Chronic kidney disease (CKD) is diagnosed when the estimated glomerular filtration rate (eGFR) is <60 mL/min/1.73 m² for 3 consecutive months, or abnormal renal structure or function other than decreased eGFR for over 3 months (1). The main cause of death among patients with CKD is cardiovascular disease, including myocardial infarction and heart failure (HF) (2–4). When HF occurs in CKD patients, the retention of sodium and fluid leads to increases in vascular tension and cardiac preload (5). The increase in ventricular pressure induces the release of biomarkers, such as natriuretic peptide (6). Therefore, biomarkers of myocardial stretching are often used for the diagnosis and prognosis of HF (7). Brain natriuretic peptide (BNP) and N-terminal pro B type natriuretic peptide (NT-proBNP) are important indicators for the diagnosis, prediction, and treatment evaluation of HF, and NT-proBNP is superior to BNP in this prognostic assessment (8–11). Although they both are important indicators of HF (12), NT-proBNP is rarely used as a diagnostic biomarker for HF in patients with end-stage renal disease (ESRD) (13), because ESRD patients without HF have high levels of NT-proBNP due to decreased renal elimination, volume overload, hypertension, and increased left ventricular hypertrophy (14). NT-proBNP is mostly eliminated by glomerular filtration, which explains the strong influence of renal function on NT-proBNP levels (15).

When using NT-proBNP for the diagnosis of HF in patients with renal insufficiency, the diagnostic cut-off value must be adjusted according to the eGFR (16). The cut-off value of NT-proBNP for the diagnosis of acute HF (AHF) in patients with CKD stages 3–5 (eGFR <60 mL/min/1.73m²) is higher than that in patients with CKD stages 1–2 (eGFR >60 mL/min/1.73m²), ranging from 1,200–6,000 pg/mL, and both points are higher than the standard “Januzzi cut-off point.” In general, the specificity and sensitivity of the NT-proBNP concentration for the diagnosis of AHF in patients with CKD stages 3–5 are low (7). To date, there have been no large-scale prospective clinical trials determining the diagnostic cut-off value of NT-proBNP for HF in CKD patients (17).

It is important to use a reliable serological indicator to predict HF in CKD patients with different eGFRs. Low molecular-weight proteins have been used to estimate the value of eGFR (18), and the most promising biomarker is cystatin C (CysC) (19). CysC is a non-glycosylated, 13-kDa basic protein that belongs to the cystatin superfamily of cysteine protease inhibitors. It is produced by all nucleated cells, unaffected by muscle mass (unlike creatinine) (20), and considered a replacement for serum creatinine (sCr) for the estimation of eGFR (21). CycC is also used for early diagnosis of renal damage (22).

In the present study, we investigated the prognostic potential of the ratio of NT-proBNP to CysC compared with NT-proBNP alone for cardiac dysfunction in Chinese CKD patients.

The baseline characteristics of 396 hospitalized patients with CKD are summarized in Table 1. Patients were divided into two groups according to their cardiac function assessed following the European Society of Cardiology (ESC) Guidelines for Chronic Heart Failure 2021: the HF group (n = 216) and the control group (n = 180).

The ROC curve showed that the cut-off value of the NT-proBNP/CysC^1.53 ratio was 847.964.

Pearson correlation analysis showed that the level of NT-proBNP was not correlated with the NT-proBNP/CysC^1.53 ratio (r = 0.376, p = 6.909; Figure 1).

The E/e', GAS, GLS, LAVI, LVMI, serum urea, sCr, hs-cTnT, NT-proBNP, and NT-proBNP/CysC^1.53 ratio in the HF group were significantly higher than those in the control group, while the LVEF was significantly lower than that in the control group (p < 0.05). No significant differences in age, CysC, LVDd, and LVDS were observed between the two groups (p > 0.05).

Then, we used ROC curve analysis to identify the predictors that were statistically different between the two groups. The ROC curves for eight potentially predictive factors (NT-proBNP/CysC^1.53, NT-proBNP, GAS, GLS, LVMI, LAVI, E/e', and LVEF) were plotted. The following AUC values for the eight factors were calculated: AUC (NT-proBNP/CysC^1.53) = 0.815, AUC (LAVI) = 0.798, AUC (E/e') = 0.747, AUC (NT-proBNP) = 0.726, AUC (LVEF) = 0.646, AUC (GAS) = 0.570, AUC (LVMI) = 0.561, and AUC (GLS) = 0.535. The AUC for the NT-proBNP/CysC^1.53 ratio was greater than those for NT-proBNP and LVEF. The diagnostic cut-off value for the NT-proBNP/CysC^1.53 ratio was 847.964, with a sensitivity of 78.24% and a specificity of 69.44%. The diagnostic cut-off value for NT-proBNP was 8,198 pg/mL, with a sensitivity of 52.31% and a specificity of 84.44% (Figure 2).

The clinical data of CKD patients from another local hospital were used to validate the new predictors and the truncation value. The baseline characteristics of 68 hospitalized patients with CKD are summarized in Table 2. The results showed a sensitivity of 70.27% and a specificity of 67.74%.

Renal function and cardiac function are interdependent in patients with CKD. HF has been identified as an independent risk factor for all-cause mortality in hospitalized CKD patients. The morbidity and mortality rates are high in CKD complicated with cardiac insufficiency and have been increasing in recent years (26). Thus, early diagnosis of cardiac insufficiency may improve the prognosis of patients with CKD.

HF can be divided into HFrEF, HFmrEF, and HFpEF. It has been reported that HFpEF accounts for more than half of all hospital admissions for HF (27). Early diagnosis of HFpEF is more difficult than that of HFrEF (LVEF ≤ 40%) and HFmrEF. The widely used New York Heart Association Classification System classifies HF based on subjective feelings of patients; therefore, the results may not be accurate. Even when the objective HFA-PEFF diagnostic algorithm is used, a cardiovascular ultrasound system (e.g., GE Vivid E90 Ultrasound System designed by GE Healthcare) is always needed for the classification of HF. However, in China, most primary and intermediate hospitals do not have such equipment. NT-proBNP is a diagnostic marker for HF, and its serum levels are closely associated with renal insufficiency and age. Therefore, the diagnostic cut-off value of NT-proBNP for HF should be adjusted based on the level of eGFR. However, no large-scale prospective clinical trials have provided an accurate cut-off value of NT-porBNP for the diagnosis of HF in patients with CKD (17).

By comparing the baseline characteristics of the HF and control groups, we found that the NT-proBNP/CysC^1.53 ratio differed significantly between the two groups. Patients with HF showed a significantly higher NT-proBNP/CysC^1.53 ratio than the control group, suggesting that a high NT-proBNP/CysC^1.53 may be related to the occurrence of cardiac dysfunction in CKD. Our results also demonstrated that the NT-proBNP/CysC^1.53 ratio was a more reliable predictor of HF than NT-proBNP, GAS, GLS, LVMI, LAVI, E/e', and LVEF in patients with CKD. According to the Kidney Disease: Improving Global Outcomes (KDIGO) 2012 clinical practice guideline for the evaluation and management of CKD (1), CysC is the best indicator of eGFR and has been used for eGFR calculation. Therefore, in this study, we used both NT-proBNP and CysC to evaluate the cardiac function of patients with CKD. To our knowledge, this is the first study to explore the potential of NT-proBNP combined with CysC as a predictor of HF in CKD patients.

A previous study (28) proposed that LVEF is an indicator of cardiac insufficiency in patients with early diagnosed CDK. However, the use of LVEF as a diagnostic marker cannot rule out cases with HFpEF. In the present study, patients with HFpEF, HFmrEF, and HFrEF were included in the HF group for the analysis of cardiac function in CKD patients, and thus, this study may provide more comprehensive results.

Gao et al. (29) proposed that the level of NT-proBNP can indicate HF in different eGFR intervals. However, whether the same cut-off value of NT-proBNP can be used for the diagnosis of HF over a wide range of eGFR values remains unknown. Moreover, they only included patients with HFrEF (LVEF ≤40%). In our study, both NT-proBNP and CysC levels were used to predict cardiac insufficiency, which may better reflect individual differences.

We further validated the results using the clinical data of patients from another local hospital. The NT-proBNP/CysC^1.53 ratio accurately determined the cardiac function of these patients.

It should be noted that the diagnostic criteria that we have derived in this study were simpler than those recently published in ESC Chronic Heart Failure Guidelines 2021 and require fewer tests. In addition, laboratory test results have less interference and are more accurate than the results of ultrasound scans, and therefore, may have a wider range of applications.

The present study has some limitations. There were only 9 (2.27%) cases with stage 1–2 CKD. Due to compliance issues, patients with symptoms or signs of HF and with a HFA score between 2 and 4 did not undergo a diastolic pressure test or invasive hemodynamic measurement. Because there were only 12 (3.03%) cases with symptoms or signs of HF and with a HFA score between 2 and 4, it might not significantly affect the overall results. We only recruited a small number of Chinese subjects from another local hospital to validate the results, which may not represent the population in other areas. Taking into account the differences in regions, populations, and examinations (e.g., altitude, ethnicity, equipment, reagent), it is recommended that the current findings be validated in a local population. In summary, our study showed that the NT-proBNP/CysC^1.53 ratio was a predictor of cardiac insufficiency in patients with CKD and might be used for the early detection of HF in this population in clinical settings.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Ethics Committee of Gezhouba Central Hospital of Sinopharm. The patients/participants provided their written informed consent to participate in this study.

SW, XW, and JL conceived and designed research. ML, CL, XY, HA, and TY collected data and conducted research. SW and YZ analyzed and interpreted data. YH and ML wrote the initial paper. SW, ML, YZ, and XL revised the paper. SW had primary responsibility for final content. All authors read and approved the final manuscript.

This study was supported by the 2017 Three Gorges University Youth Science Foundation Project (No. KJ2017A011).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.