The neurohumoral and inflammatory pathways are regarded as the main mechanisms for the progression of heart failure. We sought to investigate the prognostic value of high-sensitivity C-reactive protein (hs-CRP) and N-terminal probrain natriuretic peptide (NT-proBNP) by evaluating their relation with 12-month mortality rate in this prospective cohort study from 24 academic hospitals in Korea. In 1,608 patients with acute heart failure (AHF), the median hs-CRP and NT-proBNP values were 0.77 mg/dl (interquartile range 0.29 to 2.84) and 4,638 pg/ml (interquartile range 1,945 to 10,852), respectively. During the 12-month follow-up, 213 patients (13.3%) died. The mortality rate increased from the lowest to the highest hs-CRP quartiles (Q1 7.4%, Q2 9.5%, Q3 16.9%, Q4 19.3%, p <0.001) and NT-proBNP quartiles (Q1 7.0%, Q2 13.4%, Q3 11.6%, Q4 20.4%, p <0.001). After adjustment, both hs-CRP (hazard ratio [HR] 1.811, 95% confidence interval [CI] 1.138 to 2.882) and NT-proBNP (HR 1.971, 95% CI 1.219 to 3.187) were independent predictors of 12-month mortality among others. When combining both hs-CRP and NT-proBNP and stratifying the patients according to their median values, patients with elevation of both hs-CRP and NT-proBNP values had 2.4-fold increased hazards (HR 2.382, 95% CI 1.509 to 3.761) compared with those without elevation of both markers. In Korean patients with AHF, patients with increased levels of both hs-CRP and NT-proBNP had worse clinical outcomes. The combination of the neurohumoral and inflammatory markers may provide a better strategy for risk stratification of Asian patients with AHF.
Heart failure (HF) is a progressive disease. The two main mechanisms discussed for the development and progression of HF include the neurohumoral and inflammatory pathways. Prognostic markers in HF include plasma norepinephrine, plasma renin activity, vasopressin, atrial natriuretic peptide, brain natriuretic peptide levels, hyponatremia, serum creatinine (Cr), and others. Among inflammatory markers, C-reactive protein (CRP) is the most investigated marker, and it has been identified as a strong risk factor for development and progression of HF. Similarly, N-terminal probrain natriuretic peptide (NT-proBNP) is regarded as a discerning neurohumoral marker for detection and evaluation of HF. Patients with acute heart failure (AHF) represent a group with high disease severity with a special need for meticulous medical treatment. Thus, precise risk stratification may be helpful for the development of optimal therapeutic strategies in these patients. Combination of neurohumoral and inflammatory markers may be useful for more precise risk stratification. Furthermore, because most studies in HF have been performed in the West, data on prognostic markers in Asian patients with AHF are limited. In this nationwide HF registry study, we sought to investigate the prognostic value of high-sensitivity CRP (hs-CRP), NT-proBNP, and the combination of both markers in Korean patients with AHF.
Methods
The Korean Heart Failure (KorHF) registry is a Korean, prospective, multicenter registry designed to reflect “real-world” clinical data of Korean patients admitted for AHF. The study design and primary results of the KorHF registry have been published. The registry was founded in June 2004 and is supported by the Korean Society of Heart Failure. Twenty-four well-qualified centers participated in the registry. All consecutive patients hospitalized with an episode of AHF as the primary reason for admission were eligible for enrollment. HF was diagnosed at admission according to the Framingham criteria. Confirmed diagnosis of HF was required at discharge, too. All patients enrolled were strongly recommended for follow-up for at least 1 year, and follow-up outcome data including mortality and rehospitalization due to HF were prospectively collected by medical records and telephone contact. The institutional review board or ethics committee at each participating hospital approved the study protocol, and patients gave written informed consent before study entry.
Patient demographic and baseline characteristics, underlying diseases, clinical presentation, treatments, laboratory tests, and outcomes during admission and after discharge were recorded. Among laboratory tests, hemoglobin, blood urea nitrogen, Cr, hs-CRP, and NT-proBNP levels measured at admission and sodium, potassium, hemoglobin, blood urea nitrogen, and NT-proBNP levels measured at discharge were collected for the registry. Blood sampling and tests were conducted as routine practice by laboratories at each center certified by the Korean Association of Quality Assurance for Clinical Laboratories. Data were collected at each site by a trained study coordinator using a standardized case report form and entered into the KorHF registry database by way of a web-based electronic data capture system that included an electronic case report form. Data collection and audition were performed by the KorHF Registry Steering Committee at the Korean Society of Heart Failure.
The primary end point was the 12-month mortality from the index hospitalization for AHF according to hs-CRP and NT-proBNP levels. The patients were stratified according to the median of the hs-CRP and NT-proBNP values.
Data are presented as numbers and frequencies for categorical variables and mean ± SD for continuous variables. For comparison among groups, the chi-square test (or Fisher’s exact test when any expected cell count was <5 for a 2 × 2 table) was used for categorical variables, and unpaired Student t test or 1-way analysis of variance were applied for continuous variables. Receiver operating characteristic curves were calculated to compare the diagnostic performance of hs-CRP and NT-proBNP.
The chronological trend of outcomes is expressed as Kaplan-Meier estimates and was compared according to the hs-CRP and NT-proBNP levels. The log-rank test was used to analyze the significant indifferences in clinical outcomes. A multivariate Cox proportional hazards regression model was used to find the independent predictors of 12-month mortality. Variables associated with mortality with p <0.05 were included as confounding variables in multivariate analysis, and they were age in decade, previous myocardial infarction, previous congestive HF, peripheral artery disease, chronic lung disease, systolic blood pressure, hemoglobin, hyponatremia defined as sodium <135 mmol/L, serum Cr level, blood urea nitrogen, hs-CRP, and NT-proBNP per 100-pg/L increase. Statistical tests were performed using SPSS, version 17 (SPSS Inc., Chicago, Illinois), and MedCalc for Windows (version 12.4.0.0; MedCalc Software, Ostend, Belgium).
Results
From June 2004 to April 2009, the KorHF registry database included 3,200 patients hospitalized for AHF at 24 centers in Korea. Among them, hs-CRP measurement was performed in 2,214 patients (69.2%) and NT-proBNP in 2,169 patients (67.8%). Overall, 1,608 patients (50.2%) had both hs-CRP and NT-proBNP values. The baseline characteristics between those with and without hs-CRP and NT-proBNP results were mostly comparable ( Supplemental Table 1 ). For the present analysis, the data of the 1,608 patients were used. The mean age was 68 years, 50% were men, 46% had hypertension, 30% had diabetes, 9% had chronic kidney disease, 56% had New York Heart Association functional class III, and 23% had New York Heart Association class IV.
The median hs-CRP and NT-proBNP values were 0.77 mg/dl (interquartile range 0.29 to 2.84 mg/dl) and 4,638 pg/ml (interquartile range 1,945 to 10,852), respectively. When stratifying the patients according to the median hs-CRP and NT-proBNP values, patients with hs-CRP >0.77 mg/dl and NT-proBNP >4,638 pg/ml had more adverse baseline characteristics ( Table 1 ).
| Characteristic, n = 1,608 | CRP <0.77 mg/dl (%) | CRP >0.77 mg/dl (%) | p Value | NT-proBNP <4,638 pg/ml (%) | NT-proBNP >4,638 pg/ml (%) | p Value |
|---|---|---|---|---|---|---|
| Age (yrs) | 68 ± 15 | 69 ± 14 | 0.025 | 66 ± 15 | 71 ± 14 | <0.001 |
| Age >65 yrs | 63.5 | 70.9 | 0.002 | 63.5 | 70.9 | 0.001 |
| Men | 47 | 53 | 0.011 | 47 | 53.3 | 0.013 |
| Diabetes mellitus | 26.6 | 33.0 | 0.005 | 25.8 | 33.8 | <0.001 |
| Hypertension | 44.5 | 47.7 | 0.197 | 41.6 | 50.6 | <0.001 |
| Chronic kidney disease | 6.4 | 10.9 | 0.001 | 3.6 | 13.7 | <0.001 |
| Myocardial infarction | 15.4 | 17.0 | 0.383 | 14.4 | 17.9 | 0.059 |
| Congestive HF | 32.4 | 30.4 | 0.399 | 28.7 | 34.0 | 0.028 |
| Peripheral artery disease | 1.2 | 1.6 | 0.543 | 1.6 | 1.2 | 0.527 |
| Valvular heart disease | 12.8 | 12.2 | 0.740 | 11.1 | 13.9 | 0.105 |
| Stroke | 6.6 | 12.0 | <0.001 | 7.6 | 11.1 | 0.016 |
| Chronic lung disease | 3.1 | 5.6 | 0.019 | 4.4 | 4.3 | 0.890 |
| Causes for HF | ||||||
| Ischemia | 36.6 | 38.8 | 0.371 | 34.7 | 40.6 | 0.015 |
| Valvular heart disease | 12.4 | 11.9 | 0.769 | 11.2 | 13.1 | 0.240 |
| Myocarditis | 0.8 | 0.8 | 0.981 | 0.7 | 0.9 | 0.559 |
| NYHA functional class | 0.036 | <0.001 | ||||
| I | 6.9 | 6.7 | 7.3 | 6.3 | ||
| II | 16.5 | 12.8 | 18.0 | 11.4 | ||
| III | 56.5 | 54.5 | 56.3 | 54.7 | ||
| IV | 20.1 | 26.0 | 18.3 | 57.6 | ||
| NYHA III–IV | 76.6 | 80.4 | 0.085 | 74.6 | 82.3 | 0.001 |
| Systolic blood pressure (mm Hg) | 132 ± 31 | 130 ± 29 | 0.083 | 129 ± 29 | 133 ± 31 | 0.006 |
| Diastolic blood pressure (mm Hg) | 79 ± 19 | 78 ± 18 | 0.146 | 77 ± 17 | 80 ± 19 | 0.002 |
| Heart rate (beats/min) | 88 ± 25 | 94 ± 26 | <0.001 | 87 ± 25 | 95 ± 26 | <0.001 |
| Body temperature (°C) | 36.3 ± 0.4 | 36.5 ± 0.6 | <0.001 | 36.4 ± 0.5 | 36.4 ± 0.5 | 0.438 |
| Leukocyte (/μl) | 8,428 ± 4,177 | 9,974 ± 5,448 | <0.001 | 9,099 ± 5,186 | 9,314 ± 4,637 | 0.397 |
| Hemoglobin (mg/dl) | 12.7 ± 2.3 | 12.1 ± 2.4 | <0.001 | 13.0 ± 2.3 | 11.8 ± 2.3 | <0.001 |
| Thrombocytes (1000/μl) | 223 ± 78 | 241 ± 98 | <0.001 | 231 ± 83 | 233 ± 94 | 0.602 |
| Serum sodium (mmol/L) | 139 ± 4.9 | 137 ± 5.5 | <0.001 | 139 ± 5.0 | 137 ± 5.5 | <0.001 |
| Blood urea nitrogen (mg/dl) | 22.8 ± 13.8 | 27.1 ± 17.1 | <0.001 | 19.8 ± 10.4 | 30.1 ± 18.3 | <0.001 |
| Cr (mg/dl) | 1.31 ± 0.94 | 1.56 ± 1.20 | <0.001 | 1.14 ± 0.52 | 1.73 ± 1.39 | <0.001 |
| CRP (mg/dl) | 0.30 ± 0.20 | 5.23 ± 5.96 | <0.001 | 2.02 ± 4.01 | 3.53 ± 5.53 | <0.001 |
| Troponin I (ng/ml) | 2.49 ± 19.11 | 4.25 ± 16.32 | 0.091 | 3.30 ± 22.5 | 3.47 ± 11.7 | 0.863 |
| NT-proBNP (pg/ml) | 6,344 ± 8,045 | 10,764 ± 10,836 | <0.001 | 2,023 ± 1,333 | 15,107 ± 10,233 | <0.001 |
Supplemental Table 2 lists the baseline characteristics according to the 12-month mortality. Two hundred thirteen patients (13.3%) died during the 12-month follow-up. Patients who died had more adverse baseline characteristics along with higher hs-CRP level (2.57 vs 4.01 mg/dl, p <0.001) and higher NT-proBNP level (7,929 ± 9,237 vs 12,724 ± 12,127 pg/ml, p <0.001) compared with those who survived.
When stratifying the patients according to the quartiles of the hs-CRP level, there was a steady increase in mortality rate from the lowest to the highest quartiles (Q1 7.4%, Q2 9.5%, Q3 16.9%, Q4 19.3%, p <0.001, chi-square test for linear-by-linear association p <0.001; Figure 1 ). Similar results were observed according to the NT-proBNP quartiles (Q1 7.0%, Q2 13.4%, Q3 11.6%, Q4 20.4%, p <0.001, chi-square test for linear-by-linear association p <0.001; Figure 1 ).
Because the white blood cell (WBC) count is another well-established inflammatory marker, we also evaluated the prognostic value of the WBC count. There was a weak correlation between the WBC count and the hs-CRP level (r = 0.198, p <0.001; Figure 2 ). However, the WBC count was not associated with increased mortality (p = 0.705). In receiver operating characteristic curve analysis, the area under the curve of hs-CRP, NT-proBNP, and WBC count to predict 12-month mortality was 0.63 (95% confidence interval [CI] 0.59 to 0.67), 0.62 (95% CI 0.58 to 0.66), and 0.51 (95% CI 0.47 to 0.56), respectively ( Figure 2 ).
In Cox proportional hazards regression analysis, both hs-CRP and NT-proBNP levels were independent predictors of 12-month mortality along with advanced age, history of peripheral artery disease, low blood pressure, increased serum Cr level, and hyponatremia ( Table 2 ). When analyzing according to the quartiles of hs-CRP and NT-proBNP levels, patients in the highest quartile had 1.8-fold and twofold increased (adjusted) hazard for 12-month mortality, respectively. There existed a weak correlation (r = 0.173, p <0.001), but a significant interaction (unadjusted p for interaction = 0.013; adjusted p for interaction = 0.002), between hs-CRP and NT-proBNP. When combining both hs-CRP and NT-proBNP and stratifying the patients according to their median values, patients with elevation of both hs-CRP and NT-proBNP values had 2.4-fold increased hazards (adjusted hazard ratio [HR] 2.382, 95% CI 1.509 to 3.761) compared with those without elevation of both markers ( Table 3 and Figure 3 ).
| Variable | p Value | HR | 95% CI |
|---|---|---|---|
| Age in decade | 0.001 | 1.235 | 1.092–1.397 |
| Peripheral artery disease | 0.046 | 2.295 | 1.013–5.199 |
| Systolic blood pressure (for each 1-mm Hg increase) | <0.001 | 0.988 | 0.983–0.993 |
| Serum Cr level (for each 1-mg/dl increase) | 0.029 | 1.132 | 1.013–1.265 |
| Hyponatremia (sodium <135 mmol/L) | 0.004 | 1.603 | 1.167–2.202 |
| CRP (for each 1 mg/dl increase) | 0.028 | 1.003 | 1.000–1.005 |
| NT-proBNP level (for each 100-pg/L increase) | <0.001 | 1.003 | 1.001–1.004 |
| Quartile | Unadjusted | Adjusted | ||||
|---|---|---|---|---|---|---|
| p Value | HR | 95% CI | p Value | HR | 95% CI | |
| Hs-CRP | ||||||
| Q1 | 1 | Reference | 1 | Reference | ||
| Q2 | 0.307 | 1.285 | 0.794–2.081 | 0.421 | 1.231 | 0.742–2.041 |
| Q3 | <0.001 | 2.397 | 1.561–3.679 | 0.008 | 1.864 | 1.173–2.963 |
| Q4 | <0.001 | 2.884 | 1.892–4.398 | 0.012 | 1.811 | 1.138–2.882 |
| NT-proBNP | ||||||
| Q1 | 1 | Reference | 1 | Reference | ||
| Q2 | 0.004 | 1.955 | 1.239–3.087 | 0.051 | 1.627 | 0.997–2.654 |
| Q3 | 0.022 | 1.746 | 1.085–2.811 | 0.148 | 1.454 | 0.875–2.415 |
| Q4 | <0.001 | 3.431 | 2.242–5.250 | 0.006 | 1.971 | 1.219–3.187 |
| CRP median, NT-proBNP median | ||||||
| CRP(−), BNP (−) | 1 | Reference | 1 | Reference | ||
| CRP(−), BNP(+) | 0.005 | 2.004 | 1.238–3.243 | 0.046 | 1.683 | 1.010–2.805 |
| CRP(+), BNP(−) | <0.001 | 2.670 | 1.704–4.185 | 0.002 | 2.163 | 1.334–3.507 |
| CRP(+), BNP(+) | <0.001 | 3.579 | 2.373–5.400 | <0.001 | 2.382 | 1.509–3.761 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
