolute 10-year risk of cardiovascular death. UACR above the pre-specified gender-adjusted cut-off value of 0.73/1.06 mg/mmol (90% specificity), which was found in 120 subjects with low-moderate cardiovascular risk, identified as many as 10 CEPs with a very high negative predictive value of 98% [24]. High UACR still predicted CEP in subjects with low-moderate cardiovascular risk if younger subjects were regarded as being 60 years when calculating cardiovascular risk [24]. This suggested that primary prevention in subjects with low-moderate cardio-vascular risk may be relevant already at levels of UACR around 1 mg/mmol, which represents a practical round cut-off value close to the value at which cardiovascular risk clearly begins to increase in patients with hypertension [31]. However, others have suggested a somewhat higher cut-off value [32]. Combined use of UACR ł0.73/1.06 mg/mmol or hsCRP ł6.0/7.3 mg/l identified a larger subgroup of 228 subjects (16%) with high cardiovascular risk in which primary prevention may be advised [24] despite low-moderate cardiovascular risk based on HeartScore [2] (Figure 6). Measuring UACR and hsCRP in subjects with low-moderate CV risk seems to be a clinically relevant supplement to HeartScore as 34% are reclassified correctly versus 15% wrongly. However, in daily clinical practice, we do not suggest that these two new risk markers be measured routinely in subjects with low-moderate CV risk, but measured on an individual basis either in subjects with moderate CV risk or in subjects especially afraid of developing CV disease.
In subjects with known cardiovascular disease or diabetes, Nt-proBNP and UACR above the pre-specified 90% specificity, gender-adjusted cut-off values of 110/164 pg/ml or 0.73/1.06 mg/mmol predicted CEP with very high positive predictive values of approximately 37% and relatively high negative predictive values of 90%. Furthermore, combined use of UACR ł0.73/1.06 mg/mmol or high Nt-proBNP ł110/164 pg/ml in subjects with known cardiovascular disease or diabetes identified a larger subgroup of 228 subjects (48%) with extremely high cardiovascular risk who should be referred for specialist care in order to optimize treatment [24]. Measuring UACR and Nt-proBNP seems to be relevant in patients with known CV disease or diabetes as 49% are reclassified correctly vs. 15% wrongly. For pragmatic reasons we recommend using the threshold accepted in heart failure of 125 pg/ml [33] as the cut-off value in cardiovascular risk stratification instead of our gender-adjusted cut-off value of 110/164 pg/ml.
In patients with hypertension and electrocardiographic LV hypertrophy, blood pressure reduction was associated with a significant 30-40% reduction in UACR. However, the reduction was more marked in patients randomized to an angiotensin-II receptor blocker based antihypertensive regime compared to a β-adrenergic receptor blocker based regime [34], suggesting either a more effective reduction of the central blood pressure [35] or a blood pressure independent effect of the renin-angiotensin-aldosterone system on UACR. Furthermore, one year UACR had independent prognostic importance independently of changes in blood pressure [31] and LV hypertrophy assessed by electrocardiography [12] (Figure 7). This supports the concept that albuminuria [36-38] and LV hypertrophy [39-41] are markers of pre-clinical disease in different organs with a possible direct influence on CV risk [42]. Albuminuria reflecting generalized transvascular leakiness [43] may promote lipid insudation, atherosclerosis and thrombosis in coronary as well as cerebral arteries and thereby contribute to CV events. LV hypertrophy may through myocardial ischaemia [44] compromise LV function and increase the risk of arrhythmia [45]. Therefore, it seems likely that albuminuria and LV hypertrophy are not just markers of CV risk, but also important CV risk factors which need to be addressed directly in the future treatment of patients with hypertension to improve prognosis.
In conclusion, we as well as others have found that UACR, Nt-proBNP and hsCRP have additive predictive value and predict major cardiovascular events independently of traditional cardiovascular risk factors, whereas the clinical impact is less clear. However, in healthy subjects with a 10-year risk of cardiovascular death lower than 5% based on HeartScore, combined use of UACR ł0.73/1.06 mg/mmol or hsCRP ł6.0/7.3 mg/l seemed to identify a subgroup of 16% with high cardiovascular risk in which primary prevention with more exercise, healthy diet, smoking cessation, and blood pressure and cholesterol monitoring/lowering according to general guidelines may be advised. Whereas in subjects with known cardiovascular disease or diabetes, combined use of UACR ł0.73/1.06 mg/mmol or high Nt-proBNP ł110/164 pg/ml seemed to identify a larger subgroup of 48% with extremely high cardiovascular risk who should be referred for specialist care in order to optimize treatment. However, further studies are needed to confirm this. Antihypertensive treatment reduces UACR and the reduction carries additive prognostic information independently of changes in blood pressure and LV hypertrophy.

References
1. Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990; 322: 1561-6.
2. Conroy RM, Pyörälä K, Fitzgerald AP, et al.; SCORE project group. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J 2003; 24: 987-1003.
3. Sabatine MS, Morrow DA, de Lemos JA, et al. Multimarker approach to risk stratification in non-ST elevation acute coronary syndromes: simultaneous assessment of troponin I, C-reactive protein, and B-type natriuretic peptide. Circulation 2002; 105: 1760-3.
4. 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.
5. Wang TJ, Larson MG, Levy D, et al. Plasma natriuretic peptide levels and the risk of cardiovascular events and death. N Engl J Med 2004; 350: 655-63.
6. Ridker PM, Buring JE, Cook NR, Rifai N. C-reactive protein, the metabolic syndrome, and risk of incident cardiovascular events: an 8-year follow-up of 14 719 initially healthy American women. Circulation 2003; 107: 391-7.
7. Borch-Johnsen K, Feldt-Rasmussen B, Strandgaard S, Schroll M, Jensen JS. Urinary albumin excretion. An independent predictor of ischemic heart disease. Arterioscler Thromb Vasc Biol 1999; 19: 1992-7.
8. Olsen MH, Hansen TW, Christensen MK, et al. N-terminal pro brain natriuretic peptide is inversely related to metabolic cardiovascular risk factors and the metabolic syndrome. Hypertension 2005; 46: 660-6.
9. Olsen MH, Wachtell K, Tuxen C, et al. N-terminal pro-brain natriuretic peptide predicts cardiovascular events in patients with hypertension and left ventricular hypertrophy: a LIFE study. J Hypertens 2004; 22: 1597-604.
10. Olsen MH, Christensen MK, Hansen TW, et al. High-sensitivity C-reactive protein is only weakly related to cardiovascular damage after adjustment for traditional cardiovascular risk factors. J Hypertens 2006; 24: 655-61.
11. Olsen MH, Wachtell K, Nielsen OW, et al. N-terminal brain natriuretic peptide predicted cardiovascular events stronger than high-sensitivity C-reactive protein in hypertension: a LIFE substudy. J Hypertens 2006; 24: 1531-9.
12. Olsen MH, Wachtell K, Ibsen H, et al.; LIFE Study Investigators. Reductions in albuminuria and in electrocardiographic left ventricular hypertrophy independently improve prognosis in hypertension: the LIFE study. J Hypertens 2006; 24: 775-81.
13. Pearson TA, Bazzarre TL, Daniels SR, et al.; American Heart Association Expert Panel on Population and Prevention Science. American Heart Association guide for improving cardiovascular health at the community level: a statement for public health practitioners, healthcare providers, and health policy makers from the American Heart Association Expert Panel on Population and Prevention Science. Circulation 2003; 107: 645-51.
14. Clerico A, Carlo Zucchelli G, Pilo A, Passino C, Emdin M. Clinical relevance of biological variation: the lesson of brain natriuretic peptide (BNP) and NT-proBNP assay. Clin Chem Lab Med 2006; 44: 366-78.
15. Conen D, Pfisterer M, Martina B. Substantial intraindividual variability of BNP concentrations in patients with hypertension. J Hum Hypertens 2006; 20: 387-91.
16. Khawali C, Andriolo A, Ferreira SR. Comparison of methods for urinary albumin determination in patients with type 1 diabetes. Braz J Med Biol Res 2002; 35: 337-43.
17. Rowe DJ, Dawnay A, Watts GF. Microalbuminuria in diabetes mellitus: review and recommendations for the measurement of albumin in urine. Ann Clin Biochem 1990; 27: 297-312.
18. Olsen MH, Hansen TW, Christensen MK, et al. N-terminal pro-brain natriuretic peptide, but not high sensi-tivity C-reactive protein, improves cardiovascular risk prediction in the general population. Eur Heart J 2007; 28: 1374-81.
19. Kistorp C, Raymond I, Pedersen F, Gustafsson F, Faber J, Hildebrandt P. N-terminal pro-brain natriuretic peptide, C-reactive protein, and urinary albumin levels as predictors of mortality and cardiovascular events in older adults. JAMA 2005; 293: 1609-16.
20. Olsen MH, Hansen TW, Christensen MK, et al. Cardiovascular risk prediction by N-terminal pro brain natriuretic peptide and high sensitivity C-reactive protein is affected by age and sex. J Hypertens 2008; 26: 26-34.
21. Van Der Meer IM, de Maat MP, Hak AE, et al. C-reactive protein predicts progression of atherosclerosis measured at various sites in the arterial tree: the Rotterdam Study. Stroke 2002; 33: 2750-5.
22. Jensen JS, Borch-Johnsen K, Jensen G, Feldt-Rasmussen B. Microalbuminuria reflects a generalized transvascular albumin leakiness in clinically healthy subjects. Clin Sci (Lond) 1995; 88: 629-33.
23. Olsen MH, Hansen TW, Christensen MK, et al. Impact of the metabolic syndrome on the predictive values of new risk markers in the general population. J Hum Hypertens 2008; 22: 634-40.
24. lsen MH, Hansen TW, Christensen MK, et al. New risk markers may change the HeartScore risk-classification significantly in one fifth of the population. J Hum Hypertens 2008 [Epub ahead of print].
25. Graham IM. Guidelines on cardiovascular disease prevention in clinical practice: The European perspective. Curr Opin Cardiol 2005; 20: 430-9.
26. Pearson TA, Mensah GA, Alexander RW, et al.; Centers for Disease Control and Prevention; American Heart Association. Markers of inflammation and cardiovascular disease: application to clinical and public health practice: A statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation 2003; 107: 499-511.
27. Ridker PM, Cook N. Clinical usefulness of very high and very low levels of C-reactive protein across the full range of Framingham Risk Scores. Circulation 2004; 109: 1955-9.
28. Koenig W, Löwel H, Baumert J, Meisinger C. C-reactive protein modulates risk prediction based on the Framingham Score: implications for future risk assessment: results from a large cohort study in southern Germany. Circulation 2004; 109: 1349-53.
29. Ridker PM, Buring JE, Rifai N, Cook NR. Development and validation of improved algorithms for the assessment of global cardiovascular risk in women: the Reynolds Risk Score. JAMA 2007; 297: 611-9.
30. Danesh J, Wheeler JG, Hirschfield GM, et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med 2004; 350: 1387-97.
31. Ibsen H, Olsen MH, Wachtell K, et al. Reduction in albuminuria translates to reduction in cardiovascular events in hypertensive patients: losartan intervention for endpoint reduction in hypertension study. Hypertension 2005; 45: 198-202.
32. Redon J, Williams B. Microalbuminuria in essential hypertension: redefining the threshold. J Hypertens 2002; 20: 353-5.
33. Gustafsson F, Badskjaer J, Hansen FS, Poulsen AH, Hildebrandt PR. Value of N-terminal proBNP in the diagnosis of left ventricular systolic dysfunction in primary care patients referred for echocardiography. Heart Drug 2003; 3: 141-6.
34. Ibsen H, Wachtell K, Olsen MH, et al.; LIFE substudy. Does albuminuria predict cardiovascular outcome on treatment with losartan versus atenolol in hypertension with left ventricular hypertrophy? A LIFE substudy. J Hypertens 2004; 22: 1805-11.
35. Williams B, Lacy PS, Thom SM, et al.; CAFE Investigators; Anglo-Scandinavian Cardiac Outcomes Trial Investigators; CAFE Steering Committee and Writing Committee. Differential impact of blood pressure-lowering drugs on central aortic pressure and clinical outcomes: principal results of the Conduit Artery Function Evaluation (CAFE) study. Circulation 2006; 113: 1213-25.
36. Jensen JS, Borch-Johnsen K, Deckert T, Deckert M, Jensen G, Feldt-Rasmussen B. Reduced glomerular size- and charge-selectivity in clinically healthy individuals with microalbuminuria. Eur J Clin Invest 1995; 25: 608-14.
37. Jensen JS, Feldt-Rasmussen B, Borch-Johnsen K, Clausen P, Appleyard M, Jensen G. Microalbuminuria and its relation to cardiovascular disease and risk factors. A populationbased study of 1254 hypertensive individuals. J Hum Hypertens 1997; 11: 727-32.
38. Ravera M, Ratto E, Vettoretti S, et al. Microalbuminuria and subclinical cerebrovascular damage in essential hypertension. J Nephrol 2002; 15: 519-24.
39. Olsen MH, Wachtell K, Hermann KL, et al. Is cardiovascular remodeling in patients with essential hypertension related to more than high blood pressure? A LIFE substudy. Losartan Intervention For Endpoint-Reduction in Hypertension. Am Heart J 2002; 144: 530-7.
40. Roman MJ, Saba PS, Pini R, et al. Parallel cardiac and vascular adaptation in hypertension. Circulation 1992; 86: 1909-18.
41. Jones EC, Devereux RB, O’Grady MJ, et al. Relation of hemodynamic volume load to arterial and cardiac size. J Am Coll Cardiol 1997; 29: 1303-10.
42. Olsen MH, Wachtell K, Bella JN, et al. Albuminuria predicts cardiovascular events independently of left ventricular mass in hypertension: a LIFE substudy. J Hum Hypertens 2004; 18: 453-9.
43. Jensen JS. Renal and systemic transvascular albumin leakage in severe atherosclerosis. Arterioscler Thromb Vasc Biol 1995; 15: 1324-9.
44. Vogt M, Motz W, Scheler S, Strauer BE. Disorders of coronary microcirculation and arrhythmias in systemic arterial hypertension. Am J Cardiol 1990; 65: 45G-50G.
45. Koren MJ, Devereux RB, Casale PN, Savage DD, Laragh JH. Relation of left ventricular mass and geometry to morbidity and mortality in uncomplicated essential hypertension. Ann Intern Med 1991; 114: 345-52.