Prognostic value of plasma NT-proBNP levels in very old patients with moderate renal insufficiency in China

Background The N‑terminal pro-brain natriuretic peptide (NT-proBNP) has an important prognostic value in chronic renal insufficiency; however, most studies have been conducted in patients with end-stage renal disease (ESRD). In this study we evaluated the prognostic significance of NT-proBNP in very old patients with stage 3 chronic kidney disease (CKD) and compared its prognostic value in CKD3a versus CKD3b patients. Methods Patients (age ≥80 years old) hospitalized with stage 3 CKD from 2007 to 2010 who were eligible for this prospective study underwent follow-up examinations in 2015. The examinations included measurements of anthropometric characteristics, blood pressure, plasma NT-proBNP, creatinine, and lipids. End-point events were all-cause death and major adverse cardiac events (MACEs). Results A total of 168 patients (mean age 87.4 ± 2.9 years, range 80–99 years) were included in the analysis (CKD3a, n = 117; CKD3b, n = 51). The results showed that CKD3b was associated with lower hemoglobin levels, higher NT-proBNP levels and a higher rate of hypertension compared with CKD3a. After a median follow-up of 3.8 years (interquartile range 1.5–6.1 years), a higher NT-proBNP level was associated with a higher risk of all-cause death (hazard ratio HR 1.986, 95% confidence interval CI 1.276–2.819, p = 0.028) and MACEs (HR 2.872, 95% CI 1.241–6.644, p = 0.014) after adjusting for age, sex, and traditional risk factors; however, a subgroup comparison showed that elevated NT-proBNP levels were associated with a higher risk of all-cause death (HR 2.350, 95% CI 1.906–6.091, p = 0.039) and MACEs (HR 3.025, 95% CI 1.024–8.940, p = 0.045) in CKD3a but not CKD3b. Conclusion Levels of NT-proBNP increased with decreased renal function in very old patients with stage 3 CKD; therefore, NT-proBNP is an independent predictor for all-cause death and MACEs in these patients but has a greater prognostic value in CKD3a than in CKD3b.


Introduction
The number of old people is increasing along with the prevalence of chronic kidney disease (CKD) [1]. Zhang et al. [2] reported that the prevalence of CKD in the Chinese population was 10.8%, and Coresh et al. [3] found that the prevalence of CKD sharply rises with age, such that almost 50% of those older than 70 years are affected. Additionally, CKD is often associated with anemia, hypertension, and proteinuria, which increase the risk of cardiovascular disease, the primary cause of morbidity and premature mortality in CKD. Thus, old CKD patients have a higher risk of all-cause death or major adverse cardiac events (MACEs).
The N-terminal pro-brain natriuretic peptide (NT-proBNP) is synthesized and secreted by the ventricular myocardium in response to volume or pressure overload or myocardial ischemia [4]. Previous studies have found that NT-proBNP levels are elevated in patients with CKD, and an elevated NT-proBNP level is an independent predictor of all-cause death and MACEs in all CKD stages [5,6]; however, most studies evaluating the predictive value of NT-proBNP were conducted in patients with end-stage renal disease (ESRD) and on dialysis [5]. Patients with earlier stages of CKD also have a high rate of both fatal and nonfatal cardiovascular Qiwei Zhu and Suyan Bian contributed equally to this work. events [7]. Stage 3 CKD is classified as CKD3a if impairment of renal function is mild to moderate of CKD3b if impairment of renal function is moderate to severe. A decrease in renal function can be delayed and the risk of all-cause death and MACEs can be reduced if these patients receive comprehensive treatment; therefore, it is important to evaluate the risk of adverse events in patients with CKD stage 3. This study aimed to determine the prognostic value of NT-proBNP in very old (≥80 years old) patients with CKD stage 3 and compare the prognostic value of NT-proBNP in CKD3a versus CKD3b.

Study population
This prospective observational study included very old (age ≥80 years old) patients with stage 3 CKD who were hospitalized in the Department of Geriatric Internal Medicine in the People's Liberation Army (PLA) General Hospital, Beijing, China. Patients with severe systemic diseases, such as collagenosis, inflammation, cachexia, severe liver disease, acute heart failure, or acute coronary syndrome, and those who had undergone coronary artery bypass grafting or percutaneous transluminal coronary angioplasty in the previous 6 months, were excluded. A total of 181 participants were eligible forthe cross-sectional analysis from November 2007 to October 2010.
Follow-up visits were conducted in April 2015. The median follow-up interval for the original 181 patients was 3.8 years (interquartile range IQR 1.5-6.1). Of the patients 13 were lost to follow-up and were therefore excluded from the final analysis. Complete follow-up data were obtained from 168 patients (follow-up rate 92.8%).

Questionnaire and anthropometric measurements
Information about patient age, smoking status, and history of coronary heart disease (CHD), diabetes mellitus (DM), hypertension, atrial fibrillation (AF), and tumors was collected by the physician on admission to the hospital. The physical examination included measurement of height and weight. After the patient had been in a sitting position for ≥5 min, blood pressure was measured using a calibrated desktop sphygmomanometer (Yuyue, Armamentarium Limited Company, Jiangsu, People's Republic of China), consistent with current recommendations. Blood pressure was measured three times consecutively with ≥1 min between measurements; the mean value was used for the statistical analysis.
Echocardiography was performed within 3 days after admission by an experienced echocardiography technician, who interpreted all scans. Left ventricular ejection fraction (LVEF) was determined using the biplane Simpson's

Definition of variables
We calculated the estimated glomerular filtration rate (eGFR) using the Chinese version of the modification of diet in renal disease equation as follows [9]: eGFR (ml/min/1.73 m 2 ) = 175×standard creatinine (mg/dl) -  [11], and left ventricular mass index (LVMI) as LVMI = LVM/BSA. Hypertension was defined as (i) systolic blood pressure ≥140 mm Hg, (ii) diastolic blood pressure ≥90 mm Hg and/or (iii) use of an antihypertensive drug. Diabetes mellitus was indicated by the use of a hypoglycemic drug or insulin [12]. The diagnoses of CHD, AF, and tumors were confirmed by medical history.

Follow-up
Each patient medical record was reviewed by physicians every year of the baseline assessment.

Prognostic value of plasma NT-proBNP levels in very old patients with moderate renal insufficiency in China Abstract
Background. The N-terminal pro-brain natriuretic peptide (NT-proBNP) has an important prognostic value in chronic renal insufficiency; however, most studies have been conducted in patients with endstage renal disease (ESRD). In this study we evaluated the prognostic significance of NT-proBNP in very old patients with stage 3 chronic kidney disease (CKD) and compared its prognostic value in CKD3a versus CKD3b patients.

Statistical analysis
Continuous variables with a normal distribution are expressed as mean and standard deviation. Categorical variables are expressed as numbers and percentages. The NT-proBNP level is presented as a continuous variable (after natural logarithmic transformation). All analyses were performed at a median followup interval of 3.8 years (IQR 1.5-6.1). Continuous variables (demographic and clinical characteristics) were compared between groups by analysis of variance or Cuzick's nonparametric test for trend. Proportions were compared using the   Table 2).

Patient characteristics
Results of the ROC analysis indicated that NT-proBNP had a reasonable accuracy for predicting all-cause death and MACEs. The area under the ROC curve was 0.697 (95% CI 0.616-0.778, p < 0.001) for all-cause death (. Fig. 1a). The cutoff NT-proBNP level for predicting allcause death was 411.3 pg/ml and had a maximum Youden index of 0.360. For MACEs, the area under the ROC curve was 0.616 (95% CI 0.530-0.701, p = 0.011) (. Fig. 1b), the cut-off NT-proBNP level for predicting MACEs was 411.3 pg/ml and the maximum Youden index was 0.241. Assessment of endpoints during the follow-up period showed that in the CKD3a group, 65 patients (55.6%)died and 48 patients (41.0%) underwent MACEs. In the CKD3b group, 36 patients (70.6%) died and 19 (37.3%) suffered MACEs. Kaplan-Meier survival analysis showed that survival was lower in the CKD3b group (p = 0.037), but the incidence of MACEs did not differ significantly between the two subgroups (p = 0.563) (. Fig. 2).
Results of Cox proportional hazard model analysis showed that in patients with CKD3a, a higher NT-proBNP level was associated with a higher risk of all-cause death (HR 2.350, 95% CI 1.906-6.091, p = 0.039) and MACEs (HR 3.025, 95% CI 1.024-8.940, p = 0.045) after adjusting for multiple risk factors (model 3); however, in patients with CKD3b, NT-proBNP level had no predictive value for all-cause death and MACEs after adjusting for multiple risk factors (model 3) (. Table 2). By interaction analysis, the multiple adjusted HR values of NT-proBNP to predict allcause death (2.350 vs. 0.325, p = 0.009) and MACEs (3.025 vs. 2.457, p = 0.012) were significantly higher in patients with CKD3a compared with CKD3b.
In patients with CKD3a the ROC analysis indicated that NT-proBNP had a reasonable accuracy for predicting all-cause death but not for predicting MACEs. The area under the ROC curve was 0.699 (95% CI 0.604-0.794, p < 0.001) for all-cause death (. Fig. 3a). The cut-off NT-proBNP level for predicting all-cause death was 288.9 pg/ml and had a maximum Youden index of 0.346. In patients with CKD3b, the area under the ROC curve was 0.588 (95% CI 0.484-0.692, p = 0.106) for MACEs (. Fig. 3b). The ROC analysis indicated that NT-proBNP had no value for predicting all-cause death and MACEs in patients with CKD3b.

Discussion
The main findings of this study include the following: first, in very old patients with stage 3 CKD, the prevalence of hypertension, beta-blocker use, and NT-proBNP levels increased with the deterioration of renal function, but Hb levels decreased. Secondly, baseline NT-proBNP level was an independent predictive factor for all-cause death and MACEs in patients with stage 3 CKD, but echocardiography parameters weakened the predictive value of NT-proBNP. In particular, the risk of all-cause death and MACEs was elevated in patients with NT-proBNP levels ≥ 411.3 pg/ml in our study, suggesting this value as the cut-off NT-proBNP level for prediction of all-cause death and MACEs in old patients with stage 3 CKD. Thirdly, the risk of all-cause death was significantly higher in patients with CKD3b than in patients with CKD3a, and the prognostic value of NT-proBNP was better in CKD3a than in CKD3b.
The CKD is increasingly being recognized as a global public health problem substantially contributing to adverse clinical outcomes. Advanced age is a well-known risk factor for CKD [1], which is often associated with anemia, hypertension, and proteinuria, factors that increase the risk of cardiovascular disease and all-cause death. The increased NT-proBNP level and prevalence of hypertension, and decreased Hb level with the deterioration of renal function observed in our study, are consistent with the findings of previous studies [3]. Zhang et al. [2] found that the prevalence of CKD in the Chinese population was 10.8%, which is lower than the prevalences of 7.7% in the USA and 4.2% in Norway, and the prevalence of CKD3 was 1.6% [3,13]. Lin et al. [14] found that stage 3 CKD was more common in older patients than in younger patients (33.5% vs. 4.1%) in Asia. Quality of life is better in stage 3 CKD than in ESRD and disease progression and adverse outcomes, such as kidney failure, cardiovascular disease, and premature death can be prevented or delayed when treatment is initiated in the early stages of disease [15]; therefore, patients with stage 3 CKD should receive comprehensive treatment. The results of two metaanalyses showed that the risk of adverse kidney outcomes and all-cause mortality is substantially higher in patients with eGFR values of 30-44 ml/min/1.73 m 2 compared to those with eGFR values of 45-60 ml/min/1.73 m 2 [16,17]. This suggests that it may be appropriate to subdivide stage 3 CKD into CKD3a and CKD3b in the updated guidelines [10].
Brain natriuretic peptide (BNP) is produced as a prehormone (proBNP) by cardiac myocytes in response to ventricular wall stretching and tension [18]. Compared with BNP, NT-proBNP has a longer half-life and is more stable in vitro, making it easier to detect. Previous studies have shown that NT-proBNP is a significant predictor of cardiovascular disease and all-cause death in various populations [19,20]; however, these studieswere performed mainly in patients with ESRD on hemodialysis [5,6]. The present study is the first to report that the NT-proBNP level is an independent predictive factor for all-cause death and MACEs in very old (≥80 years old) patients with stage 3 CKD. These results indicate that monitoring NT-proBNP level is helpful for assessing the risk of all-cause death and MACEs, correcting risk factors, and delaying the deterioration of renal function in this patient population.
Thereareseveralpossibleexplanations for the association between elevated NT-proBNP level and risk of all-cause death and MACEs in our study. Firstly, an elevated plasma NT-proBNP level is associated with a variety of factors including old age, anemia, and combined hypertension and coronary heart disease, which are the risk factors for all-cause death and MACEs [21]. Secondly, increased plasma NT-proBNP level may indicate more advanced renal dysfunction, and deterioration of renal function is associated with all-cause death and MACEs [22]. Thirdly, elevated levels of plasma NT-proBNP denote impaired cardiac function, including latent structural heart diseases, cardiac volume overload, and myocardial damage [23]. In this study, NT-proBNP was not an independent predictor of all-cause death and MACEs after adjusting for echocardiography parameters (LVEF, LAD, and LVMI). Both abnormal echocardiography parameters and elevated plasma NT-proBNP level denote impaired cardiac structure and function, but no published studies have compared the prognostic value of NT-proBNP and echocardiography parameters.
The prognosis based on the NT-proBNP level has not been thoroughly studied in CKD patients, and there is no standard cut-off value for NT-proBNP to predict adverse events in this population. Fu et al. [24] reported that NT-proBNP predicted death with a cut-off value of 369.5 pg/ml in non-CKD patients and a cut-off value of 2584.1 pg/ml in CKD patients. In the present study, the risks of all-cause death and MACEs were particularly increased in patients with NT-proBNP levels ≥ 411.3 pg/ml, suggesting this value as an appropriate cut-off NT-proBNP level for prediction of all-cause death and MACEs in very old patients with stage 3 CKD. The cutoff value for predicting all-cause death suggested by Fu et al. was significantly higher than that of the present study (2584.1 pg/ml vs. 411.3 pg/ml). Possible explanations for this discrepancy include differences in the percentages of patients with heart failure, AF, and anemia and levels of NT-proBNP, which were higher in the study by Fu et al. [24]. In addition, the study population of Fu et al. consisted of patients with coronary artery disease, whereas patients with coronary artery disease accounted for 79.1% of the patients in our study. It is difficult to determine an explicit cut-off value for predicting all-cause death and MACEs in patients with CKD because the range of plasma NT-proBNP values is wide, especially in the old; therefore, further study is needed to determine the best cut-off value for plasma NT-proBNP in very old patients with stage 3 CKD.
Patients with CKD are at increased risk of mortality, cardiovascular disease, and less commonly, progression to ESRD [25] and these risks are increased with deterioration of renal function [26]. In the present study, the mortality rate was significantly higher in patients with CKD3b than in those with CKD3a (70.6% vs. 55.6%, p = 0.037) during follow-up, but the incidence of MACEs did not significantly differ (41.0% vs. 37.3%, p = 0.563). This discrepancy may be due to death occurring before the occurrence of MACEs in patients with CKD3b. In addition, the prognostic value of NT-proBNP for predicting mortality and MACEs was better in CKD3a than in CKD3b. The reason for this discrepancy is not clear but may be associated with the small sample size in this study. There is a different mode of clearance of the NT-proBNP in patient with CKD3b, and many other factors that have not been discussed in this study affect the prognosis of patients with CKD3b. Further studies are needed to compare the prognostic value of NT-proBNP in CKD3a versus CKD3b.
Several limitations should be considered when interpreting our results. Firstly, the study population was relatively small; therefore, large studies are needed to confirm our results. Secondly, in this patient population, numerous factors can affect plasma NT-proBNP levels and patient outcomes. Although the results were adjusted for multiple covariates that may be associated with circulating NT-proBNP levels and outcomes, there is a possibility of residual confounding factors. Thirdly, in this study, an inclusion criterion was eGFR < 60 ml/min/1.73 m 2 calculated by the MDRD (Modification of Diet in Renal Disease) formula, which may be less representative of actual eGFR in very old patients because of reduced muscle mass [27].

Conclusion
The NT-proBNP level is a strong and independent prognostic factor for all-cause death and MACEs in very old patients with CKD stage 3. Furthermore, patients with a NT-proBNP level ≥ 411.3 pg/ml have a significantly higher risk of allcause death and MACEs, suggesting the potential usefulness of this NT-proBNP level as a cut-off value for predicting adverse events in this population. In addition, the prognostic value of NT-proBNP was better in CKD3a than in CKD3b.